Hydrodynamics and Sedimentation in Wave-Dominated Coastal -- B_ Greenwood and R_A_ Davis (Eds_) -- Developments in Sedimentology 39, 1984 -- Elsevier -- 9780080869537 -- 75d36386eed2b37eb5e4bb0e8dc1b259 -- Anna’s Archive

DEVELOPMENTS IN SEDIMENTOLOGY 39
HYDRODYNAMICS AND SEDl MENTATION

IN WAVE- DOMINATED COASTAL ENVl RONMENTS
FURTHER TITLES IN THIS SERIES
VOLUMES 1.2.3.5.8 and 9 are out of print
4 F.G. T I C K E L L
THE TECHNIQUES O F SEDIMENTARY MINERALOGY
6 L. V A N D E R P L A S
T H E IDENTIFICATION O F DETRITAL FELDSPARS
7 S. D Z U L Y N S K I and E.K. W A L T O N
SEDIMENTARY FEATURES O F FLYSCH AND GREYWACKES
10 P.McL.D. D U F F , A. H A L L A M and E.K. W A L T O N
CYCLIC SEDIMENTATION
11 C.C. R E E V E S Jr.
INTRODUCTION T O PALEOLIMNOLOGY
12 R.G.C. B A T H U R S T
CARBONATE SEDIMENTS AND THEIR DIAGENESIS
13 A.A. M A N T E N
SILURIAN R E E F S O F GOTLAND
14 K.W. G L E N N I E
DESERT SEDIMENTARY ENVIRONMENTS
15 C.E. W E A V E R and L.D. P O L L A R D
THE CHEMISTRY O F CLAY MINERALS
16 H.H. R I E K E I l l and G . V . C H I L I N G A R I A N
COMPACTION O F ARGILLACEOUS SEDIMENTS
11 M.D. P I C A R D and L.R. HIGH Jr.
SEDIMENTARY STRUCTURES O F EPHEMERAL STREAMS
18 G.V. C H I L I N G A R I A N and K.H. W O L F , Editors
COMPACTION O F COARSE-GRAINED SEDIMENTS
19 W. S C H W A R Z A C H E R
SEDIMENTATION MODELS AND QUANTITATIVE STRATIGRAPHY
20 M . R . W A L T E R , E d i t o r
STROMATOLITES
21 B. V E L D E
CLAYS AND CLAY MINERALS I N NATURAL AND SYNTHETIC SYSTEMS
22 C.E. W E A V E R and K.C. B E C K
MIOCENE O F THE SOUTHEASTERN UNITED STATES
23 B.C. H E E Z E N , Editor
INFLUENCE O F ABYSSAL CIRCULATION O N SEDIMENTARY
ACCUMULATIONS IN SPACE AND TIME
24 R.E. G R I M and G U V E N
BENTONITES
25A G . L A R S E N and G . V . C H I L I N G A R , Editors
DIAGENESIS IN SEDIMENTS AND SEDIMENTARY ROCKS, I
26 T . S U D O and S. S H I M O D A , Editors
CLAYS AND CLAY MINERALS O F JAPAN
21 M.M. M O R T L A N D and V.C. F A R M E R , Editors
INTERNATIONAL CLAY CONFERENCE 1918
28 A . N I S S E N B A U M , E d i t o r
HYPERSALINE BRINES AND EVAPORITIC ENVIRONMENTS
29 P. T U R N E R
CONTINENTAL R E D BEDS
30 J.R.L. A L L E N
SEDIMENTARY STRUCTURES
31 T . S U D O , S. S H I M O D A . H . Y O T S U M O T O and S . A I T A
ELECTRON MICROGRAPHS O F CLAY MINERALS
32 C.A. N I T T R O U E R , E d i t o r
SEDIMENTARY DYNAMICS O F CONTINENTAL SHELVES
33 G.N. B A T U R I N
PHOSPHORITES ON THE SEA FLOOR
34 J.J. F R I P I A T , E d i t o r
ADVANCED TECHNIQUES FOR CLAY MINERAL ANALYSIS
35 H. V A N O L P H E N and F. V E N I A L E , Editors
INTERNATIONAL CLAY CONFERENCE 1981
36 A. I I J I M A , J.R. H E I N and R. S I E V E R . Editors
SILICEOUS DEPOSITS IN THE PACIFIC REGION
31 A. S I N G E R and E. G A L A N , Editors
PALYGORSKITE-SEPIOLITE: OCCURRENCES, GENESIS AND USES
38 M.E. B R O O K F I E L D and T.S. A H L B R A N D T , Editors
EOLIAN SEDIMENTS AND PROCESSES
DEVELOPMENTS IN SEDIMENTOLOGY 39
HYDRODYNAMICS AND
SEDIMENTATION IN
WAVE- DOMINATED
COASTAL ENVIRONMENTS
Edited by
B. GREENWOOD and R.A. DAVIS, Jr.

Departments of Geography and Geology, Scarborough College, University
of Toronto, Scarborough, Ont. M1 C 1A4 (Canada)
Department of Geology, University of South Florida, Tampa, FL 33620
(U.S.A.)
Reprinted from Marine Geology, Vol. 60 (1-4)
ELSEVIER
Amsterdam - Oxford - New York - Tokyo 1984
ELSEVIER SCIENCE PUBLISHERS B.V.
1 Molenwerf
P.O. Box 211,1000AE Amsterdam, The Netherlands
Distributors f o r the United States and Canada.
ELSEVIER SCIENCE PUBLISHING COMPANY INC.

52,Vanderbilt Avenue
New York, N.Y. 10017
Library or ('ongrr\s Cataloging in Publication Data

Main entry under title:
Hydrodynamics and sedimentation in wave-dominated

coastal environments.
(Developments in sedimentoloa ; 39)

Papers from a symposium held at the Eleventh Inter-
national Congress on Sedimentology which met at Hamilton,
Ontario in Aug. 1~82.
"Reprinted f r o m Marine geology, vol. 60 (1-4)."
Bibl5ography: p.
1. Coast changes--Congresses. 2. Sedimentation and
deposition--Congresses. 3. Wdrobamics--Congresses.
I. Greenwood, B. (Brian) 11. Davis, Richard A.
(Richard Albert), 1937- .
111. International
Congress on Sedinentology. IV. Marine geology.
V. Series.
GB450.%.H96 1984 551.4'57 84-10321
ISBN 0-444-42400-8 (U.S.)
ISBN 0-444-42400-8
(Vol. 39)
ISBN 0-444-41238-7
(Series)
@ Elsevier Science Publishers B.V., 1984

All rights reserved. N o part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, mechanical, photocopy-
ing, recording or otherwise without the prior written permission of the publisher, Elsevier
Science Publishers B.V./Science & Technology Division, P.O. Box 330,1000 AH Amster-
dam, The Netherlands.
Special regulations for readers in the USA - This publication has been registered with the
Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be ob-
tained from the CCC about conditions under which photocopies of parts of this publica-
tion may be made in the USA. All other copyright questions, including photocopying
outside of the USA, should be referred to the publisher.
Printed in The Netherlands

CONTENTS
Preface ................................................... VII
Waves, long waves and nearshore morphology

A.J. Bowen and D.A. Huntley (Halifax, N.S., Canada) . . . . . . . . . . . . . . . . . . 1
Spatial and temporal variations in spectra of storm waves across a barred nearshore
R.G.D. Davidson-Arnott and D.C. Randall (Guelph, Ont., Canada) . . . . . . . . . . 15
Waves, currents, sediment flux and morphological response in a barred
nearshore system
B. Greenwood and D.J. Sherman (Scarborough, Ont., Canada). . . . . . . . . . . . . 31
Landward migration of inner bars
‘T. Sonamura and I. Takeda (Ibaraki, Japan) . . . . . . . . . . . . . . . . . . . . . . . . . 63
Sediment flux and equilibrium slopes in a barred nearshore
B. Greenwood (Scarborough, Ont., Canada) and P.R. Mittler (Vancouver,
B.C., Canada). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Sediment transport and morphology a t the surf zone of Presque Isle, Lake Erie,
Pennsylvania
D. Nummedal, D.L. Sonnenfeld (Baton Rouge, La., U.S.A.) and K. Taylor
(Erie, Pa., U.S.A.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . :. . . . . . . . 99
Sedimentology and morphodynamics of a microtidal beach, Pendine Sands,
SW Wales
C.F. Jag0 (Gwynedd, U.K.) and J. Hardisty (Bristol, U.K.) . . . . . . . . . . . . . . . 123
High-frequency sediment-level oscillations in the swash zone
A.H. Sallenger, Jr. and B.M. Richmond (Menlo Park, Calif., U.S.A.) . . . . . . . . . 155
Wave-formed structures and paleoenvironmental reconstruction
H.E. Clifton and J.R. Dingler (Menlo Park, Calif., U.S.A.) . . . . . . . . . . . . . . . . 165
Boundary roughness and bedforms in the surf zone
D.J. Sherman (Woods Hole, Mass., U.S.A.) and B. Greenwood (Scarborough,
Ont., Canada). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Tidal-cycle changes in oscillation ripples o n t h e inner part of an estuarine sand flat
J.R. Dingler and H.E. Clifton (Menlo Park, Calif., U.S.A.) . . . . . . . . . . . . . . . . 219
Bedforms and depositional sedimentary structures of a barred nearshore system,
eastern Long Island, New York
R.C. Shipp (Orono, Me., U:S.A.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Beach and nearshore facies: southeast Australia
A.D. Short (Sydney, N.S.W., Australia). . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Structures in deposits from beach recovery, after erosion by swell waves around
the southwestern coast of Aruba (Netherlands Antilles)
J.H.J. Terwindt (Utrecht, The Netherlands), C.H. Hulsbergen (Emmeloord,
The Netherlands) and L.H.M. Kohsiek (The Hague, The Netherlands) . . . . . . . . 283
What is a wave-dominated coast?
R.A. Davis, Jr. (Tampa, Fla., U.S.A.) and M.O. Hayes (Columbia
S.C.,U.S.A.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Shoreface morphodynamics o n wave-dominated coasts
A.W. Niedoroda (Houston, Texas, U.S. A. ), D.J.P. Swift (Dallas, Texas, U.S.A. ),
T.S. Hopkins (Upton, N.Y ., U.S.A.) and Chen-Mean Ma (Houston, Texas, U S . A . ) 331
Control of barrier island shape b y inlet sediment bypassing: East Frisian Islands,
West Germany
D.M. FitzGerald (Boston, Mass., U.S.A.), S. Penland and D. Nummedal
(Baton Rouge, La., U.S.A.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
Coarse clastic barrier beaches: A discussion of the distinctive dynamic and
morphosedimentary characteristics
R.W.G. Carter (Londonderry, Northern Ireland) and J.D. Orford (Belfast,
Northern Ireland) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Shoreface translation and the Holocene stratigraphic record: Examples from Nova
Scotia, the Mississippi Delta and eastern Australia
R. Boyd (Halifax, N.S., Canada) and S. Penland (Baton Rouge, La., U.S.A.). . . . 391
Holocene sedimentation of a wave-dominated barrier-island shoreline : Cape
Lookout, North Carolina
S.D. Heron, Jr. (Durham, N.C., U.S.A. ), T.F. Moslow (Baton Rouge, La., U.S.A.),
W.M. Berelson (Los Angeles, Calif., U.S.A.), J.R. Herbert (Houston, Texas, U.S.A.),
F.A. Steele I11 (Casper, Wyo., U.S.A.) and K.R. Swanson (San Francisco,
Calif., U.S.A.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Reconstruction of paleo-wave conditions during the Late Pleistocene from marine
terrace deposits, Monterey Bay, California
W.R. Dupr6 (Houston, Texas, U.S.A.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Reconstruction of ancient sea conditions with an example from the Swiss Molasse
P.A. Allen (Cardiff, U.K.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
VII
PREFACE
This volume evolved from a symposium held at the Eleventh International

Congress on Sedimentology which met at Hamilton, Ontario in August 1982
under the auspices of the International Association of Sedimentologists.
The symposium entitled “Coastal Environments Dominated by Waves”
was convened jointly by Brian Greenwood (representing the primary sponsors,
the Geological Association of Canada and Canadian Society of Petroleum
Geologists) and Skip Davis (representing the Society of Economic Paleontol-
ogists and Mineralogists, the co-sponsor).
The objective of the symposium was to provide a forum for discussion of
the links between hydrodynamics and sedimentation in both modern and
ancient wave-dominated coastal environments. Of the 36 original papers 23
are presented in this volume: they range in content from theory and experi-
ment on recent sedimentation under waves (papers 2-9), to. the interpreta-
tion of bed forms and sedimentary structures under a range of environmental
constraints (papers 10-15), to macro-scale sedimentation, morphodynamics
and the stratigraphic record (papers 16-21) and finally to explicit recon-
struction of ancient environmental constraints on sedimentation revealed in
the rock record (papers 22 and 23). Each paper is an original research con-
tribution and has been selected to indicate the current levels and directions
of research on sedimentation in wave-dominated coastal environments.
We would like to thank the authors for their patience during the produc-
tion of this volume and the many referees who gave their time freely to rigor-
ously review the manuscripts.
BRIAN GREENWOOD and RICHARD A. DAVIS, Jr. (Editors)

This Page Intentionally Left Blank
Marine Geology, 60 (1984) 1-13 1
Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
WAVES, LONG WAVES AND NEARSHORE MORPHOLOGY
A.J. BOWEN and D.A. HUNTLEY

Department o f Oceanography, Dalhousie University, Halifax, N . S . B3H 451 (Canada)
(Received August 4, 1983; revised and accepted November 1, 1983)
ABSTRACT
Bowen, A.J. and Huntley, D.A., 1984. Waves, long waves and nearshore morphology. In:
B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in
Wave-Dominated Coastal Environments. Mar. Geol., 60: 1-13.
Recent field measurements o n beaches of different slopes have established that wave
motion at periods substantially longer than the incident waves dominates the velocity
field close to the shore. Analysis of a number of extensive data sets shows that much of
this long wave motion is in the form of progessive edge waves, though forced wave
motion, standing edge waves and free waves propagating away from the shore may also
contribute t o the energy.
Theoretically, the drift velocities in bottom boundary layers due to edge waves show
spatial patterns of convergence and divergence which may move sediment to form either
regular crescentic o r cuspate features when only one edge wave mode dominates, or a
bewildering array of bars, bumps and holes when several phase-locked modes exist to-
gether.
Convincing field demonstration of the link between nearshore topography and edge
waves only exists for the special case of small-scale beach cusps on steep beaches, formed
by edge waves at the subharmonic (twice the period) of the incident waves. At longer
periods the link is proving more difficult to establish, due to the longer time-scales of
topographic changes, the interaction between pre-existing topography and the water
motion, and the observation'of broad-banded edge wave motion which is not readily
linked to topography with a well-defined scale.
These ideas are, however, central to the study of nearshore processes, as most of the
plausible alternate hypotheses d o not seem to lead to quantitative predictions. Clearly,
further theoretical and observational work is essential.
INTRODUCTION
I t is clear from the most superficial examination that the nearshore zone
is a very active area of sediment transport. This activity is observed on both
long time scales, seasonal changes in beach profiles being evident in many
parts of the world, and on time scales of a few hours, during a major storm
for example. I t was natural that the early attempts to explain these pro-
cesses focussed on the obvious driving mechanism, the waves moving in from
deep water. The belief that these incident waves are the only important
hechanism was not often explicitly stated, perhaps because it was already
known that the strong, wavedriven currents occurred in and just outside the
0025-3227/84/$03.00 0 1984 Elsevier Science Publishers B.V.

2
surf zone. Shepard et al. (1941) had pointed to the possible geological
importance of rip currents. However, an expectation that the waves them-
selves, not the nearshore currents, are the dominant process was implicit
in the large number of laboratory experiments in long, narrow wave flumes
that sought to explain the shape of the beach (Rector, 1954) or the exist-
ence of longshore bars (Johnson, 1949; Iwagaki and Noda, 1963).
An interesting consequence of the assumption that the incoming waves are
the dominant process was the very limited set of parameters that seemed to
be involved in determining characteristics of the beach such as the beach
profile. In the idealized laboratory experiment, a single wave frequency u
(or wavelength L ) and height H characterized the waves, while the sand was
apparently adequately described by a density and a typical grain size d. Any
further property of the system was then only dependent on a very few non-
dimensional numbers, H / L and H / d particularly. The criteria for the forma-
tion of a longshore bar given by Iwagaki and Noda (1963) are typical of this
approach.
An interesting question is why these very sensible ideas did not lead to a
satisfactory model for on-offshore processes on a beach. One way of focus-
sing on this question is to look at conditions well inside the surf zone. To a
very good approximation, breakers inside the surf zone are “saturated”, that
is, the wave height at any point is limited by the local depth (Fig.1). This is
what might be expected on dimensional grounds (Longuet-Higgins, 1972)
and what is found both in laboratory experiments using monochromatic
waves and in detailed field measurements in conditions of both broad and
narrow spectral distributions of wave energy (Thornton and Guza, 1982).
1201
Hrms= 0.42 h NOV Time
/
/ A
18A%
8 I I 09211
+ 2 2 IOU6
0 100 200 300 400
h (cm)
Fig.1. Saturation: the limitation of wihre height H in the surf zone by the local depth, h
(after Thomton and Guza, 1982).
3
If breakers are saturated, the local wave height is proportional to the depth,
regardless of conditions offshore. How then does the beach face “know”
that there are large waves offshore? To address this question it is useful to
look more closely at the nature of the wave motion very close to the shore,
in the run-up for example.
WAVES AND LONG WAVES
There has recently been considerable interest in the direct measurement

of the swash movement, the run-up, at the shoreline using both run-up
meters, essentially wave staffs laid parallel to the beach, and time-lapse
photography (Holman and Guza, 1984). The observations show that in the
region very close to the shore the dominant wave motion is not normally at
frequencies typical of the incident waves. On steep beaches the run-up shows
significant energy at the subharmonic frequency of the incident waves
(Huntley and Bowen, 1973). On gently sloping beaches the frequencies are
generally much lower and not obviously related to those of the incoming
waves. However, the amplitude of the low-frequency motion is very clearly
related to size of the incident waves. Figure 2‘ shows that the amplitude of
the movement at the shoreline at low frequencies increases linearly with
I I I I I I I 1
0 40 80 I20 I60
H, ( c m )
Fig. 2. The horizontal excursion in the swash at the shoreline RY as a function of the
significant waveheight offshore H, . The open numbers show the magnitude of the signifi-
cant excursion at periods longer than 25 s, those circled at periods less than 25 s (after
Guza and Thornton, 1982).
4
increasing incident wave height while the amplitude at the incident fre-
quency remains constant, as would be expected from the idea of saturation.
The amplitude of the incident waves at the shoreline increases with breaker
height for very small incident waves but reaches a constant value a t a small
value of E (Guza and Bowen, 1976), where
E =
HuZ
-
2gP2
and 0 is the beach slope. Thus we would expect shoreline amplitudes t o be
independent of offshore wave conditions when e is large, which will generally
be the case at incident wave frequencies where the waves are normally large
enough t o break. E is, in fact, an important dimensionless parameter, a surf
similarity measure, characterizing many properties of the system including,
for example, the type of breaking.
The occurrence of low-frequency waves is not unexpected. Munk (1949)
observed similar waves outside the surf zone and suggested that they were
associated with the wave groups which naturally occur with the beating of
several incident wave frequencies. This led t o the adoption of the generic
term “surf beat” for all motions in a frequency range from 30 s t o several
minutes. Despite the early evidence for the existence of such waves, they
did not attract much interest. In retrospect this is surprising because it
should have been evident that the wavelengths associated with these motions
are of the same order as those of major multiple bar structures. There seems
to have been a general impression that these waves were not only long, but
also of low amplitude.
However, the results of Guza and Thornton (1982) shown in Fig.2 suggest
that such low-frequency motions are dominant over a significant region in
the surf zone and that this dominance increases in very severe conditions,
precisely the conditions in which the most active sediment transport is
expected t o occur. It may be useful to emphasize that this dominance is not
of one small quantity relative to another. The orbital velocities associated
with these low frequency motions are of the same order as the currents
associated with the incident waves and the wave-induced nearshore circula-
tion, with typical values of the order of 1m s-’. This is clearly an important
contribution to the total velocity field. T o understand the way in which this
motion influences the nearshore morphology, we need t o know more about
the onshore and longshore structure of this low-frequency activity.
SURF BEAT AND EDGE WAVES
Figure 3 schematically illustrates the principle types of waves that may

contribute to the low-frequency motion in terms of the wave frequency u
and the longshore wavenumber h. The solid lines denote a set of edge-wave
modes whose amplitude is largest at the shoreline and dies away seaward so
that the wave is trapped to the shoreline. Mathematically, these solutions
require the dispersion relation:
5
n = 0 1 2 3
0
Fig.3. The dispersion relationship between wave frequency a and longshore wave number,
h, for a beach slope p = 0.12. n is the modal number of the edge waves ( e ) .1 indicates the
region o f leaky modes and f illustrates areas of forced modes.
u2 = g h sin(2n + 1)0, n = 0, 1, 2 . . . (2)

where n is the mode number shown in Fig.3. If (2n + 1)0 > 7 ~ / 2there are no
trapped solutions, and waves having these (u, h ) values may propagate to, or
from, deep water and are consequently known as “leaky” modes. They
include the normal incident waves.
The spaces between the edge wave modes are combinations of (u, h ) values
which do not satisfy the dispersion relation, eq.2. Any motion here is a
forced wave and the response will theoretically be weaker than at the free
modes for which the response is “resonant”. There is an additional set of
forced modes which do not satisfy the necessary dispersion relationship for
free modes in the on-offshore direction. These waves have a frequency and
wavenumber that do not satisfy the normal conditions for surface gravity
waves. Examples are the harmonics of the incident waves and the set-down
under groups of incident waves (Longuet-Higgins and Stewart, 1962).
There have been a number of observations of wave conditions near the
shore that suggest that both leaky waves and edge waves are important
components of this low-frequency motion (Suhayda, 1974; Huntley, 1976;
Wright et al., 1979; Holman, 1981). However, in order t o separate modes
when a number of different modes, trapped and leaky, occur at the same
frequency a large longshore array is necessary. Using data from the 520 m
long longshore array at the first Nearshore Sediment Transport Study site,
Torrey Pines Beach, California, Huntley et al. (1981) were able t o determine
the observed low-frequency energy spectra as functions of both frequency
6
and longshore wavenumber, and hence plot the observed energy on a dia-
gram like Fig.3. Figure 4 shows an example for longshore currents. Clearly
most of the energy here occurs along the expected edge-wave dispersion
curves. Their analysis also suggests that other forms of motion, possibly
forced waves driven by incoming wave groups, also contribute to the energy.
However, results like those shown in Fig.4 provide unambiguous confirma-
tion that progressive edge waves are indeed present at surf-beat frequencies.
- 0.030
-0.025
-0.020
-rl
XI
m
0
c
m
-0.015 Z
0
<
I
N
-0.010
-0.005
a 6 4 2 2 4 6 8
SOUTHWARDS NORTHWARDS
WAVENUMBER ~ I O - ~ r n - l
NOVEMBER 21st, 1978
Fig.4. Contours of energy in the longshore component of velocity plotted on a ( 0 , A) dia-
gram, with negative and positive wavenumbers representing waves moving southwards or
northwards, respectively, along the beach (after Huntley et al., 1979).
77
LONG WAVES AND NEARSHORE MORPHOLOGY
As previously mentioned, surf beat, whether a leaky mode or an edge

wave, provides length scales which are very much of the same order as many
of the morphological features actually observed (Bowen and Inman, 1971;
Short, 1975; Bowen, 1980). Leaky modes are primarily seen in the on-off
shore direction and are therefore most likely to play a role in the formation
of shore parallel features such as longshore bars. Edge waves, on the other
hand, have a welldefined longshore length scale and are therefore a possible
cause of rhythmic longshore features. However, a single progressive edge
wave provides no longshore variability on the average. It is seen, at different
locations in the longshore direction, as exactly the same wave, simply
arriving at slightly different times. To produce longshore variation at least
two modes of the same frequency are necessary. Bowen and Inman (1971)
showed that a standing edge wave, which can be regarded as two edge waves
of the same frequency, mode and amplitude propagating in opposite direc-
tions, might generate a nearshore morphology of cusps and crescentic bars
whose wavelength would be half that of the edge waves (Fig.5). The drift
velocities, calculated from Hunt and Johns (1963), seemed to provide a
pattern which would form the observed structure.
__-----
-?
J
-10 _ _ _ _ - - - -
I
U
0
u v
u . 0 Y
I
I
.
I
.. ...
..-.
I : !
ORBITAL I VELOCITY D R I F T VELOCITY
I
I
I
I
Fig.5. Theoretical patterns of orbital and bottom drift velocity for a mode n = 1 standing
edge wave whose shoreline elevation ~ ( 0is) shown above. L is the edge wave wavelength.
The solid arrows indicate the orbital velocity at the time when t) = 0 everywhere, the
dashed arrows the velocity half a wave period later (after Bowen and Inman, 1971).
8
Holman and Bowen (1982) have recently extended these ideas to look at
any combination of edge-wave modes of the same frequency. The waves can
be of different modes, different amplitudes and may propagate in the same
direction or in opposite directions. In every case, provided there is some
coherence between the modes, a longshore pattern will result. For example,
the combined elevation of two modes m and r is given by
q = a m X m(x)sin(h,y - a t ) + a,X,(x)sin(h,y - ot + a) (3)
where a m , a, are the shoreline amplitudes of the two modes, Q is the phase
angle between the waves and X , and X , give the dependence of the wave
amplitudes on the offshore co-ordinate x. If we assume that Q is constant,
i.e. that the waves are “phase-locked”, and look at a region in which X m ( x )
- X,(x), then eq.3 can be rewritten as
q = (a, -a,)X(x)sin(h,y - o t ) + 2a,X(x)sin[0.5(hm + h,)y - o t + a/2]
COS[O.~(X, - h,)y - a/2] (41
If a, = a, and A, = -Ar, this result reduces to the standing edge wave
studied by Bowen and Inman (1971), the negative wavenumber showing a
wave propagating in the negative y direction. However, for any two modes
the cosine term in eq.4 produces a longshore modulation which is fixed in
space, provided (Y is constant. This modulation has wavelength r/(Am- h,)
so the wavelength tends to be longer if the waves move in the same direction
than if they move in opposite directions. Again, using expressions derived by
Hunt and Johns (1963), the drift velocity pattern can be computed for any
combination of modes. Holman and Bowen (1982) introduced a simple
sediment transport model to compute the equilibrium slope for any pattern
of drift velocities. Figure 6 shows both the drift velocities and the computed
topography generated by the interaction of edge waves of modes 1 and 2
moving in the same direction. The waves were assumed to be of equal ampli-
tude at the shoreline for the purposes of the illustration. In this case the drift
velocity pattern generates both topographic features and a residual transport
of sediment up the coast.
There is an important difference between the pattern of transport and the
shape of the morphology. The morphology shown is assumed to be in
equilibrium with the wave and drift velocity fields. The condition which
must then be satisfied is that the divergence of the sediment transport vector
is zero. Sediment is moving through the system, but everywhere the trans-
port into any small area is balanced by the transport out. Even, for the case
of crescentic bars where there is no net longshore drift velocity, there re-
mains a local circulation of the sediment over the bar system (Fig.7), a circu-
lation very similar to that deduced from studies of sedimentary structure
and the movement of sediment tracers by Greenwood and Davidson-Arnott
(1979) and Greenwood and Hale (1982).
One interesting result found by Holman and Bowen (1982) was that the
theoretical topography computed in their model did not look familiar. It
9
45 0 r--
t I I I I
2
0.0 5.0 10.0 15.0 2O;O h
h u
UP
i 2.0
X
X I
Fig.6. T h e dimensionless drift velocity, u,/ul, and derived equilibrium bottom contours
for t h e interaction o f two edge waves, modes 1 and 2, of the same frequency, both propa-
gating in the direction of 9 positive. O n e complete wavelength of t h e pattern is shown.
T h e coast is to the left, and 32 is the offshore coordinate (after Holman and Bowen, 1982).
only took on a familiar form, a crescentic bar for example, if it was added on
t o an overall beach slope. This has important implications when making
measurements of sand bars in the field. I t is natural to think of a bar crest
as being at the point where the slope of the bottom is zero, i.e. where the
local depth is a minimum. However the position of this point is dependent
on the general slope of the beach in relation to the size of the perturbation
creating the bar.
Furthermore, in the theoretical model there is no difference conceptually
between a perturbation which gives a clear bar form and a smaller perturba-
tion which produces a low tide terrace, a flattening of the profile rather
than a bar per se. In fact it is clear that if the slope'of the perturbation is
10
0.0
1 I I I I
2.O 4.0 6.O 8.0 10.0
A
X
N
11
nowhere as large as the beach slope, no “bar” will be formed, while on a

gentler beach the same perturbation will produce a bar.
These points may be readily illustrated by a simple model of a bar on a
linearly sloping beach. Consider, for example, a Gaussian-shaped bar, centred
at xo, on a beach of slope 0. The depth h can be written
h =-ox + a exp -(x - 3 ~ ~ ) ’ (5)
The crest of the bar, defined to be where ahlax = 0, will be where
p/2a = ( x - x , ) exp --(x - x0)* (6)
The minimum value of bar amplitude, a , for which a crest will occur will be
where
a(p/%)/a(x-xo) = 0 i.e. (x - x o ) = -0.71 (7)
and this occurs for p/% = 0.43.Larger values of this ratio will not cause a bar
crest to be formed, merely a change in beach slope. Figure 8 shows some
examples and illustrates how the bar “crest” can be significantly displaced
from the location of maximum perturbation. This kind of displacement can
make theoretical interpretation of measured bar locations difficult.
This clearly illustrates how low-frequency motion may influence the
Fig.8. The beach profile described by eq.5, showing the effect of reducing the amplitude
of the bar a. The position of the point of minimum depth, dhldx = 0, is shown by the
arrow.
Fig.7. a. Nearshore topography in the form of a crescentic bar generated by the interac-
tion of two mode 1 edge waves illustrated in Fig.5. b. T h e residual sediment flux, ip, over
this bar system. Transport is in at the horns of the bar, out over the cusp (after Holman
and Bowen, 1981).
12
beach profile in a subtle way without necessardy providing obvious length

scales. Nevertheless, there is a conceptual link between increasing wave-
height, leading to increasing low frequency activity, larger perturbations in
the topography and the formation of bars. This is entirely consistent with
the conclusions of Iwagaki and Noda (1963) that large waves make bars, but
the physical model is very different.
CONCLUSIONS
It is now clear that the velocities dominating the motion in the inner surf
zone and at the shoreline are often those at surf-beat frequencies rather than
incident wave frequencies. Recent measurements also show that the ampli-
tude of this surf beat increases essentially linearly with the amplitude of the
incident waves in deep water. The magnitude of these surf-beat velocities is
on the order of 1m s-’.
Complete identification of this surf-beat energy has not yet been achieved,
but it is clear that edge waves, trapped to the shoreline and propagating
parallel t o the shore, form a significant proportion of the total energy.
Numerical models of drift velocities and sediment transport show that
regular topographic features are expected t o be formed under a single long-
period edge-wave mode, while complex patterns of bars, bumps and holes
can occur when several modes, phase-locked together, occur simultaneously.
Thus many of the observed topographic features on natural shorelines could
be the result of edge-wave activity.
There remain, of course, a number of questions which are the subject of
continuing research. Firstly, how d o growing bars influence the edge waves
which might produce them? Holman and Bowen (1982) ignore any feed-
back between t h e topography and the edgewaves but recent work by Kirby,
Dalrymple and Liu (1981) suggests one approach to this problem. Secondly,
d o edge waves occur in nature over a narrow enough frequency range to
explain the specific topographic length scales observed? Equation 2 shows
that, o n plane beach slopes, the wavelength depends strongly on the fre-
quency. However, there are several possible mechanisms which might lead
t o a narrowing of the wavelength range, amongst them the feedback mech-
anism just mentioned. Thirdly, are different edge-wave modes sufficiently
phase locked in nature to allow the range of features predicted by Holman
and Bowen? Analysis is continuing in an attempt t o answer this.
Despite these remaining uncertainties, the observations and conceptual
ideas presented here lead to quantitative predictions of many features of
bar shapes and scales and these must be tested against accurate measure-
ments in the field. The simple model of a bar on a sloping beach (Fig.8)
shows that this will require more than simply measuring the location of the
bar “crest”.
13
REFERENCES
Bagnold, R.A., 1963. Mechanics of marine sedimentation. I n : M.N. Hill (Editor), The Sea,
Vol. 3. Wiley-Interscience, New York, N.Y., pp.507-528.
Bowen, A.J., 1980. Simple models of nearshore sedimentation : beach profiles and long-
shore bars. I n : S.B. McCann (Editor), T h e Coastline of Canada, Geol. Surv. Can.,
Ottawa, Ont., Pap. 80-10: 1-11.
Bowen, A.J. and Inman, D.L., 1971. Edge waves and crescentic bars. J. Geophys. Res.,
74: 8662-8671.
Greenwood, B. and Davidson-Arnott, R.G.D., 1979. Sedimentation and equilibrium in
wave formed bars: a review and case study. Can. J. Earth Sci., 1 6 : 312-332.
Greenwood, B. and Hale, P.B., 1982. Lagrangian sediment motion in a crescentic near-
shore bar under storm-induced waves and currents. Can. J. Earth Sci., 1 9 : 424-433.
Guza, R.T. and Bowen, A.J., 1976. Resonant interactions f o r waves breaking o n a beach.
I n : Proc. 15th Conf. o n Coastal Engineering. Am. SOC. Civ. Eng., New York, N.Y.,
pp. 560-579.
Guza, R.T. and Thornton, E.B., 1982. Swash oscillations o n a natural beach. J. Geophys.
Res., 8 7 : 483-491.
Holman, R.A., 1981. Infragravity energy in t h e surf zone. J . Geophys. Res., 8 6 : 6442-
6450.
Holman, R.A. and Bowen, A.J., 1982. Bars, bumps and holes: models f o r the generation
of complex beach topography. J. Geophys. Res., 8 7 : 457-468.
Holman, R.A. and G u m , R.T., 1984. Measuring run-up o n a natural beach. Coastal Eng.,
8 : 129-140.
Hunt, J.N. and Johns, B., 1963. Currents produced by tides and gravity waves. Tellus, 15:
343-3 5 1.
Huntley, D.A., 1976. Long period wave motion o n a natural beach. J. Geophys. Res., 8 1 :
6 4 4 1-6 4 4 9.
Huntley, D.A. and Bowen, A.J., 1973. Field observations of edge waves. Nature, 243:
160-162.
Huntley, D.A., Guza, R.T. and Thornton, E.B., 1981. Field observations of surf beat, 1,
progressive edge waves. J. Geophys. Res., 8 3 : 1913-1920.
Iwagaki, Y. and Noda, H., 1963. Laboratory study of scale effects in t w o dimensional
beach processes. In : Proc. 8 t h Conf. o n Coastal Engineering. Am. Soc. Civ. Eng., New
York, N.Y., pp.194-210.
Johnson, J.W., 1949. Scale effects in hydraulic models involving wave motion. Trans. Am.
Geophys. Union, 3 0 : 517-525.
Kirby Jr., J.T., Dalrymple, R.A. and Liu, P.L.-F., 1981. Modification of edge waves by
barred beach topography. Coastal Eng., 5 : 35-49.
Longuet-Higgins, M.S., 1972. Recent progress in t h e study of longshore currents. In: R.E.
Meyer (Editor), Waves o n Beaches and Resulting Sediment Transport. Academic Press,
New York, N.Y., pp.203-248.
Longuet-Higgins, M.S. and Stewart, R.W., 1962. Radiation stress and mass transport in
gravity waves, with application to surf beats. J. Fluid Mech., 1 3 : 481-504.
Munk, W.H., 1949. Surf beats. EOS, Trans. Am. Geophys. Union, 3 0 : 849-854.
Rector, R.L., 1954. Laboratory study of equilibrium profiles of beaches. U.S. Army
Corps Eng., Beach Erosion Board, Tech. Memo, 41, 38 pp.
Shepard, F.P., Emery, K.O. and Lafond, E.C., 1941. Rip currents: a process of geological
importance. J. Geol. 4 9 : 337-369.
Short, A.D., 1975. Multiple offshore bars and standing waves. J. Geophys. Res., 80:
3838-3840.
Suhayda, J.N., 1974. Standing waves o n beaches. J. Geophys. Res., 7 9 : 3065-3071.
Thornton, E.B. and Guza, R.T., 1982. Energy saturation and phase speeds measured o n a
natural beach. J. Geophys. Res., 8 7 : 9499-9508.
Wright, L.D., Chappell, J., T h o m , B.G., Bradshaw, M.P. and Cowell, P., 1979. Morphody-
namics of reflective and dissipative beach and inshore systems: Southeastern Australia.
Mar. Geol., 3 2 : 105-140.
Marine Geology, 60 ( 1 9 8 4 ) 15-30 15
SPATIAL AND TEMPORAL VARIATIONS IN SPECTRA OF STORM

WAVES ACROSS A BARRED NEARSHORE
ROBIN G.D. DAVIDSON-ARNOTT and DAVID C. RANDALL

Department of Geography, University of Guelph, Guelph, Ont. N I G 2 W1 (Canada)
(Received April 1 6 , 1 9 8 3 ; revised and accepted December 15, 1983)
ABSTRACT
Davidson-Arnott, R.G.D. and Randall, D.C., 1984. Spatial and temporal variations in
spectra of storm waves across a barred nearshore. In: B. Greenwood and R.A. Davis,
Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated Coastal Environ-
ments. Mar. Geol., 60: 15-30.
Wave staffs and electromagnetic current meters were deployed on a profile across a
two-bar system at Wendake Beach, southern Georgian Bay. This paper examines spatial
and temporal changes in the characteristics of wave form, and the spectra of surface
elevation and on-offshore current motion, during one storm. Non-linear effects of wave
shoaling and breaking across the bars result in the appearance of secondary waves and
both the wave and on-offshore current spectra have significant harmonic peaks during
most of the storm. Significant low-frequency energy occurs only during the peak of the
storm. While the peak frequency remains constant across the bar system, the proportion
of energy in the primary peak is greatest in the troughs and lowest over the bar crests
and there are similar changes in the proportion of energy in the first harmonic. However,
in both surface elevation spectra and on-offshore current spectra, the greatest proportion
of energy is found in frequencies related t o the incident wind waves.
INTRODUCTION
On many gently sloping sandy coastlines the nearshore area is charac-

terized by the presence of sand bars which assume a variety of forms, with
most being aligned parallel to the shoreline (Greenwood and Davidson-Arnott,
1979). In some areas these are essentially permanent features of the near-
shore profile (e.g. Greenwood and Davidson-Arnott, 1975; Davidson-Arnott
and Pember, 1980; Hale and McCann, 1982) while in others they appear to
be related to the occurrence of steep storm waves and may be destroyed
during other times (e.g. Wright et al., 1979). It is evident that the existence
of a barred, in contrast to a planar, profile reflects the establishment of a
condition of dynamic equilibrium, resulting from a particular combination
of processes controlling the rate and pattern of sediment transport in the
breaker and surf zones. A number of controlling mechanisms have been
suggested to account for the occurrence of the bars and their overall mor-
phology and location, including standing waves (e.g. Short, 1975; Bowen,
1980), edge waves (Bowen and Inman, 1971; Holman and Bowen, 1982),
0025-3227/84/$03.00 0 1984 Elsevier Science Publishers B.V

16
interaction between breaking waves and rip cell circulation (Greenwood and
Davidson-Arnott, 1979) and harmonics generated during wave shoaling
(Boczar-Karakiewicz et al., 1981). Ultimately, the occurrence of a non-
barred or a barred profile is probably controlled by the characteristics of the
incident waves and factors determining the type of shoaling deformation,
breaking, and the amount of wave reflection.
Numerous studies have reported on the characteristics of wave trans-
formation during shoaling, breaking and in the surf zone on planar slopes
(e.g. Galvin, 1968; Divoky et al., 1970; Longuett-Higgins, 1976; Suhayda
and Pettigrew, 1977 ; Shemdin et al., 1980) and on water motion associated
with breaking waves (e.g. Thornton et al., 1976; Thornton, 1979; Battjes
and Van Heteren, 1980). There have been fewer studies of wave transforma-
tion over barred profiles (e.g. Byme, 1969; Wood, 1970;McNair and Sorensen,
1970; Busching, 1976; Wright et al., 1979; Mizuguchi, 1980) where the
processes are complicated by multiple breaking and an irregular surf zone
profile, with deeper water in the troughs leading t o zones of wave reformation.
This paper reports on the form and spectral characteristics of waves
over a barred system. In particular, attention is focussed on changes in the
number and form of waves in the zones of wave shoaling, breaking and
reformation, on spatial and temporal changes in the energy distribution
within the wave spectra, and on the relationship between spectra from the
water surface profile and the horizontal component of water motion below
the surface. While consideration is given primarily to the effects of the bar
topography on the incident waves, it is evident that some of the phenomena
described here must in turn be considered in modelling overall nearshore
sediment transport and, thus, ultimately in the formation of the bars
themselves.
STUDY AREA
The study was carried out at Wendake Beach at the southeast end of
Georgian Bay, Lake Ontario (Fig.1). The study site is located near the middle
of a 5 km stretch of sandy beaches which form part of the head of
Nottawasaga Bay. The beach and nearshore zone consist primarily of
well-sorted fine to medium sand with occasional outcrops of till or bed-
rock, and is characterized by the presence of two or three bars. A nearly
continuous, straight to sinuous outer bar extends parallel to the shore for
most of the stretch of beach and there are one or two inner bars which are
generally crescentic in form and broken by rather poorly defined rip channels.
Three bars were present along the profile instrumented for the study (Fig.2).
The outer bar was located about 100 m offshore with water depth over the
crest of 1.6 m and in the trough 2.0 m. The second bar was located about
50 m offshore with water depth 1 m over the crest and there was a small,
poorly developed bar just lakeward of the step. The outer bar was nearly
symmetric in shape while the lakeward slope of the second bar was much
steeper than the shoreward slope. Over the study period the outer bar
17
Fig.1. Location of study area.

18
0-
WI9
1-
2-
3-
4-
INNER BAR VY 1.08

5- OUTERBAR v48 Or%
v.7 OD48
1
20 40 60 80 IW 120 140 160 180 200
DISTANCE FROM BASELINE iml
Fig.2. Profile across nearshore bars showing location of wave staffs and current meters
used in study. Because of a failure in the electronics only frequency data is available for
wave staff 8.
remained relatively stable, while the second bar grew in height and migrated
about 15 m landward.
Prevailing winds are offshore from the southwest, with waves affecting
the bars being generated by winds from the W, NW, and N, blowing over
fetches of 40,150 and 5 km, respectively, during the passage of depressions.
During these storms, waves reaching Nottawasaga Bay have significant wave
heights of 1-2 m and periods of 4-7 s (Davidson-Amott and Pollard, 1980).
The presence of shorefast ice and ice in Georgian Bay restricts wave processes
for a 3-4 month period from mid-December. More details of the nearshore
morphology and wave climate are given in Randall (1982).
METHODOLOGY
Water surface elevations and water motion were measured using continuous
resistance wave staffs and biaxial electromagnetic flow meters. For the mea-
surements reported here, nine wave staffs and six flow meters were placed
along a shore normal profile in the configuration shown in Fig.2. Due to
instrumentation problems data were not obtained from wave staffs 5 and 8.
Pre-storm calibration indicated that wave staff linearity was good except for
the lower 25 cm. Wind speed and direction were recorded using a D.C.
wind generator and vane mounted on a 4 m high mast jetted into the sand on
the beach. Analogue signals from each sensor were hard-wired to a Hewlett-
Packard Data Aquisition System with analogue to digital scanner converter,
and micro computer. Data were collected over 9-min spans with a sampling
interval of 0.5 s. Spectral analysis was carried out for the first 1024 data
points using a lag of 50. More details on instrumentation can be found in
Greenwood and Sherman (1982).
19
STORM CHARACTERISTICS
A meteorological depression moving northeastward above Georgian Bay

during May 31 and June 1 1980 generated storm waves affecting the study
area over a period of about 1 9 h. During this time a total of ten sets of
records were taken. Wind speed and direction recorded at the study site
are presented in Fig.3 along with the significant wave height and peak
period recorded at the outer wave staff (staff 9). Wave build-up occurred
primarily in response t o increasing wind speed and a clockwise shift in wind
direction from the SW at about 18.00 h on May 31 to the NW at 22.00 h.
A t the peak of the storm average wind speeds recorded at the beach exceeded
9 m s-l with corresponding wave height greater than 1.0 m and wave period
greater than 5.0 s. After 1.00 h on 1 June, wind speed began to decrease
gradually while wind direction shifted slowly towards the north, and cor-
respondingly shorter fetch lengths, resulting in a slow decrease in both wave
height and period.
- PEAK.
60
5.0
4.0
P-
n
e
3.0
a
Y
4
M
0 ' ,
1800
1
l 7 M
I
l9CQ 21
I
W
I
23 M
I
*w
I
300
I
5 M
I
700
1
9w ItW
20
I 0
0.0
,4
a
-k May31 = Id June 1
-I
Fig.3. Trends for wind speed, wind direction, significant wave height (4V,where V 2 is the
variance of the wave record) and peak period during the storm monitored here. Wave data
are derived from spectra for wave staff 9.
20
SPATIAL VARIATION IN WAVE CHARACTERISTICS
Wave height
Changes in significant wave height across the bars are shown in Fig.4 for
two runs (15.00 and 22.00 h on May 31) which reflect conditions during
the growth and peak of the storm, respectively. The dashed lines indicate
changes in wave height from one staff to the next over the bar and trough
systems while the vertical lines show the changes in height at each staff in
the interval between the two runs. Also shown in Fig.4 are values for the
ratio of wave height to water depth (y) changes in which reflect primarily
the increase in wave height but also include the effects of increased water
depth due to set-up.
The data for both runs show a similar spatial pattern with greatest heights
occurring over the bar crest and a considerable reduction occurring in the
trough landward of the bar, reflecting height loss through breaking on the
bar, and the effect of deeper water in the trough. Even during the growth
stage there is considerable breaking over the inner bar because of its shallow
depth and steep lakeward slope, though during this time breaker type was
I = Hslh
0.18 0.36
0.38 0 15
0.68 0.98
0% 0.40
0.0 ,
I
1 , 1 I I I I I I
20 40 60 80 1W 120 140 160 180
DISTANCE FROM BASELINE Iml
Fig.4. Spatial variations in significant wave height at the beginning and peak of the storm.
The solid lines indicate the increase in wave height through time and the dashed lines
indicate changes in height across the bar trough system.
21
observed to be primarily in the spilling range. A t the peak of the storm

there is a large increase in wave height on the inner bar, breaking is intense
and occurs as a mixture of spilling and plunging types. The inshore breaker
type parameters of Galvin (1968) and Battjes (1974) were calculated using
significant wave height measured at staffs 9 and 3, and average slope values
for the lakeward slope of the outer and inner bars, respectively. Values for
the outer bar were well within the spilling range, but those for the inner
system at the peak of the storm are close t o the transition from spilling to
plunging [0.79-0.85 compared to the transition value of 0.68 using Galvin’s
(1968) parameter]. If H 1/10 is used rather than H 1/3 then the values at
the peak of the storm fall into the plunging category. Values of y at staff 3
range from 0.68 to 1.06 over the storm and are thus in the range conven-
tionally assumed for a highly dissipative surf zone. The intensity of breaking
over the inner bar is reflected in the large reduction in significant wave
height between staffs 3 and 2. The fact that there is little change in height
and the value of y at staff 2 over most of the storm is indicative of a satur-
ated inner surf zone; and observations show the formation of surf bores
rather than wave reformation in many instances.
Over the outer bar, while there is considerable peaking due to shoaling,
breaking is confined to the largest waves during the growth phase and is of
the spilling variety. During the peak of the storm breaking was more intense
but still primarily within the spilling range with breaking ceasing as the waves
moved into the trough. Thus values of y recorded at staff 7 range from 0.12
to 0.27 and are much less than recorded for the saturated conditions land-
ward to the inner bar. A puzzling feature is the magnitude of the height
decrease in the trough (staffs 7 and 6) landward of the outer bar, particularly
during the growth phase. It is difficult to account for this decrease on the
basis of the observed breaker intensity and comparatively deep water over
the bar crest, particularly as staffs 9 and 7 are located in approximately the
same water depth.
Wave form and number
The change in wave characteristics across the nearshore profile can be seen
visually by comparison of wave records at the peak of the storm (Fig.5). The
water surface profiles from each staff have been time shifted so that the
transformation of individual waves can be traced vertically. The effects of
wave shoaling and breaking on the outer (staff 9) and inner (staffs 4 and 3)
bars is seen in the increased wave height and steepness, and the peaked form
of both the trough and the crest. A number of small crests are present,
particularly on the crests and backs of larger waves. Some of these are small
wind waves, reflecting continuous wave generation up to the beach during
the storm, while others are probably secondary waves (solitons) resulting
from breakdown of the primary waves during and just before breaking
(e.g. Galvin, 1968; Byme, 1969; Thornton et al., 1976). In the outer trough
(staffs 7 and 6) the considerable reduction in wave height is readily app&ent,
22
SECONDS
0
STAFF 9
150 1
STAFF 7
50
STAFF 6
50
50
STAFF4
04
STAFF3
STAFF2
50
0’ I 1 I I
10 20 30 40 50 60 70 80 90 100
SECONDS
Fig. 5. Portion of the analogue record from wave staffs 9, 7 , 6, 4, 3 and 2 takenat 00.30 h
o n June 1. The records have been time shifted so that individual waves can be traced
vertically.
23
waves are less steep and troughs are more rounded. There is also, particularly
at staff 7, a noticeable reduction in the number of smaller peaks and these
re-appear as the waves shoal and break across the inner bar. However, land-
ward of the inner bar at staff 2, while there is an obvious height reduction
due t o breaking, there does not appear to be the same reduction in the
number of secondary crests. Actual counts of both visible peaks and zero
crossings for two sets of records at the peak of the storm confirm this
general pattern and show approximately a 30% reduction in wave crests
between staffs 9 and 7. There is, however, little change in the number of
wave crests between staffs 3 and 2.
Wave and current spectra
Spatial and temporal changes in the magnitude and distribution of energy

associated with both water surface elevations and with the onshore-offshore
component of water motion were examined by comparison of their spectra.
Examples of the spectra recorded during the peak of the storm are shown
in Fig.6.
The spectra for water surface elevation are typical of those associated with
storm waves in fetch limited conditions, with most of the energy in a rela-
tively narrow primary peak, a rapid decay towards lower frequencies and a
more gentle slope towards the higher frequencies reflecting saturation in
the equilibrium range (Thornton, 1979). There is considerable temporal and
spatial variation in the shape of individual spectra both in terms of the size
and number of peaks. Most of the spectra contain a single significant primary
peak and many exhibit one or more secondary peaks at frequencies both
higher (harmonic) and lower (sub-harmonic)than the primary peak frequency.
With very low waves at the beginning and end of the storm a prominent
primary peak is absent. Otherwise it ranged from 0.16 to 0.49 Hz (2.0-6.0 s)
with changes reflecting the changes in wind speed and fetch described earlier.
As has been found elsewhere (e.g. Thornton et al., 1976) there was no
detectable change in the peak frequency across the breaker zones.
Secondary peaks at frequencies approximately twice that of the primary
peak, the first harmonic, are most pronounced at the height of the storm.
This peak is generally associated with non-linearities generated during
shoaling and breaking and with the appearance of solitons. As would be
expected, therefore, these peaks are best developed on the lakeward slopes
and bar crests (staffs 9 , 4 and 3) and are generally less apparent in the troughs
(staffs 7, 6, 2; see Fig.6). They also occur for a greater proportion of the
storm in the inner system. N o significant peaks are present at higher order
harmonics, but this is not surprising in view of the relatively high frequency
of the primary peak.
In general there appears to be only limited energy at frequencies lower
than the peak frequency in the wave record. However, during the height of
the storm when wave breaking is most intense, a significant peak does
develop at a frequency of about 1/4 of the primary peak. This is most
24
WENDAKE BEACH MAY 31,1982 22:OO HRS
loeooo 1 10,000 -
i - ws3
L
a -ws2
Ni 1,000 -__
CM 2
0
N
v)
I
' N '
5
N
10-
'--
PERIOD (SECS) PERIOD (SECS)
1 4 2
- 1
1 1
0.25 6.50 0.75 1.00 0.25 0.50 0.i5 100
FREQUENCY (HZ) FREQUENCY (HZ)
lo,ooo 1 - ws 4
lorooo 1
- WS 6
- - _CM 6
PERIOD (SECS)
4 2 1
i I
0.25 0.50 0.75 1.00 0.25 0.50 0.75 1.00
FREQUENCY (HZ) FREQUENCY (HZ)
10,000 1 10,GOO 1
5 10- \ A
* L
PERIOD (SECS)
y; '"I\
4 2 1
Fig.6. Spectra from wave staffs and X (onshore-offshore) component of electro magnetic
current-meter records.
25
evident in the outer trough (staffs 7 and 6) and in the inner system (staffs
3 and 2) and is not significant in the breaker zones on the lakeward side
of the bar crests (staffs 9 and 4).The low-frequency peak is discussed in
more detail by Greenwood and Sherman (1984,this volume).
The spectra of onshore/offshore water motion measured at mid-depth
are similar in general shape to that of surface water elevations, with most
of the energy centred around the primary peak (see Fig.6). The secondary
peak at the first harmonic is also well-developed. Thus, as has been noted in
other areas, (e.g. Thornton et al., 1976) with spilling breakers water motion
below the wave trough appears to be dominated by the incident wave field
rather than by turbulence due to breaking. However, there is generally more
energy at low frequencies in the current meter record and instead of a
pronounced peak at 1 / 4 of the primary peak, the energy density increases
linearly.
Temporal and spatial changes in energy distribution
Temporal changes in peak period are shown in Fig.2 and, as noted earlier,
the peak period remained constant across the bar-trough system at all times.
However, there were noticeable changes in the distribution of energy within
the spectra both spatially and temporally, and an attempt was made to
examine changes in the proportion of energy within the peak frequency
( V f ) and in the first harmonic ( V , f ) .Several methods of defining the limits
of the peak (e.g. Huntley, 1980) were tested but did not prove consistent,
and instead a fixed bandwidth of 0.067 Hz centred on the primary peak
was used. The upper and lower limits of the first harmonic were then de-
termined at twice the upper and lower cut-off frequencies of the prima-
ry peak. Except at the height of the storm, where a small portion extend-
ed beyond the fixed bandwidth, the technique appeared to fit the prima-
ry peak well, and where a significant harmonic peak was present in the
record, this also appeared, to coincide well with the defined limits. These
results are summarised in Table I.
The proportion of energy in the primary peak ranges from under 30%
at the beginning and end of the storm to about 45% at the height of the
storm when the primary peak is particularly well defined. Temporal varia-
tions in the proportion of energy in the first harmonic are much smaller
but there is a similar, if not as well defined, pattern.
There appears to be a distinct pattern t o spatial variations in the energy
in the primary peak. Except at the height of the storm the proportion of
energy is much greater in the trough (staffs 7 and 6) than on the outer
bar (staff 9), and it again falls over the inner bar. Landward of the inner bar,
the proportion is lower during most of the storm but higher than on the
preceding crest at the beginning and end when waves are lowest. The spatial
variations in V,f are again much smaller and the pattern is less distinct. Over
much of the storm the proportions of energy in the first harmonic are greatest
at staff 2, the inner trough, and they are frequently higher in the outer
trough as well although this is not consistent and the differences are small.
26
TABLE I
Variance in primary peak and first harmonic as a percentage of the total variance within
each spectrum through the storm
Wave staff Time

15:OO 16:15 17:15 18:30 20:OO 22:OO 00:25 02:20 06:20 10:15
Primary peak
9 * 26.2 30.5 22.7 24.4 45.0 35.0 26.8 16.0 *
7 * 42.0 40.1 28.9 34.1 43.1 37.3 32.9 30.2 *
6 * 43.5 36.0 30.1 36.0 44.3 36.2 34.2 35.5 *
4 31.8 38.1 31.9 26.2 31.2 42.8 39.2 28.9 29.1 24.5
3 32.0 34.9 25.4 22.3 30.1 38.6 34.7 27.8 26.0 21.2
2 36.3 40.8 30.6 19.3 25.9 33.8 29.8 30.5 34.0 28.0
First harmonic
9 * 7.2 11.8 8.6 14.8 14.7 14.5 13.7 14.6 *
7 * 8.5 10.4 9.8 12.7 14.8 16.3 13.5 16.8 *
6 * 5.4 10.0 7.7 16.0 15.0 18.3 11.1 12.9 *
4 10.2 8.5 12.8 9.3 12.7 13.5 12.6 11.0 14.9 6.1
3 10.3 12.7 14.3 11.0 15.3 15.5 14.7 11.7 18.9 5.4
2 7.4 9.0 16.6 14.3 19.2 12.7 20.3 15.8 11.5 7.4
* N o significant primary peak.
DISCUSSION
It is evident that, under storm conditions, the portion of the nearshore

zone studied is typically dissipative rather than reflective (e.g. Guza and
Bowen, 1976; Chappell and Wright, 1978; Wright et al., 1979). Waves break
some distance offshore as spilling and plunging breakers and there is consider-
able energy transfer to higher harmonics (Thornton, 1979). The presence of
the bars results in a segmentation of the surf zone into areas of high turbu-
lence (the bar crests) and areas of lower turbulent dissipation (the troughs)
with the intensity varying with incident wave height and between the outer
and inner bar system.
The lakeward slope and crest of both bars is an area of wave shoaling
and breaking. The shallower water over the inner bar crest and landward
trough, however, results in greater intensity of wave breaking over the inner
bar and some differences between conditions in the outer and inner troughs.
At staff 2, landward of the inner bar, conditions appear similar to that of a
saturated inner surf zone with frequent generation of surf bores and wave
height controlled by turbulent dissipation (e.g. Horikawa and Kuo, 1966;
Suhayda and Pettigrew, 1977). The inner bar acts as an effective filter
for wave height over much of the storm and this is reflected in the nearly
constant value fory .The value of 0.4 recorded for y is similar to that reported
by Wright et al. (1982).
Over the outer trough the greater water depth and reduced breaker
intensity leads to much lower values for y. Wave breaking generally decreases
27
as the wave moves into the trough and surf bores are not formed. This zone
is frequently referred to as one of wave reformation though, under these
conditions it is perhaps better described in terms of cessation of breaking
since the primary wave form is never actually destroyed. The large decrease
in wave height landward of the inner bar crest can be accounted for by wave
breaking. However, the decrease over the outer bar between wave staffs 9
and 7, particularly during the early and late stage of the storm when there
was little breaking there, is more difficult to explain. Since staffs 9 and 7
are in approximately the same water depth the changes cannot be explained
purely on the basis of shoaling and changes in celerity. Some energy loss may
be due to bed friction over the shallow bar crest. Another possible explana-
tion may be related to spatial changes in the relative position of primary
waves and the secondary waves generated during breaking and intense shoal-
ing. In the trough these would be out of phase leading to an overall decrease
in the amplitude of water surface fluctuations. As incident waves catch up with
secondary waves generated by the preceding wave the overall amplitude of
water motion would increase. The general increase in wave height at staff 6,
although there is a slight increase in water depth suggests that this may in
fact be occurring.
The transfer of energy to higher harmonics, which can be seen in the
well developed peaks in the spectra of both water surface elevations and
the on-offshore water motion, has been reported from a number of field and
laboratory studies. Examination of actual wave profiles such as those shown
in Fig. 5, suggests that the secondary waves generated are not phase locked,
and thus travel more slowly through the trough than the primary waves
from which they were derived. Interaction of these secondary waves with
succeeding waves can produce spatial variations which are dependent on the
recurrence length and which can therefore influence the location of breaker
zones and the spacing of bars (e.g. Hulsbergen, 1974; Bijker et al., 1976;
Boczar-Karakiewicz et al., 1981). Because of the presence of considerable
incident wind wave energy a t the first harmonic frequency and the genera-
tion of a second set of secondary waves as waves shoal over the inner bar,
the behaviour of the secondary waves seen here is not easily traced and
further work is necessary to determine their importance in this bar system.
Although the spectra of water surface elevations and on-offshore water
motion show close agreement a t the peak frequency and at higher harmonics,
they differ somewhat at the low-frequency end of the spectrum. The surface
elevation spectra show a fairly pronounced peak at about 1/4 the incident
frequency while the current meter spectra show an essentially continuous
increase in energy at frequencies below 0.125 Hz. The low-frequency peak
is best developed in the trough and landward edge of the bar (staffs 8, 7 , 6
and 3, 2), and occurs only at the peak and beginning of the decay phase of
the storm when there is fairly intense breaking on the outer bar. There is a
number of possible mechanisms for generating this sub-harmonic peak.
Numerous investigations have reported on the generation of standing edge
waves (e.g. Guza and Davis, 1974; Bowen and Guza, 1978; Huntley and
28
Bowen, 1978; Huntley et al., 1981). Chappell and Wright (1978) found
standing edge waves at about 1/4 of the incident frequency were dominant
in a bar trough system at the highest energy levels but were subordinate t o
edge waves at 1/2 of the incident frequencies with lower incident waves.
However, the peak here is not well-developed in the current meter spectra
and does not decrease offshore very rapidly as would be expected for a zero
mode standing edge wave [see Greenwood and Sherman, (1984, this volume)
for further discussion]. A second possibility is a resonant response of water
levels in the trough t o intense wave breaking over the bar crest, possibly
related to groupiness in the incident wave record (e.g. Symonds and Huntley,
1980; Symonds et al., 1982).
Finally, a number of mechanisms has been proposed to explain bar
formation, location and form which are related to phenomena associated
with both higher harmonics (e.g. Boczar-Karakiewicz et al., 1981) and
sub-harmonics (e.g. Holman and Bowen, 1982). From the analysis to date
it is not possible to comment on the validity of these mechanisms in relation
to the bar-trough systems studied. However, three points are worthy of
note: (1) in both surface elevation spectra and on-offshore spectra the
greatest proportion of energy is found in frequencies related to incident
wind waves; (2) there are considerable temporal variations in the proportion
of harmonic and sub-harmonic energy present and changes in the significance
of these is much greater than the change in bar morphology during the storm;
and (3) it seems likely that the bar morphology is the primary control on
the spacing of the breaker zones and on the nature of the resonant interac-
tion generating sub-harmonic energy (e.g. Huntley, 1980). The bars studied
here are therefore macroscale features reflecting the cumulative effects of a
number of storm events of different magnitudes rather than a bedform res-
ponding instantaneously to changing wave conditions. It also seems likely
that no single mechanism, such as edge waves of a particular frequency, or
recurrence length, can be used to explain the final morphological form.
Rather, it may reflect an equilibrium resulting from the interaction of
several mechanisms controlling the spatial and temporal patterns of sediment
transport with the primary processes being related to the incident wind
waves.
ACKNOWLEDGEMENTS
This paper results from a study carried out jointly with Brian Greenwood
and Doug Sherman of Scarborough College, University of Toronto. Paul
Cristilaw, Leslie Joynt and Ross Sutherland provided valuable field assistance.
We are grateful t o the support staff in the Geography Department, Univer-
sity of Guelph, for their technical, secretarial, cartographic and computing
help. The paper has benefited from comments by David Huntley, Ed Thornton
and Brian Greenwood. The work was supported by a N.S.E.R.C. operating
grant (R.D.A.) and a University of Guelph Graduate Scholarship (D.R.).
29
REFERENCES
Battjes, J.A., 1974. Surf similarity. Proc. 14th Coastal Engineeringconference, Copenhagen,
pp.466-480.
Battjes, J.A. and Van Heteren, J., 1980. Field measurements of wind wave kinematics.
Proc. 17th Coastal Engineering Conference, Sydney, N.S.W., pp.347-362.
Bijker, E.W., Van Hijum, E. and Vellinga, P., 1976. Sand transport by waves. Proc. 15th
Coastal Engineering Conference, Honolulu, Hawaii, pp.1149-1167.
Boczar-Karakiewicz, B., Paplinska, B. and Winiecki, J., 1981. Formation of sand bars by
surface waves in shallow water. Laboratory experiments. Rozpr. Hydrotech., 4 1 :
111-125.
Bowen, A.J., 1980. Simple models of nearshore sedimentation: beach profiles and long-
shore bars. In: S.B. McCann (Editor), The Coastline of Canada. Geol. Surv. Can.,
Pap. 80-10: 1-11.
Bowen, A.J. and Guza, R.T., 1978. Edge waves and surf beat. J. Geophys. Res., 83(4):
1913-1920.
76(36): 8662-8671.
Busching, F., 1976. Energy spectra of irregular surf waves. Proc. 15th Coastal Engineering
Conference, Honolulu, Hawaii, pp.539-559.
Byrne, R.J., 1969. Field occurrences of induced multiple gravity waves. J. Geophys.
Res., 74( 10): 2590-2596.
Chappell, J. and Wright, L.D., 1978. Surf zone resonance and coupled morphology. Proc.
16th Coastal Engineering Conference, Hamburg, pp.1359-1377.
Davidson-Arnott, R.G.D. and Pember, G.F., 1980. Morphology and sedimentology of
multiple bar systems, Southern Georgian Bay, Ontario. In: S.B. McCann (Editor),
The Coastline of Canada. Geol. Surv. Can., Pap. 80-10: 417-428.
Davidson-Arnott, R.G.D. and Pollard, W.H., 1980. Wave climate and potential longshore
sediment transport patterns, Nottawasaga Bay, Ontario. J. Great Lakes Res., 6( 1 ) :
54-67.
Divoky, D., Le Mehaute, B. and Lin, A., 1970. Breaking waves on gentle slopes. J. Geo-
phys. Res., 75(9): 1681-1692.
I Galvin, C.J., 1968. Breaker type classification o n three laboratory beaches. J. Geophys.
Res., 73(12): 3615-3659.
Greenwood, B. and Davidson-Arnott, R.G.D., 1975. Marine bars and nearshore sedi-
mentary processes, Kouchibouguac Bay, New Brunswick. In: J. Hails and A. Carr
(Editors), Nearshore Sediment Dynamics and Sedimentation. Wiley, New York,
N.Y., pp. 123-150.
wave-formed bars: a review and case study. Can. J. Earth Sci., 16(2): 312-332.
Greenwood, B. and Sherman, D.J., 1983. Shore-parallel flows in a barred nearshore.
Proc. 18th Coastal Engineering Conference, Cape Town, pp.1677--1696.
Greenwood, B. and Sherman, D.J., 1984. Waves, currents, sediment flux and morpho-
logical response in a barred nearshore. In: B. Greenwood and R.A. Davis, Jr. (Editors),
Hydrodynamics and Sedimentation in Wave-Dominated Coastal Environments. Mar.
Geol., 60: 31-61 (this volume).
Guza, R.T. and Bowen, A.J., 1976. Resonant interactionsfrom waves breaking on a beach.
Proc. 15th Coastal Engineering Conference, Honolulu, Hawaii, pp. 560-579.
Guza, R.T. and Davis, R.E., 1974. Excitation of edge waves by waves incident on a
beach. J. Geophys. Res., 79: 1285-1291.
Hale, P.B. and McCann, S.B., 1982. Rhythmic topography in a meso-tidal, low-wave-
energy environment. J. Sediment. Petrol., 52: 415-429.
Holman, R.A. and Bowen, A.J., 1982. Bars, Bumps and Holes: Models for the generation
of complex beach topography. J. Geophys. Res., 87: 457-468.
30
Horikawa, K. and Kuo, C., 1966. A study o n wave transformation inside the surf zone.
Proc. 10th Coastal Engineering Conference, Tokyo, pp.217-233.
Hulsbergen, C.H., 1974. Origin, effect and suppression of secondary waves. Proc. 14th
Coastal Engineering Conference, Copenhagen, 392-411.
Huntley, D.A., 1980. Edge waves in a crescentic bar system. In: S.B. McCann (Editor),
The Coastline of Canada. Geol. Surv. Can., Pap. 80-10: 111-121.
Huntley, D.A. and Bowen, A.J., 1978. Beach cusps and edge waves. Proc. 16th Coastal
Engineering Conference, Hamburg, pp.1378-1393.
Huntley, D.A., Guza, R.T. and Thornton, E.B., 1981. Field observations of surf beat 1 :
Progressive edge waves. J. Geophys. Res., 85: 6451-6466.
Longuett-Higgins, M.S., 1976. Recent developments in the study of breaking waves.
Proc. 15th Coastal Engineering Conference, Honolulu, Hawaii, pp.441-460.
McNair, E.D. and Sorensen, R.M., 1970. Characteristics of waves broken by a longshore
bar. Proc. 13th Coastal Engineering Conference, Washington D.C., pp.4 15-434.
Mizuguchi, M., 1980. An heuristic model of wave height distribution in surfzone. Proc.
17th Coastal Engineering Conference, Sydney, N.S.W., pp. 278-289.
Randall, D.C., 1982. Changes in the form and energy spectra of storm waves across a
nearshore bar system, Wendake Beach, Nottawasaga Bay, Ontario. Unpubl. M.Sc.
thesis, Univ. of Guelph, Guelph, Ont., 1 3 3 pp.
Shemdin, O.H., Hsiao, S.V., Carlson, H.E., Hasselmann, K. and Schulze, K., 1980. Mech-
anisms of wave transformation in finite-depth water. J. Geophys. Res., 85(9):
50 12-501 8.
Short, A.D., 1975. Multiple offshore bars and standing waves. J. Geophys. Res., BO(27):
3838-3840.
Suhayda, J.N. and Pettigrew, N.R., 1977. Observations of wave height and wave celerity
in the surf zone. J. Geophys. Res., 82(9): 1419-1424.
Symonds, G. and Huntley, D.A., 1980. Waves and currents over a nearshore bar system.
Proc. of the Canadian Coastal Conference. Natl. Res. Council Can., Burlington, Ont.,
pp. 64-78.
Symonds, G., Huntley, D.A. and Bowen, A.J., 1982. Two-dimensional surf beat: Long
wave generation by a time-varying break point. J. Geophys. Res., 87: 492-498.
Thornton, E.B., 1979. Energetics of breaking waves within the surf zone. J. Geophys.
Res., 84: 4931-4938.
Thornton, E.B., Galvin, J.J., Bub, F.L. and Richardson, D.P., 1976. Kinematics of break-
ing waves. Proc. 15th Coastal Engineering Conference, Honolulu, Hawaii, pp.461-476.
Wood, W.L., 1970. Transformation of breaking wave characteristics over a submarine bar.
Dept. of Natural Science, Michigan State University, Tech. Rep. 4.
Wright, L.D., Chappell, J., Thorn, B.G., Bradshaw, M.P. and Cowell, P., 1979. Morpho-
dynamics of reflective and dissipative beach and inshore systems: Southeastern Australia.
Mar. Geol., 32: 105-140.
Wright, L.D., Guza, R.T. and Short, A.D., 1982. Dynamics of a high-energy dissipative
surf zone. Mar. Geol., 45: 41-62.
Marine Geology, 60 (1984) 31-61 31
WAVES, CURRENTS, SEDIMENT FLUX AND MORPHOLOGICAL

RESPONSE IN A BARRED NEARSHORE SYSTEM
BRIAN GREENWOOD and DOUGLAS J. SHERMAN*

Departments of Geography and Geology, Scarborough Campus, University of Toronto,
Scarborough, Ont. M I C IA4 (Canada)
Department o f Geography, Scarborough Campus, University o f Toronto, Scarborough,
Ont. M l C l A 4 (Canada)
(Received April 16, 1983; revised and accepted August 31, 1983)
ABSTRACT
Greenwood, B. and Sherman, D.J., 1984. Waves, currents, sediment flux and morph-
ological response in a barred nearshore system. In: B. Greenwood and R.A. Davis, Jr.
(Editors), Hydrodynamics and Sedimentation in Wave-Dominated Coastal Environ-
ments. Mar. Geol., 60: 31-61.
A shore-normal array of seven, bi-directional electromagnetic flowmeters and nine

surface piercing, continuous resistance wave staffs were deployed across a multiple barred
nearshore at Wendake Beach, Georgian Bay, Canada, and monitored for a complete storm
cycle. Time-integrated estimates of total (ITVF) and net (INVF) sediment volume flux
together with bed elevation changes were determined using depth-of-activity rods.
The three bars, ranging in height from 0.10 to 0.40 m accreted during the storm
(0.03 m), and the troughs were scoured (0.05 m). Sediment reactivation depths reached
0.14 m and 12% of the nearshore control volume was mobilized. However, the INVF
value for the storm was less than 1%of the control volume revealing a near balance in
sediment volume in the bar system. Landward migration of the inner, crescentic and
second, sinuous bars occurred in association with an alongshore migration of the bar form
itself; the outermost, straight, shore-parallel bar remained fixed in location.
The surf zone was highly dissipative throughout the storm ( E = 3.8 X 102--192 X 10’)
and the wave spectrum was dominated by energy at the incident frequency. Spectral peaks
at frequencies of the first harmonic and at one quarter that of the incident wave were
associated with secondary wave generation just prior to breaking and a standing edge
wave, respectively. The former spectral peak was within the 95% confidence band for the
spectrum while the latter contributed not more than 10% to the total energy in the surface
elevation spectrum even near the shoreline.
During the storm wave height exceeded 2 m (H,) and periods reached 5 s ( T p k ) :
orbital velocities exceeded 0.5 m 5 - l (urns)and were above the threshold of motion for
the medium-to-fine sands throughout the storm. Shore-parallel flows in excess of 0.4 m 5-l
were recorded with maxima in the troughs and minima just landward of the bar crest.
The rate and direction of sediment flux is best explained by the interaction of anteced-
ent bed slopes with spatial gradients in the mean and asymmetry of the shore-normal
velocity field. These hydrodynamic parameters represent “steady” flows superimposed on
the dominantly oscillatory motion and assumed a characteristic spatial pattern from the
*Present address: Department of Geography, University of Southern California, Los

Angeles, CA 90007, U.S.A.

32
storm peak through t h e decay period. Increases spatially in t h e magnitudes of both the
mean flows and flow asymmetries cause a n increasing net transport potential (erosion);
decreases in these values spatially cause a decreasing net transport potential and thus
deposition. These transport potentials are increased or decreased through t h e gravity
potential induced by t h e local bed slope. Shore-parallel flow was important in explaining
sediment flux and morphological change where orbital velocities, mean flows and flow
asymmetries were a t a minimum.
INTRODUCTION
Since the first scientific description of wave-formed bars by Elie de

Beaumont in 1845, considerable effort has been devoted to understanding
their origins and dynamics. A large literature exists on the morphology, and
more recently the sedimentology, of these ubiquitous geomorphological
features. Many questions remain, however, concerning the complex inter-
actions between the nearshore fluid and sediment motions, and the topog-
raphy itself (for a review see Greenwood and Davidson-Arnott, 1979; and
Greenwood, 1982). Knowledge of these interactions is important, since bars
frequently form stable, equilibrium, bathymetric configurations in a wide
range of coastal environments, often under conditions of high alongshore
sediment transport (Greenwood and Davidson-Arnott, 1979; Greenwood and
Mittler, 1984, this volume). In other places, in contrast, bars are formed and
destroyed as conditions change (Goldsmith et al., 1982; Bowman and Gold-
smith, 1983).
Most studies of the hydrodynamics and sediment dynamics of barred near-
shores have been restricted to theoretical and laboratory research with the
associated problems of scale in the latter. Only recently has instrumentation
been available allowing comprehensive experiments in the prototype, capable
of providing data for the testing of theoretical models. Since sediment flux
and bar morphodynamics are a direct response to wave-induced currents and
since the latter (at least at present) are the more easily measured, then exam-
ination of such fluid motions should provide insights into sediment and bar
dynamics. This paper documents an experiment designed t o monitor the
incident and secondary wave characteristics, wave-induced orbital and shore-
parallel flows, and the local sedimentary response during a single storm event
in a barred nearshore in the Canadian Great Lakes.
LOCATION OF STUDY
The study site is at Wendake Beach in Southern Georgian Bay, Ontario

(Fig.1). It is tideless, and waves are generated over discrete time intervals
during the passage of meteorological depressions, causing rapid changes in
the magnitude and direction of incident wave energy. Fetch lengths vary,
with maxima to the WNW (84 km) and W (51 km). With limited fetch and
local storm winds, wind forcing of the waves is continuous to the shoreline
under most wave states.
In the Wendake Beach system 3 bars are present on a mean slope of 0.015
in medium to fine sands (0.43-0.13 mm mean diameter). The outer bar is
33
Im
?ach
Wym bol wood

Beach
SITE
Not tawasaga
metres
110 m from the shoreline, approximately 0.50 m high, relatively straight,

near-symmetrical and shore-parallel (Fig.2). The inner bars, at approximately
54 and 10 m from the shoreline, respectively, range in height from 0.10 to
0.35 m and are asymmetrical in section and sinuous to crescentic in plan-
form. There is a reduction in spacing of the bars landward and, although
some alongshore periodicity in form is evident in the inner bars, it is difficult
to isolate: the dominant periodicity in form is in the shore-normal direction
and thus the process(es) controlling sedimentation are most probably periodic
in the same direction. While the bars are distinct bathymetric configurations,
the nearshore slope deviates little from a planar configuration: a least squares
linear fit explains more than 90% of the observed variance in relief.
34
LINE 1 NORTH SWL
w1 W b W7 WE WV
I I LINE 0 SWL
-5.Ot r r r r r r r
zu i . .
-l
2
w
-1.0
-4.0
-5.0
0 20 40 60 80 100 120 140 160 180 200
DISTANCE FROM BASELINE (rn)
Fig. 2. Nearshore profiles and instrument deployment prior to the storm, Wendake Beach,
1980:05:31. Line zero is the instrument array transect; lines 1 North and 1 South are
spaced 30 m either side of the zero line. W = wave wire; C = current meter; r = depth of
activity rod. Note that along the instrument transect the depth of activity rods were offset
1 m to avoid interference by the wave wire supports.
EXPERIMENTAL DESIGN
The field experiment was designed specifically to examine the two-

dimensional nature of the nearshore hydrodynamics and sedimentation in a
plane normal to the shoreline. It involved the deployment of a shore-normal
array of wave and current sensors cable-linked to a high-speed data acquisition
and mass storage system, Morphological changes were monitored using con-
ventional profiling with level and staff in shallow water and echo sounding
offshore; more detailed measurements of sediment flux and bed elevation
change were made using depth-of-activity rods. More detail on the fluid
monitoring system can be found in Greenwood (1982) and Greenwood and
Sherman (1983);the depth-of-activity rods are described fully in Greenwood
and Hale (1980),Greenwood et al. (1980) and Greenwood and Mittler (1984,
this volume). Only a synopsis of methods will be presented here.
Wave and current sensors
Two types of instruments were used to monitor the fluid motions. Water
surface elevation changes associated with waves were measured by surface
35
piercing, continuous resistance wave staffs following the basic design of

Truxillo (1970). Helically wound, steel wires set in 2 or 3 m long, 18.8
mm diameter PVC pipes were mounted by insulated brackets on 37.5 mm
galvanized steel pipes. The latter were mounted on two metre long bases
jetted into the bed. Staffs were field calibrated individually to specific
oscillator-detector circuits and were linear except for the lowermost 0.25 m.
Shore-normal and shore-parallel horizontal flows were monitored using
biaxial electromagnetic flow meters designed by Marsh-McBirney. These
instruments have a time constant of 0.2 s and measure flows up to 3 m s-'.
Considerable work on the accuracy of these meters has been undertaken (see
review of Huntley, 1980) and, while some problems still exist, the level of
accuracy is thought to be significantly better than 10%. The meters were
mounted in a specially designed bracket that allowed rotation in both the
vertical and horizontal planes for accurate orientation of the probes. This
assembly was mounted on a stainless steel support. A galvanized steel pipe,
with fin to prevent rotation, was jetted into the bed to provide the base
support.
Figure 2 documents the instrument locations during the experiment. Of
particular importance were the locations of the current meters, since near-bed
flows were to be related to sediment transport. Unfortunately our knowledge
of the structure of turbulent oscillatory boundary layers over rough beds is
still rudimentary, especially with respect to combined waves and currents
(see Grant and Madsen, 1979). In any case the thickness of the bottom
boundary layer will be highly variable depending upon local flow and bed
roughness conditions. With the necessity of a fixed position in the shallow
water depths of the Wendake Beach surf zone, and the likelihood of a
logarithmic law being important even in the oscillatory boundary layer
(Jonsson, 1980) it was decided to place the sensors at the same relative water
depth (h/'2) in all cases. In this way at least the same segment of the water
column was monitored.
Data acquisition and analysis
All sensors were hardwired to the shore-based data acquisition system,

which consisted of a mini-computer controlled, high-speed multiplexer and
voltmeter giving potential analogue-to-digital conversion rates up to 1000
channel readings per second. Typically the instrument array was scanned
every 0.5 s over a period of 9 min and the data stored on magnetic tapes.
Record lengths were, of necessity, short t o maintain time series stationarity
in such a volatile environment. This did, however, restrict the scale of any
low-frequency effects which could be detected.
The 1100 data points per sensor per sampling were truncated to 1024
points for spectral analysis and the computation of descriptive statistics, using
Biomedical Computer Packages (BMD03T and BMDPBD), after Dixon (1971)
and Dixon and Brown (1979). Wave and current spectra were computed with
a unit bandwidth of 0.016 Hz (64 lags), a detrended but unfiltered time series,
36
and a fast fourier transformation to determine the spectral densities. Phase

and coherence of cross spectra were also calculated. Simple descriptive
statistics of the velocity field were computed using moments; for on-offshore,
for example:
1
Mean = U = -
N
i = l
Standard deviation = urmS= [j$ N
i = l
I
(ui -U ) 2
lI2
Skewness = u,k = - (3)

N i = l
where u is the on-offshore velocity and N the sample size.
Monitoring sediment f l u x and local morphological response
Although sediment monitoring equipment has advanced rapidly recently,

no single instrument can accurately measure both suspended load and bedload
and associated bed elevation changes on a continuous basis. A simple, well-
tested technique is used in this study to estimate sediment flux and morph-
ological response. This involves depth-of-activity rods: these are 1 or 2 m
long steel rods (0.5 cm diameter) with a free sliding washer that migrates
vertically down the rod during sediment erosion and is buried at the reactiva-
tion limit by any subsequent deposition. Measurements by Scuba divers are
taken prior to and at the end of the transport event to record the thickness
of the active layer and the bed elevation changes. This allows estimates to be
made of the time-integrated total and net volume flux through a storm as
well as the net bed elevation changes for a series of discrete points. At
Wendake Beach, depth-of-activity rods were deployed across the nearshore
zone at the locations shown in Fig.2. Owing to incomplete installation prior
t o the storm, only bed elevation changes were recorded along line 0 (the
instrument transect), but both depth-of-activity and bed elevation changes
were recorded along line IN.
WAVE-GENERATED CURRENTS AND SEDIMENT TRANSPORT POTENTIAL
Sediment motion in the nearshore bars must be related to both shore-

normal currents induced directly by primary wave oscillations, and to shore-
parallel currents (longshore currents) resulting from the radiation stress
associated with wave breaking and additional stresses induced by wind and
alongshore pressure gradients. Secondary waves such as leaky mode or edge
waves may be superimposed on these primary components. Since orbital
velocities under waves are generally larger than any unidirectional current
37
within the surf zone, the component of boundary shear stress due to waves
will be significantly greater (Grant and Madsen, 1979) and therefore most
important in the initiation of sediment movement. However, waves are
inefficient transporters of sediment unless the oscillation is asymmetrical
and/or combined with a superimposed current. Since the primary periodicity
in sediment accumulation is normal to shore at Wendake Beach, then trans-
port by shore-normal currents induced by primary and secondary waves is
likely of greatest importance in both bar generation and maintenance in this
system.
Waves within the nearshore are reasonably approximated using linear wave
theory, where sinusoidal water motion is implicit (e.g. Gaughan and Komar,
1975). Waves in the surf zone, however, are markedly non-linear and water
motions no longer symmetric. Sediment transport in this zone will be a func-
tion not only of the absolute velocity near the bed (a function of wave
height), but also the frequency distribution of the velocity vectors. Of para-
mount importance in the oscillatory flow field, where the range of magni-
tudes of current vectors is large, is the asymmetry of the distribution of
these current vectors, since sediment transport is not related linearly to the
velocity, but more probably to the cube or fourth power of the absolute
velocity. Quite small asymmetries can therefore be more important than any
net (Lagrangian type) steady drift, even where the latter may be quite high.
The importance of this was clearly recognized by Inman and Bagnold (1963),
Inman and Frautschy (1966) and Wells (1967) and most recently in the trans-
port models proposed by Bowen (1980), Bailard (1981) and Bailard and
Inman (1981).
STORM EVENT 1980 : 05 :3 1-1 980:06:0 1
Figure 3 illustrates the temporal pattern of wind speed and direction and
the angle of wave approach through the storm. Storm waves were initiated
by 1500 h (EDST) on May 31 with winds of 3 m s-l from the WSW, which
increased t o a maximum of 8 m s-l from the W at 2200 h. With the passage
of the frontal system, winds veered to the NW at 2230 h reaching speeds of
7 m s-'; a gradual reduction in speed occurred as a further switch to the N
took place at 0320 h. The storm winds dropped below 3 m s-l at 0650 h as
the winds backed towards the westerly direction again and the storm ended
by 1015 h on June 1. This dramatic change in wind and wave direction is
typical of the passage of meteorological depressions in this region.
Incident and secondary waves
The general pattern of growth of the incident wave spectrum is best illus-
trated by characteristics measured at wave wire nine at the lakeward margin
of the three-bar-system (see Fig.2). It should be stressed, however, that during
the storm peak, the surf zone width reached some distance t o lakeward;
theory indicated that depth controlled breaking occurred u p to 229 m from
the shoreline a t this time.
38
8-
WIND SPEED
7 - r7L
6 -
RI'
5 -
4 -
&/Y r/A-$-' \
3 - *4
/
2 -
1 -
1 I I I I I I I
0
75 - - ssw
60 - WIND DIRECTION - sw
45 - -
-
- -
I I I I
- 1600 1800 2000 2200

I
2400 0200 0400 0600-
- TIME (h)
45 -
60 - NNW
75 i
N
W
WAVE DIRECTION (SPECTRAL PEAK)
0--+-*,
I I
'* - ..*- - - -
I
+\, 1
'\+-----J
1600 1800 2000 2200 24%? 0200 0400 0600
TIME (h)
--. -----*
--.*-- WNW $
Fig.3. Wind speed, direction and angle of wave approach to the shore normal, Wendake
Beach, 1980:05:31:15:00-1980:06:01:10:15.
Wave heights, periods and directions of approach
Table I summarizes the important incident wave parameters. Significant

wave heights in excess of 2 m and with periods of 5 s occurred at the storm
peak. Spilling breakers were the dominant wave form, with most intense
breaking on the bar crests. Although wave reformation did occur in the
39
TABLE I
Incident wave parameters, Wendake Beach, 1980:05:31-06:Ol
Time Heieht' Period Andel Breaker3 index Scaling4 parameter Surf zone' width
E
(X 10')
1500 0.59 0.82 3.4 10 0.404.56 8 74

1615 0.68 0.95 3.4 10 0.46-0.64 6.5 124
1715 0.76 1.06 4.0 10 0.37-0.52 6.4 143
1830 0.85 1.19 2.3 9 1.26-1.76 192 133
2000 0.78 1.09 4.3 7 0.334.46 5.9 127
2200 1.52 2.12 5.0 8 0.48-0.67 5.9 202
0025 0.98 1.37 5.0 5 0.314.43 4.8 190
0320 0.76 1.06 5.0 10 0.24-0.33 3.8 142
0620 0.62 0.87 4.0 6 0.30-0.43 4.8 121
'Wave heights determined from the total variance of the record; 'angle based upon mea-
sured orbital vector (after Sherman, 1983); 3Breaker Index, Bb = Hb/gmT* (after Galvin,
1972) where beach slope ( m )= 0.013 and values for both H,, and H, are shown; %caling
Parameter ( c ) = abwP/gtanPm(after Guza and Inman, 1975) where ab is taken as the root-
mean-square amplitude at breaking following Wright et al. (1979); 'estimated based upon
solitary wave breaking criterion (orb = 0.78).
troughs observations indicated that during the storm peak and well into the
decay period the whole of the surf zone was saturated with spilling breakers
propagating across the troughs. Plunging breakers were occasionally observed
on the second bar during the later stages of the storm and only rarely were
true bores developed, occurring within a few metres of the beach face. Waves
dissipated finally by collapsing at the foot of the beach.
The surf zone was in a highly dissipative state at all times and the scaling
parameter (Table I ) varied over two orders of magnitude. Large angles of
wave approach (5"-10" ) throughout the storm (Fig.3) gave rise to significant
shore-parallel currents.
WAVE AND CURRENT SPECTRA
Figure 4 indicates the form and temporal changes evident in the incident
spectra. In an environment such as this, where wind forcing is continuous
into the surf zone, the spectra exhibit considerable energy at a wide range of
frequencies above the peak. In all cases the latter is marked by a very sharp
truncation at the lower frequency end. As expected, spectral growth was
accompanied by a consistent shift in the spectral peak t o lower frequencies
(Fig.4a and b), which remained right through the decay phase.
During the most intense part of the storm a markedly bimodal spectrum
appeared (Fig.4b), with a second peak at twice the frequency of the incident
peak. Energy at this frequency (the first harmonic) has long been recognized
in active surf zones. It can appear as an artifact of harmonic analysis of the
strongly non-linear wave form, but also through the generation of secondary
waves as very rapid shoaling and energy conservation takes place just prior to
40
FREQUENCV (Hz)
1205 805 2 35
+ + ++
465405
+ 1
01
-
3
--
001
0 001
b
I I I , 1 )OOOI
0 125 0 250 0 375 0 500 0 625
FREQUENCY (Hz)
m
I-
8 - - 0001
% 10
C
1
I I 1 I 0 0001
U 0 125 0 250 0 375 0500 0625
41
breaking (Thornton et al., 1974). The presence of secondary waves at this

frequency was confirmed in the surface elevation analogue record (see also
Davidson-Arnott and Randall, 1984, this volume), but overall the energy at
this frequency was low (Table 11).
There is very little evidence at the outer limits of the surf zone for struc-
ture in the spectra at frequencies lower than that of the incident waves
(Fig.4), although under highly dissipative conditions many workers have
noted a dominance of low-frequency energy (e.g. Holman et al., 1978;
Holman, 1981; Wright etal., 1982). A t Wendake this may reflect the distance
offshore of wave wire 9 (136 m). Examination of spectra closer inshore (wave
wire 4, 6 5 m offshore; see Fig.2) reveals an energy peak at a frequency one
fourth that of the incident wave (Fig.5). It was, however, statistically signifi-
cant only at the peak of the storm but remained significant until the very
end. It was associated with a highly dissipative state (Table I) and the largest
incident wave heights and periods. It does not, however, contribute more
TABLE I1
Contributions of differing frequencies of oscillation to shore-normal and shore-parallel

flows in the surf zone, Wendake Beach, 1980:01:00:25
Location offshore Frequency Current speeds*

(m) (Hz) Shore-normal flows urmS Shore-parallel flows uImS
(cm s-I) (cm s-I)
0.188 4.2 4.4

14
{ 0.047 5.0 5.4
0.188 8.0 5.6
38
{ 0.047 4.5 3.2
0.188 8.7. 4.4
57
{ 0.047 5.5 2.6
0.188 12.0 3.6
64
{ 0.047 4.5 3.4
0.188 8.3 2.7
94
{ 0.047 2.6 1.5
0.188 8.7 3.1
106
{ 0.047 2.8 1.3
0.188 10.2 2.6
118
{ 0.047
~~~ ~~~~~ ~
3.6 1.3
Simple calculations based upon variances over one unit bandwidth a t the respective spec-
tral peaks. The peaks were identified using surface elevation spectra.
Fig.4. Wave spectra a t outer margin of the surf zone through the early part of the storm
(a), during the storm peak ( b ) and during the period of storm decay from the peak (c).
These and all following spectra are based on 1024 data points. Unit bandwidth is 0.016 Hz
and is given by the horizontal bar: the 95% confidence band around the spectral estimates
is given by the vertical bar. All spectra have 32" of freedom.
N
- Ip
SPECTRAL ENERGY (cm2Aff-'l
Fig.5. Wave spectra from the inner part of the surf zone through the early part of the
storm (a), the storm peak (b), and the period from storm peak to decay (c).
43
than 10% t o the total variance of the surface elevation even at its maximum
(Table 111). It appears of more importance in the velocity field: the maximum
contribution of this low frequency band to the total energies of the on-
offshore and alongshore motions was found to be 22 and 25%, respectively.
Although energy at the incident frequency dominated the on-shore motions,
the low frequency component was of equal importance in alongshore oscilla-
tions (Table 111). Conclusive identification of the mode of this oscillation is
not possible with the present dataset. Nevertheless, there does appear to be a
spatial pattern t o the energy at this frequency. Figure 6 documents the shore-
normal variability in the spectra from a number of wave staffs and Fig.7 illus-
trates the spatial distribution of peak energies at both incident period and
the longer period component (0.047 Hz) for on-offshore ( u ) and alongshore
( u ) velocity spectra. While there is a general tendency for energy at 0.047 Hz
to decrease with distance from the shoreline (Fig.7 and Table 11) it is impor-
tant to note that a strong peak in u occurs over the crest of bar 2 in coinci-
dence with a trough in the values of u. This suggests the trapping of energy
near the shore as might be expected if the 21 s oscillation were an edge wave,
and also the presence of a nodal point at the bar 2 location if the edge wave
was in a standing mode. Evaluation of phase relationships between surface
elevation ( n ) and the two components of the horizontal velocity field (u and
u ) , for the incident and low-frequency spectral peaks at positions close to the
shoreline, support the presence of edge wave motions at 0.047 Hz (Table IV).
In no case however, did this low-frequency component dominate the spec-
trum. In all cases incident wave energy was more important and only in two
cases were the low-frequency peaks in the velocity spectra statistically
significant (Fig.8).
TABLEIII
Surf zone spectral energies, Wendake Beach, 1 9 8 0 : 0 6 : 0 0 : 2 5
Location offshore Total energy (X lo3) Percentage of total*

(m) n U U n U U
14 (TI)** - 15.0 31.2 - -

38 ( T 2 ) 4.9 41.1 13.6 8 41 15 48 24 10
57 (B2) 26.9 44.0 23.1 10 47 22 43 23 24
64 ( B 2 ) 15.0 61.4 24.8 6 46 14 57 25 18
94 (T3) 8.0 33.8 6.5 4 43 9 53 23 27
106 (B3) 4.3 3.8 10.1 3 43 7 54 22 36
118 ( B 3 ) - 59.1 7.8 - -
3 53 20 11
136 (B3) 20.6 - - 9 4 5 - - -
*Values determined using the variance in unit bandwidth at these peak frequencies.
**Tl,
T2, T3 refer to the trough locations, and B2, B3 to the bar locations.
44
FREQUENCY (Hz)
10,ooc
21.0s 4.6s 2.5s
.) .) c
1,000
1
--
1
6 3
-E
& 100
01
1
a
w
z
w
_I
a
a
+
0
W
% 10
30 1
I I
1 I I I t 3001
0 0 125 0 250 0 375 0 500 0 625
FREQUENCY (Hz)
Fig.6. Wave spectra across the surf zone at the storm peak (0025 h). Note the strength of
the low frequency oscillation at 21s in the inner surf zone (a) compared to further off-
shore (b). Wave wire numbers refer to the locations in Fig.2.
u u
45
Bar 2 Crest Bar 3 Crest
Trough Trough
2.0
1.8 v)
v
rn
1.6 5
52)
1.4
rn
2
1.2 g
0
<
’
1.0 y
w
0.8
0
0.6 3
, v)
0.4 -,
p”
- -
0.2
0
0 10 20 30 40 50 60 70 80 90 100 110 120
DISTANCE LAKEWARD (m)
Fig.7. Spatial distribution of spectral energies f o r t h e incident peak, 5 s (0.188 Hz) and
the low-frequency peak, 2 1 s (0.047 Hz) in t h e shore-normal ( u ) and shore-parallel ( u )
velocity spectra. F o r simplicity t h e value of t h e spectral density over o n e unit bandwidth
at the peak is shown. Right-hand scale refers to the low-frequency peak.
TABLE IV
Phase relationships between surface elevation (I)), onshore velocity ( u ) and longshore
velocity ( u ) , Wendake Beach, 1980:06:00:25 and 06:25
Time Location offshore Frequency Phase (coherence) in degrees

(h) (m) (Hz) I) vs. u u vs. u ’
14 + 8 0 (0.50) + 14 (0.47)
0025 { 38 0.047
{ 0.188
-140 (0.60)
+ 120 (0.78)
-178 (0.98)
-15 (0.75)
157 (0.30)
+ 94 (0.58)
14 0.047 +88 (0.36) + 3 (0.32)
{ 0.250
0620 { 38 0.047
{ 0.250
-130 ( 0 . 7 5 )
+ 6 0 (0.30)
-165 (0.98)
0 (0.18)
-180 (0.12)
-20 (0.08)
46
2 3s
.)
r
0
Y
m
a
10 -
I
I I I I I b
Fig. 8 . Wave and current spectra at the storm peak ( 0 0 2 5 h ) : ( a ) 38 m offshore; ( b ) 57 m

offshore. C2 and C3 are current meter designations and X refers t o the shore-normal
and Y to the shore-parallel directions. W2 and W3 are wave wire designations. Locations
are shown in Fig.2.
47
Spatial and temporal variability of wave-orbital motions
Early in the storm (1500 and 1715 h), wave orbital velocities were already
sufficient to initiate sediment motion (Fig.9). Taking a conservative velocity
of 0.2 m s-l for the threshold of entrainment, it is evident that the Wendake
Beach sands would be in motion across the whole of the three-bar system.
Maximum instantaneous currents a t these times were in excess of 0.7 and
1.3 m s?, respectively. The mean flows (Fig.10) were offshore, but did not
exceed 0.05 m s-l and were therefore close to the instrument accuracy. At
BAR BAR BAR

10
ln
3
1
v)
W
v)
z o
z
Y
v)
-1 0
DISTANCE FROM BASELINE (m)

Fig.9. Spatial variability of wave amplitude, orbital velocity and orbital asymmetry across
the surf zone, Wendake Beach, 1980:05:31:15:00 and 1980:05:31:17:15.Negative skew-
ness is offshore. Dots mark bar crest locations and the vertical dashed line indicates the
first breaker line based upon the depth-controlled breaking criterion.
48
a
m
U
Fig.lO. Spatial variability of mean orbital velocities across the surf zone, Wendake Beach, 1980:05:31:15:00-1980:06:01:06:20.
LL
a
m
a
m
4
m N - 0 - N N - O - N m N - O - N
0 0 . 0 0 0 0 0 0 0 0 0 0 0 0
I I l l I I
m N - 0 -
0 0 0 0
m y - 0 -
0 0 0 0
m N - 0 - m y - 0 -
0 0 0 0 0 0 0 0
I I
49
1500 h wave breaking was initiated on the second bar and by 1715 h the surf
zone width had expanded lakeward of the third bar (Fig.9). Relatively small
variation in orbital velocity was measured across the surf zone, but strong
orbital asymmetries were present in the troughs. Near symmetrical motion
occurred on the bar crests. The basic pattern was one of lakeward asymmetry
(negative skewness) in the innermost trough and landward asymmetry (posi-
tive skewness) in the second trough; an initial lakeward asymmetry changing
to a landward asymmetry was evident in the outer trough (Fig.9).
As wave height increased during the storm so did the orbital velocities: at
the storm peak (2200-0025 h) urmSreached maxima in excess of 0.60 m s-'
(Fig.ll), with peak orbital flows of 1.7 m s-'. As expected, maximum
currents were found in the shallow water over the bar crests. However, since
height loss through breaking was now occurring well-lakeward of the outer
bar and across the full surf zone width, the largest velocities were on the
outer bar with decreasing values landward (Fig.11).
After 2200 h a strong, steady lakeward drift was superimposed on the
orbital velocities across virtually the whole of the surf zone. This drift reached
values of 0.4 m s-' in the second trough and across the crest of the second
bar at 0025 h (Fig.10). Only in the very shallow zone landward of the first
bar were mean flows landward (Fig.lO), as an apparent response to bore
development near the beach face. A t 2200 h asymmetry in the oscillatory
motion was landward across the whole of the surf zone (Fig.l2), and could
be capable of producing net transport opposed to that of the mean drift.
Thus sediment transport differentials could be controlled by this asymmetry
superimposed on the steady drift.
During the peak of the storm (0025 h), a pattern of mean flow and flow
asymmetry developed which remained for the duration of wave activity
(Figs.11 and 13). It was still present at 0620 h when wave heights and absolute
orbital velocities had dropped to values almost identical to those at the
beginning of the storm (cf., Figs.9 and 11).
On-offshore sediment transport, net sediment flux and morphological
change will depend upon the interactions of mean flows with the flow asym-
metries, provided flows are great enough to initiate transport. If, therefore,
one can assume that most sediment transport and morphological change will
be associated with the highest orbital velocities when bedload transport would
be the highest (the storm peak), and with the decaying limb of the storm as
any suspension settling accelerates, then the period 0025-0620 h becomes
critical in any explanation of sediment flux.
The primary flow characteristics at this time were (Figs.10 and 13):
(a) Mean flows were offshore everywhere except landward of the first bar.
(b) Mean flows reached maxima on the landward slope and across the bar
crest of the second bar (0.4 m s-I), decreasing both landward into the trough
and lakeward across the outer trough and bar crest. In the outer part of the
surf zone, flows were less than 0.1 m s-'.
(c) Flow asymmetries were onshore (positive uSk)landward of the central
part of the second trough and lakeward of the centre of the outer trough.
50
BAR BAR BAR
Y
In
-1 0 ll I/
20 30 40 50 60 70 80 90 100 110 120 130 140 150
DISTANCE FROM BASELINE ( m )
Fig.11. Spatial variability of wave amplitude, orbital velocity and orbital asymmetry
across the surf zone, Wendake Beach, 1980:05:31:22:00, 1980:06:01:00:25 and 1980:
06 :0 1 :06 :20.
Across the landward slope and crest of bar two, however, flow asymmetries
were offshore with largest negative skewness values occurring in the part of
the trough immediately landward of the bar slope and crest.
(d) Symmetrical flow fields must therefore have occurred in the second
and outer troughs.
With respect to the contribution of secondary oscillations to the fluid
motion described above it has already been noted that a very large difference
exists between the spectral densities of the water surface elevation at the
incident frequency and the low-frequency peak at 21 s; this difference is not
as great with respect to the two components of the velocity field especially
51
W E N D A K E B E A C H SURF Z O N E
10 BAR BAR BAR
1Lo
3
I
m
v)
$ 0
5
Y
m
-1 0
BAR BAR BAR

10
1Lo
3
I
m
m
$ 0
5
Y
m
-1 0
10 BAR BAR BAR
Lo
1
3
I
(0
w
v,
z o
z
Y
m
-1 0
Fig.12. Spatial variability of orbital asymmetry across the surf zone, Wendake Beach,
1980:0 5 :31 : 15:OO-22:OO.
52
WENDAKE BEACH SURF ZONE
.I U- BAR BAR BAR
x
2
I
m
m
Y o
2
Y
m
-1 0
BAR BAR
10
rn
Y
3
I
m
m
W
z 0
5
Y
cn
-1 0
Fig.13. Spatial variability of orbital asymmetry across the surf zone, Wendake Beach,
1980: 06 :01:00:25-06 :20.
close to the shoreline (Fig.8). However, the low-frequency peaks in the

velocity structure are poorly defined and overall the surf zone is dominated
by oscillatory currents at the incident wave frequency. Table I1 shows the
relative contribution to nearshore currents of energy at the peak incident
frequency and at 0.047 Hz during the time of maximum total energy and
most significant low frequency spectral peak. Except very close to the shore-
line, velocities associated with the incident waves are higher by a factor of
1.5 to 3. If the total energies in the two peaks were considered this difference
would be greater still. Contrary to the results of similar studies, the currents
at Wendake Beach are dominated by oscillations at the incident frequency.
53
Spatial and temporal variability of shore-parallel currents
During the early part of the storm (Fig.14) the time-averaged longshore
component of motion in the surf zone was somewhat non-coherent, generally
less than 0.20 m s-l and thus close to or below the threshold of sediment
motion. However, these flows were superimposed on the shore-normal oscilla-
tions, which would have set sediment in motion at this time and thus long-
shore advection of sediment could occur.
By 1830 h a strongly coherent flow pattern was established and maintained
until the peak of the storm. Maximum velocities occurred on the lower lake-
ward slopes of the bars and reached a value in excess of 0.40 m s-' in the
BAR BAR
-
-- 15100
16115
BAR
-*I.- 17:15
.1.1.1.1.1.
--
"1'-
20:oo
22:oo
2 0.4 -00 25
0.3
0.2
0.1
0
0.1
0.2
0.3
0.4
20 30 40 50 60 70 80 90 100 110 120 130
DISTANCE F R O M BASELINE (m)
Fig. 14. Spatial variability of shore-parallel currents across the surf zone, Wendake Beach,
1980:05:31:15:OO-06 :20.
54
second trough a t 2200 h. Minima occurred just lakeward of the bar crests in
both second and third trough-bar systems. Between 2200 and 0025 h a
change in the direction of wind and wave approach caused a bi-directional,
disequilibrium response in the longshore current (Fig.14). It will be recalled,
however, that a strongly coherent pattern of shore-normal flows existed at
this time. Upon reversal of the longshore current, maximum flows occurred
in the second trough and the lowest velocities were now occurring in the
outer trough. As waves decayed so did the shore-parallel flows until velocities
across the surf zone were again less than 0.20 m s-l by 0620 h, although the
orbital velocities (urms)were still well above this value.
Sediment flux and morphological response
From the preceding discussion it is apparent that the potential for signifi-
cant sediment motion was present over most of the nearshore zone throughout
the storm and that net transport in the on-offshore direction could have been
forced by both the mean flow and asymmetries associated with the orbital
motions. Furthermore, spatial gradients in these flows (giving rise to net
water and sediment flux) and local slopes (controlling the gravity potential
on transported sediment) would interact to determine whether erosion
(accelerating transport rates), deposition (decelerating transport rates) or no
change (constant transport rates) would occur. Shore-parallel flows would
become increasingly important where orbital velocities were low, near-
symmetrical, or both. Any shore-normal gradient in these longshore currents
could then produce transport differentials leading t o shore-normal patterns
in sediment erosion and accretion.
Examination of pre- and post-storm profiles along the instrument transect
and at 30 m either side revealed bars present a t both times, and with rela-
tively small differences in morphology. Thus the bars are not totally destroyed
by a single episode of storm-waves, but rather are in some form of equilibrium,
either with a single storm or a series of storm events. Recent analysis of
depth of activity and bed elevation data using a grid of 51 rods for a series of
six storms at this location suggest strongly the existence of a steady state, at
least in the overall sediment budget (Greenwood, in prep.).
In this storm, morphological changes were evident both from profile data
and the depth-of-activity rods. In general there was a landward shift of the
first and second bar crests ( 5 and 10 m, respectively), while the outer bar
crest remained stable. Accretion of both second and third bars occurred,
with erosion of the adjacent troughs. Landward movement of bars 1 and 2
was associated with a southerly alongshore shift in the sinuous form of the
two bars. The latter is important, since southward flowing currents only
occurred at the storm peak and during the decay period, after the major
switch in the direction of wave approach at approximately 0000 h.
Figure 1 5 illustrates the local depths of activity and bed elevation change
across the bar system together with the spatial distribution of the primary
hydrodynamic parameters a t the peak of the storm (0025 h). This particular
55
data set was used because: (1)at this time both orbital velocities and asym-
metries were greatest and thus likely to reflect the most important sediment
transport phase of the storm; and (2) this general pattern was consistently
present for the longest part of the storm (at least six hours).
Sediment reactivation was, as expected, controlled strongly by the anteced-
ent morphology: in the shallow water over the bar crests, where orbital
velocities were greatest, depth of activity reached a maximum. In contrast,
the outer trough and lakeward slopes of the third bar with deeper water,
lower wave heights and lower orbital velocities were marked by minima. In
the trough landward of bar 2 a relatively high value for sediment reactivation
was obtained because of the shallowness of the water.
It is rather surprising, however, given the measured currents, that the depth
of activity relative to the pre-storm surface was not larger. Maxima were
only 1 4 cm on both second and third bars. This indicates that with high,
relatively uniform, bed stress, sediment entrainment reaches a maximum,
which cannot be increased regardless of the duration of applied stress unless
considerable net transport of sediment occurs. The latter was clearly not the
case: although considerable volumes of sediment were set in motion during
the storm (the average depth of activity was 9 cm), the net flux was less than
two percent of the nearshore control volume (Greenwood, in prep.).
Bed elevation change indicated accretion of the bars and scouring of the
troughs, and thus bar growth during storms: however, these changes were
small, reaching maxima of only a few centimeters (Fig.15). This would
suggest that conditions of near-equilibrium transport existed over the locally
varying nearshore slopes for much of the storm. This could have been
achieved through a continuous, but uniform, sediment transport rate over
the bar system or, more probably, by some form of oscillating equilibrium,
where a balance existed between the net oscillatory flows (mean and asym-
metry), the local slope and sediment inertia.
Accretion on the landward slope'and crest of bar 2 combined with erosion
of the adjacent trough would account for the observed landward migration
of the bar crest (Fig.15). This is best explained by increasing landward flows
from the centre of this trough landward, and a similar increase in lakeward
flows from the centre of the trough lakeward fFig.15). Such a pattern was
evident in both the mean flow (ti) and flow asymmetry ( U , ~ ) .The trough
thus represents an erosional node. Decreases in both the mean flows lake-
ward and the lakeward asymmetries toward and across the landward slope
and crest of bar 2 would explain the deposition here; the relative steep land-
ward slope (3-7" ) would provide further restraint on the lakeward transport.
Enhancement of this pattern would be unlikely as a result of shore-parallel
flows since the shore-parallel currents were lower in this second trough than
on the landward slope and crest of the adjacent bar (see Fig.15).
Decreasing lakeward mean flows and flow asymmetries lakeward of 'the
crest of the middle bar, would suggest continuing accretion into the centre
of the outer trough (albeit somewhat less than on the landward side and
crest of the bar). Also orbital asymmetries either side of the outer trough
56
BAR 3
BAR 2
TROUGH
t
i
3
-1
57
suggest convergence of sediment transport paths towards a depositional node

in the centre of this outer trough. In fact erosion (3-5 cm) characterizes this
location. It would appear, therefore, that with a reduction in u,,,, 0 and Usk,
and the consequent reduction in total load (corroborated in part by the
depth of activity data), the longshore currents become the primary advecting
agents (Fig.15) and caused the observed erosion.
Across the crest of the outer bar, bed elevation changes were positive on
the landward side indicating accretion and negative on the lakeward side
indicating erosion (Fig.15). At this location mean orbital flows were low, as
indeed were mean longshore flows; it would seem that the best correlation is
between the strong landward asymmetries in the orbital velocity field and a
landward movement of sediment across the bar crest (Fig.15).
DISCUSSION AND CONCLUSIONS
Sediment transport and morphological change in the multiple bar system

at Wendake Beach are most satisfactorily explained by the interrelationship
of spatial gradients in the means and asymmetries of shore-normal oscillatory
currents with the local slopes. Orbital asymmetries integrated over time
represent a steady sediment transport potential which, together with the
mean flows, is superimposed on the initial entrainment induced by the orbital
motions themselves.
Figure 1 6 summarizes the patterns of fluid flow, erosion, deposition and
morphological response that can be inferred from the field measurements.
Sediments accumulate on the bars and are scoured from adjacent troughs
during storms to cause bar growth: the alignment of scour and accretion rela-
tive to the pre-storm topography may cause a shift in bar location lakeward
or landward. Sediment deposition occurs, paradoxically, in areas with the
highest absolute orbital speeds: it results from spatial decreases in the orbital
asymmetry and thus in the steady sediment drift. The gravitational resistance
to sediment in transport is enhanced wherever local bed slopes oppose the
steady drift and in some cases may aid accretion, as for example on the land-
ward slope of bar 2 (Fig.16). Erosion results also form gradients in the orbital
asymmetries, but in this case the asymmetries increase spatially and thus the
sediment transport potential increases. Reversals (in the directional sense) of
the orbital asymmetries can create either erosional nodes, as in the case of
the middle trough, or depositional nodes, as in the case of outer trough
(Fig.16).
Shore-parallel steady currents contribute to the net alongshore advection
of sediment. They are of prime importance to the shore-normal patterns of
Fig.15. Spatial distribution of bed elevation change, depth of activity, orbital velocity
(u,,, and u ) , orbital asymmetry ( u s k ) and longshore current velocity (urmS and V ) across
the outer 2 bar-trough systems, at Wendake Beach. Hydrodynamic data are from the storm
peak, 1980:06:01:00:25. u is offshoreatallpointsand V is t o the south at all points. With
respect t o bed elevation change and depth of activity, the solid line represents data from
line IN and the dotted line, data from line 0.
58
Fig.16. Generalized fluid flows, sediment flux and morphological response over the near-
shore bars at Wendake Beach inferred from measurements 1980:05:31-1980:06:01. The
relative magnitude of near bed, shore-normal oscillatory flow is shown by t h e solid double
arrow and the direction and magnitude of t h e net flows (resulting from mean flows and
flow asymmetries) by t h e broad open arrow. T h e relative magnitude of t h e shore-parallel
flow is shown by t h e pairs of thin arrows. Cross-hatching represents areas of erosion and
dots areas of accretion. A solid line indicates the post-storm profile.
sedimentation where absolute orbital velocities are small and/or symmetrical.

The relatively strong longshore current in the outer trough is the most likely
cause of erosion here since the orbital asymmetries indicate a depositional
node.
The energy driving both shore-normal and shore-parallel currents at
Wendake Beach was primarily from incident, wind-generated waves. Second-
ary waves were present during the peak of the storm, but contributed rela-
tively little energy to the total spectrum. Waves at the frequency of the first
harmonic of the incident wave were associated with breaking, and a low-
frequency oscillation at a period four times that of the incident period was
significant very close to the shoreline. The latter would appear to be a stand-
ing edge wave formed during the storm peak and remaining for the duration
of the storm. The strong mean on-offshore flows and orbital asymmetries
were coincident in time with the appearance of the edge wave and it is tempt-
ing to conclude a causal relationship as suggested by previous workers (e.g.
Bowen and Inman, 1971; Holman and Bowen, 1982). If indeed the latter is
correct then the edge wave node located over bar 2 would be consistent with
Symonds and Huntley’s (1980) observation of a resonant edge wave con-
trolled by bar spacing. Furthermore, it would suggest that bars form under
nodes of long waves (Carter et al., 1973) and thus result primarily from bed-
load transport rather than suspension load (Bowen, 1980). Evidence from
primary sedimentary structures (Greenwood, in prep.) indicates a predomi-
nance of bedding produced by oscillation ripples, megaripples and sheet flow.
N o good example of massive bedding as might be produced by suspension
fallout was ever observed. Secondary waves at a frequency of the first har-
monic also coincided with the storm peak, in association with extensive wave
breaking. If, as seems probable, such waves are dispersive (Thornton, 1979)
then interactions with the primary wave may be capable of producing the
orbital patterns observed. Certainly such wave-wave interactions have been
59
proposed for bar genesis (e.g. Hulsbergen, 1974; Bijker e t al., 1976; Van de
Graaf and Tilmans, 1980; Boczar-Karakiewicz and Bona, 1982).
The present paper provides only a qualitative interpretation of the relation-
ships between fluid and sediment motions, and their integration into a quan-
titative model remains for future work. At this time a plethora of theories
exist for predicting shore-normal transport (some purely analytical others
based on results from model experiments) and only detailed comparisons
along the lines of those of Bowen (1980), Seymour and King (1982) and
Bailard (1983) will determine the most appropriate predictor for sediment
transport in a barred system such as Wendake Beach.
ACKNOWLEDGEMENTS
This study forms part of a continuing research programme in Nearshore

Hydrodynamics and Sedimentation supported by equipment (E39218,
E6614) and operating (A7956) grants awarded t o B.G. from the Natural
Sciences and Engineering Research Council of Canada. A University of
Toronto Open Fellowship, an Ontario Graduate Scholarship and a Post-
doctoral Scholarship a t Woods Hole Oceanographic Institution assisted D.J.S.
Computer costs were defrayed by both Scarborough College and the Depart-
ment of Geography, University of Toronto. The Academic Workshops a t
Scarborough College provided instrument design and construction and the
Graphics and Photography Department assisted with the illustrations. Invalu-
able field and laboratory assistance was provided by M. Rollingson, P.
Christilaw and R. Sutherland (Scarborough College) and R.G.D. Davidson-
Arnott and D.C. Randall (University of Guelph). The writing of this paper
was greatly assisted by the facilities and hospitality received by the senior
author during research leave visits to the Universities of Uppsala, Sweden and
Sydney, Australia. In particular Professor J.O. Norrman, Goran and Bodil
Albjar (Uppsala), and Professor M. Daly and Dr. Andy Short (Sydney) are to
be thanked. Drs. J. Bailard and A.H. Sallenger reviewed the manuscript and
are thanked for their comments. The results and interpretations presented in
this paper are, however, solely the responsibility of the authors.
REFERENCES
Bailard, J.A., 1981. An energetics total load sediment transport model for a plane sloping
beach. J. Geophys. Res., 86: 10,938-10,954.
Bailard, J.A., 1983. Modeling offshore transport in the surf zone. Proc. 18th Coastal
Engineering Conf., Cape Town, pp.1419-1438.
Bailard, J.A. and Inman, D.L., 1981. An energetics bedload model for a plane sloping
beach: local transport. J. Geophys. Res., 86: 2035-2043.
Bijker, E.W., Van Hijum, E. and Vellinga, P., 1976. Sand transport by waves. Proc. 15th
Coastal Engineering Conf., Honolulu, Hawaii, pp.1149-1167.
Boczar-Karakiewicz, B. and Bona, J.L., 1982. Shallow water waves over mildly varying
topography. Abstracts, 18th International Conf. on Coastal Engineering, Cape Town,
pp. 215-276.
60
Bowen, A.J., 1980. Simple models of nearshore sedimentation; beach profiles and long-
shore bars. In: S.B. McCann (Editor), The Coastline of Canada. Geol. Surv. Can., Pap.
80-10, pp.1-11.
76; 8862-8871.
Bowman, D. and Goldsmith, V . , 1983. Bar morphology of dissipative beaches: an empir-
ical model. Mar. Geol., 51: 15-33.
Carter, T.G., Liu, P.L.-F. and Mei, C.C., 1973. Mass transport by waves and offshore sand
bedforms. J. Waterways, Harbors Coasts, Am. SOC.Civ. Eng., WW2: 165-184.
Davidson-Arnott, R.G.D. and Randall, D.C., 1984. Spatial and temporal variations in
spectra of storm waves across a barred nearshore. In: B. Greenwood and R.A. Davis,
J r . (Editors), Hydrodynamics and Sedimentation in Wave-Dominated Coastal Environ-
ments. Mar. Geol., 60: 15-30 (this volume).
De Beaumont, E., 1845. Lecons de G6ologique Practique. Paris.
Dixon, W.J., 1971. BMD: Biomedical Computer Programs. University of California Press,
Berkeley, Calif., 600 pp.
Dixon, W.J. and Brown, M.B., 1979. BMD P-79: Biomedical Computer Programs, €'-Series,
University of California Press, Berkeley, Calif.
Galvin, C.J., 1972. Wave breaking in shallow water. In: R.F. Meyer (Editor), Waves on
Beaches and Resulting Sediment Transport. Academic Press, New York, N.Y.,
pp.4 13-4 55.
Goldsmith, V., Bowman, D. and Kiley, K., 1982. Sequential stage development of cres-
centic bars: HaHoterim beach, southeastern Mediterranean. J. Sediment. Petrol., 52:
233-249.
Grant, W.P. and Madsen, O.S., 1979. Combined wave and current interaction with a
rough bottom. J. Geophys. Res., 84: 1797-1808.
Greenwood, B., 1982. Bars. In: M.L. Schwartz (Editor), Encyclopaedia of Beaches and
Coastal Environments. Dowden, Hutchinson and Ross, Stroudsberg, Pa., pp.135-139.
Greenwood, B., 1983. Hydrodynamics and sedimentation in barred nearshores: a review
of work by the Scarborough Campus Coastal Research Group. In: N.A. Rukavina
(Editor), Proc. Third Workshop on Great Lakes Coastal Erosion and Sedimentation.
Natl. Water Research Institute, Burlington, Ont., pp.45-48.
wave-formed bars: a review and case study. Can. J. Earth Sci., 1 6 : 312-332.
Greenwood, B. and Hale, P.B., 1980. Depth of activity, sediment flux and morphological
change in a barred nearshore environment. In: S.B. McCann (Editor), The Coastline of
Canada: Littoral Processes and Shore Morphology. Geol. Surv. Can., Pap. 80-10,
pp.89-109.
Greenwood, B. and Mittler, P.R., 1984. Sediment flux and equilibrium slopes in a barred
nearshore. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedi-
mentation in Wave-Dominated Coastal Environments. Mar. Geol., 60: 79-98 (this
volume).
Greenwood, B. and Sherman, D.J., 1983. Shore-parallel flows in a barred nearshore. Proc.
18th Coastal Engineering Conf., Cape Town, pp.1677-1696.
Greenwood, B., Hale, P.B. and Mittler, P.R., 1980. Sediment flux determination in the
nearshore zone: prototype measurements. In: Workshop on Instrumentation for
Currents and Sediments in the Nearshore Zone, National Research Council of Canada,
Associate Committee for Coastal Erosion and Sedimentation, Ottawa, Ont., pp.99-120.
Guza, R.T. and Inman, D.L., 1975. Edge waves and beach cusps. J. Geophys. Res., 80:
299 7-3 0 12.
Holman, R.A., 1981. Infragravity energy in the surf zone. J. Geophys. Res., 86: 6442-
6450.
Holman, R.A. and Bowen, A.J., 1982. Bars, bumps and holes: models for the generation
of complex topography. J. Geophys. Res., 87: 457-468.
Holman, R.A., Huntley, D.A. and Bowen, A.J., 1978. Infragravity waves in storm condi-
tions. Proc. 16th Coastal Engineering Conf., Hamburg, pp.268-284.
61
Hulsbergen, C.H., 1974. Origin, effect and suppression of secondary waves. Proc. 14th
Coastal Engineering Conf., Copenhagen, pp.392-411.
Huntley, D.A., 1980. Electromagnetic flow meters in nearshore field studies. In: Work-
shop on Instrumentation for Currents and Sediments in the Nearshore Zone, National
Research Council of Canada, Associate Committee for Coastal Erosion and Sedimenta-
tion, Ottawa, Ont., pp.47-60.
Inman, D.L. and Bagnold, R.A., 1963. Littoral processes. In: M.N. Hill (Editor), The Sea,
Inman, D.L. and Frautschy, J.D., 1966. Littoral processes and the development of shore-
lines. Proc. Coastal Engineering Speciality Conference, Santa Barbara, Calif., pp.511-
536.
Jonsson, I.G., 1980. A new approach to oscillatory rough turbulent boundary layers.
Ocean Eng., 7: 109-152.
Seymour, R.J. and King Jr., D.B., 1982. Field comparisons of cross-shore transport models.
J. Waterway, Port, Coastal and Ocean Div., 108: 163-179.
Symonds, G. and Huntley, D.A., 1980. Waves and currents over a nearshore bar system.
Proc. Canadian Coastal Conf., Natl. Res. Council Can., pp. 64-78.
Thornton, E.B., 1979. Energetics of breaking waves within the surf zone. J. Geophys. Res.,
84: 4931-4938.
Thornton, E.B., Galvin, J.J., Bub, F.L. and Richardson, D.P., 1974. Kinematics of break-
ing waves. Proc. 15th Coastal Engineering Conf., Honolulu, Hawaii, pp.461-476.
Truxillo, S.G., 1970. Development of a resistance-wire wave gauge for shallow-water wave
and water level investigations. Unpublished manuscript, Coastal Studies Institute,
Louisiana State University, Baton Rouge, La.
Van de Graaf, J. and Tilmans, W.M.K., 1980. Sand transport by waves. Proc. 17th Coastal
Engineering Conf., Sydney, N.S.W.
Wells, D.R., 1967. Beach equilibrium and second-order wave theory. J. Geophys. Res.,
77: 497-504.
Wright, L.D., Chappell, J., Thom, B.G., Bradshaw, M.P. and Cowell, P., 1979. Morpho-
dynamics of reflective and dissipative beach and inshore systems: Southeastern Australia.
Mar. Geol., 32: 105-140.
Wright, L.D., Guza, R.T. and Short, A.D., 1982. Dynamics of a high-energy dissipative
surf zone. Mar. Geol., 45: 41-62.
Marine Geology, 6 0 ( 1 9 8 4 ) 63-78 63
Elsevier Science Publishers B.V., Amsterdam - Printed in T h e Netherlands
LANDWARD MIGRATION OF INNER BARS
TSUGUO SUNAMURA and ICHIROU TAKEDA

Institute of Geoscience, University of Tsukuba, Ibaraki (Japan)
(Received December 3 0 , 1982; revised and accepted April 1 2 , 1 9 8 3 )
ABSTRACT
Sunamura, T. and Takeda, I., 1984. Landward migration of inner bars. I n : 3.Greenwood
and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated
Coastal Environments. Mar. Geol., 6 0 : 63-78.
A field investigation was carried o u t to collect data of inner bar migration. Profiles
were measured once o r twice a week f o r a two-year period a t Naka Beach, Ibaraki Prefec-
ture, Japan. I t was found that t h e onshore migration of inner bars could be described by
two dimensionless quantities as: 5D/(Hb),,, < (Hb)max/gT&x < 2oD/(Hb)max where
(H&, is t h e maximum value of daily average breaker height during one interval between
surveys, T,, is t h e average wave period of t h e day giving (Hb),=, D is t h e mean size of
the beach sediment, and g is t h e acceleration d u e t o gravity. Analyses based o n surfzone
sediment dynamics yields E / ( w D / b )= 2 X lj-” where E is t h e average speed of
onshore bar-migration, b is t h e bar height, Hb is t h e average breaker height, and w is t h e
fall velocity of t h e beach sediment. Nomographs for t h e speed of landward migrating bars
are also presented.
INTRODUCTION
Storm waves transport beach material offshore causing beach erosion and
form a sand bar in the surfzone as a temporal sediment reservoir. Post-storm
waves gradually move the sand bar onshore. The bar eventually emerges from
the water level and welds onto the beach face. Such migrating bars have been
called “ridge and runnel” topography by North American sedimentologists
(e.g., Davis et al., 1972; Hayes, 1972; Owens and Frobel, 1977), but, because
the application of this terminology has been questioned by Orford and Wright
(1978), the present paper uses the term “inner bar”.
Many researchers have considered the landward migration of inner bars
(e.g., Evans, 1939; King and Williams, 1949; Sonu, 1969, 1973; Davis and
Fox, 1972a, b, 1975; Davis et al., 1972; Hayes, 1972; Greenwood and
Davidson-Arnott, 1975; Owens and Frobel, 1977; Fox and Davis, 1978;
Short, 1978, 1979; Hine, 1979; Sasaki, 1982). However, few quantitative
studies have been performed on this problem in connection with surfzone
sediment dynamics. The purpose of the present study is to relate the landward
migration of the inner bar t o nearshore wave parameters and sediment
properties.

64
STUDY AREA AND DATA ACQUISITION
Naka Beach, Ibaraki Prefecture, Japan, facing the Pacific Ocean was
selected as the study area (Fig.1). The beach, located between Tokaimura
and Ajigaura, is an approximately straight, north-south oriented, open coast
with a 5-km shoreline length. The shoreline is stable on a long-term basis
(Tanaka et al., 1973). The beach sediment is composed of coarse sand in the
northern part of the beach and nearly fine sand in the southern part.
An outer bar is located 200-300 m offshore, nearly parallel to the shore-
line; the water depth at the outer bar crest is 2-3 m below MSL. Nearshore
bottom contours are approximately parallel to the shoreline (Fig.1). The
average bottom gradient is 0.011 to a water depth of 20 m, and is almost
constant along the beach. No significant alongshore difference in incident
wave characteristics has been observed. The study area experiences a maxi-
mum tidal range of 1.4 m and a mean of 1m.
Three monitoring sites were established along this beach (Fig.1). They are
North, Central, and South sites, the first having an alongshore length of
300 m and the other two a length of 500 m each. Beach profile surveys were
conducted at each site at an interval of once or twice per week for a period
I
Tokat mura
-5 -10 -15 -20 -25m
Fig.1. Study area and three monitoring sites.

65
of two years beginning August 28, 1980. Survey lines, drawn perpendicular
to the general shoreline trend, were established at an interval of 20 m at the
North site and 50 m at the Central and the South sites. The survey, conducted
using a telescopic level, a surveyor's rod, and tape, was extended to the limit
of wading, approximately 1.5 m deep.
Sediment samples were collected from the mid-foreshore at each site on
October 23,1980, April 5,1981, and September 19,1981. The time-averaged,
mean grainsize of the beach sediment was found to be 0.76 mm (0.39 @ )at
the North site, 0.66 mm (0.60 @ )at the Central site, and 0.26 mm (1.9 @ )at
the South site (Takeda, 1983).
Twenty-minute wave measurements were made every two hours by an
ultrasonic-type wave gage installed 10 km south of the study area (at a water
depth of 21 m). Daily averages of deep-water significant wave height and
period are obtained.
RESULTS AND DISCUSSION
Condition f o r landward bar-migration
Figure 2 shows some representative examples of onshore migration

sequences of inner bars. The bar migrates onshore with a marked slip-face, as
reported widely from the coasts of other parts of the world (e.g., Hoyt, 1962;
Davis et al., 1972; Hayes, 1972; Greenwood and Davidson-Arnott, 1975;
Owens and Frobel, 1977; Hine, 1979; Hunter et al., 1979; Dabrio and Polo,
1981). The onshore migration is brought about by onshore sediment trans-
port (e.g., Komar, 1976, p.298). Such distinct beach accretion as shown in
Fig.2b, d, and e usually took place during shoreward bar-migration, although
minor beach erosion was sometimes observed.
Figure 3 shows a few examples of offshore migration of inner bars. Large
waves often hindered the, nearshore profile survey. On such occasions, the
position of the bar crest was roughly determined by wading when it was
possible. The dashed line in Fig.3 shows the approximate bar profile. Davis
and Fox (1975) have also observed offshore migration under stormy condi-
tions. Offshore sediment transport causes the seaward bar-migration, giving
rise to beach erosion (Fig.3).
Because the direction (onshore or offshore) of bar migration seems to be
closely related to the shoreline change (accretion or erosion), the relation-
ship which allows a demarcation of beach accretion and erosion could be
applied to the delimitation of bar migrating directions. Such an accretion-
erosion relationship (Sunamura, 1980), originally derived from wave-tank
experiments (Sunamura and Horikawa, 1974), is described by:
where H,, and Lo are the deep-water wave height and wavelength, respectively,
tan 0 is the average nearshore bottom slope to a water depth of 20 m, D is
66
Onshore migration
Sept. 19
M.S.L.
North site \
-c. --
Sept. 15, 1981
Sept. 2
Central stte ',, ..

-- - - _ _ _
\
Aug. 30. 1980
I I I I I I
0 10 20 30 40 50 m
Fig.2. Onshore movement of inner bars.
67
Offshore migration
-- . Sept. 20, 1980
\
SeDt. 23
't
\
Sept. 2 5
I I
0 10 20 30 40 50 m
Fig.3. Offshore movement of inner bars.
the grain size of beach sediment, and C is an empirically determined constant

(C= 18). Accretion (or erosion) takes place when the right-hand side of eq. 1
is greater (or less) than the left-hand side.
A relationship among wave breaker height, deep-water wave parameters,
and nearshore bottom slope is approximated explicitly by (Sunamura and
Horikawa, 1974; Sunamura, 1982):
where Hb is the breaker height. The deep-water wavelength, Lo, is related to

the wave period, T , by linear theory:
Lo=-gT2 (3)
2n
where g is the acceleration due to gravity.
Using these three equations, the following relationship can be derived:
68
where K is a dimensionless coefficient. In this derivation, such approximations

as 1.35 4/3, 0.34 N 1/3, and 0.67 2/3 were used. Equation 4 is basically
the same as eq. 1 , but eq. 4 is convenient t o use in the field, because Hb and
T can be easily measured using a surveyor's rod and stopwatch (e.g., Bascom,
1964, p.173; Hoyt, 1971, p.21), although these measurements give only
approximate values. In this study, Hb was estimated by substituting the data
of the daily average wave parameters and tan ?!, = 0.011 into eq. 2.
Because temporal changes in wave climate exist even in a day, and because
the topographic change is most sensitive t o larger waves during one interval
between surveys, the demarcation of bar migrating directions is better
indexed by:
where (Hb)maxis the maximum value of daily average breaker height during
one interval between surveys, TmaXis the average wave period of the day
giving (Hb)max,and h is a dimensionless coefficient. Figure 4 shows the
demarcation of bar migrating directions (Table I). Kimura's (1976) data,
obtained at Tatado Beach, Shimoda, Japan, are also plotted. Although some
overlapping of the data points is seen, the demarcation can be reasonably
described by the solid line:
10':
D/(Hb)max
Fig.4. Delimitation of inner-bar migrating directions.

69
TABLE I
Data of bar migrating directions (notation explained in text)
Site D Tmax Migrating direction Remarks

(mm) (m) (s)
North 0.76 1.53 7.6 Onshore 9115-9/19/81

4.02 10.2 Offshore 9 / 25-9/29/81
Central 0.66 2.24 9.3 Onshore 8128-8/30/80
2.16 9.6 Onshore 8/3+9/2/80
1.63 8.7 Onshore 912-9/6/80
1.49 6.5 Onshore 9/6-9/9/80
1.44 6.9 Onshore 9112-911 5/80
1.42 7.5 Onshore 9122-9/25/80
South 0.26 1.73 7.1 Onshore 917-9/10/80
1.13 7.3 Onshore 9/ 13-9/16/80
1.70 6.3 Offshore 9116-9/ 2 0 / 8 0
1.62 7.8 Offshore 9/2+9/23/80
4.02 10.2 Offshore 9126--10/4/80
1.36 9.1 Onshore 1018-1 0111 / 8 0
1.77 6.9 Onshore 10111-1 0/14/80
2.08 8.1 Offshore 10114--1 O/ 15/80
2.55 8.0 Offshore 10123- 10/28/80
1.15 7.6 Onshore 1117--11/12/80
1.52 6.1 Onshore 11112-1 1/17/80
2.37 11.6 Offshore 2122-2/28/81
1.80 8.7 Onshore 513-511 1181
1.97 6.2 Offshore 51 11---5/17/81
3.30 9.7 Offshore 5117--5124181
1.62 7.8 Offshore 611- 6 / 7 / 8 1
1.65 7.9 Onshore 6/7-61 1 5 / 8 1
2.28 9.5 Offshore 6 / 15-61 2 1181
1.00 6.7 Onshore 719-7/14/81
1,.6 5 8.1 Onshore 9111-9/16/81
1.53 7.6 Onshore 9116-9/20/81
1.62 7.8 Onshore 10/ 13-10/18/81
3.65 11.4 Offshore 10118-10/24/81
Tatado Beach 0.23 1.34 6.3 Onshore 8/14-8/15/76
(Kimura, 1976) 1.40 8.0 Onshore 8117-811 9/76
1.15 8.0 Onshore 8119-8/21/76
1.01 6.8 Onshore 8 / 21-8/22/76
1.03 6.4 Onshore 8122-8/23/76
1.36 6.4 Onshore 8125-8/29/76
The dashed line would probably demarcate onshore migration and no migra-
tion of bars, although the data showing no migration have not been obtained.
The area below the dashed line indicates lower-energy waves which are too
small to move bars onshore effectively. The dashed line is written as:
70
Figure 4 indicates that the necessary condition for onshore bar-migration

should be located in the area between the solid and the dashed lines.
Bar migration speed
Rates of onshore migration of bars have been measured at various beaches;

for example, a rate of 1 inch 30 min-' (1.2 m day-') has been reported from
the coast of Lake Michigan by Evans (1939), 4 f t h-' (29.3 m day-') from
Nags Head, North Carolina by Sonu (1969), 9.23 cm h-' (2.2 m day-') from
Ajigaura, Japan by Hashimoto and Uda (1976), 13.7 cm h-' (3.3 m day-')
from the same location by Hashimoto and Uda (1977), 0.83-10 m day-' from
the coasts of Magdalen Islands, Quebec by Owens and Frobel (1977), 1-5 m
day-' from the Oregon coast by Fox and Davis (1978), and 4-5 m day-' from
Dai-nigori-zawa Beach, Japan by Sasaki (1982). Hayes (1972) stated that the
rate of migration depends upon sediment grain size, nearshore gradient, wave
conditions, length of wave duration, and tidal range. However, no studies
have been performed to quantitatively relate the bar migration speed t o the
controlling factors.
When a bar is migrating onshore, wave breaking always occurs over the
gently seaward-sloping surface of the bar. Waves after breaking form bores
which advance over the bar. Such bores transport sediment onshore across
the gently sloping surface, primarily in a bed-load manner, and eventually
the material is moved to the steep landward edge of the bar where it is depos-
ited on the slip-face of the bar, as is schematically illustrated in Fig. 5. This
type of sediment transport and deposition on the slip-face causes onshore
migration of bars.
Because bed-load is dominant for this type of sediment transport, Shields
parameter would be useful:
where Jr is Shields parameter, T~ is the bottom shear stress, s = p s / p , p s is the

sediment density, and p is the fluid density. Shear stress, T ~ can
, be expressed
by:
Plunge point Break point
S.L.
Fig.5. Schematic diagram showing wave transformation and sediment transport over a bar.
71
70 = 1P f W 4 (9)
where u o is the maximum bottom velocity on the bar crest (Fig.6) and f , is
the wave friction factor. Assume that the wave height at the bar crest, H ,
(Fig.G), is linearly related to the breaker height, Hb, as:
H, =3Hb (10)
where B is a dimensionless coefficient ( B< 1).Because H , is a strongly depth-
controlled quantity due to the shallow water depth, the following relationship
(e.g., Komar, 1976, p.56) is employed:
5 = 0.78
h*
where h , is the water depth at the bar crest (Fig.6). Linear shallow-water
wave approximation yields (see Appendix):
2
u0=%&
Using eqs. 8 through 12, Shields parameter on the bar crest is related to the
breaker height as:
= 0.098 B f w Hb
(s - 1)D
Part of the sediment transported onshore across the bar crest is deposited on
the landward steep slope. The remaining part is transported alongshore by
longshore currents which develop in the trough. Denoting q , as the deposi-
tional rate on the steep slope and qo as the net onshore sediment transport
rate on the bar crest (Fig. 7) :
Q, = A q o (14)
where A is a dimensionless coefficient ( A < 1).This equation holds when
two-dimensionality of sediment transport and of the resultant bar migration
is maintained. It would be valid to assume that the depositional area, which
is shown by the hatched area in Fig.7, can be expressed by the parallelogram
€or a short period of time. Then:
A S = bAl (15)
Fig.6. Definition sketch.

72
Fig.7. Definition sketch.
where A S is the cross-sectional area, A1 is the bar migration distance, and b is

the bar height. Dividing both sides of this equation by the time interval, A t ,
we have :
The left-hand side of eq. 16, A S / A t , should be equal to q , which is the depo-
sitional rate, and AZ/At in the right-hand side is the bar migration speed.
Namely:
q , = bv (17)
where u is the bar migration speed. Equations 14 and 17 lead to:
bu
40 = 2
The dimensionless sediment transport rate, @, is described as:
where w is the fall velocity of sediment. According to the work of Madsen

and Grant (1976a, b):
where c is a dimensionless coefficient. Using eqs. 13, 18, 19, and 20, the
wy(%r
normalized bar migration speed, u / (w D / b ) ,is:
u (
- Ac 0.098Bf
m- s-1
Assuming that all the coefficients involved in eq. 21, i.e., A, B, c, f,, ands,
are constants, we have:
w u m = K’ (3)’
where K’ = Ac[0.098Bfw/(s - l)] which is an unknown constant. Equation
22 shows that the rate of onshore bar-migration is expressed as a function of
13
the breaker height, the bar height, and the sediment properties (grain size
and fall velocity). Because of the existence of temporal changes in wave
climate, however, eq. 22 was replaced by:
ij -K ( 5 ) 3
w m-
where at, is the time-averaged breaker height during one interval between
surveys, 0 is the average speed of bar migration, and K is an unknown con-
stant. The tidal effects have been neglected in the present study.
The speed of onshore bar-migration was examined using eq. 23. Due t o
assumptions involved in deriving eq. 23, only the data which satisfy the
following two conditions were applied: (1)migrating bars have distinct two-
dimensionality; and (2) the bar height does not significantly change on two
successive beach profiles. The average bar-migration speed, 0, was obtained
from the beach profile records. The bar height, b , was determined from the
first of two successive surveys. The time-averaged, beach sediment grain size
at each monitoring site was used for D, i.e., D = 0.76 mm for the North site,
D = 0.66 mm for the Central site, and D = 0.26 mm for the South site. The
fall velocity of sediment, w , was obtained from the D - w relationship (e.g.,
U S . Army Coastal Engineering Research Center, 1973, fig.4-31).
Selected data are listed in Table 11, which includes Kimura's (1976) data
obtained a t Tatado Beach, Shimoda, Japan. Figure 8 gives a plot of these
data. Although some scatter in the data points is found, the average speed of
onshore migrating bars can be well described by:
Possible reasons for the data scatter are: (1)the sediment grain size, D, is
treated as a constant, but this does vary slightly in time; and (2) the effect of
tides upon the bar migration speed (Davis et al., 1972) is not considered.
Nomographs f o r bar migration speed
Rewriting of eqs. 6 and 7 on a daily basis gives the following relation which
describes the condition for the onshore migration of inner bars:
where i;r, and are the daily average breaker height and wave period, respec-
tively. Similarly, eq. 24 can be rewritten as:
where i7 is the daily average migration speed.

Using eqs. 25 and 26, nomographs giving the speed of onshore migrating
bars were plotted in Figs.Sa, b, and c for the beaches with a sediment grain
74
TABLE 11
Data of onshore bar-migration speed (notation explained in text)
Site D gb b V Remarks
(mm) (m) (m) ( m day-')
North 0.76 1.41 0.2 2.3 9115-91 19181

1.41 0.2 2.5 9115-9/19/81
Central 0.66 2.16 0.8 8.5 a/ 2 8-813018 0
2.16 0.7 11.5 8128-8/30/80
1.94 0.5 3.3 8130-9/2/80
1.14 0.2 3.3 9112-9/15/80
1.14 0.4 2.3 9/ 12-9/15/80
South 0.26 1.50 0.8 4.3 917-9/10/80
1.06 0.3 5.0 9113-9/16/80
1.06 0.4 3.0 9113-91 16/80
1.06 0.4 3.3 9/13-9/16/80
1.28 1.1 3.7 1O/8-1 O / 11/80
1.28 0.5 3.0 1018-10/11/80
1.36 0.9 3.7 10/11-10/14/80
1.36 0.8 2.7 10/11-10/14/80
0.94 0.2 2.2 1117-11/12/80
0.94 0.2 2.4 1117-11/12/80
0.94 0.4 1.2 11/7-11/12/80
1.29 0.3 2.2 11112-1 1/17/80
1.41 0.3 3.0 10113-1 0/18/81
1.41 0.4 2.6 10113-10/18/81
Tatado Beach 0.23 1.34 0.4 3.5 8114-81 15/76
(Kimura, 1976) 1.21 0.6 5.0 8/17-81 19/76
1.03 0.6 5.0 8/ 19-8/ 2 1/ 7 6
1.01 0.5 3.5 8121-8/22/76
1.03 0.5 3.0 8122-8/23/76
1.13 0.4 3.8 8125-8/29/76
size of 0.2, 0.4, and 0.8 mm, respectively. These figures were constructed for
the bars having a relative height of 0.5 and 1m.
Assume, for example, that waves having a daily average breaker height of
2 m and a wave period of 10 s act on a beach composed of 0.4-mm sand. Use
Fig.9b for this case. An intersecting point of two lines, i.e., fib = 2 m and =
1 0 s, is located near the line showing 5 m day-' for a l-m bar height case or
10 m day-' for an 0.5-m case. Namely, under such wave and sediment condi-
tions, the average speed of onshore bar-migration is estimated at 5 m day-' if
the bar height is 1 m, or 1 0 m day-' if the bar height is 0.5 m.
CONCLUDING REMARKS
Demarcation of migrating directions of inner bars is shown in Fig.4. A

necessary condition for landward bar-migration should be located in the area
between the solid line (eq. 6 ) and the dashed line (eq. 7) in this figure. The
-
-
0 N m l h stle
A Central site
o South site
] Naka BeachIPresent sludyl I
-
v Talado Beach iKlmura.19761
1o21 I I I I I I I I I I I I I11111 I I I I I I I I I
10' 1O'O 10" 10l2

( Hb/0 )3
Fig.8. Dimensionless plot of average speed of onshore bar-migration.
average speed of onshore migrating bars is described by eq. 24, which is a

function of the wave breaker height, the grain size and the fall velocity of
the beach sediment, and the bar height. Nomographs for the migration speed
are presented for the selected sediment grain size and bar height (Figs. 9a-c).
ACKNOWLEDGMENTS
Part of this study is financially supported through the Science Research

Fund of the Ministry of Education (56020001, Principal investigator:
Professor A. Ozaki, Hokkaido University) and the Kajima Foundation's
Research Grant (Principal investigator: Associate Professor H. Nishimura,
University of Tsukuba). Permission t o use the wave data was given by the
Sanpin Kowan Office of Ibaraki Prefecture. H. Tsujimoto, T. Mori, and T.
Shimizu assisted in the field survey.
APPENDIX
Linear wave theory gives the maximum orbital velocity of the water particle near the
bottom, u,, as:
u, =
nH
T sinh ( 2 n h / L ) (A-1)
where H and L are the wave height and the wavelength at a water depth of h , respectively,
and T is the wave period. Using shallow-water wave approximations such as:
76
2nh 2nh
sinh - =
L
-L
and
L
= &E (= wave velocity)
transformation of eq. A-1 leads t o :
u, =F&
where u , is the maximum bottom velocity in a shallow water region.
sec
20
I I u
0 1 2 3 4rn
Breaker height. &,
se6
20
15
0 1 2 3 4 5rn
Breaker height, fib

77
0 1 2 3 4 5 6rn
Breaker height, u b
Fig.9. Nomographs for daily average speed of onshore migrating inner bars; %b and ? =
daily average breaker height and wave period, respectively, D = sediment grain size, and
b = bar height.
REFERENCES
Bascom, W., 1964. Waves and Beaches. Doubleday, Garden City, New York, 2 6 1 pp.
Dabrio, C.J. and Polo, M.D., 1981. Flow regime and bedforms in a ridge and runnel
system, S. E. Spain. Sediment. Geol., 28: 97-110.
Davis, R.A. and Fox, W.T., 1972a. Coastal processes and nearshore sand bars. J. Sedi-
ment. Petrol., 42: 401-412.
Davis, R.A. and Fox, W.T., 1972b. Four-dimensional model for beach and inner near-
shore sedimentation. J. Geol., 80: 484-493.
Davis, R.A. and Fox, W.T., 1975. Process-response patterns in beach and nearshore sedi-
mentation: I. Mustang Island, Texas. J. Sediment. Petrol., 45: 852-865.
Davis, R.A., Fox,W.T., Hayes, M.O. and Boothroyd, J.C., 1972. Comparison of ridge and
runnel systems in tidal and non-tidal environments. J. Sediment. Petrol., 42: 413-421.
Evans, O.F., 1939. Mass transportation of sediments on subaqueous terraces. J. Geol., 47:
3 2 5-3 34.
Fox, W.T. and Davis, R.A., 1978. Seasonal variation in beach erosion and sedimentation
on the Oregon Coast. Geol. SOC.Am. Bull., 89: 1541-1549.
Greenwood, B. and Davidson-Arnott, R.G.D., 1975. Marine bars and nearshore sedimen-
tary processes, Kouchibouguac Bay, New Brunswick. In: J. Hails and A. Carr (Editors),
Nearshore Sediment Dynamics and Sedimentation. Wiley, London, pp.123-149.
Hashimoto, H. and Uda, T., 1976. Shore processes at Ajigaura Beach. Proc. 23rd Japan.
Conf. Coastal Eng., pp.245-249 (in Japanese).
Hashimoto, H. and Uda, T., 1977. Shore processes a t Ajigaura Beach (2). Proc. 24th
Japan. Conf. Coastal Eng., pp.216-220 (in Japanese).
Hayes, M.O., 1972. Forms of sediment accumulation in the beach zone. In: R.E. Meyer
(Editor), Waves on Beaches. Academic Press, New York, N.Y., pp.297-356.
Hine, A.C., 1979. Mechanisms of berm development and resulting beach growth along a
barrier spit complex. Sedimentology, 26: 333-351.
Hoyt, J.H., 1962, High-angle beach stratification, Sapelo Island, Georgia. J. Sediment.
Petrol., 32: 309-311.
78
Hoyt, J.H., 1971. Field Guide to Beaches. Houghton Mifflin, Boston, 44 pp.
Hunter, R.E., Clifton, H.E. and Phillips, R.L., 1979. Depositional processes, sedimentary
structures, and predicted vertical sequences in barred nearshore systems, southern
Oregon coast. J. Sediment. Petrol., 49: 711-726.
Kimura, R., 1976. Change of beach profile - in the case of Tatado Beach, Shimoda,
Shizuoka Prefecture, Japan. Unpubl. BS Thesis, Dept. Geogr., Tokyo Kyoiku Univ.,
53 pp. (in Japanese).
King, C.A.M. and Williams, W.W., 1949. The formation and movement of sand bars by
wave action. Geogr. J., 112: 70-85.
Komar, P.D., 1976. Beach Processes and Sedimentation. Prentice-Hall, Englewood Cliffs,
N.J., 429 pp.
Madsen, O.S. and Grant, W.D., 1976a. Sediment transport in the coastal environment.
Tech. Rep., Dept. Civ. Eng., M.I.T., 209, 1 0 5 pp.
Madsen, O.S. and Grant, W.D., 1976b. Quantitative description of sediment transport by
waves. Proc. 15th Int. Conf. Coastal Eng., pp.1093-1112.
Orford, J.D. and Wright, P., 1978. What’s in a name? - Descriptive or genetic implications
of “ridge and runnel” topography. Mar. Geol., 28: M1-M8.
Owens, E.H. and Frobel, D.H., 1977. Ridge and runnel systems in the Magdalen Islands,
Quebec. J. Sediment. Petrol., 47: 191-198.
Sasaki, T., 1982. Three-dimensional topographic changes on the foreshore zone of sandy
beaches. Unpubl. Doct. Thesis, Inst. Geoscience, Univ. Tsukuba, 172 pp.
Short, A.D., 1978. Wave power and beach-stages: a global model. Proc. 16th Int. Conf.
Coastal Eng., pp.1145-1162.
Short, A.D., 1979. Three dimensional beach-stage model. J. Geol., 87: 553-571.
Sonu, C.J., 1969. Collective movement of sediment in littoral environment. Proc. 11th
Int. Conf. Coastal Eng., pp.373-400.
Sonu, C.J., 1973. Three-dimensional beach changes. J. Geol., 81: 42-64.
Sunamura, T., 1980. Parameters for delimiting erosion and accretion of natural beaches.
Annual Report, Inst. Geoscience, Univ. Tsukuba, 6: 51-54.
Sunamura, T., 1982. Determination of breaker height and depth in the field. Annual
Report, Inst. Geoscience, Univ. Tsukuba, 8 : 53-54.
Sunamura, T. and Horikawa, K., 1974. Two-dimensional beach transformation due to
waves. Proc. 14th Int. Conf. Coastal Eng., pp.920-938.
Takeda, I., 1983. Beach changes by waves. Unpubl. Doct. Thesis, Inst. Geoscience, Univ.
Tsukuba, 86 pp.
Tanaka, N., Ozasa, H. and Ogasawara, A,, 1973. Note of the investigations on changes of
shorelines in Japan, Part 1. Tech. Note, Port and Harbour Research Institute, Ministry
of Transport, 163, 95 pp. (in Japanese).
U S . Army Coastal Engineering Research Center, 1973. Shore Protection Manual, Vol. 1.
U S . Govt. Printing Office, Washington, D.C.
Marine Geology, 6 0 ( 1 9 8 4 ) 7 9 - 9 8 79
SEDIMENT FLUX AND EQUILIBRIUM SLOPES IN A BARRED

NEARSHORE
BRIAN GREENWOOD and PETER R. MITTLER*

Departments of Geography and Geology, Scarborough Campus, University of T o r o n f o ,
Scarborough, Ont. M l C l A 4 (Canada)
Department of Geography, Scarborough Campus, University of Toronto, Scarborough,
Ont. M l C l A 4 (Canada)
(Received April 1 8 , 1 9 8 3 ; revised and accepted July 31, 1 9 8 3 )
ABSTRACT
nearshore. I n : B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedi-
mentation in Wave-Dominated Coastal Environments. Mar. Geol., 60: 79-98.
Estimates of time-integrated values of total ( I T V F ) and net (INVF) sediment volume

flux and the associated changes in bed elevation and local slope were determined for a
crescentic outer nearshore bar in Kouchibouguac Bay, New Brunswick, Canada, f o r eight
discrete storm events. A 100 X 1 5 0 m grid of depth-of-activity rods spaced a t 10 m inter-
vals was used to monitor sediment behaviour o n the seaward slope, bar crest and landward
slope during t h e storms, a t which time winds, incident waves and near-bed oscillatory
currents were measured. Comparisons between storm events and between these events
and a longer-term synthetic wave climatology were facilitated using hindcast wave para-
meters. Strong positive correlations between storm-wave conditions (significant height
and total cumulative energy) and total volume flux contrasted strongly with the zero
correlation between storm-wave conditions and n e t volume flux. I T V F values ranged u p
to 1646 m 3 for t h e experimental grid and were found t o have power function relations
with significant wave height (exponent 1 2 ) and cumulative wave energy (exponent = 0.4);
values of INVF ranged from 0 u p to 100 m3 for t h e same grid indicating a balance of sedi-
ment volume in t h e bar f o r m through time. Sediment reactivation increased linearly with
decreasing depth across t h e seaward slope and bar crest reaching maxima of 20 cm for the
two largest storms; bed elevation, and thus slope, changes were restricted t o t h e bar crest
and upper landward slope with near zero morphological change o n t h e seaward slope. The
latter represents a steady-state equilibrium with null n e t transport of sediment under
shoaling waves. Measurements of t h e asymmetry of orbital velocities close t o the bed
show that the energetics approach to predicting beach slope of Inman and Bagnold (1963)
is sound. Gradients predicted vary from 0.01 to 0.03 for a range of angles of internal
friction appropriate to t h e local sediment ( t a n cp = 0.3-0.6). These compare favorably
with the measured seaward slope of 0.015 formed under average maximum orbital veloc-
ities of 1.12 m 5-l (landward) and 1.09 m s-' (seaward) recorded during t h e period of the
largest storm waves.
*Present address: Dasco Data Products, 304-8495 Ontario Street, Vancouver, B.C. V5X
3E8, Canada.
0025-3227/84/$03.00 0 1 9 8 4 Elsevier Science Publishers B.V

80
INTRODUCTION
In many coastal environments dominated by waves the nearshore slope is

characterized by one or more bars (Greenwood, 1982), which remain as
stable bathymetric configurations throughout the annual cycle of wave
climate (for example see Greenwood and Davidson-Arnott, 1975) and there-
fore appear in equilibrium with nearshore wave and current processes. How-
ever, the magnitude and frequency of occurrence of the sedimentation
processes controlling bar growth and dynamics are still ill-defined and con-
siderable debate continues concerning the mode and forcing of the sediment
transport processes involved (Greenwood and Davidson-Arnott, 1979).
Clarification of these issues requires knowledge of both fluid, sediment and
morphological dynamics in the prototype over reasonably long periods to
include especially periods of intense sediment transport. Furthermore, since
the bar form consists of a very large number of slope facets each responding
to local stresses, there is a need for knowledge of the spatial variability of
process and form at this scale.
At the present time technological limitations prevent the necessary experi-
ments t o adequately answer the questions outlined above. However, recently
it has been possible t o obtain measures of both the sediment flux and bed
elevation changes integrated over a storm cycle for a series of closely spaced
locations within the nearshore zone (Greenwood and Hale, 1980; Greenwood
et al., 1980). In this paper the sediment flux and morphological response in
a nearshore bar is analysed for a series of storm-wave events of known
frequency of occurrence and the data used to test one of the basic approaches
to the prediction of local bed slope under wave activity.
EQUILIBRIUM NEARSHORE SLOPES
Currently no fully adequate theory of nears,hore equilibrium exists

although our general understanding of this equilibrium was well defined in
the highly deductive and qualitative statement of Johnson as early as 1919:
“At every point the slope is precisely of the steepness required to enable t h e amount
of wave energy there developed to dispose of t h e volume of sediment there in transit.”
Even earlier Cornaglia (1898) had proposed a null point hypothesis empha-
sizing the concept of the balance of forces (wave-generated currents, gravity,
inertia) controlling the stability of single particles on nearshore slopes. This
theory was quantified and tested experimentally by Ippen and Eagleson
(1955), Eagleson and Dean (1961) and Eagleson et al. (1963), but met with
only limited success. Qualitative models of the nearshore slope as a balance
between wave-induced transport of sand landward and a seaward return via
rip-current activity (Grant, 1943; Cook, 1970; Greenwood and Davidson-
Arnott, 1979) have been supported by observation, but lack a quantitative
form capable of being tested rigorously. Perhaps the most significant develop-
ment in understanding coastal equilibrium under wave motion was that due
81
to Inman and Bagnold (1963) based on Bagnold’s (1963, 1966) energetics-

based bedload transport model originally derived for unidirectional flow. In
this formulation a local slope of zero net transport develops at equilibrium as
a result of a balance of forces induced by asymmetry in the on-offshore
velocities, the angle of internal friction of the sediment and the tangential
component of gravity controlled by the slope itself. In essence the slope was
defined by the ratio of offshore t o onshore energy dissipation under oscilla-
tory flow and the angle of internal friction of the sediment:
where tan p = beach slope; tan Q, = coefficient of internal friction; and c is

defined:
offshore energy dissipation
C=
onshore energy dissipation
Thus, under shoaling waves, the increasing beach slope landward was a res-
ponse to the increased total dissipation due t o increased velocities at the bed
and the relative difference in dissipation in the landward and seaward direc-
tions.
In 1979 the senior author proposed a conceptual model for equilibrium in
nearshore bars whereby the seaward slope was maintained by the asymmetries
in transport associated with the shoaling waves (Greenwood and Davidson-
Arnott, 1979). If such is the case, then the Inman and Bagnold relationship
should provide a prediction of the nearshore slope in this zone, if the energy
dissipation ratio can be measured or predicted. Inman and Frautschy (1966),
following Bagnold (1963), proposed that the energy dissipation ratio, c , was
proportional to the third power of the ratio of the relevant onshore and off-
shore orbital velocities:
c= [Urn
U,
- offshore
- onshore I (3)
where U , is the maximum orbital velocity. This is in fact a measure of asym-

metry in the orbital velocity field. Support for the basic mechanisms sug-
gested by Bagnold (1963) and Inman and Bagnold (1963) can be seen in
more recent sediment transport models which have adopted this approach.
In 1980 Bowen proposed a quantitative model for predicting equilibrium
beach slopes, which included a consideration of both suspended and bed
load under oscillatory flows. An attempt to incorporate the effects of long-
shore currents is seen in the transport model of Bailard (1981, 1983) and
Bailard and Inman (1981).
It is not possible in this study t o provide a test of these later models, but
rather a simple evaluation of the basic energetics approach is made by testing
the predictor equation for slopes proposed in the original formulation.
82
LOCATION
The study was carried out in Kouchibouguac Bay, which is located a t the
western end of Northumberland Strait on the New Brunswick coast of Canada
(Fig.1). Extensive detail on the form, structures, textures and bar dynamics,
together with the general environmental constraints on the bar systems, have
been documented previously (Greenwood and Davidson-Arnott, 1975,1979;
Davidson-Arnott and Greenwood, 1976; Greenwood and Hale, 1980) and
only a brief synopsis need be presented here.
The area is a low-to-medium energy, micro-tidal (1.25 m maximum spring
tide), storm-wave dominated environment, where sediment flux occurs as a
highly discrete process during the passage of meteorological depressions
(Greenwood and Hale, 1980). The nearshore bathymetry consists of a two-
bar system with local slopes typified by those illustrated in Fig.2. The inner
system is planimetrically variable with straight, oblique and crescentic forms
cut through by well defined rip channels in places. In contrast the outer bar
is generally continuous and crescentic in character (average wavelength =
500 m; average amplitude = 35 m) and ranges in height from 1.5 t o 2.5 m.
The outer bar, of particular relevance to this study is built in well-sorted,
medium-to-fine sands (mean diameter 0.56-0.14 mm). Furthermore, it is a

83
6.
8.
0
.
100
I
200
.
300
.
400 500
-a
600
Distance Offshore
(metres)
Fig.2. Representative profile across t h e inner and outer bar systems of Kouchibouguac Bay.
highly stable feature in an environment of relatively high annual alongshore

transport. Table I documents the sediment volumes and transports. It is clear
that although large volumes of sediment exist in the bar system, collective
transport alongshore due t o bar migration contributes relatively little to the
annual littoral drift.
A simple conceptual model of the equilibrium nature of these bars is based
on a landward flux of sediments under asymmetric orbital motions due to
waves, balanced by a seaward return of sediment in rip currents as part of a
cellular nearshore circulation (Greenwood and Davidson-Amott, 1979).
Implicit to this hypothesis was that the seaward slope was essentially an
equilibrium transport surface maintained by orbital asymmetries under the
shoaling waves. Important questions therefore arise as t o the mobility of the
bar sediments (rates, spatial variability, relationship to wave energy, etc.),
the morphological response t o this mobility and the type of morphological
equilibrium established (static, steady, dynamic, etc.). These will be addressed
in this paper.
TABLE I
Sediment volumes and gross sediment flux Kouchibouguac Bay
Source Amount
Bar form: Total bar’ -2x 1 0 5 ~ 3

: Per m length - 4 1~0 2 m 3
: Total bay2 - 6 X 106m3
Net annual littoral drift3 -1 x 10’m3 yr-l
Collective transport by
bar migration4
- 1 x 10’ m 3 yr-l
IBased o n single crescentic unit, 500 m long, 1.5-3 m high; 215 km of bar; 3based o n
averaging dredging data, growth of inlet shoals and prediction based o n CERC method
1973; 4migration rate based o n repeat surveys.
a4
EXPERIMENTAL DESIGN
A t the present state of technology no satisfactory instrument exists for

monitoring simultaneously both suspended and bedload, and the morph-
ological changes associated with a net sediment flux, across any large area in
the neaushore. In this study a simple device is used which gives a measure of
sediment transport (total and net) as well as local changes in elevation of the
transport surface. This depth-of-activity rod (Greenwood and Hale, 1980;
Greenwood et al., 1980) is a simple adaptation of the vigil network erosion
pin and can be deployed in numbers sufficient to cover large areas. It consists
of a round steel rod (0.5 cm diameter, 1-2 m in length) driven vertically
into the sand by a Scuba diver until 0.45 m is left exposed; a loose fitting
washer (0.6 cm internal diameter) is placed over the rod and allowed to fall
to the bed. The rod is tagged with a fluorescent streamer tape and stamped
with an identifier to assist re-location. Figure 3 illustrates the pre- and post-
storm measurements made by a Scuba diver. These measurements are used to
make estimates of the sediment flux taking place through discrete periods of
time, i.e. the duration of the storm event. Assuming that the changing washer
location indicates the maximum depth of reactivation (see Greenwood and
Hale, 1980, for tests), then the minimum total volume of sediment in trans-
port is also given. Since the rod measurements are time-integrated estimates
of this transport rate (i.e., a volume) through the local control volume, this
has been termed the integrated total volume flux (ITVF). In a similar manner
the bed elevation changes measured at each rod represent the net volume
transports integrated over the storm duration (again a volume measure): this
has been termed the integrated net volume flux (INVF). The depth of activity
and bed elevation measures record the response of the local slope to wave
activity at the storm peak and the total sediment transport event.
INVF = iSt, - S t 2 ] .L2

I TVF = [Wtz- W t l 1 e L2
or = [WtZ-StZ 1 .L’ if ~t
S = surface elevation relative to top of rod
W = washer elevation relative to top of rod
tz,tZ= time intervals
Fig.3. Measurements made with the depth-of-activity rod and computations for sediment
flux: I T V F = time-integrated total volume flux, ZNVF = time-integrated net volume flux.
Both I T V F and I N V F are volume measures.
85
To measure sediment flux and bed elevation change across the outer bar
system, a reference grid (100 X 150 m) of depth-of-activity rods was estab-
lished. Figure 4 illustrates the configuration of the grid, the primary rod
coordinates and the relationship of the grid t o a shore-normal profile. Rods
were spaced a t 10 m intervals and a subset was used to monitor the seaward
slope. Control volumes for each grid were computed simply by extrapolating
a line from the break-of-slope a t the foot of the seaward slope t o the trough
and drawing vertical lines from the grid margins (Fig.5); a mean profile was
assumed for the total grid t o simplify calculations of these volumes poten-
tially available for entrainment.
Fluid motions near the bed were recorded with an electromagnetic flow-
meter (Marsh-McBirney Model 551) located in the centre of the seaward
slope grid (Fig.4). The flowmeter sensor was mounted initially 0.2 m above
the bed; however, depth of activity and bed elevation changes indicated
possible variations in height of the sensor during storms between 0.15 and
0.38 m.
A continuous resistance wave gauge was deployed 700 m offshore in 7 m
of water t o record incident wave conditions, which provided a yardstick for
storm magnitude. Winds, measured a t the beach face with a Type 4 5 B
Anemometer, were used to hindcast waves when direct measurements were
unavailable. Hindcasting of waves using winds from the meteorological
station at Chatham (Fig.1) allowed individual storms t o be placed in the
DEPTH OF DISTURBANCE GRID

Distance offshore (m)
3 1
li
I
/ \,\ l O , l O
:.
-5.10/' 0.10
. a f = m m !.
.. .. ..
-5.0 bar crest 0.0 10.0
Fig.4. Deployment of experimental grids of depth of activity rods, Kouchibouguac Bay,

1977 and 1978. The 1977 grid was restricted to the seaward slope, row 0 to row 10 and
this subset was used t o monitor the seaward slope in both 1977 and 1978. Note the loca-
tion of the flowmeter.
m
0,
1
TOTAL CONTROL VOLUME 27.750 m3
2
-E
1
I
z 3
a
n
5
4w 4
m
I
I-
$ 5
n
7
100 200 300 400 500 600 700 800
DISTANCE OFFSHORE (mi
Fig. 5. Control volume determinations for the experimental grids.

87
context of a synthetic wave climatology already established for the Bay (Hale
and Greenwood, 1980). In this way the return periods and thus relative
significance of the storms could be assessed.
Wave and current sensors were typically sampled a t 2 Hz for periods of
5-7 min every hour during storms, with occasional record lengths up t o
30 min. Wave spectra were computed using the University of California Bio-
medical Computer Package BMDO2T (Dixon, 1971) t o determine character-
istic wave height (Hmo = 4 X standard deviation of the wave record) and
peak period (TPk= period corresponding to the frequency of the maximum
energy density). Flowmeter records were resolved to produce true flow vectors
(speed and orientation) and scatterplots used t o describe the oscillatory
motion at a particular period of time.
SEDIMENT FLUX AND EQUILIBRIUM SLOPES
Sediment flux
Table I1 summarizes data on the primary wave parameters and sediment

flux for eight discrete storm events during 1977 and 1978, and reveals quite
clearly the nature of the sediment balance. Although the percentage of the
control volume mobilized by a single storm is n o t large it is significant, rang-
ing from 1 t o 12% with grid average maxima of 0.3 m3 per square metre of
seaward slope in absolute terms. However, the time-integrated net volume
flux (INVF) is extremely small, being less than one percent of the control
volume in all cases. Small net additions of sediment and small net losses of
sediment occur in approximately equal numbers.
To evaluate the response of the system t o events of differing magnitude,
which could then be placed within a correct time perspective, the time inte-
grated total volume flux (ITVF) values were plotted against both measured
TABLE I1
Sediment flux during storm events over the outer bar (seaward slope), Kouchibouguac Bay
Date Duration Hmo Tpk ITVF Percent* mobilized INVF Percent

(h) (m) (s) (m3) (%) (m’) change (5%)
1977 :07 :06 8 1.30 6.4 661 5 110 -0.8

1977 :07 123 11 1.17 6.0 319 2 36 -0.2
1977 : 0 8 :24 15 1.00 5.5 208 1 20 + 0.1
1977:08:30 17 1.06 5.3 182 1 61 -0.4
1978:05:23 5 N o data 297 2 1 0
1978:06:01 10 N o data 880 6 98 -0.7
1978:07:01 45 1.94 8.2 1734 12 29 -0.2
1978:07:24 17 1.61 7 1657 11 103 f 0.7
Average 5 -0.2
*Estimated control volume = 14,500 m 3 .

88
and hindcast wave data for each storm. Figure 6a, b and c illustrates double
logarithmic plots of ITVF against both the maximum significant wave height
generated during the storms and the cumulative energy density (hind-
cast) for the duration of the storms. Both wave height and cumulative energy
density are used here t o express two characteristics of the storm: the former
is a measure of the maximum energy input per unit time, while the latter is
indicative of the total work done during a storm.
Strong positive correlations are evident in all cases with the percentage of
explained variance ranging from a low of 45% in the case of hindcast
lo3 t
1021
lo4
' "'"'.~
105
I .......106 I * * . . a nJ
10'
E (Jm-' )
m 50 ACCRETION
E
u> o
EROSION
100 0 .
150 I I I I I I I I I I
0 200 400 600 800 1000 1200 1400 1600 1800
ITVF ( m 3 )
Fig. 6 . Relationships between time-integrated total volume flux (ITVF), and (a) measured
maximum significant wave height (Hmo);( b ) hindcast maximum significant wave height
(c) hindcast cumulative energy density ( E ) ; (d) time-integrated net volume flux
(INVF). Values for discrete storm events are plotted.
89
cumulative energy density t o a high of 88% in the case of the measured

significant wave height. Correlations with wave height are significant at the
0.05 level, while correlations with cumulative energy density are significant
at the 0.10 level. The least squares regression lines suggest that a power
function relationship exists between sediment response and wave energy.
While it would be inappropriate to attach too great a significance to the
absolute values of the exponents at this time it should be noted that sedi-
ment transport has always been considered as a power function of the
forcing parameter. Indeed transport rates in the alongshore direction have
been empirically established as being proportional to the square of the wave
height (Galvin and Vitale, 1976). Certainly it is clear that the estimates of
sediment reactivation are reasonably well recorded by the depth of activity
rods and that maximum significant wave height is the better of the two wave
parameters for predicting total sediment flux.
Of greater importance in the present context are the near-zero values for
the INVF (Table 11), even though the ITVF values range over an order of
magnitude (Fig.6d). It would appear, therefore, that, at least in terms of
sediment volume, the seaward slope exists in a state of steady equilibrium
with the prevailing wave climate. If it can be established that the bar form
remains stable both throughout storms and from storm to storm, then a
steady morphodynamic equilibrium would be established.
Spatial Variability of sediment flux and bar morphodynamics
Figure 7 illustrates the depths of activity and bed elevation changes for
the two largest storms monitored (1978:Ol and 1978:24). Table I11 docu-
ments the general storm characteristics. These two events had recurrence
intervals close to 1.5 yr based upon hindcast wave parameters and might be
expected to indicate bar response, to storms equal to that of the most prob-
able annual maximum. Although differing in intensity t o some degree it is
evident that the sedimentary response was similar. In both cases maximum
reactivation occurred on the bar crest with similar absolute values of 28 and
32 cm; the average values for reactivation were also comparable (14 and
13 cm). A general tendency for decreasing depth of activity with increasing
water depth is also evident as might have been expected, but the seaward
slope was subjected to considerable sediment motion even at its outer margin.
It should be noted, however, that the data are somewhat “noisy” due to
sampling variability, experimental error and variations in the bedforms
generated. Bed elevation changes are somewhat less “noisy” overall and illus-
trate again a comparable response of the bar slope to the two storms. Large
areas of the grid in both cases exhibit less than ?1cm change, which is close
to the limit of measurement. Changes that do occur are more prevalent on
the landward side of the bar crest and the landward slope; the seaward slope
in contrast remained essentially stable in both cases even though large volumes
of sediment were in motion. In order to generalize the shore-normal vari-
ability in the two measured parameters, grid row averages were calculated:
90
a LANDWARD -5.10 - 5 0 LANDWARD 5.10
0 lo SEAWARD 1010 10.0 SEAWARD 10.10

C
-8.0 LANDWARD -5.10 - S O LANDWARD -5.10
10 0 SEAWARD 10 10 100 . SEAWARD 10 l o
Fig.7. Spatial distribution of depth of activity (a and b) and bed elevation change ( c and d)
over t h e o u t e r bar during t h e storms of 1978:07:01 (a and c ) and 1978:07:24 (b and d).
Contours are a t 4 c m intervals (a and b) and 2 cm intervals (c and d). The grid coordinates
are check marked along t h e edges.
91
TABLE 111
Peak storm characteristics Kouchibouguac Bay, 1978:07:01 and 1978:07:24
Storm characteristics Storm event

1978:07:01 1978:01:24
Measured Predicted Measured Predicted
Maximum wind speed 58 kph 32 kph 48 kph 26 kph

Wind direction E NE NE N
Wind duration 45 h 37 h 17 h 23 h
Characteristic wave height (Hmo) 1.94 m 0.72 m 1.64 m 0.62 m
Characteristic wave period ( Tpk) 8s 4s 7s 3.2 s
Maximum orbital velocity* 1.26 m s-’ - 1.20 m s-’ -
0.42 m s-l - 0.39 m s-l -
Average orbital velocity*
Recurrence interval** : H,13 1.45 yr (0.14 yr) 1.3 yr (0.10 yr)
:ZE 1.1yr (0.40 yr) 1 Y* (0.20 yr)
*Determined from resolved flowmeter vectors; sample period 5 min.

**Figures determined from both the annual maximum series and, in parentheses, the
partial duration series of storm wave predictions.
this gave an overall view of the sedimentary responses along a profile by

reducing much of the “noise” in the data. Figure 8 illustrates these data and
confirms that:
(a) As waves shoal and orbital velocities near the bed increase so there is
an increase in the rates of sediment motion. An almost linear increase in the
depth of activity is suggested here which follows the almost linear seaward
slope. Although there is a slightly greater variation associated with the larger
of the two storms there appears to be no difference in the average response
of sediment transport. This is reasonable in that although measured wave
heights were greater in the ,July 1 storm, the near-bed currents measured on
the seaward slope were almost the same (Table 111).
(b) With the reduction in orbital velocities as a result of height loss through
breaking on the bar crest and with the increase in water depth, the lower
landward slope and trough experienced reduced rates of sediment motion.
Again a linear decrease in the depth of activity is suggested down the land-
ward slope.
(c) In general bed elevation and thus slope changes on the seaward slope
are extremely minor even under intense sediment reactivation by the shoal-
ing waves. Most of the variability in the averaged values is less than +1 cm.
(d) The bar crest and landward slope experience the greatest redistribution
of sediment and morphological change. This is to be expected since these
areas are subject to wave breaking and wave reformation with high orbital
velocities and strong flow asymmetries. Intense turbulence and interactions
between waves and secondary flows, such as seaward flowing rip currents,
are common. In the July 1 storm it appears that on average the bar was
shifted seaward with erosion of the landward slope and accretion on the bar
92
DISTANCE OFFSHORE (m)

300 400 500 600
I I
Inner Shelf
- 4 1
- +* 2-
E 4.--
;; u 8 - A s-7- ? .
g ;&=Y
----
x.>p*p x
5 2 - O X x X
x 516-
4
20 -
- .'x -
Landward Slope ;Bar Crest; Seaward Slope i Inner Shelf
Fig.8. Average depth of activity and bed elevation change f o r t h e storms of 1978:07:01
(crosses) and 1978:07:24 (dots) across t h e outer bar. The profile is t h a t taken along t h e
centre line of the grid o n 1978:07:14. T h e numbered arrows indicate locations of box
cores.
crest: in the following storm of July 24, with only slightly lower wave heights,
accretion on the landward slope and thus landward migration of the bar was
the average condition. It should be stressed that these are averages, since
along a 100 m length of bar crest re-orientation of the crest location could
produce landward migration in one section and seaward migration in another.
This has been observed frequently on these bars (Greenwood and Davidson-
Arnott, 1975) with the crest oscillating about some mean position in associa-
tion with the slow alongshore migration of the crescentic form and the
transient positions of seaward flowing rip currents.
Two important aspects of the sediment flux patterns during these storms
cannot, however, be addressed directly with the depth of activity rod data.
The paths of net transport associated with bed elevation changes are not
expressed, for example, nor is it possible to say whether the morphologically
stable seaward slope is maintained by high but uniform rates of transport or
whether a true oscillating equilibrium exists in the sediment transport.
Examination of sedimentary structures preserved in the reactivated layer can
provide some indication of transport paths (Greenwood and Davidson-Arnott,
1975; Davidson-Arnott and Greenwood, 1976; Greenwood and Mittler, 1979;
Greenwood and Hale, 1980) as in the case of the 1978:07:24 storm. Epoxy
peels of box cores from the bar crest and upper landward slope along the
central profile (locations 1and 2 in Fig.8) of the experimental grid all exhibit
strong landward dipping cross-stratification indicative of landward migrating
lunate megaripples (Fig.9a and b). This would suggest that the accretion of
the landward slope indicated by the depth of activity rods results from
93
sediment transferred landward from the crest where the rods suggest some
slight degradation (Fig.8). Cores from the bar crest for a distance of 90 and
120 m either side of the grid all reveal similar structures supporting both a
general transfer of sediment and a displacement of the bar in a landward
direction (Mittler, 1981). Structural indices from the reactivation layer of
the morphologically stable seaward slope (locations 3 and 4 in Fig.8) reveal
a preponderance of planar stratification (Fig.9c and d). Thus the high sedi-
ment transport indicated by the depth of activity and the near-zero net sedi-
ment flux indicated by the bed elevation change were associated primarily
with sediment motion under flat bed conditions: no directional information
is provided by these structures but continuity considerations would suggest
that during the phase of high transport rates the sediment was likely in a
state of near oscillating equilibrium rather than one of high uniform net
transport.
The most striking aspect of these data is, however, the morphological
stability of the seaward slope with high rates of sediment motion. This
clearly supports the thesis that this unit is a surface of null net transport and
thus in a state of steady equilibrium with the orbital velocity field associated
with the shoaling waves.
Sea ward slope equilibrium
From the preceding analysis it would seem that a balance of forces involv-
ing wave energy dissipation and the morphological gradient in the manner
suggested by Inman and Bagnold (1963) would be appropriate to explain the
steady state of the seaward slope of the bars in Kouchibouguac Bay. To test
this theory, measured orbital velocities on the seaward slope were analyzed.
In order to satisfy constraints imposed by the model, the flow field must be:
(1)solely the result of the onshore-offshore motion of waves approaching
along the shore-normal (line of maximum topographic gradient); (2) be
representative of the lower'boundary layer; and (3) great enough not only to
produce motion but to ensure transport under flat bed conditions since, as
was pointed out by Inman and Bowen (1962), transport over rippled beds
under oscillatory flow produces a complex phase-dependent process involving
suspension associated with vortex generation. Indeed, as the basic model
requires bedload transport, a flat bed sheet flow model would seem most
appropriate.
Current records were obtained from a number of the storms (Mittler,
1981), but in only one case did the velocity field satisfy the above mentioned
criteria. Figure 1 0 illustrates this velocity field, recorded on July 1 at the
peak of the largest storm monitored. The ellipses represent the limits of the
distribution of current vectors determined from the electromagnetic flow-
meter records. The.principal axes of the ellipses, determined by least squares,
illustrate the shore-normal approach of the waves at this time, the strength
of the wave oscillations, and, most importantly, the landward asymmetry of
the maximum orbital velocities. During other storms, either the wave
95
a b
Offshore Offshore
4
4
--
\
H -H 1 : :
;t /
Onshore
...................... Principal AXIS of Oscdlatory Flow
Fig.lO. Elliptical distribution of near bed flow vectors on the seaward slope of the outer
bar, Kouchibouguac Bay, during the peak of the storm, 1978:07:01; (a) 1520 h, (b)
1627 h.
approach (and thus the primary orbital velocity vector) was not normal to
shore, or the measured maximum offshore orbital velocity was greater than
its onshore equivalent. Even with the very strong shore-normal currents of
the July 1 storm, a slight displacement of the centroid of the ellipse suggests
a superimposed shore-parallel flow (Fig.10). Detailed analysis of similar off-
sets in a large number of discrete flowmeter records shows such flows to be
of tidal origin ('Mittler, 1981). For sand of the size common on the seaward
slope (mean diameter -0.18 mm), even the mean orbital velocities (Table 111)
at this time would theoretically produce flat bed. Using the criterion of
Dingler (1974) as expressed by Clifton (1976) for the transition velocity
between rippled and flat bed this grain size would require a velocity of
-0.26 m s-'. This is far exceeded by the mean velocities measured and would
indicate flat bed sediment transport at this time. Corroboration for this is
provided by post-storm box cores from the seaward slope (locations 4 and 5
Fig.9. Epoxy peels of box cores and the associated depths of activity (in parentheses) after
the storms of 1978:07:24 (a, b, c, d) and 1978:07:01 (e, f). The horizontal bar is 0.10 m;
the arrow indicates the direction of the shoreline. (a) bar crest (0.15 m); ( b ) upper land-
ward slope (0.15 m ) ; (c) upper seaward slope (0.13 m); (d) mid-seaward slope (0.14 m);
(e) mid-seaward slope (0.13 m); ( f ) lower seaward slope (0.13 m).
96
in Fig.8) which illustrate dominance of planar stratification in the most

recent active layer (Fig.Se, f).
Table IV documents the maximum onshore and offshore velocities, the
ratio of offshore to onshore energy dissipation as defined by Inman and
Frautschy (1966) and the predicted equilibrium beach slopes for a range of
values of the angle of internal friction appropriate to the sand materials
present. Also given is the average slope along the profile measured after the
storm from aposition 10 m landward t o a position 10 m seaward of the flow-
meter location. From the earlier analysis of the depth of activity it is evident
that differences in bed elevation would not have exceeded 2 cm even at the
storm peak over this section of slope and this could have no effect on the
slope value used in this comparison. Predicted slope values are given for both
the maximum instantaneous velocity recorded, and the average maximum
velocity based on three discrete sample records spaced 1 h apart during the
storm peak. The correspondence between measured and predicted values is
good, particularly, if the average maximum velocity for the storm peak is
used. The latter is a more reasonable figure to use for sediment transport in
any case, since transport is a time integrated phenomenon; further, the
instantaneous maxima in a record could be the result of single waves.
CONCLUSIONS
Time-integrated values of total sediment flux over a large outer crescentic

bar in Kouchibouguac Bay, Canada, were found to be positively correlated
with measured and predicted wave energy parameters: power function rela-
tionships with exponents of -2 t o -4 for significant wave height and -0.4 for
cumulative energy values provide statistically significant explanation. Time-in-
tegrated values of net sediment flux, in contrast, were close to zero and bear
no relationship t o the total reactivation of sediment. Local bed elevation
changes were only significant over the bar crest and landward slope, with the
seaward slope existing in a state of steady equilibrium with the wave processes.
Continuity considerations suggest a state of oscillating equilibrium for the large
TABLE IV
Measured maximum orbital velocities, dissipation coefficients, predicted and measured

nearshore slopes, Kouchibouguac Bay, 1978:07:01
Maximum orbital velocity Dissipation coefficient Nearshore slope

(U,) (c) tan p
Onshore Offshore Predicted Measured
(cm s-') tan d 4 = 0.30 0.45 0.60
128' 123' 0.89 0.018 0.026 0.035

0.0153
1122 1092 0.92 0.012 0.019 0.025
'Maximum instantaneous velocity; 2average maximum during storm peak; 3average slope
from post-storm survey; 4tan 0 = angle of internal friction.
97
volumes of sediment set in motion on this slope by storm waves. Morpholog-

ical changes on the bar crest and landward slope result from the number and
variability of processes occurring here: asymmetric oscillatory currents assoc-
iated with propagating surface gravity waves are present, but also turbulence
associated with wave breaking and interactions between these and secondary
longshore and rip currents.
Results of this study support both the basic energetics concept inherent in
the Inman and Bagnold (1963) model for equilibrium nearshore slopes and
the predictor equation based on the asymmetry in the maximum orbital
velocities (Inman and Frautschy, 1966). Predicted values for the seaward
slope of 0.012-0.025 compare favourably with the measured value of 0.015
using measured average maximum orbital velocities of 1.12 m s-l (landward)
and 1.09 m s-l (seaward), and a range for the appropriate angle of internal
friction of 0.3-0.6.
REFERENCES
Bagnold, R.A., 1963. Mechanics of marine sedimentation. In: M.N. Hill (Editor), The Sea,
Vol. 3. Wiley-Interscience, New York, N.Y.
Bagnold, R.A., 1966. An approach to the sediment transport problem from general
physics. Prof. Pap. 422-1, US. Geol. Surv., Washington, D.C., 37 pp.
Bailard, J.A., 1981. An energetics total load sediment transport model for a plane sloping
beach. J. Geophys. Res., 86: 10,938-10,954.
Bailard, J.A., 1983. Modeling on-offshore transport in the surfzone. Proc. 18th Coastal
Engineering Conf., Cape Town, pp.1419--1438.
Baiiard, J.A. and Inman, D.L., 1981. An energetics bedload model for a plane sloping
Bowen, A.J., 1980. Simple models of nearshore sedimentation; beach profiles and iong-
80-10, pp.1--11,
Clifton, H.E., 1976. Wave-formed sedimentary structures: a conceptual model. In: R.A.
Davis, Jr. and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SOC.
Econ. Paleontol. Mineral., Spec. Publ., 24: 126-148.
Cook, D.O., 1970. Models for nearshore sand transport. Proc. 13th Conf. on Great Lakes
Research, Buffalo, N.Y., pp.210-216.
Cornaglia, P., 1898. On beaches. Accadamia Nazionale dei Lincei Atti, Classe di Scienze
Fisiche, Matematiche e Naturali, Mem. 5, Ser. 4, pp.284-304. Reproduced in trans-
lation in: J.S. Fisher and R. Dolan (Editors), Beach Processes and Coastal Hydro-
dynamics. Benchmark Pap. in Geol., Stroudsberg, Pa., 39: 11-26.
Davidson-Arnott, R.G.D. and Greenwood, B., 1976. Facies relationships on a barred coast,
Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davis, Jr. and R.L. Ethington
(Editors), Beach and Nearshore Sedimentation. SOC.Econ. Paleontol. Mineral., Spec.
Publ., 24: 149-168.
Dingler, J.R., 1974. Wave-formed ripples in nearshore sands. Unpubl. Ph.D. Diss., Univer-
sity of California, San Diego, Calif., 136 pp.
Dixon, W.J., 1971. BMD: Biomedical Computer Programs. University of California Press,
Berkeley, Calif., 600 pp.
Eagleson, P.S. and Dean, R.G., 1961. Wave-induced motion of bottom sediment particles.
Trans. Am. SOC.Civ. Eng., 126(1): 1161-1189.
Eagleson, P.S., Glenne, B. and Dracup, J.A., 1963. Equilibrium characteristics of sand
beaches. J. Hydraul. Div. Am. SOC.Civ. Eng., 89(HY1): 37-57.
98
Galvin, C.J. and Vitale, P., 1976. Longshore transport prediction - SPM 1973 Equation.
Proc. 15th Coastal Engineering Conf., Honolulu, Hawaii, pp.1113--1148.
Grant, US., 1943. Waves as a sand transporting agent. Am. J. Sci., 241: 117-123.
Greenwood, B., 1982. Bars. In: M.L. Schwartz (Editor), Encyclopaedia of Beaches and
Coastal Environments. Dowden, Hutchinson and Ross, Stroudsberg, Pa., pp.135-139.
tary processes, Kouchibouguac Bay, New Brunswick, Canada. In: J. Hails and A. Carr
(Editors), Nearshore Sediment Dynamics and Sedimentation. Wiley, New York, N.Y.,
pp.123--150.
wave-formed bars: a review and case study. Can. J. Earth Sci., 16: 312-332.
Canada: Littoral Processes and Shore Morphology. Geol. Sum. Can., Pap. 80-10,
pp.89--109.
Greenwood, B. and Mittler, P.R., 1979. Structural indices of sediment transport in a
straight wave-formed nearshore bar. Mar. Geol., 32: 191-203.
nearshore zone: prototype measurements. In: Workshop on Instrumentation for Cur-
rents and Sediments in the Nearshore Zone, National Research Council of Canada,
Associate Committee for Research on Shoreline Erosion and Sedimentation, Ottawa,
Ont., pp. 99-120.
Hale, P.B. and Greenwood, B., 1980. Storm wave climatology: a study of the magnitude
and frequency of geomorphic process. In: S.B. McCann (Editor), The Coastline of
Canada: Littoral Processes and Shore Morphology. Geol. Surv. Can., Pap. 80-10,
pp.70-88.
Inman, D.L. and Bagnold, R.A., 1963. Littoral processes. In: M.N. Hill (Editor), The Sea,
Inman, D.L. and Bowen, A.J., 1962. Flume experiments on sand transport by waves and
currents. Proc. 8th Coastal Engineering Conf., Mexico City, pp.137-150.
lines. Proc. Coastal Engineering Speciality Conf., Santa Barbara, Calif., pp.511-536.
Ippen, A.T. and Eagleson, P.S., 1955. A study of sediment sorting by waves shoaling on a
plane beach. Beach Erosion Board, U.S. Army Corps of Engineers, Tech. Mem., 63,
8 3 PP.
Johnson, D.W., 1919. Shore processes and shoreline development. Columbia University,
Facsimile Ed., Hafner, New York, N.Y., 1972, 584 pp.
Mittler, P.R., 1981. Storm related sediment flux and equilibrium in a barred nearshore,
Kouchibouguac Bay, New Brunswick, Canada. Unpubl. Ph.D. Diss., University of
Toronto, Scarborough, Ont., 419 pp.
Marine Geology, 6 0 ( 1 9 8 4 ) 99-122 99
SEDIMENT TRANSPORT AND MORPHOLOGY AT THE SURF ZONE OF

PRESQUE ISLE, LAI(E ERIE, PENNSYLVANIA
DAG NUMMEDAL’, DAVID L. SONNENFELD’* and KENT TAYLOR*

‘Department of Geology, Louisiana State University, Baton Rouge, L A 70803 (U .S.A.)
’Department of Geology, Mercihurst College, Erie, P A 16546 (U.S.A.)
(Received June 1, 1983; revised and accepted September 1, 1 9 8 3 )
ABSTRACT
Nummedal, D., Sonnenfeld, D.L. and Taylor, K., 1984. Sediment transport and morph-
ology a t the surf zone of Presque Isle, Lake Erie, Pennsylvania. In: B. Greenwood and
R.A. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated
Coastal Environments. Mar. Geol., 60: 9 9 - 1 2 2 .
A four-year investigation of surf zone sedimentation a t Presque Isle, Pennsylvania,

was undertaken in preparation for the design of a segmented breakwater system. Sedi-
ment transport calculations were based o n hind-cast annual wave power statistics and
“calibrated” by known accretion rates a t the downdrift spit terminus. 30,000 m3 of
sediment reaches the peninsula annually from updrift beaches. The transport volume
increases downdrift due t o shoreface erosion and retreat of the peninsular neck. A t the
most exposed point o n Presque Isle ( t h e lighthouse) t h e annual transport is 209,000 m’.
East of the Iighthouse is a zone of net shoreface accretion as the Iongshore transport rate
progressively decreases.
The downdrift variation in sediment supply, combined with increasing refraction and
attenuation of t h e dominant westerly storm waves produce a systematic change in prevail-
ing surf zone morphology. Storms produce a major longshore bar and trough along the
exposed peninsular neck. The wave energy during non-storm periods is too low to sig-
nificantly alter the bar which consequently becomes a permanent feature. The broad
shoreface and reduced wave energy level east of the lighthouse produce a morphology
characterized by large crescentic outer bars, transverse bars, and megacusps along the
beach. A t the sheltered and rapidly prograding eastern spit terminus the prevalent beach
morphology is that of a ridge and runnel system in front of a megacuspate shore.
The morphodynamic surf zone model developed for oceanic beaches in Australia is
used as a basis for interpretation of shoreface morphologic variability a t Presque Isle. I n
spite of interference by major shoreline stabilization structures, and differences between
oceanic and lake wave spectra, the nearshore bar field a t Presque Isle does closely cor-
respond t o t h e Australian model.
INTRODUCTION
Presque Isle is a 10.5 km long, compound recurved spit off Erie,

Pennsylvania. Originally a natural eastward-migrating “flying” spit, Presque
*Present address: Louisiana Geological Survey, Louisiana State University, Baton Rouge,
LA 70803, U.S.A.

100
Isle has over the last 150 years been subjected to increasingly intensified
stabilization efforts. Presque Isle’s dual role as a state park with the best
beaches along the U.S. shores of Lake Erie and as a shelter for the industrial
Erie harbor provides more than adequate economic incentives for the
increasingly expensive stabilization efforts. The investigations summarized
in the paper were undertaken in direct response t o the need for better in-
formation about the patterns of nearshore sediment movement prior to the
design and construction of a new large-scale lake shore protection system
planned for the 1980’s (U.S. Army, Corps of Engineers, 1979).
This phase of design-related studies consisted of field monitoring of
changes in the Presque Isle beach profiles and nearshore bathymetry (Fig.1).
The field surveys were supplemented by vertical aerial mapping photography.
Both ground surveys and aerial photography were done on a seasonal basis.
Field surveys lasted from the spring of 1978 through the fall of 1980. Annual
reports to the US. Army Corps of Engineers (Nummedal, 1979, 1980,
1981) present all relevant survey data as well as preliminary analyses. The
emphasis in this paper is on the patterns and mechanisms of sediment
transport as deduced from the surveys.
The two main topics of this paper are:
(1)The sediment budget f o r Presque Isle. The budget is based on the
premise that the relative transport rates along different segments of Presque
Isle can be calculated based on the longshore wave power distribution
during dominant southwesterly wave approach. Absolute values for longshore
transport are determined by “calibrating” the computed relative rates by
ERIE PA. 0 1 2
Fig. 1. Location map of Presque Isle, Pa. Numbered sites designate beach profile locations
reoccupied seasonally during the 3-year shoreline monitoring program (1978-1980).
Circled numbers designate sites where both beach and nearshore bathymetric profiles
were obtained. These profiles are designated with their “PI numbers” throughout the text.
101
using known rates for maintenance dredging a t the harbor entrance and his-
torical rates of accretion a t Gull Point.
(2) Bar morphology along the lake shore of Presque Isle. Observed sea-
sonal changes and longshore variability in the bar field are documented.
The morphologic variability is explained in terms of the derived sediment
budget and the inferred longshore changes in absolute wave power and
nearshore current regime.
LONGSHORE VARIABILITY IN SEDIMENT TRANSPORT RATES
Wave climate
N o long-term wave gauge has been operating in eastern Lake Erie. There-
fore, data on the lake shore wave climate are based on Summaries of Synoptic
Meteorological Observations (SSMO-data from the US. Naval Weather
Service Command, 1975); hindcast wave data from meteorological reports
(Saville, 1953; Resio and Vincent, 1976); and occasional periods of LEO
observations.
The most severe weather disturbances affecting the Great Lakes region
are extratropical cyclones. Analysis of a large number of synoptic charts
have demonstrated that common cyclone paths take the center of the low-
pressure system on an eastward course passing to the north of Lake Erie.
The strongest associated winds, therefore, commonly blow o u t of the west-
southwest (Nummedal et al., 1976). Because this corresponds t o the direc-
tion of maximum fetch with respect t o the shore of Presque Isle one would
expect t o find that the dominant waves and the maximum wave power reach
Presque Isle from the west. Saville’s (1953) wave hindcast (Fig.2) and
the derived wave power distribution diagram (Fig.3) demonstrate that
such is indeed the case. Observations summarized in the SSMO data files
are generally consistent with those hindcast by Saville (1953).
Resio and Vincent (1976) conducted a more recent hindcast study of
wave conditions on Lake Erie. Resio and Vincent (1976) applied a numerical
hindcasting technique developed by Pierson and Moskowitz (1964) which is
better suited t o the short and variable fetches encountered a t Lake Erie,
than is the Sverdrup-Munk-Bretschneider approach (Coastal Engineering
Research Center, 1973) as used by Saville. Comparison of the results in-
dicate that the SMB-technique tends t o overpredict the Great Lakes wave
heights by a factor of 10-20%. Although significant in design wave con-
siderations, this level of discrepancy does not appear to seriously impair the
applicability of Saville’s (1953) results in studies of nearshore sediment
transport.
Lake Erie has a seasonal wave climate. Hindcasting demonstrates that
winter is the stormy season (Fig.4). However, Lake Erie is often frozen
between late December and March. Thus, most of the wave energy effective
in sediment transport hits the Presque Isle beaches in the months of April,
May and November. The total average hindcast, deep water, wave power at
Erie, Pennsylvania, is 566 W m-‘.
102
WAVE DlAGRAMS
Erie, PA
,-
-- --
I I\
/NNW/-~NNE \\
Calm or less than

15 cm. waves: 83.42%
wsw \
Wave height in meters
Fig.2. Summary diagram of the annual wave height distribution in eastern Lake Erie
based o n data obtained from hindcasting performed by Saville (1953). The hindcast is
for deep water conditions for the whole year, including those months which some years
have a lake-wide ice cover.
Long periods of calm interrupted by short-lived high-energy events charac-

terize a storm wave environment such as Lake Erie. As will be demonstrated
below, bar dynamics and beach stability at Presque Isle correspond closely to
what has been documented on micro-tidal, storm-dominated oceanic shores.
Longshore distribution of wave power
The longshore sediment transport rate can be calculated for the lake
shore of Presque Isle based on the wave power distribution in Fig.3. The
assumptions used in the calculations are: (a) the waves responsible for the
bulk of the annual transport have a period of 4 s and they break in 1m of
water (see Nummedal et al., 1976, for field data supporting this assumption);
and (b) the bathymetry consists of simple shore-parallel contours. This
permits analytical determination of the refracted wave angle at breaking
from Snell's law (Komar, 1916).
103
WAVE ENERGY
Erie, PA
W Total 566
wsw 0
Buffalo, NY
I L
Wlm
Fig. 3. Directional distribution of mean annual wave power in eastern Lake Erie. The dis-
tribution has been calculated from hindcast data of Saville (1953). The listed total (e.g.
566 W m-’ at Erie) designates the mean annual wave power at that station irrespective
of wave approach direction. Deep-water data.
The deep-water wave power from a given direction is designated Po ( m ) .

Its longshore component can be shown to be P I (m),
where:
PI ( m )=Po ( m )sin a b cos a. (1)
This equation is derived in Walton (1973). Equation 1 properly accounts
for refraction and shoaling but assumes no friction or bottom percolation.
Angle a. is determined by the local shoreline orientation and ab is calculated
using Snell’s law.
The results of the computations are summarized in Table I. The peninsula
was divided into five straight shoreline segments (Fig.5). There are distinct
breaks in shoreline trend at groin 11, the big groin by the lifeguard station,
the lighthouse, the eastern end of Beach 10, and at Gull Point. These points
were used as segment boundaries.
Gross longshore wave power (sum of components to the right and to the
left along a shoreline) is a fairly good measure of the total sediment trans-
port potential along a shoreline segment. As seen in Table I, the gross long-
shore power is essentially invariant between the neck of the peninsula and
Beach 10. The gross longshore power is, as expected, much less in the Gull
104
- - W l
ICE FREE
ICE COVERED
80
70
60
50
I
40
30
20
10
Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec.
Fig.4. Annual variation in wave characteristics. Maximum storm activity occurs in March,
generally associated with the break-up of lake ice. May and November have about the
same storm frequency. Short-lived summer thunderstorms, which do generate high waves,
are not adequately represented in the hindcast statistics. From Nummedal et al. (1976).
TABLE I
Wave power calculations for individual segments of Presque Isle. See Fig.5 for location
of segments, and Nummedal(1983) for a more detailed account of the computations
Aza Po P,IC P,I F P , IIIC P,IVC P,V C
wsw 61.5 15.1 14.3 9.2 offshore offshore

W 172.0 29.7 37.3 42.0 23.4 offshore
WNW 69.0 0 4.9 12.7 16.6 1.5
NW 87.9 -15.1 -10.0 1.6 17.8 16.4
NNW 45.6 -10.8 -10.4 -7.1 2.1 10.9
N 22.5 -3.6 -4.6 -6.1 -2.9 3.3
NNE 55.1 0 -3.8 -9.8 -12.7 1.1
NE 52.4 offshore offshore -0.86 -10.5 -9.6
Net - 16.2 27.8 43.0 33.8 21.2

power
Gross 566 74.3 85.2 88.9 86.0 42.8
power
=Direction of incident waves; bdeep-water wave power in W m* (from Fig.3); ‘3ongshore

component of wave power within segments I, 11, 111, I V , V. Negative sign indicates move-
ment to the left (westward) along the beach.
105
Point area because the shoreline here faces away from the dominant incoming
westerly waves.
Net longshore wave power is uniformly directed toward the east along
the entire shore but the rates vary greatly. There is a continuous down-
drift increase in longshore power from the neck all the way to the lighthouse,
then a reduction from there eastward. This pattern is consistent with net
erosion along the entire peninsular neck and net deposition to the east of
the lighthouse for the shoreface as a whole. Individual beaches may erode or
accrete within both of these areas in response to local, in part man-made,
shoreline perturbations. The erosion along the neck will be most severe
where there is a sudden increase in the net longshore wave power. This
occurs where there are abrupt changes in shoreline trend, as for example, in
the area between groin 9 and Beach 6. This has, in recent years, been an area
of rapid beach erosion.
L A K E ERIE
ERIE H A R B O R
kilometers
Rates in 1 0 4 8 y r .
Fig.5. Calculated rates of sediment transport, erosion and accretion along the Lake Erie
shore of Presque Isle. The budget was derived from the distribution of wave power,
historical accretion rates at Gull Point and Erie harbor dredging records.
106
Sediment transport rates
The wave power values in Table I are used only to determine the relative
rates of longshore sediment transport along the Presque Isle beaches. The
assumption is made that the volume rate of sediment transport is linearly
related to wave power (e.g. eq. 4-40 in the Shcre Protection Manual, CERC,
1973). Sediment delivery rate to the updrift end of Presque Isle has been
estimated from bluff retreat rates measured by Carter (1976) by the U.S.
Army, Corps of Engineers (1979). Data on the downdrift accretion were de-
rived from historical records of shoreline and bathymetric change, and Erie
harbor dredging records, in two independent budget estimates (Nummedal,
1979, 1983; US. Army Corps of Engineers, App. C, 1979). The results of
these budget estimates, for average conditions over the 1955-1978 time
period (which covers periods of heavy artificial nourishment), are sum-
marized in Table IIA.
Knowing sediment input and downdrift accretion rates, as well as the
relative magnitude of the wave power along the five segments of Presque
Isle (Table I), one can write a set of simple linear equations to solve for the
actual transport rates between the individual shoreline segments. Along
shorelines where there is an increase in the longshore transport rate the
supplied material comes from erosion of the shoreface, where there is a
diminishing transport rate the shoreface accretes. The results of these com-
putations are summarized in Table IIB and Fig.5.
The computations document shoreface erosion along the neck of Presque
Isle; the rate of erosion being the highest in segment 111. Segments IV and V
are seen to accrete. This is consistent with the observed shallow water
depths and multiple bars in segment IV, and net spit progradation in
segment V.
TABLE I1
Input data and results of calculations of sediment transport along the lake shore of
Presque Isle. All rates in m3 yr-'
A . Input data
Net accretion at Gull Point (seg. V ) : 55,000
Delivery of littoral material from Presque Isle
beaches to Erie harbor entrance: 110,000
Supply of littoral material from beaches to the
west of Presque Isle: 30;OOO
B. Results of computations
Transport between segments I and 11: 74,000
Net loss from segment I : 44,000
Transport between segments I1 and 111: 135,000
Net loss from segment 11: 61,000
Transport between segments I11 and IV: 209,000
Net loss from segment 111: 74,000
Transport between segments IV and V : 165,000
Net gain in segment IV: 45,000
107
Beach nourishment and shoreline erosion east of the lighthouse
The beach east of the lighthouse and east of Beach 10 is characterized by

large-scale shoreline megacusps. The megacusps typically migrate yet main-
tain their identifiable form over periods ranging from months to years. The
area immediately east of the lighthouse was closely monitored during the
time period Oct. 1977 through Sept. 1980 (Fig.6). Because the road was
threatened by erosion, artificial nourishment was initiated in the spring of
1977. By fall of that year some of the nourished sand had moved eastward
and begun building a megacusp between profiles PI 11 and PI 1 2 (Fig. 6).
Continued nourishment at profile PI 10, combined with sediment bypassing
of the lighthouse groin from beaches nourished farther updrift, led to dra-
matic widening of the beach at PI 10 by 1980. The associated megacusp
migrated t o the east side of profile PI 1 2 (Fig.6C) and the intervening beach,
centered at PI 11,became subject to intense erosion. An ironic situation had
developed: after filling of about 124,000 tons of beach material east of the
lighthouse groin between 1977 and 1980 (Nummedal, 1981), the beach at
PI 11,600 m downdrift, was eroding faster than before the nourishment.
Sediment dispersing eastward from the lighthouse fill area migrates along
PI 10
PI 1 1
Pi 10
---_ . ,
. .. .. Sepi 12. 1980

0 100 200 300
meter5
Fig.6. Shoreline changes downdrift of the lighthouse groin between October 1977 and
September 1980. Incipient beach fill led to rapid megacusp development and beach accre-
tion between profiles PI 11 and PI 1 2 (panel A). Continued nourishment caused down-
drift migration of the megacusp and the beginning of an erosion phase at PI 11 (panel B).
By 1980 erosion a t PI 11 had become severe; the bulk of the megacusp had now migrated
to the east of PI 12, a total distance of 500 m downdrift from the site of its incipient
formation in 1978 (panel C). See Fig.1 for location.
108
an arcuate bar system. Some of this re-enters the beach at the bar-attachment
point on the PI 1 2 megacusp, some sediment continues farther eastward.
The sole source of sediment t o the megacusp is the bar; there is a totally in-
significant amount moving along the beach. As a consequence of this bar
bypassing the beach at PI 11 did erode. Eastward extension of the artificial
beach fill in 1981 alleviated the erosion problem at PI 11.
NEARSHORE BARS
A detailed echo-sounder survey of 87 bathymetric profiles conducted in

the summer of 1979 by the US. Army Corps of Engineers, Buffalo Dis-
trict (drawings no. 79S-PIP-2/4.7 sheets), combined with vertical aerial
photography obtained on July 16th, 1979, a day with excellent water
clarity, form the basis for a set of figures depicting the three-dimensional
geometry of the entire Presque Isle bar field (Fig.7A-F). This is probably
the most extensive continuous bar field ever mapped in such detail. The
maps form the basis for the following description of the component bars.
Bar conditions at Presque Isle are a function of location and wave condi-
tions preceding the observation period. The spit neck is characterized by a
dual bar system: a transient, highly variable inner bar which typically crests
in 0.5-1.5 m of water about 50 m offshore, and a permanent outer bar
with a crest in 3-4 m of water about 150-200 m offshore. East of the
lighthouse, the bar field is more complex with strongly developed shore-
normal (transverse) bar components. In segment V (Gull Point) the only
bars present are small and close to shore.
Bars along the peninsular neck
Inner bars. The inner bars in the groin field (segment I) are linear or
crescentic in plan form. The geometries seen in Fig.8 are most commonly
encountered. When present, the crescents vary in degree of concavity. The
most arcuate ones extend about 100 m offshore and have well developed
complete crescents. Other bars have a rather flat longshore form with broad
landward-pointing horns and poorly developed rhythmicity. Bar relief and
cross-sectional profile varies widely. At some profiles, the inner bar is fairly
symmetrical with a relief of more than a meter (Fig.7B, profile 14);at other
profiles, the bar is expressed as a gentle shoulder.
Selected bathymetric profiles were measured on a seasonal basis over
three years (April 1978-November 1980). Because of the long time separa-
tion between successive surveys it was impossible to tie changes in bar relief
t o specific wave conditions. However, the repetitive pattern each year was
characterized by maximum bar relief at the spring survey, a moderate or
non-existent bar in the summer and only a gentle shoulder in the fall. This
suggests that the high-energy events which are most common in early winter
and spring are directly responsible for the growth of the inner bar.
In areas devoid of regularly spaced large groins, the inner bar system is
A.
Us bo*d D lkra h.
b. -1. U
% 16, 1979
Ld.I-)
0 1rnmmIm
d"
\\-
DEPTH PROFILES
- 14
1-9 k.rm. 1979
..
I 0 1rnmmIm
DEPTH PROFILES 61-62 July 26, 1979

63-72 AuC 25. 1979
F. July 16, 1979
II -:-:--
'I 76
1w 200
\ .wd 1wk
3w zw
B 3w 4w
'El
'p
I 0
w
co
I
fig. 7 . Bar configurations on Presque Isle. Map is baaed on air photos and Army Corps of Engineers surveys during June and b
w
July, 1979. Panels A through F sequentially cover the whole lake shore of Presque Isle. N
113
LAUE FUlF
FUlf H
I
I
I
I
II
No
I DAlA
I
0 xa I
I
M
A 28MAY81
LAKffRlE
GI0 7
I DNA
I
0 300 I
M
c 18AUGUSTBl
Fig.8. Morphology of the inner bars in the groin field along the neck of Presque Isle
during the summer of 1981. (A)Bars after more than one month of essentially calm lake
conditions. (B) Bars one week after a summer storm with 1 m high primary breakers.
( C ) Bars after a long period of essential calm interspersed with small summer storms.
From Sonnenfeld (1983).
generally straight. These bars dominate within segments I1 and 111. The bar
crest typically lies less than 50 m offshore in a water depth of 1-1.6 m. Near
all groins the bar is invariably deflected lakeward. Welding of the bar to the
beach has never been directly observed, but the seasonal reduction in bar
relief generally coincides with periods of beach accretion suggesting that the
bar does supply some of the sediment.
The beach along the peninsular neck is generally straight or cuspate with-
out any large-scale rhythmic features.
Outer bar. Segments I, 11, and I11 are characterized by one continuous
outer bar. This bar disappears as a single continuous feature to the east of
114
the lighthouse (between bathymetric profiles no. 61 and 63, Fig.7E). The
outer bar merges with the inner one at the updrift attachment point of
Presque Isle (Fig.7A). Along the adjacent Lake Erie mainland shore there
is only one bar. Downdrift from the point of merger the outer bar gradually
moves offshore until an apparent “equilibrium” distance of between 150 and
200 m is attained in the vicinity of bathymetric profile no. 7 (Fig.7A). The
bar crest typically lies in 3-4 m of water; its associated landward trough
attains depths of between 4 and 5 m. The transverse bar profile is quite
variable. It may be distinctly asymmetric with a steep flank toward land
(Fig.7A, profiles 5 , 7 and S), or symmetrical (Fig.7A and B, profiles 6, 1 2
and 14). In plan view the outer bar is gently crescentic with a “wavelength”
on the order of 400-1000 m. This bar rhythm displays no relationship to
any obvious shoreline features.
A longshore structure is apparent in the crest depth and the relief of
the outer bar (Fig.9). The water depth at the bar crest rapidly increases as
the bar moves farther offshore downdrift from its point of merger with the
inner bar. Bar relief, measured as the ratio between trough and crest depth,
has an average value of 1.305 for all the 6 0 profiles. This relief is low com-
pared to earlier published bar relief (Komar, 1976, p.301).
The relief exceeds the average along the first 33 profiles, it is less than the
average along the rest of the peninsula. The high-relief section roughly cor-
responds t o the major groin field (segment I).
Bathymetric profiles across the outer bar at PI 4,PI 7 and PI 9 demon-
strate seasonal changes in bar relief (Fig. 10). The most distinctive pat-
tern is one of maximum relief a few days after major storm events. Storm
magnitude is measured by the associated water level set-up recorded on the
NOS gage in Erie harbor (Fig.11). The largest storm during the 3-year
monitoring period occurred on April 6, 1979. The shoreface profiles along
the neck of Presque Isle recorded four weeks after this event displayed
the greatest bar relief of all the nine surveys conducted. PI 7 (Fig. 10) is a
typical example. The ratio between the trough depth and the crest depth
was 1.38 compared t o an average for all nine surveys at this location of 1.19.
The profile with the second highest relief ratio (1.37) was recorded on
.....CREST o w n
- RELIEF ITIC)
20- ’._
D -
GROIN I BEACH 8 UGHTHOUSE

SEGMENT I SEGMENT I1 SEGMENT Ill
Fig.9. Crest depth and relief of the outer bar along the neck of Presque Isle. Relief is
defined as the ratio between the trough and crest water depth. Numbers along the x-axis
(1-60) refer to Army Corps of Engineers’ bathymetric profiles shown in Fig.7.
115
Nov. 3, 1980, only a few days after a storm on October 25th (Fig.11). All
other shoreface profiles were recorded after long periods of fair weather.
During fair weather the bar relief is reduced through combined scour of the
crest and infill of the associated landward trough (Fig.10, 1979 set of pro-
files). The storm of October 25th, 1980, caused other profile changes as well.
The inner bar was reformed from the pre-existing shoulder which prevailed
through the summer and early fall and a third bar of low relief appeared in
deeper water more than 250 m offshore. Evidence for this third bar can also
be seen in the profile of May 3rd, 1979.
These seasonal bathymetric profiles demonstrate that onshore sediment
transport with attendant shoreface shoaling is the rule during the summer
and fall. Much of this shoaling is a direct consequence of onshore bar migra-
tion; in each of the three survey years the bar crest moved about 60 m land-
ward between the spring and fall surveys.
Meters
April 27
July 21
..... Nov. 3
MetepC’
25
“l
..... July
Oct. 19
Meters
100 200 300
April 23
July 30 1980
..... Nov. 3
Fig.10. Seasonal changes in shoreface profile PI 7 during the three survey years. Note the
generally higher bar relief at the spring surveys. The biggest storm of the study period
occurred about 4 weeks prior to the recording of the May 1979 profile. See Fig.1 for
profile location.
116
576
0Monthly Mean Lake Level 175 5
575
- Monthly M a x take Level
Survev Dates
7
--
574 175
*
c
-
aJ
1,
4 573
8 I 1 1 1 1
P T I TI I l l 170 5
572
57 1
c
i i 1I
I I I1
I
170
570
I977 I978 I979

+ 7!--7--
1980
F i g . l l . Water levels a t t h e Erie, NOS gage during the three year study period. Arrows
designate survey dates., Note the 175.6 m one-day water level for the April 6th, 1979
s t o r m ; 4 weeks prior to that year's spring survey.
Bars east of the lighthouse
As demonstrated in Fig.7E and F these bars have a very complex mor-

phology. A shore-parallel component is clearly present. In profiles 68-70
one can trace four longshore bars; farther west or east the pattern is more
confused. This segment of Presque Isle's shoreface (Fig.5, segment I V )
has a significantly gentler slope than the others (Table 111). This is consistent
with the budget-prediction of shoreface accretion within this segment. As a
consequence of this high rate of sediment supply, a multibarred shoreface
is developed. There is a clear relationship between the nearshore slope and
the prevailing bar types. Where the slope is very steep, as at Gull Point (east
end of segment V), the shore is characterized by one, or at most two, simple
longshore bars close t o shore. Slightly gentler slopes characterize the area
immediately west of Gull Point and the whole peninsular neck. The slope
of these is about 1" (Table 111). This shore is fronted by a transient inner
bar, a permanent outer bar and a third (discontinuous?) deep-water bar.
Finally, the gentle shoreface in segment IV has multiple parallel longshore
bars superimposed on transverse bar components.
TABLE 111
Nearshore slopes along the lake shore of Presque Isle. The listed slope is t h e average
between the water line and 5 m depth contour. Slope in degrees
Segment I I1 I11 IV V
Mean slope 1.05 0.97 0.97 0.67 0.91
Range 0.82-1.3 0.86-1.14 0.82-1.19 0.53-0.77 0.71-1.6
117
DISCUSSION
Morphodynam ics
Nearshore bars, in general, respond to storms by a decrease in relief, off-

shore crest migration, and subsequent rapid recovery. The Presque Isle shore-
face bars follow this same pattern. Short (1979) and Wright et al, (1979)
documented the detailed morphological changes occurring in an oceanic
surf zone bar system during storm and post-storm recovery cycles. In this
progressive change one can recognize a set of six morphodynamic stages
characterized by distinct morphology, wave forcing mechanisms and current
structure. According t o Wright et al. (1982), the morphologic changes go
from low-amplitude straight bars and troughs at the storm peak, through
crescentic and transverse bar stages to a non-barred straight beach with a
well-developed berm at the end of the completed recovery cycle (Table IV).
During storms the surf zone generally is dissipative (incident waves dissipate
their energy through the process of repetitive breaking; Guza and Inman,
1975), and the energy spectrum, at least in oceanic settings, is dominated
by long-period infragravity oscillations. The surf zone velocity field is
stratified with offshore residual currents near the bed. Morphodynamic
stages 2-5 (Table IV), associated with the progressive post-storm recovery
of the surf zone bed, generally have dissipative conditions at the bars and
reflective conditions at the beach face. Rather than being vertically strati-
fied, the surf zone current field is horizontally segregated into a series of rip
current circulation cells. On the fully accreted beach the wave processes are
dominated by reflection of the incident and subharmonic components of the
wave spectrum.
The completion of a recovery cycle of an oceanic beach may take many
weeks; therefore, a cycle is rarely brought to completion before the advent
of a new storm. Observed over long time, any beach is likely t o occupy the
whole range of possible morphodynamic stages. The most frequently en-
countered stage is referred to as the “modal beach stage” (Short, 1979).
Years of observation at Presque Isle have demonstrated that the configura-
tions described earlier in this paper are the modal stages for this system. The
modal stages progressively change along the Presque Isle beaches in response
to differences in prevailing surf zone processes.
The objective here is t o explain this observed variability in the Presque
Isle bar field in terms of the inferred modes of surf zone motion. The selec-
tion of such modes is a function of the deep water wave spectrum, and the
shallow water refraction and energy attenuation. These factors, in turn,
are controlled by the lake wave climate, large-scale bathymetry and sediment
supply.
118
TABLE IV
Characteristics of the surf zone morphodynamic stages in the model used in this paper.
Data from Wright et al. (1982) and other sources
Stage Morphology Dynamics Dominant spectral band Currents
1 Low-amplitude Dissipative Long-period Vertically

straight bars infragravity stratified.
and troughs oscillations Offshore
(T: 1-3 min) flow near bed
2 Straight/crescentic Dissipative Broad band. Weak rips
well-developed bars. across bar. Low-period
Straight beach' Reflective infragravity
at beach (T: 30-60 s)
and subharmonic
oscillations
3 Highly crescentic Dissipative Same as 2 Moderate
bar. Megacusps at bar. rip currents
on beach Reflective
at beach?
4 Transverse bars Dissipative Same as 2 Strong rip
and rip at bars. currents
channels Dissipative
or reflective
in rip channels
5 Ridge and runnel Dissipative Same as 2 Small rip
or low-tide at low water currents
terrace level.
Reflective at
high water
level.
6 Well-developed Reflective Incident wave No hori-
berm on straight frequency and zontal o r
beach first subharmonic vertical
segregation
Ismall cusps may be present in the swash zone for all stages 2-6.
lDissipative conditions across transverse bars induce water level set up, driving rip current
circulation.
Lake Erie mainland shore
Because of refraction and attenuation of the prevailing westerly waves

before they reach this mainland shore, the wave power is relatively low.
Persistent longshore currents to the east effectively remove sand and gravel
supplied by the bluffs preventing the building of a wide shoreface. Conse-
quently, this segment of shore has only one transient nearshore bar. After
storms the bar quickly attains the welded and transverse bar stages (Table IV,
stages 3 and 4). Continued fair-weather bar migration forms a straight (or
cuspate) berm after a few weeks.
119
The neck of Presque Isle
Storm waves commonly arrive nearly perpendicular to this section of the

Presque Isle shore. A large proportion of their energy is transferred into
storm water-level set-up against the beach; the surf zone current pattern
becomes stratified with net lakeward flow near the bed as documented
by fluorescent tracer studies (Sonnenfeld, 1983;Taylor, 1983). Sand moving
lakeward from the beach and inner bar system under such conditions appear
t o feed the outer bar, which typically is moved lakeward to an “equilibrium”
distance of 200 m during storms.
Historical data are also consistent with the interpretation that the adjacent
beach is the primary source of bar sediment. The outer bar along the Presque
Isle neck attained its maximum recorded relief shortly after completion of a
major beach nourishment project in 1955-1956. The average outer bar
relief (trough depthicrest depth) in a 1957 survey was 1.74 (Nummedal,
1981).
The outer bar remains in morphodynamic stage 2 (or 3) (Table IV) during
fair-weather periods throughout the year. Because of the absence of long
period swell conditions in Lake Erie, the storm-generated bar remains as a
“relic” form rather than migrating onshore during “intermediate” energy
conditions. This is probably the explanation for the permanence of the outer
bar along many exposed Great Lakes shorelines (Short, 1979).
The inner bar along the neck of Presque Isle is directly affected by the
groins. Standing waves between the shore-normal groins may interfere
with components of the incident wave spectrum to produce a crescentic
pattern of bottom sediment convergence (Bowen and Inman, 1971).
After long periods of low summer waves the inner bar evolves through
the “low-tide” terrace stage (stage 5, Table IV) and welds onto the beach
as a (frequently cuspate) berm.
Observations show that the inner bar is removed during major storms.
Tracer dispersal patterns and drogue studies of the currents suggest that
sediment is moving lakeward both by bottom return flow and groin-associated
rip currents (Sonnenfeld, 1983).
The essential difference between the modal bar configuration in segments
11 and I11 and that in segment I is that the inner bar in the former segments
generally has no crescentic components. This suggests that the current field
lacks longshore structure. Such structure is generally induced by groins.
Bars east of the lighthouse
As documented in the transport section, sediment for this shoreface

region is largely derived from sources updrift rather than the local beach.
Sediment abundance has built a permanently multi-barred, dissipative surf
zone. Because the storm-wave energy level is lower than farther west this
surf zone maintains a morphodynamic modal stage different from the
peninsular neck. Crescentic bars and megacusps (stage 3, Table IV) are the
120
norm. Megacusps slowly migrate eastward and beach erosion is tied to rip
current embayments between the megacusps. In spite of abundant sediment
on the shoreface, strong rip circulation keeps these beaches erosion-
prone. This explains why beach nourishment immediately east of the light-
house groin did not stop downdrift erosion (Fig.6). The nourished sand
moved lakeward by rip current action onto the crescentic bar.
The complex bar morphology in this area may be explained as follows.
The outer bar becomes a post-storm “relic” feature in stage 2 (or 3)
(analogous to the outer bar farther west). The multiple inner bars are
progressively altered through lower (more reflective) morphodynamic stages
during the post-storm recovery phase thus generating distinct transverse
bar components. The entire bar-rip channel-megacusp system remains strongly
skewed to the east because of strongly eastward-directed incident waves.
Gull Point area
This beach (segment V, Fig.5) receives the lowest annual wave power
along the entire outer shore of Presque Isle. The modal morphodynamic
stages are 4 and 5 (Table IV). Because of strongly eastward-directed incident
waves the rip channels and transverse bars are skewed t o become nearly
shore-parallel (Fig.7F). As such, they are hard t o differentiate from ridge
and runnel systems. Associated megacusps maintain a typical spacing of
700-800 m. The surf zone is narrow; the shoreface profile is the steepest
one along the peninsula (Table 111).This profile is maintained by the prevail-
ing onshore transport. Storms with incident waves from the northeast are
the only ones which could generate sufficient energy for offshore transport
and shoreface broadening. Such storms are infrequent and moderate in
strength. Gull Point has a high rate of sediment influx from updrift and is
therefore rapidly accreting. The wave climate is such that this sediment will
be kept close to shore maintaining a steep prograding shoreface profile.
CONCLUSIONS
Volume rates of sediment transport for the Lake Erie shore of Presque Isle
were determined from hindcast wave power distributions and historical
records of shoreline change and dredging volumes at the downdrift end of
the transport system. The calculations show that: (1)the maximum net
longshore transport occurs in the vicinity of the lighthouse where the annual
volume rate is 209,000 m3; (2) due to downdrift changes in sediment trans-
port rate, the shoreface along the neck of Presque Isle is eroding whereas
the shoreface to the east of the lighthouse is accreting; and (3) the erosional
or accretionary state of the shoreface is not, by itself, adequate information
t o determine the stability of the adjacent beach. Beach stability depends on
the morphodynamic stage of the shoreface. Specifically, the beaches east of
the lighthouse experience erosion in mega-cusp embayments in spite of large-
scale beach nourishment immediately updrift. Only at Gull Point do the
beaches experience sustained accretion.
121
The morphodynamic surf zone model developed for oceanic beaches by

Short (1979) and Wright et al. (1979) is used as a basis for discussion of
shoreface morphologic variability at Presque Isle. In spite of interference by
major shoreline stabilization structures and expected differences between
oceanic and lake wave climates the nearshore bar field at Presque Isle does
closely correspond t o that model. The Presque Isle lakeshore can be divided
into three sections with distinctly different morphodynamics. These are:
(1)The neck. A permanent outer bar is maintained by storm waves in
stages 2 and 3 (Table IV). The inner bar may occupy any morphodynamic
stage in response t o varying wave conditions. It is commonly crescentic
because of waves induced by the groin field.
( 2 )Shore east of the lighthouse. Abundant sediment supply has produced
a multi-barred shoreface of gentle slope. Strongly skewed transverse bar
components maintained by rip current circulation are generally present.
( 3 ) GulZ Point. Prevailing low incident waves maintain this shore in the
ridge and runnel morphodynamic stage.
ACKNOWLEDGEMENTS
This work has been supported by the U.S. Army Corps of Engineers,
Buffalo District, through contracts DACW49-78-C-0020, DACW49-79-C-
0055 and DACW49-80-(3-0047. The authors acknowledge valuable discus-
sions with Joan Pope. Thorough reading by two anonymous reviewers
greatly helped improve the manuscript.
REFERENCES
Bowen, A.J. and D.L. Inman, 1971. Edge waves and crescentic bars. J. Geophys. Res., 76:
86 6 2-8 67 1.
Carter, C.H., 1976. Lake Erie shore erosion, Lake County, Ohio: setting, processes and
recession rates from 1876 to 1973. Ohio Geol. Surv., Rep. of Invest. 99, 105 pp.
Coastal Engineering Research Center, 1973. Shore Protection Manual, 3 vols.
Guza, R.T. and D.L. Inman, 1975. Edge waves and beach cusps. J. Geophys. Res., 80:
2997-3012.
Holman, R.A., 1981. Infragravity energy in the surf zone. J. Geophys. Res., 86: 6442-
6450.
New Jersey, 429 pp.
Messinger, D.J., 1977. Form and change of a recurved sand spit, Presque Isle, PA. M.Sc.
thesis, State University of New York, Fredonia, N.Y., 1 2 5 pp.
Nummedal, D., 1979. Monitoring of shoreline change, Presque Isle, Pa. Annual Report to
Buffalo District, US. Army Corps of Engineers, for contract no. DACW49-78-C-
0200,52 pp.
Buffalo District, U.S. Army Corps of Engineers, for contract no. DACW49-79-C-0055,
72 PP.
Buffalo District, US. Army Corps of Engineers, for contract no. DACW49-80-C-0047,
62 PP.
122
Nummedal, D., 1983. Sediment transport and surf zone morphodynamics of Presque
Isle, Pennsylvania. Coastal Zone '83, Am. SOC.Civ. Eng., 111: 2612-2629.
Nummedal, D., Hayes, M.O. and Fahnestock, R.K., 1976. Littoral processes and sedimen-
tation in the Cattaraugus Embayment, N.Y., Appendix E. In: General Design Memo-
randum, Sunset Cattaraugus Harbor, N.Y., 246 pp.
Pierson, W.J. and Moskowitz, L., 1964. A proposed spectral form for fully developed
wind seas based on the similarity theory of S.A. Kitaigorodskii. J. Geophys. Res., 69:
518W5190.
Resio, D.T. and C.L. Vincent, 1976. Design wave information for the Great Lakes; report
1, Lake Erie, U.S. Army, Waterways Experiment Station, Tech. Rep. H-76-1, 54 pp.
Saville, T., 1953. Wave and lake level statistics for Lake Erie. Beach Erosion Board, Tech.
Memo. 3 7 , 2 4 pp.
Seabergh, W.T., 1983. Design for prevention of beach erosion of Presque Isle beaches,
Erie, Pa. - Hydraulic model investigation. U S . Army, Waterways Experiment Station,
Vicksburg, MS, Draft Report to Buffalo District, U.S. Army Corps of Engineers.
Short, A.D., 1979. Wave power and beach stages: a global model. Proc. 16th Int. Conf.
Coastal Engineering, ASCE, 11: 1145-1162.
Sonnenfeld, D.L., 1983. Inner bar sediment dynamics, Presque Isle, Pa. M.Sc. thesis,
Louisiana State University, Baton Rouge, La.
Sonu, C.J., 1973. Three-dimensional beach changes. J. Geol., 81: 42-64.
Taylor, K.B., 1983. Sand dispersal patterns on the outer bar, Presque Isle, Pa. M.Sc. thesis,
State University of New York, Fredonia, N.Y.
U.S. Army Corps of Engineers, 1953. Presque Isle Peninsula, Erie, Pa., Beach erosion con-
trol study, House Document no. 231,83rd Congress.
US. Army Corps of Engineers, 1979. Presque Isle Pennsylvania, Erie, Pa., Draft Phase I,
General Design Memorandum, Dept. of the Army, Buffalo District.
U.S. Naval Weather Service Command, 1975. Summary of synoptic meteorological obser-
vations for Great Lakes area. National Climatic Center, Asheville, N.C., 201 pp.
Walton, T.L., 1973. Littoral drift computations along the coast of Florida by means of
ship wave observations. Coastal and Oceanographic Engineering Laboratory, Univ. of
Fla., Tech. Rept. 15, 9 6 pp.
Wright, L.D., Thom, B.G. and Chappell, J., 1979. Morphodynamic variability of high
energy beaches. Proc. 16th Int. Conf. Coastal Engineering, ASCE, 11: 1180-1194.
Wright, L.D., Short, A.D. and Nielsen, P., 1982. Morphodynamics of high energy beaches
and surf zones: a brief synthesis. Tech. Rept. 82/5, Coastal Studies Unit, Univ. of
Sydney, 6 4 pp.
Marine Geology, 60 (1984)123-154 123
SEDIMENTOLOGY AND MORPHODYNAMICS OF A MACROTIDAL

BEACH, PENDINE SANDS, SW WALES
C.F. JAG0 and J. HARDISTY

Department of Physical Oceanography, Marine Science Laboratories, U.C.N. W., Menai
Bridge, Gwynedd, LL59 5EY (U.K.)
Department of Geology, University of Bristol, Bristol (U.K.)
(Received March 14, 1983;revised and accepted January 4, 1984)
ABSTRACT
Jago, C.F. and Hardisty, J., 1984. Sedimentology and morphodynamics of a macrotidal
beach, Pendine Sands, SW Wales. In: B. Greenwood and R.A. Davis, Jr. (Editors),
Geol., 60: 123-154.
The foreshore of Pendine Sands forms the seaward part of an extensive, sandy coastal
barrier in a shallow Carmarthen Bay, SW Wales. The sedimentological features of the
macrotidal foreshore reflect a tide-induced modification of nearshore wave characteris-
tics. As the tide ebbs, the breaker height may decrease, the surf zone widens and becomes
increasingly dissipative, and swash/backwash velocities diminish. A concomitant change
from plunging to spilling breakers and increasingly symmetrical swash zone flows are
associated with a decreasing beach gradient.
A zero net transport model demonstrates that the beach profile is self-stabilising in the
short-term, and periodic levelling has shown that the beach is in long-term equilibrium
with prevailing conditions, though this does not preclude a significant dynamic response
to changing tides and waves.
The flow regimes of wave-generated currents decline as the tide ebbs, and normal
beach processes do not usually affect the lower foreshore. Accordingly, there is an
overall seaward-fining of the primary framework component of the sands. In more
detail, this framework component displays a slight seaward-coarsening across an upper
foreshore dominated by high water swash and surf; a rapid seaward-fining across the mid-
foreshore in response to the ebb-attenuating swash zone flow velocities; and a slight
seaward-fining across the lower foreshore under the action of nearshore shoaling waves.
Bedforms vary from a swash/backwash emplaced flat bed across the upper foreshore to
the small ripples of nearshore asymmetric oscillatory flows across the lower foreshore.
The surface sediment veneer is not representative of the subsurface sediments which
form in response partly to fairweather conditions, partly to storms. The upper foreshore
is characterised by swash/backwash emplaced plane bedding in fine sands frequently
disrupted by bubble cavities. The mid-foreshore is composed of coarser-grained shelly
traction clogs arranged as landward- and seaward-dipping large-scale cross bedding and/or
plane bedding; these are probably storm breakerlsurf deposits. The lower foreshore,
though partially and sometimes totally bioturbated, shows landward-dipping small-scale
cross bedding in very fine sands sorted by nearshore shoaling waves.
Tide- and storm-induced modification of the nearshore flow regimes therefore pro-
duces a distinctive shore-normal array of sedimentary facies. Each facies is characterised
by diagnostic textural and structural signatures. A prograding sequence of such macro-
tidal deposits would be similar to, but more extensive than, a comparable microtidal
sequence.
124
INTRODUCTION
The nearshore environment can be envisaged as a shoreward progression

of distinct dynamic zones - shoaling waves, breaker, surf, swash - each
characterised by a particular mode and intensity of sediment mobility (see,
for example, Komar, 1976, for summary). Such a progression of wave
effects develops a shore-normal sequence of bedforms and syndepositional
sedimentary structures arranged in more-or-less linear bands parallel to the
strandline (Clifton et al., 1971; Davidson-Arnott and Greenwood, 1976).
This arrangement of sedimentary features must respond to fluctuating
nearshore dynamics and water levels. Thus, on macrotidal shorelines, while
wind-generated surface waves remain the principal, if variable, source of
energy, dissipation of this energy may be controlled by tidal processes. As a
consequence of the tide, the nearshore dynamic and sedimentary zones
sweep the foreshore t o a degree that depends on the tidal range. This must
create a certain migration of sedimentary facies with each tidal cycle; Clifton
et al. (1971) briefly described just such a migration across a mesotidal
foreshore.
The aim of the study outlined below is firstly to examine the shore-normal
intertidal textural gradient and array of Sedimentary structures, and thence
to derive a simple dynamic model that incorporates grain-size, bedforms,
beach slope and wave characteristics on a depositional macrotidal shoreline.
PENDINE SANDS
The beach at Pendine Sands forms the seaward part of an embayed

Quaternary beach/dune barrier which extends some 10 km from a rocky
headland in the west to the confluence of three small estuaries in the east
(Fig.1). Carmarthen Bay is rather shallow, a mere 1 0 m deep some 6 km sea-
ward of the dunes at low spring tide. A reconnaissance study of the bay has
established that it is floored with abundant offshore sand waves and with
nearshore bars and intertidal sandbanks (particularly off the eastern end of
the barrier). Historical evidence, field observations and aerial photographs
emphasise the mobility of the nearshore features in response to the rapidly
changeable dynamics of the bay. The predominant sedimentological trend
is depositional in the northern and eastern nearshore areas (Fig.1) coupled
with a progressive and rapid movement of marine sand into the estuaries
(Jago, 1980). The area is therefore of considerable geological interest as a
model of coastal sedimentation.
Carmarthen Bay has a southwesterly aspect and so is exposed to an
oceanic swell with a fetch of over 5000 km. Draper (1972) estimated that
the highest 50 year storm wave in the Celtic Sea should be 30 m. Analysis
of waves logged during 1968 by the St. Gowan Lightvessel at the western
entrance of the bay gives: 6 = 1.2 m, = 2.0 m, TI,,, = 6 s; while
Darbyshire’s (1963) analysis of data during 1960/61 from the “Helwick
Lightvessel” at the eastern entrance of the bay gave a modal wave height of
125
0 5 krns
Fig.1. Carmarthen Bay and Pendine Sands, showing working transect at the western end
of the beach.
3.0 m. However, our observations at Pendine Sands suggest that the energy
loss as waves sweep shoreward across the shallow bay is such that few large
waves reach the shoreline. I t may ,be predicted (Hydraulics Research Sta-
tion, 1978) that, for the prevailing waves that approach from 203",4 s waves
will be reduced in height by 50% and 10 s waves by 70% as they shoal from
the 40 to the 5 m isobaths off the eastern end of the barrier. Constriction
of the shelf tidal wave in the Bristol Channel produces semi-diurnal tides in
Carmarthen Bay of mean spring range 7.5 m and mean neap range 3.7 m.
Extreme springs are of 10.0 m range. Surface currents reach 1.0 m s-' in the
middle of the bay (R.J. Uncles, pers. commun., 1980). With such a tide, and
a gentle foreshore/shoreface slope (<lo), the intertidal zone becomes very
wide at low water springs - 1500 m or more at the widest part of the barrier.
FIELD AND ANALYTICAL PROCEDURES
Seven shore-normal transects, spaced at approximately 1.4 km intervals

along the foreshore, were regularly levelled from the dune to the low water
mark of a spring tide at 3 monthly intervals during 1968-1970. The shoreface
was profiled once by echo sounder along seaward extensions of the beach tran-
sects using a small boat positioned by theodolites from the shore.
126
Surficial sediment samples were collected along the transects at low

tide, each sample consisting of a composite of four closely-spaced sub-
samples of the upper 3 cm of the beach sediment. These composites were
sieved at 0.25 phi unit intervals. I t was established that the between-sample
station variation was significantly greater than any sampling/sieving errors
which could therefore be ignored. Textural parameters (M,= mean grain
size, uI = graphic standard deviation) were calculated after Folk and Ward
(1957). Senckenberg box cores were taken at intervals across the transects
at low tide and impregnated with either nitrocellulose lacquer or epoxy
resin (after Bouma, 1969). Additional grain-size analyses were made of
individual laminae within selected box cores.
The results discussed below are mostly limited to the tsansect at the
western, non-barred end of the barrier (Fig.1) where we have additional
annual surveys during 1980-1982. Concurrent with the sedimentological
procedures, certain breaker, surf and swash characteristics were monitored
along this transect at 1 5 min intervals during several tidal cycles. These
measurements included breaker height (using a hand-held staff), surf and
swash zone widths (with measuring tape), times of uprush (of surf and
swash), swash/backwash velocities (using hand-held Ott and Braystoke
flowmeters), and swash/backwash bedload transports (using a simple sedi-
ment trap).
MORPHODYNAMICS OF THE FORESHORE
Beach morphology
All the beach profiles at the western end of the barrier have a uniform
concave-upward shape with a marked absence of either longshore features,
such as ridges and runnels, or rhythmic structures such as cusps, crescentic
bars, etc. The foreshore is backed by established foredunes (ca. 6 m high)
and a backshore of variable width (0-16 m depending on the tide). Except
at the eastern end of the barrier, where offshore shoals are important, the
foreshore profile continues smoothly seaward at a diminishing gradient
(Fig.2). The foreshore can be divided into three zones on morphological
and sedimentological grounds: the upper foreshore, from highest spring
to mean neap high tide marks, has the steepest gradient (though still, on
average, less than 1.5"); the middle foreshore, between mean neap high and
low tide marks, is of intermediate gradient (= 0.7"); and the lower foreshore,
below the mean neap low tide mark, has the lowest gradient (= 0.4"). The
profile continues offshore with a gradient of <0.1".Obviously the frequency
of exposure of the intertidal zone during a lunar half cycle diminishes from
upper to lower foreshores.
Tidal variation of wave characteristics
Carmarthen Bay is up to 10 m shallower at low water than it is at high

water. Because of the concave-up intertidal/subtidal profile, the shoreface
127
Distance seawards krn

0 4 8 I2 16 20 24
I 1 I I I I I
-.I0
0
E
-10 g
L
O
--20 4
W
- -30
--LO
*6-
E *&-
= .2-
.-
2
(I
5 0-
w
-2-
-4-
-6-
1 I I , I I I I
0 200 4 00 600 800 1000 1200 1400 1600
D i s t a n c e s e a w a r d s . rn
Fig.2. Nearshore and foreshore morphology. HWHT = high water highest tide; L W L T =
low water lowest tide; MHWNT and M L W N T = mean high water and mean low water
neap tide, respectively.
(Le., that part of the seafloor affected by waves) may be twice as wide at
low water as at high water (Fig.3). Since the frictional dampening of wave
energy is a function of water depth and shoreface width, the progressive
shoreward modification of shoaling waves must vary according to the stage
of the tide. Wave attenuation will increase, and effective nearshore wave
energy decrease, as the tide ebbs.
The morphodynamic variables are reflected in the changing character
of the nearshore waves during the tidal cycle. Thus, for example, the breaker
coefficient (B, = Hb/ g P t a n a, where Hb= breaker height, tan a! = beach
T = 10s L = 156m
HWHT -
f
Fig.3. Variable shoaling modification of a 10 s wave. Significant shoaling begins at L/4.

HWHT and LWLT = high water highest tide and low water lowest tide, respectively.
128
slope, T = wave period) after Galvin (1968): plunging breakers, with BO <
lo-', over the upper foreshore; spilling breakers, with Bo > lo-', over the
middle and lower foreshores (Fig.4). The breaker height may diminish as the
tide ebbs (Fig.4).
There are concomitant changes in the surf zone as the profile slope
changes. The surf scaling parameter, E = 4nZBo/tana, after Guza and Inman
(1975) and Wright et al. (1979), is always high and increases during the
ebb (Fig.4). Hence the flat lower foreshore is highly dissipative under any
conditions; the uppermost foreshore could be very moderately reflective
with respect to exceptionally low waves, but such conditions were not
observed. The phase difference (P = tb/T, where tb = duration of surf and
swash flows) of Kemp (1961) increases with the ebb as surf and swash zones
wid'en and swash/backwash currents decline (Fig.5).
.G
Mid - lo r
L O W P ~ foreshore
L 1 I , I 1 1 I ~ ' ' ~ ' ' ~
0 200 400 GOO 800 1000 1200 1400 ltOO
Distance seawards, rn
Fig. 4. Variation, o n an ebbing tide, of breaker height ( H b ) , breaker coefficient ( B , ) and

surf scaling parameter ( 6 ) . E is displaced landward since surf is landward of breaker.
129
-L-
6-
Lower foreshore
L , I I , " I ' m , ,
0 200 LOO GOO 800 1000 1200 1400 !GOO
Distance seawards, rn
Fig.5. Variation, o n an ebbing tide, of swash width (Zs), surf width (Ib), phase difference
( P ) and mean swash/backwash velocity (U,).
Beach stability
Taken as a whole, the seven surveyed transects showed a mean vertical

erosion of 0.5 m during the period 1968-1970 (this seems very little but
nevertheless represents a total volume of 0.46 X lo6 m3). However, much
of this erosion occurred in the vicinity of the estuary due to movement of
estuarine channels. The western end of the beach was remarkably stable
(Fig.6). The greatest change between surveys of this westernmost transect
was a mean erosion of 0.32 m (amounting to 96 m3 m-l width of profile)
in response to a westerly gale. After nine surveys, the standard deviation of
volumetric variation of the beach prism (above mean spring tide low water
130
Accretion = 1860 m 3 / m
a,= 30m31rn
2 6 8 - 5 68
5 68 - 9 68
9 68-11 68
11 6 8 - 2 69
2 69-669
6 69 - 8 69
8 6 9 - 1 1 69
11 69 - 2 70
Fig.6. Periodic beach profile changes, 1968-1970. V = mean beach prism volume above
mean low water spring tide mark. u v = standard deviation of V.
mark) was 30 m3 m-' (cf.Sho~i,

1980). The stability of thepmfjjlesuggests
that the foreshore was in long-term equilibrium with the prevailing condi-
tions over the period of study. The beach is similar, in this respect, to micro-
tidal dissipative beaches which rarely experience substantial profile changes
because of the low sediment exchange between nearshore, surf and subaerial
zones (Short, 1980).
SEDIMENTOLOGY
Interpretation of grain-size distributions
Like most grain size frequency distributions, the Pendine Sands sediments
plot on probability paper not as straight lines but as S-shaped curves. There
is little doubt that these curves represent composite distributions and that
they can be interpreted as mixtures of three or more subpopulations. These
subpopulations are the result of bed-building and depositional processes (Moss,
1972) or of hydraulic and transportprocesses (Middleton, 1976) or of both.
Many of the Pendine sands can be dissected into three components
using a simple graphical technique (Cassie, 1954). A typical example is
shown in Fig.7A. These will be referred to as contact, framework and
131
0 units
0 units
Fig.7. Graphical dissection of grain size distributions, (A) single sedimentation unit; (B)
dual sedimentation unit.
interstitial, after Moss (1972), while acknowledging that they probably

relate also to different modes of transport, after Middleton (1976) (although
it should be noted that both Moss and Middleton were concerned with
steady, unidirectional flows and not with rapidly reversing or oscillatory
flows).
As shown in Fig.7, the framework component is lognormal and very well
sorted. This indicates the degree of selectivity of the bed-building process
and/or the efficiency of the intermittent suspension transport mechanism.
132
I t is not clear whether the remaining components form log-normal or trun-

cated populations since in many cases the data permit either solution. The
precision of the method (sieving plus graphical presentation) is too low to
resolve this question but, on balance, we favour truncation as this is most
compatible with Moss’s (1972) field observations.
Not all samples resolve into three components. Some contain a sizeable
fourth component (Fig.7B). It would be expected that any number of com-
ponents could be present if the sample intersects several depositional hori-
zons and this seems to be the case here. The mixing reflected in the grain-
size distribution is therefore a sampling ‘error’, and reflects the difficulty
of recognising and sampling discrete sedimentation units in the field. This
difficulty was resolved by sampling the Senckenberg box cores used to
examine sedimentary structures (see below). After removal of the impreg-
nated face, the sand remaining in the box retains a perfect record of the
structure and can be sampled at any points of interest. Our intention was to
sample individual laminae but this proved impossible in the fine-grained,
often bioturbated, sands. Where feasible, samples of visually homogeneous
units of several laminae were taken at about 3 cm intervals down the core.
After this more precise sampling we find that the grain-size distributions of
the sands can always be resolved into three components.
G ra in-size characteristics
The surficial sediments of the Pendine foreshore are mostly fine and very
fine sands ( M , > 2.50 phi units) and usually very well sorted (uI < 0.35 phi
units). While textural variations within the 30 cm vertical sections of the box
cores are generally small, there are pronounced shore-normal variations, so
that the upper, mid, and lower foreshore sands are distinctive.
Upper foreshore
On the upper foreshore, the subsurface sands are virtually identical to the
surficial sands (Fig.8). The framework component (with uF = 0.27 phi units)
is dominant; the contact component contributes less than 3% and gives the
grain size distribution a short tail; the interstitial component is variable,
0.1-11.0%, and greatest at high water mark where the grain size curve can
show a marked gradient change toward its fine end. Where well-developed,
the interstitial component consists mostly of heavy minerals deposited at
the upper limit of swash action.
Mid-foreshore
Over the mid-foreshore, the subsurface sands are again frequently uniform
but differ from the surface layer (Fig.8). The subsurface sands contain three
components: a framework component (uF = 0.40 phi units); a large contact
component, up to 7096, which is poorly sorted (a, = 0.70-1.35) and whose
mean is about 3.0 phi units coarser than the framework; a negligible inter-
stitial component (<0.1%). These sands make up almost the entire mid-
133
-
20 50 80 99 'I,
F F O R E S HOR E
- 20 50 80 99 01.
F FORESHORE
30'
Fig.8. Vertical variation of textural parameters in box cores. C, F and I refer to the con-
tact, framework and interstitial component, respectively.
foreshore section. They are distinctive because of their large contact com-
ponent which consists of ill-shaped shell debris, whole shells and some
pebbles. This material is found only in the mid-foreshore deposits. We have
yet to observe any swash zone characteristics that would account for such a
mid-tide phenomenon. A likely explanation is that it represents a storm lag
produced under the breaker/surf at high water still-stand. If we take the
breaking criterion as Hblhb = 0.78 (where Hb = breaker height and h , =
water depth at break point), then a 2 m breaker will break in 2.6 m of water.
134
Corresponding to the limits of mean spring and mean neap high tides,
breakers of this size would produce a sweep zone which matches quite
closely the observed range across the mid-foreshore of this coarse grained
component (Fig.9). This implies that coarse-grained material may concen-
trate close to the breaker during storms. That such a mechanism can occur
has been demonstrated by Ingle (1966) who showed, using fluorescent
tracers, that coarse grains move into the breaker from the surf zone under
non-equilibrium conditions. I t is likely that, on a macrotidal beach, the
breaker and surf zones are too mobile for a significant deposit to form
except at high water (and perhaps at low water). The resulting deposits
resemble traction clogs (Moss, 1972) which may form when accumulating
grains of the contact component control bed-building processes. The sands
are often cross-bedded (see below) which suggests that some of the coarse
material may accumulate as lags in the troughs of large bedforms
(B. Greenwood, pers. commun., 1983). Micro-layering of alternate coarser
and finer sands occurs within the major cross-bedded units and this may
result from the strongly asymmetrical oscillatory flows of the breaker zone.
The traction clogs are missing on the eastern half of the Pendine barrier
where the nearshore shoals apparently protect the beach from storm breakers.
The surface layers of the mid-foreshore differ in that they contain aframe-
work component that is finer grained than the subsurface frameworks; a
smaller, but still important (up to lo%), contact component; and an inter-
stitial component that is insignificant (<0.5%), though greater than in the
subsurface. As a result, the mean grain size is finer in the surface layers than
in the subsurface (Fig.8). The surface layers must be derived, at least in part,
from reworking of the uppermost portion of the subsurface. Blackley and
Heathershaw (1982) have shown that selective transport of different grain
sizes may take place in both alongshore and shore-normal directions on a
comparable macrotidal beach. It is likely that reworking by swash and/or
surf of the framework component, and a little of the contact component, of
the storm deposits creates the finer-grained framework and contact com-
ponents of the surface layers.
Lower foreshore
The lower foreshore sands are identical throughout the vertical section
(Fig.8). This is partly the result of bioturbation, and the subsurface bio-
turbated sands are texturally indistinguishable from the surficial cross
bedded sands. The framework (uF = 0.20 phi units) is again the major
component; the contact component, though <6%, is poorly sorted so that
the parent grain size curve has a long coarse tail (Fig.13); and the inter-
stitial component makes up t o 2% of the curve.
Dual sedimentation units

Sampling methods which cut through bundles of laminae will clearly be
sampling several sedimentation units. Samples may represent a range of
transport processes and depositional events created by the tidal sweep of
135
3 0-
c
y1
'526-
\9
1
2 2.
18-
x Admixed component
20-
. 0 Contact
-Interstitial
10-
0-
P
015
2L
7 B r e a k point at M H W N T for T Z 1 0 s
.. ... ,::;. .. : ,. ...- _. ... , . . ., . .

Lower foresn o r e . . ...I.- -:: . . ...... ::1
"
, I
0 200 400 600 800 1000 1200 1400 1600

Distance s e a w a r d s , m
Fig.9. Shore-normal variations of mean grain size (M,) and sorting (uI); population
frequencies;framework mean grain size (MF) and sorting ( uF ).
136
the foreshore. It is noteworthy that the surface veneer over the mid-
foreshore differs from, and is consequently unrepresentative of, most of the
mid-foreshore deposits. Comparable units have been observed on micro-
tidal (Otvos, 1965) and mesotidal beaches (Williams, 1971). Graphical
dissection of the grain-size distributions of the composite samples from the
Pendine mid-foreshore often produces four components (Fig.7B). These
are dual sedimentation units consisting of: (1)the usual three components
(surface layers); and (2) an admixed component (subsurface). The relative
proportion of the two units depends solely on the depth of sampling rela-
tive to the depth of reworking of the storm deposits by fairweather swash
and surf. King (1951) found that on a Carmarthen Bay beach of similar
grain size and slope, the depth of disturbance was about 1 cm for every
30 cm of wave height. Wave conditions prior to sediment sampling at
Pendine should have disturbed the upper and mid-foreshores down to
2-3 cm and the lower foreshore to about 1 cm or less. With a sampling
depth of 3 cm, the mid-foreshore samples contain up to 24% of sediment
undisturbed by reworking (Fig.7B).
These dual sedimentation units appear only on the mid-foreshore. The
laminae of the upper foreshore sands, while clearly deposited at different
times and subsequently reworked, are always exposed to the same kind of
process - i.e. swash and backwash. The laminae are accordingly similar in
texture. The lower foreshore sands are frequently bioturbated and suffer
minimal reworking by waves (see below). Little selective sorting of grains
seems to occur (once deposited), so the sands are remarkably similar in
texture in a vertical section (Fig.8).
Shore-normal textural variations
From the foregoing, it is obvious that the grain-size distributions of the

sands will be influenced by the choice of sampling procedure. Hence, com-
posite samples, which cut through more than one depositional unit, show
a pronounced coarsening and become poorly sorted over the mid-foreshore
(Fig.9). Dissection of the grain-size curves shows that this is Iargely due to
the presence of a coarse admixed component - from the storm-emplaced
subsurface layers. This component can therefore be isolated by analysis
of the curves of composite samples or by sampling of discrete laminae
(or bundles of laminae) in box cores. Both methods have been used here.
Considering first the surface sands: these are dominated by a very well
sorted framework component which becomes finer grained down the fore-
shore, This seaward-fining gradient parallels the seaward-attenuating swash
zone velocities. In more detail, the shore-normal framework gradient has
three components (Fig.9): (1)a slight seaward-coarsening across the upper
foreshore; this must be established at high water when surf energy and swash
zone velocities decline from the break point to the landward limit of the
upmsh (cf. Evans, 1939);(2) arapid seaward-fining across the mid-foreshore,
matching the progressive ebb-attenuation of surf energy and swash zone
137
velocities; and (3) a slight seaward-fining across the lower foreshore; swash
zone flows are too subdued to entrain these sands during most conditions
and the textural gradient here is probably generated by nearshore shoaling
waves (cf. Komar, 1976; Jag0 and Barusseau, 1981). The lower foreshore
framework components are the best sorted - perhaps because rates of depo-
sition are slower under shoaling waves than in the swash zone. Repeated
sampling of the transect has shown that minor temporal variations can occur
as tides and waves change, but the above trends are invariably present.
The interstitial component is barely represented in these sediments since
fine grained material is removed by surf zone turbulence. I t is sometimes
present at the upper limit of swash action (where it consists mostly of heavy
minerals) and is marginally more abundant on the low energy lower fore-
shore than elsewhere (Fig.9).
The contact component also is unimportant except over the mid-
foreshore where it makes up to 10%of the surface sands (Fig.9). Its increased
frequency here is presumably because it can draw from material in the under-
lying storm deposits - the result, therefore, of a ‘source’ control.
The shore-normal trends - in the surface sands - therefore reflect the
passage of the fairweather tide across the foreshore: sorting by swash/surf
action across the upper and mid-foreshores, by shoaling waves across the
lower foreshore.
The subsurface sands show the added influence of storms over the mid-
foreshore. The shore-normal textural gradients in the subsurface are there-
fore disrupted by an abrupt coarsening across the mid-foreshore. I t appears
therefore that the upper foreshore sands are emplaced by swash and surf
action, the mid-foreshore by storm breakers, and the lower foreshore by
shoaling waves. Since these subsurface sands are the accumulating beach
deposits, this subsurface trend is obviously of the most geological signifi-
cance. The grain-size curves are distinctive but hardly individually diagnostic
of the beach environment. But the shore-normal, ultimately vertical, sequence
of textures may be a useful ‘indicator of ancient beaches (Fig. 13).
Bedforms and sedimentary structures
The foreshore surface changes in appearance depending on conditions.

The shore-normal sequence commonly begins with a flat surface and/or
antidunes (length = 0.5 m, height <0.02 m ) on the upper part of the profile,
followed further down by small near-symmetrical ripples, then small asym-
metric landward-facing ripples across the lower part (length N 0.05 m,
height <0.05 m). After storms the sequence consists almost entirely of flat
beds with small asymmetric ripples near low water mark. Both sequences
indicate a seaward-diminishing flow regime. Larger-scale bedforms, like
those reported by Clifton et al. (1971) on Oregon beaches and by Hawley
(1982) at Rhosilli beach (also in Carmarthen Bay), have not been observed
on the exposed foreshore under any conditions.
138
Upper foreshore
Subsurface structures vary consistently across the foreshore. Plane bedding
is ubiquitous on the highest part of the upper foreshore (Fig.lOA). Whole
shells of subtidal organisms are sometimes embedded in the sands but do
not much disturb the laminations. Postdepositional cavity or bubble struc-
tures are as commonly developed as the plane beds they disrupt (Fig.lOB).
To seaward, the upper foreshore displays a greater variety of structures.
Plane bedding always characterises the near-surface layers, but both small
and large-scale cross bedding may occur below. The small-scale sets are
<0.02 m in thickness, festoon-shaped (Fig.10D and F) and frequently ob-
scured by bubble cavities (Fig.10C and D). The large-scale sets (= 0.05 m
thick) are both planar and trough bedded. Both scales of cross beds always
have a landward-dipping component (Fig.lOC and E). While the small-scale
sets are usually parallel to the strand, the large-scale sets have longshore com-
ponents with dips to both east and west (Fig.lOD). Bioturbation sometimes
destroys the primary bedding after long calm spells but is confined to the
subsurface layers (Fig.lOD, E and F). These upper foreshore structures are
found shoreward of the high water breaker zone and must therefore form in
the surf or swash zones. The plane beds are clearly swash zone products
and are typical of swash/backwash flows (Clifton, 1969). The cross beds
probably form in the surf zone and the large-scale sets probably indicate
surf-generated longshore currents. However, the megaripples which generate
the large-scale structures are never exposed by the ebbing tide and must be
washed out by swash and backwash.
Mid-foreshore
This is invariably dominated by shelly traction clogs arranged as either
plane beds or large-scale cross beds (Fig.ll). The cross beds, usually planar,
and up t o 0.08 m in thickness, have both landward- and seaward-facing sets
(Fig.llA, C and E) and longshore dips t o both east and west (Fig.llB and
D). Sharp erosional contacts are characteristic. These are the presumed
storm breaker/surf deposits mentioned above and are analogous to the struc-
tures formed at the break-point of microtidal beaches. Whether the bimodal
cross sets represent variable surf-zone current velocities or superimposed
reversing tidal currents or both is not yet known. The large-scale bedforms
that must be responsible for these structures are not to be seen on the
exposed foreshore. Reworking of the surface layers on the tide ebbs can
give rise to the plane beds of the swash zone (Fig.11A). Or the surface layers
can be remoulded t o landward-dipping small-scale cross beds (0.01-0.02 m
thick) during low flow regime phases (Fig.llC, note preserved ripple form).
The latter do not appear t o be swash or surf zone features but instead form
under the shoaling waves seaward of the breaker and then retain their
identity despite the ebbing surf and swash. Bioturbation modifies, though
generally does not totally obliterate, the primary structures during prolonged
calms (Fig.11, all cores).
139
Fig. 10. Box cores from the upper foreshore. Left-hand column, shore-normal. Right-
hand column, shore-parallel. Scale in em.
Fig.11. Box cores from the mid-foreshore. Left-hand column, shore-normal. Right-hand
column, shore-parallel. Scale in cm.
141
Lower foreshore
The lower foreshore sands almost always exhibit small-scale, trough
cross beds (usually <0.02 m thick) a t and near the surface (Fig.12A and B).
They are landward-facing structures, usually parallel to the strand. Climbing
ripples are not uncommon and foresets sometimes develop offshoots which
pass the troughs and reach adjacent flanks (Fig.12A and C). These structures
apparently develop under the asymmetric oscillatory flows seaward of the
breaker; similar structures are to be found a t the mid-foreshore surface dur-
ing calm periods (see above). The asymmetric ripples are sometimes just
modified by the ebbing backwash to give a form-discordant internal struc-
ture (Fig.12A). Rather larger cross sets of thickness 0.04 m and with a
pronounced longshore dip t o either east or west (Fig.12B and D) can occur,
and this must be an indication of longshore currents and/or tidal flows.
Rarely, the upper part of the lower foreshore is free of small ripples and then
the upper layers are plane bedded (Fig.12F). The couplets of small ripple
cross beds and planar truncation surfaces near the top of Fig.12F may be
either storm- or tide-induced. Few primary structures are preserved more
than 10 cm below the beach surface across the lower foreshore. Benthos
burrows and small shells can be distinguished and frequently the sands are
completely bioturbated (Fig.12, especially 12E). Both the physical and the
biogenic structures of the lower foreshore would suggest a dominance of
subtidal rather than intertidal processes.
Shore-normal sequence of structures
The shore-normal sequences of bedforms and sedimentary structures show

that, while swash and surfibreaker processes control the upper and mid-
foreshore, respectively, subtidal shoaling waves dominate the lower fore-
shore (which is therefore better described as a low tide terrace rather than as
a beach). The structural suite contains elements of day-to-day activity (con-
fined mostly to the surface layers) but is dominated by extremes of activity
(plane bedding, large-scale cross bedding) and inactivity (bioturbation). A
prograding sequence (Fig.13) should produce a succession that commences
with very fine grained sands mostly bioturbated but with some small-scale
cross bedding, passes up through coarser grained, cross-bedded sands also
partially bioturbated, and ends with fine-grained sands either plane bedded
or structureless (from bubble cavities). Small-scale ripple bedding may have
low preservation potential since it is destroyed both by hydraulic processes
during storms and by biological processes during calms.
Bedforms and flow regime
Under steady, unidirectional flows the bedforms of a sandy bed undergo

a sequential transformation as the fluid power increases (Simons et al.,
1965). For fine sands the sequence is: no movement-small ripples-
dunes - plane bed. The bedforms arranged on a mesotidal foreshore were
142
Fig.12. Box cores from t h e lower foreshore. Left-hand column, shore-normal. Right-hand
column, shore-parallel. Scale in cm.
143
Shoreward
c- ~~~
(
Om
Upper
foreshore
Mid
f o r e s ho r e
5m
- 2 - 1 0 I 2 3 4
0 units
0 units 0 units
Fig.13. Sedimentary facies of a prograding macrotidal foreshore.
accordingly qualitatively assessed by Clifton e t al. (1971) in terms of a flow

regime model. More recent flume studies (Southard, 1975) suggest that, for
very fine sands, the dune stage is left out. In Fig.14 we have plotted some
swash and backwash flow data from Pendine Sands on the depth-velocity
diagram given by Southard (1975). Note that we have used maximum in-
stantaneous velocities and not the mean velocities of the original flume
data. Clearly there is some accord between the Pendine and the flume bed-
form sequences: diminishing swash zone velocities and depths as the tide
ebbs leave plane beds and backwash antidunes on the upper foreshore and
small ripples on the lower; movement ceases altogether near the low water
mark.
Obviously, a macrotidal beach and a flume differ in many major respects.
The nearshore zone is subjected t o interacting flows of several kinds: sym-
metrical oscillatory flow, asymmetrical oscillatory flow and unidirectional
flow (Clifton, 1976; Davidson-Arnott and Greenwood, 1976); a macrotidal
shoreline is also subjected to reversing tidal flow. Furthermore, on a
macrotidal foreshore, a sequence may not start from a flat bed stage with no
144
F l o w v e l o c i t y . crnls
Fig. 14. Depth-velocity bedform diagram (after Southard, 1975) with Pendine data super-
imposed.
movement. As the tide sweeps the foreshore, a dynamic zone will be be-
queathed a bed configuration from the preceding zone and hence the flow
regime and bedform will be temporarily out of phase; the bed phase will in
turn impose a rhythmic roughness and create a flow separation. Thus the
unidirectional flow model correctly predicts ripples over much of the mid
and lower foreshores but in fact the ripples are formed not by the swash/
backwash but by shoaling waves. While the swash zone flows may be com-
petent t o form ripples, they do not do so because the ripples are already
there; instead the backwash may superficially modify the existing bedforms
t o give form-discordant internal structures. Such a bedform disequilibrium
must be common during storms when megaripples/dunes create the large-
scale cross beds in the coarser sands of the mid-foreshore.
Moreover, such is the speed of lateral migration of the tide across the
foreshore (an average 4 cm 8' on a mean spring) that it seems likely that
the growth of bedforms may lag so far behind flow conditions that the
appropriate bedforms may not form at all. A particular flow regime sequence
may therefore be incomplete.
There is very little information on disequilibrium bedforms (except Allen,
1973, 1974; Lofquist, 1978), and the flow regime beach model of Clifton
et-al. (1971) may not be appropriate on a macrotidal shoreline. But there
seems little doubt that the upper foreshore plane beds represent Clifton's
inner planar fucies which is typical of swash zone activity (Clifton, 1969;
Davidson-Arnott and Greenwood, 1974). The landward-facing cross beds
which sometimes contribute to the upper foreshore have something in
common with structures occasionally observed by Clifton in his inner rough
fucies. We have still to see the larger storm-generated bedforms which produce
145
the mid-foreshore cross beds, but tentatively assign these to the outer rough
fucies as the structures are similar to those described by both Clifton et al.
(1971) and Davidson-Arnott and Greenwood (1976) in the breaker zone.
The lower foreshore small-scale cross bedding is clearly the asymmetric
ripple fucies. The plane beds that can occur from time to time on any part
of the foreshore may be the result of symmetrical or asymmetrical oscillatory
flow, or unidirectional flow, or disequilibrium effects.
The sedimentary facies of Pendine Sands are therefore recognisably similar
to those elsewhere on microtidal and mesotidal beaches and are apparently
established more by waves than by tidal currents. However, without further
observation it would be imprudent to attempt a more detailed flow regime
interpretation , especially if some of the structures are generated by disequilib-
rium bedforms.
MORPHODYNAMIC MODEL
A more complete understanding of these particular sedimentological

associations and variations on the macrotidal profile depends upon a com-
prehension of the processes which lead to this sedimentary accumulation
and also of the relationships between these processes and the local tide
and wave regimes.
The levelling data have highlighted the longer-term morphological equilib-
rium of the beach and yet the bedform survey has revealed quite consider-
able disturbance on a shorter time scale. A simple, and very general,
morphodynamic model will now be developed and tested against the field
data in an attempt t o reconcile these observations and to clarify the im-
portant processes.
Definition o f the flow ratios
Three ratios may be employed to characterise the dynamics of the beach

system. Consider that, at a point on the bed beneath a single wave, the
onshore flow has an effective mean velocity uin which persists for a time
ti, and transports sediment at a rate j , producing a net shoreward sediment
mass transport of Jh. Define the corresponding parameters for the subse-
quent offshore flow as u e x , t,,, j , , and J,,, respectively. The ratio of the
flow velocities is the ‘velocity ratio’, V,:
Velocity ratio = V, = uin/u,, (1)
This is analogous to Kemp’s (1975) ‘velocity magnitude asymmetry’. A large
velocity ratio indicates that the onshore flow exceeds the offshore flow, a
value of unity indicates that the two are symmetrical, and a value of less
than one indicates that the offshore flow exceeds the onshore. The flow
durations are characterised by the ‘duration ratio’, D,:
Duration ratio = D, = t&,, (2)
146
This is analogous to Kemp’s (1975) ‘velocity time asymmetry’. The two

ratios may be usefully combined by neglecting the effect of residual shore-
normal flow components so that continuity of water allows the equation of
the onshore and offshore discharges. That is uhth = ueXtexso that:
D, = l/V, (3)
The ratio of the net sediment mass transports is defined as the ‘transport
ratio’ J,:
Transport ratio = J , = Jin/Jex (4)
This ratio will now be used to develop a simple beach equilibrium concept
which will be tested with Pendine Beach data.
Equilibrium sediment dynamics
A state of morphological equilibrium can result from either of two general

conditions, known as ‘zero transport’ and ‘zero net transport’ (Hardisty,
1981). The former state, wherein flows remain below threshold values, so
that no sediment is moving, has been applied to beaches in the ‘null point
hypothesis’ (for example, Johnson and Eagleson, 1966). The latter state,
wherein equal amounts of sediment enter and leave the system, was applied
theoretically to beaches by Bagnold (1963), Inman and Frautschy (1966),
Bowen (1980) and Bailard and Inman (1981). This more promising approach
t o the problem can be employed with the flow ratios defined above to
present a simple general beach model.
Zero net transport occurs on the beach when the quantity of sediment
carried shoreward by the wave, Jh, is precisely balanced by the quantity
carried seaward by the returning flow, J,,. The transport ratio (eq. 4) in
this equilibrium condition is then equal to unity:
J, 1
= Jin/Jex = (5)
Swash zone measurements were collected throughout a number of tidal
cycles on the profile at Pendine to test this equilibrium concept. A differen-
tial bedload trap (Fig.15) was buried in the beach and the top was smoothed
level with the sediment surface. After submergence by the rising tide the lid
was removed and, after the passage of a single wave, the trap was removed.
The net onshore sediment transport and the net offshore sediment transport
were thus collected separately and later dried and weighed. These data (Fig.
16) show that the beach was close to the equilibrium condition defined by the
transport ratio model. The scatter is probably due to the fluctuating nature
of the flows from one wave to another, highlighting the need for longer-
term measurements.
These experimental results show that the foreshore was in short term
equilibrium during the period of study. Our levelling data (discussed above)
suggest that the foreshore (at this western end of the barrier) is also in
long-term equilibrium with the nearshore flows. However, the sedimentary
147
On shore ,Swinging vane
,Swash trap
Fig.15. Bedload trap.
structures of the upper and mid-foreshore show that this is indeed a ‘zero
net transport’ dynamic equilibrium and that considerable disturbance of the
beach sand does occur under certain conditions.
Equilibrium beach slope
The transport ratio equilibrium concept outlined and tested above can be
extended to relate the surface slope of the beach sediments to the measure
of the flow asymmetry provided by the velocity ratio. For this a relationship
between the flow speeds and the sediment transport rate is required.
Bagnold’s (1963,1966) equation has proven useful (Langhorne, 1982) and
more accurate than others in the marine environment (Heathershaw, 1981).
The equation correlates the bedload transport rate with the cube of the
flow velocity (Hardisty, 1983;Greenwood and Mittler, 1984,this volume).
The onshore, upslope transport rate (Bagnold, 1963)is:
j, =k uh3/(tan 4 + tan a)
where 4 and a are the angle of internal friction and the beach gradient,
148
J, , kg. wt / m
Fig.16. Measured bedload transport Ji, and J,,.
respectively, and h is an empirical constant. The (tan 4 + tan a) term sen-

sibly reduces the transport rate with increasing bed slope to the limiting
value of a = @, at which the bed fails. The net onshore mass transport, Ji, =
jinti, is therefore:
Ji, = h u&th/(tan 4 + tan a)
Similarly the offshore, downslope transport rate (Bagnold, 1963) is:
j,, = k u$/(tan # - tan a)
which sensibly increases the rate with increasing downslope gradient to the
limiting value of a = at which the bed fails and offshore transport con-
tinues regardless of the flow. The net offshore transport is therefore:
Jex= h u:,t,,/(tan @ - tan a)
These two net transport equations are potentially very useful for beach
investigations but an assessment of the calibration coefficient k is presently
difficult due to the paucity of published beach transport data. The limited
measurements reported here do, however, yield a value for the coefficient of
about 10 kg m-4 s - ~when the appropriate units are used and uh and u,, are
taken as the mean flow velocities.
Combining these two formulations the transport ratio (eq. 4)becomes:
h uin 3ti, (tan 4 - tan a)
J, =
k u,, t,, (tan @ + tan a)
Substituting for V , (eq. 1)and D, (eqs. 2 and 3) yields:
149
J, = (tan @ - tan a) v,'/(tan 4 + tan a)

The zero net transport equilibrium slope, tan a,, which was discussed above
as occurring when J, = 1,is given by solving this equation for tan a as:
t a n @(V,' - 1)
-
EquiIibrium slope, tan a, =
(V," + 1)
This relationship is plotted in Fig.17 for a range 01 values of the velocity
ratio showing that the more asymmetric the flow then the steeper the
equilibrium gradient that is required to maintain a balance between the
onshore and the offshore sediment transports.
Furthermore the diagram appears to correctly display the self-stabilising
nature of such systems. Two unstable situations may be identified. Firstly
the region above the equilibrium line where the bed gradient is too steep and
the velocity ratio is too low. Here the higher offshore velocity combines with
the steeper slope to increase the offshore sediment transport, moving material
down the beach and thus flattening the profile until equilibrium is re-
established. Alternatively within the region below the equilibrium line, the
higher onshore velocities and flatter slopes combine to increase the onshore
transport thus steepening the profile gradient until again equilibrium is re-
established. In morphodynamic terms the beach profile is therefore self-
stabilising and any perturbation from the equilibrium induces a response
which opposes the change and returns the beach toward the equilibrium
condition.
The onshore and offshore flow velocities were monitored at various posi-
tions along the profile t o test these slope predictions. The flow velocities
were used to calculate the velocity ratio (eq. 1)and thence the theoretical
equilibrium slope (eq. 6). These theoretical values are compared with the
actual beach slopes from the levelling data as shown in Fig.18. There is
clearly considerable agreement between the theoretical and observed values
for this limited data set.
The model suggests that the beach gradient steepens across the profile
Fig.17. Equilibrium slope as a function of the velocity ratio.

150
Measured tan a
Fig. 18. Theoretical and measured equilibrium beach slopes.
from low to high water because the flow velocities become more asymmetric.
This is in line with an earlier argument (Hardisty, 1981)where the inherent
asymmetry of the Stokes Wave Theory was preferred to the symmetry of the
Airy Wave Theory and was related to the occurrence of different breaker
types. At Pendine the rising tide submerges steeper sections of the profile;
this produces a gradation from dissipative spilling breakers, which have
relatively symmetrical onshore and offshore flow components and hence low
sediment surface gradients, over the lower foreshore, up to the relatively
asymmetric flows and hence steeper beach gradients associated with the
narrow surf zones and plunging breakers over the upper foreshore. The
breaker variations caused by the macrotidal range at Pendine result in in-
creasingly asymmetric flows and equilibrium beach gradients toward high
water. These in turn produce the shore-normal changes in sedimentological
characteristics described earlier.
CONCLUSIONS
The foreshore of Pendine Sands is exposed to wind-generated waves and

swell and is subject to a large tidal range. The sedimentological characteris-
tics of this macrotidal beach are comparable to those of mesotidal and
microtidal beaches but there are important differences. These differences
arise partly because the daily to-and-fro sweep of the tide across the fore-
shore causes variations in breaker, surf and swash zone characteristics and
energy dissipation. Furthermore, the large tidal range creates a variable
shoaling modification of unbroken nearshore waves such that the breaker
height can change significantly during the tidal cycle. The result is that, as
the tide ebbs, the breaker height decreases, swash and surf zones widen, the
surf zone becomes increasingly dissipative and swash zone velocities diminish.
These changes are related to breaker type, beach slope, and swash/backwash
velocity asymmetry. The plunging breakers and narrow surf zone at high
water generate relatively asymmetric swash zone flows and are associated
151
with a steeper beach gradient while the spilling breakers and dissipative surf
at low water produce more symmetrical swash zone flows and hence a gentle
beach gradient.
The important sedimentological consequence of these tide-induced modi-
fications of wave-induced currents is that the hydraulic flow regimes of the
nearshore circulation diminish as the tide ebbs. This is reflected in the
overall seaward-fining textural gradient of the primary framework popula-
tion of the surficial sands. In more detail, the shore-normal framework
gradient has three components: slight seaward-coarsening (upper foreshore);
rapid seaward-fining (mid-foreshore); slight seaward-fining (lower foreshore).
These gradients are established by the high water swash/surf, the ebbing
swash/surf, and nearshore shoaling waves, respectively. There is a corre-
sponding change down the beach profile from plane beds to small ripples.
The plane beds are a product of unidirectional swash zone flows, the small
ripples of asymmetric oscillatory flows in the wave build-up zone.
Subsurface sedimentary structures reflect this tidal variation with plane
bedding characteristic of the upper foreshore and landward-dipping small-
scale cross bedding of the lower foreshore. The upper foreshore structures
are frequently modified by bubble cavities, the lower foreshore structures
destroyed by bioturbation. Variations in wave energy result in cross-bedding,
both large- and small-scale, associated with longshore currents, across the
lower part of the upper foreshore. The mid-foreshore subsurface is distin-
guished by coarse shelly and lithic traction clogs arranged as plane beds and
both seaward- and shoreward-inclined large-scale cross beds presumably
deposited under storm breakers and highly dissipative surf at high water still-
stand. The storm deposits become bioturbated during calms. These sequences
of sedimentary structures suggest that swash zone processes establish the
upper foreshore facies, storm breakers and surf the mid-foreshore facies, and
shoaling waves the lower foreshore facies. Whereas the texture of the surface
veneer of sediments displays a short-term equilibrium with foreshore dynam-
ics, the subsurface deposits reflect long-term responses t o both storm and
fairweather conditions. Consequently, the upper, mid and lower foreshore
facies have distinctive, and diagnostic, textural and structural signatures. A
composite vertical section through a regressive sequence would be, from
bottom to top: shoreward-dipping small-scale cross bedding in very fine
sands, much bio turbated ; seaward- and shoreward-dipping, large-scale cross
bedding in shelly coarse and medium sands; plane bedded fine sands; aeolian
dune cross bedding. Such a sequence would be similar to that of a microtidal
shoreline but should be laterally and vertically more extensive.
Periodic levelling of beach profiles has shown that the foreshore is in long-
term equilibrium with prevailing meteorological and dynamical conditions.
A zero net transport model for the beach shows that, in morphodynamic
terms, the beach profile is self-stabilising in the short-term. The sedimentary
structures, however, indicate a depth of disturbance of at least 30 cm over
much of the foreshore. A long-term dynamic equilibrium is therefore not
incompatible with a significant short-term mobility.
152
ACKNOWLEDGEMENTS
The bulk of the beach profiling programme was carried out by C.F.J.
during the tenure of a research studentship, from the Natural Environmen-
tal Research Council, in the Department of Geology, Imperial College,
London. Dr. Graham Evans initiated the Carmarthen Bay programme, ad-
vised during the early part of this study, and critically reviewed an early
draft of the paper. John Malcolm, Peter Bush, and especially the late Peter
Gee, provided invaluable assistance in the field, as did many undergraduates
too numerous to name. Mr. F. Dewes drafted the figures and Mr. W.
Rowntree photographed the box cores. Everyone’s help is most gratefully
acknowledged.
NOTATION
Bo : breaker coefficient t: flow duration

H : deep-water wave height t,: flow duration, surf and swash
HI,, :mean height of highest one- U: mean flow velocity
third of the waves u* mean swash/backwash velocity
H,: breaker height e: surf scaling factor
h,: water depth a t break point a: beach slope, degrees
J: net bedload transport @: angle of repose of sand
j: bedload transport rate ( ) c : denotes contact component
L : deep-water wavelength ( )F: denotes framework component
1,: width of surf zone ( )I: denotes interstitial component
1,: width of swash zone ( )in:denotesswash
P : phase difference ( )ex: denotes backwash
T: wave period ( ) e : denotes equilibrium
Tl,lo:mean period of longest one-
tenth of the waves
REFERENCES
Allen, J.R.L., 1973. Phase differences between bed configurations and flow in natural
environments, and their geological significance. Sedimentology, 20: 323-329.
Allen, J.R.L., 1974. Reaction, relaxation and lag in natural sedimentary systems: general
principles, examples and lessons. Earth-Sci. Rev., 10: 263-342.
Bagnold, R.A., 1963. Mechanics of marine sedimentation. In: M.N. Hill (Editor), The
Sea, Vol. 3. Wiley, New York, N.Y., pp.507-582.
Bagnold, R.A., 1966. An approach t o the sediment transport problem from general
physics. U.S.Geol. Surv., Prof. Pap., 422-1.
Bailard, J.A. and Inman, D.L., 1981. An energetics bedload model for a plane sloping
Blackley, M.W.L. and Heathershaw, A.D., 1982. Wave and tidal-current sorting of sand on
a wide surf-zone beach. Mar. Geol., 49: 345-356.
Bouma, A.H., 1969. Methods for the Study of Sedimentary Structures. Wiley, New York,
N.Y., 458 pp.
153
Bowen, A.J., 1980. Simple models of nearshore sedimentation: beach profiles and off-
shore bars. Proc. Conf. on Coastlines of Canada, Halifax, 1978, Can. Geol. Surv.,
Pap. 80-10: 1-11.
Cassie, R.M., 1954. Some uses of probability paper in the analysis of size frequency dis-
tributions. Aust. J. Mar. Freshwater Res., 5 : 513-522.
Clifton, H.E., 1969. Beach lamination- nature and origin. Mar. Geol., 7: 553-559.
Clifton, H.E., 1976. Wave formed sedimentary structures - a conceptual model. In: R.A.
Davis and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SOC.Econ.
Paleontol. Mineral., Spec. Publ., 24: 126-148.
Clifton, H.E., Hunter, R.E. and Phillips, R.L., 1971. Depositional structures and processes
in the non-barred high energy nearshore. J. Sediment. Petrol., 41: 651-670.
Darbyshire, M., 1963. Wave measurements made by the National Institute of Oceanog-
raphy. In: R.C. Vetter (Editor), Ocean Wave Spectra. Prentice-Hall, Englewood Cliffs,
N.J., pp. 258-291.
Davidson-Arnott, R.G.D. and Greenwood, B., 1974. Bedforms and structures associated
with bar topography in the shallow-water wave environment, Kouchibouguac Bay,
New Brunswick, Canada. J. Sediment. Petrol., 44: 698-704.
Davidson-Arnott, R.G.D. and Greenwood, B., 1976. Facies relationships on a barred
coast, Kouchibouguac Bay, New Brunswick. In: R.A. Davis and R.L. Ethington
Publ., 24: 149-168.
Draper, L., 1972. Extreme wave conditions in British and adjacent waters. Proc. 13th
Conf. Coastal Engineering, Vancouver, B.C., 1 : 157-165.
Evans, O.F., 1939. Sorting and transportation of material in the swash and backwash. J.
Sediment. Petrol., 9: 28-31.
Folk, R.L. and Ward, W.C., 1957. Brazos river bar: A study in the significance of grain
size parameters. J. Sediment. Petrol., 27: 3-26.
Galvin, C.J., 1968. Breaker type classification on three laboratory beaches. J. Geophys.
Res., 73: 3651-3659.
nearshore. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedi-
mentation in Wave-Dominated Coastal Environments. Mar. Geol., 60: 79-98 (this
volume).
Guza, R.T. and Inman, D.L., 1975. Edge waves and. beach cusps. J. Geophys. Res., 80:
2 99 7- 30 12.
Hardisty, J., 1981. Sediment dynamics on the intertidal profile. Unpubl. Ph.D. thesis,
University of Hull, Hull, 194 pp.
Hardisty, J., 1983. An assessment and calibration of formulations for Bagnold’s bedload
equation. J. Sediment. Petrol., 5 3 : 1007-1010.
Hawley, N., 1982. Intertidal sedimentary structures on macrotidal beaches. J. Sedi-
ment. Petrol., 52: 785-795.
Heathershaw, A.D., 1981. Comparison of measured and predicted sediment transport
rates in tidal currents. Mar. Geol., 42: 75-104.
Hydraulics Research Station, 1978. Coast erosion at Pendine: a wave refraction study.
Rep. No. EX 809, Wallingford, 22 pp.
Ingle, J.C., 1966. The Movement of Beach Sand. Elsevier, Amsterdam, 211 pp.
lines. Proc. Spec. Conf. Coastal Engineering, Am. SOC. Civ. Eng., Santa Barbara, Calif.,
pp. 511-536.
Jago, C.F., 1980. Contemporary accumulation of marine sand in a macrotidal estuary,
Southwest Wales. In: A.H. Bouma, D.S. Gorsline, C. Monty and G.P. Allen (Editors),
Shallow Marine Processes and Products. Sediment. Geol., 26: 21-49.
Jago, C.F. and Barusseau, J.P., 1981. Sediment entrainment on a wave graded shelf,
Roussillon, France. In: C.A. Nittrouer (Editor), Sedimentary Dynamics of Continental
Shelves. Mar. Geol., 42: 279-299.
154
Johnson, J.W. and Eagleson, P.S., 1966. Coastal processes. In: A.T. Ippen (Editor),
Estuary and Coastline Hydrodynamics. McGraw-Hill, New York, N.Y ., pp.405-493.
Kemp, P.H., 1961. The relationship between wave action and beach profile characteris-
tics. Proc. 7th Conf. Coastal Engineering, Am. SOC.Civ. Eng., pp.262-277.
Kemp, P.H., 1975. Wave asymmetry in the nearshore zone and breaker area. In: J. Hails
and A. Carr (Editors), Nearshore Sediment Dynamics and Sedimentation. Wiley,
London, pp.47-68.
King, C.A.M., 1951. Depth of disturbance of sand on beaches by waves. J. Sediment.
Petrol., 21: 131-140.
N.J., 429 pp.
Langhorne, D.N., 1982. A study of the dynamics of a marine sandwave. Sedimentology,
29: 571-594.
Lofquist, K.E.B., 1978. Sand ripple growth in an oscillatory-flow water tunnel. C.E.R.C.,
Tech. Pap. 78-5.
Middleton, G.V., 1976. Hydraulic interpretation of sand size distributions. J. Geol., 8 4 :
4 05-4 2 6.
Moss, A.J., 1972. Bed load sediments. Sedimentology, 18: 159-219.
Otvos, E.G., 1965. Sedimentation-erosion cycles of single tidal periods on Long Island
Sound beaches, J. Sediment. Petrol., 35: 604-609.
Short, A.D., 1980. Beach response to variations in breaker height. Proc. 17th Int. Conf.
Coastal Engineering, Am. SOC.Civ. Eng., Sydney, pp. 1016-1035.
Simons, D.B., Richardson, E.V. and Nordin, C.F., 1965. Sedimentary structures gen-
erated by flow in alluvial channels. In: G.V. Middleton (Editor), Primary Sedimentary
Structures and their Hydrodynamic Interpretation. SOC.Econ. Paleontol. Mineral.,
Spec. Publ., 1 2 : 34-52.
Southard, J.B., 1975. Bed configurations. In: Depositional Environments as Interpreted
from Primary Sedimentary Structures and Stratification Sequences. SOC.Econ. Paleon-
tol. Mineral., Short Course No. 2, Dallas, Texas, pp.5-44.
Williams, A.T., 1971. An analysis of some factors involved in the depth of disturbance of
beach sand by waves. Mar. Geol., 11: 145-158.
dynamics of reflective and dissipative beach and inshore systems: Southeastern
Australia. Mar. Geol., 32: 105-140.
Marine Geology, 60 (1984)155-164
155
Elsevier Science Publishers B.V., Amsterdam -Printed in The Netherlands
HIGH-FREQUENCY SEDIMENT-LEVEL OSCILLATIONS IN THE SWASH

ZONE
ASBURY H. SALLENGER, Jr. and BRUCE M. RICHMOND

U S . Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025 (U.S.A.)
(Received June 14, 1983, revised and accepted January 16, 1984)
ABSTRACT
Sallenger Jr., A.H. and Richmond, B.M., 1984. High-frequency sediment-level oscilla-
tions in the swash zone. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydro-
dynamics and Sedimentation in Wave-Dominated Coastal Environments. Mar. Geol.,
60: 155-164.
Sediment-level oscillations with heights of about 6 cm and shore-normal lengths of

order 10 m have been measured in the swash zone of a high-energy, coarse-sand beach.
Crests of oscillations were shore parallel and continuous alongshore. The oscillations were
of such low steepness (height-to-length ratio approximately 0.006) that they were dif-
ficult to detect visually. The period of oscillation ranged between 6 and 15 min and de-
creased landward across the swash zone. The sediment-level oscillations were progressive
landward with an average migration rate in the middle to upper swash zone of 0.8 m
mini. Migration was caused mostly by erosion on the seaward flank of the crest of an
oscillation during a period of net seaward sediment transport. Thus, the observed migra-
tion was a form migration landward rather than a migration involving net landward sedi-
ment transport. The observed sediment-level oscillations were different than sand waves
or other swash-zone bedforms previously described.
INTRODUCTION
Most previous studies on beach-profile changes have focussed on changes

measured at intervals of weeks, days, or hours. We know of only one
previously published study (Waddell, 1973) that focussed on high-frequency
changes, changes measured at intervals approaching the swash period.
On a medium-sand low-energy beach, Waddell (1973) measured sediment
level at two locations in the upper half of the swash zone after the backwash
of each wave swash. He found significant oscillations of sediment level with
periods of 40 s and longer and presented evidence that the sediment-level
oscillations had some characteristics of sand waves. Discussing the same data,
Waddell (1976) hypothesized that the bed oscillations were caused by
ground-water oscillations of the same frequency. The rising and falling ground
water would cause areas to erode and accrete similar t o the ground-water
control over tidal cycle sedimentation discussed by Duncan (1964). When
the water table is high, swash infiltration into the beach is relatively low and
erosive backwash is enhanced. When the water table is low, backwash is
0025-3227/84/$03.00 Q 1984 Elsevier Science Publishers B.V.

156
diminished due t o swash infiltration into the beach and accretion results.
The low frequency ground-water oscillations were thought to be caused by
waves of the same frequency in the surf zone.
In the present study, we measured changes in sediment-level at numerous
locations across the swash zone of a coarse-sand high-energy beach. We will
show that sediment level changed in a surprisingly rapid and well-organized
manner. We found sediment-level oscillations which were low-amplitude,
landward-progressive, and had a unique mode of migration. We will show
that these oscillations had characteristics different than sand waves or the
several types of swash-zone bedforms previously described in the literature.
The processes of formation are presently unclear, although we point out that
ground-water oscillations could not explain a critical characteristic of the
observed sediment-level oscillations.
EXPERIMENT DESCRIPTION
Our experiment was conducted during January 1981 at Fort Ord,

California. Fort Ord is located on the shore of southern Monterey Bay about
150 km south of San Francisco. Average foreshore slope during our study
was 7.5". The foreshore was composed of coarse sand with amedian diameter
of 0.8 mm. During the experiment, waves were of normal incidence, breaker
heights were 3.0-4.0 m, and the surf zone width was approximately 100 m.
Wind speeds were low during the experiment although local strong winds
associated with squalls occurred both before and after the experiment. Low-
energy sea waves associated with the passing squalls were present, but inci-
dent waves were dominantly of the swell type with a period of about 16 s.
Stakes, 1 cm in diameter and 2 m long, were driven into the swash zone at
the locations shown in Fig.1. The shore-normal array of seven stakes had 3-m
spacings between stakes and the shore-parallel array of five stakes had 4-m
spacings. Stake numbers for the shore-normal array refer t o the distance
from the landward stake (for example, the landward stake is called 0 and the
stake 15 m seaward is called 15). Stakes in the shore-parallel array are refer-
enced by the letters indicated on Fig.1.
Stake heights above the bed, initially about 0.6 m, were measured using
modified meter sticks. A circular base plate (15-cm diameter) was attached
to one end of a meter stick to inhibit settling into the bed. A movable pointer,
free to slide over the length of the meter stick, was used to determine
stake heights. Repeated measurements of a stake not exposed to the swash
showed that the accuracy of the technique was 1-2 mm. In the upper swash
zone, the accuracy of our measurements approached that of our test case.
In the lower swash zone, where measurements needed to be made rapidly
and the base plate tended t o sink into the bed more than in the upper swash
zone, accuracy was about 5 mm.
During our experiment, stake heights were measured after a backwash
when the swash zone was subaerial. Heights were read t o the closest milli-
meter. Three persons measured stake heights and one person recorded the
157
0 9
0 78
0 75
0 9
0 6
/
’2
03
L A N D W A R D STAKE 0 0 -
0
METERS
3
w
3---
Fig.1. Stake locations and foreshore contours. Contour interval is 1 m; the vertical datum
is arbitrary.
stake-height data and time. Persons measuring stakes were careful to stand
as far as possible in a longshore direction away from the stakes. Scour holes
caused by persons standing in the swash did not extend t o the stakes. The
stakes were of sufficiently small diameter, 1cm, to prevent significant scour
around themselves. Stakes in the lower swash zone were measured on the
average every 40-50 s or about every third swash. Stakes in the extreme
upper swash zone were measured only after a swash reached their location;
they were not reached by every wave swash. Stakes were measured for a
period of about 70 minutes midway through a flooding tide. Although the
diurnal tide range for the area is 1.6 m, the range from low to high tide on
the day of our experiment was 0.6 m. Tide range during our experiment was
0.09 rn which caused a 0.7 m landward shift of the swash zone.
The mean swash position was between stakes 12 and 15. Due to the
powerful swash, it would have been very difficult to obtain data lower in
the swash zone than stake 18. In fact, the person who measured stake 18
was frequently knocked down by the upwash. The maximum landward
excursion of upwash during our experiment was 1-2 m landward of the
landwardmost stake, stake 0. The seaward excursion of backwash was typical-
ly 3-fj m seaward of the seawardmost stake, stake 18, although some back-
washes reached as much as 9 m seaward of stake 18.
158
RESULTS
Significant sediment-level oscillations are apparent in the time series from

the shore-normal array (Fig.2). These oscillations showed a shift in frequency
across the swash zone. At the lower stakes (15 and 18) which were below
the mean swash position, the period of oscillation was roughly 6 min,
whereas at the middle stakes (9 and 12) period increased to 10-15 min.
Oscillations disappeared, or were not as apparent, in the extreme upper
swash zone.
A measure of the height of sediment-level oscillations is twice the stan-
dard deviation of a sediment-level time series with the linear trend removed.
For stakes in the seaward end and middle of the shore-normal array, heights
were uniformly about 5.5 cm (Fig.3 and Table I). Heights decreased signifi-
cantly to about 1.5 cm for stakes in the landward end of the swash zone.
Significant negative and positive elevation trends, indicating net changes
to the swash zone profile, are apparent in the records of Fig.2. The amount
of net vertical change due to these trends (over a period of 67.7 min) was
plotted versus distance offshore in Fig.3 (tabulated in Table I). The lower
swash zone underwent net erosion whereas the upper swash zone underwent
net accretion. The volumes, however, were not equal; the amount of erosion
was significantly greater than the amount of accretion.
Sediment-level data for each stake in the shore-parallel array are plotted in
Fig.4. Oscillations measured at stakes in the shore-parallel array appear to
be in phase. Heights and net sediment-level changes were computed for these
records as before. Heights ranged from 4.7 to 6.0 cm, but did not vary
systematically alongshore (Table I). The in-phase relationships among records
and the lack of systematic change in heights suggest the oscillations were
o - - :
-6
I T , 3 I 3 ! , I I I ,
0 10 20 30 40 50
TIME (minutes)
Fig.2. Time-series of sediment-level for the shore-normal array.
159
-151
i5j*
w
1 0 r
= /--
0 3 6 9 12 15 18
DISTANCE SEAWARD FROM LANDWARD STAKE (m)
Fig.3. In the upper plot, net elevation change (over 67.7 min) is plotted versus location
for the shore-normal array. Net changes were computed from the linear slope of each
record. In the lower plot, heights of sediment-level oscillations are plotted for the shore-
normal array.
TABLE I
Statistics on sediment elevation records

~~
Stake Net change* (cm) Heights** (cm)
0 4.5 1.3
3 6.5 1.4
6 0.5 1.7
9 9.0 6.0
12 16.0 6.0
15 15.0 6.8
18 10.2 5.2
12A 17.1 5.6

12B 17.4 5.2
12 16.0 5.0
12c 10.2 4.7
12D 8.5 5.5
~~ ~
*Computed over 67.7 min; **this is twice the standard deviation of the record with the
trend and mean removed.
160
0 10 20 30 40 50 60
TIME (minutes)
Fig.4. Time-series of sediment-level for the shore-parallel array.
shore parallel and continuous alongshore, at least over the 16 m length of

the shore-parallel array. However, net changes varied systematically along-
shore. The amount of erosion decreased significantly to the south along the
shoreline (Table I). The reason for this pattern of net change is unknown.
The 6-min oscillations apparent in records obtained at stakes 15 and 18
appear to be migrating landward (Fig.2). The cross-spectrum between
records at stakes 15 and 18 was calculated in order to confirm this visually
apparent landward migration. Unfortunately, in order to resolve 6-min peaks
we had to use such a narrow bandwidth to smooth raw spectral estimates
that little confidence could be placed in the resolved peaks. The landward
migration of longer period oscillations measured more landward in the
swash zone is confirmed below in another manner.
Figure 5 shows a time sequence of shore-normal profiles of bed elevations.
These profiles were plotted at one minute intervals using records with the
mean and linear trend removed. In essence, an individual profile shows the
difference in bed elevation from a mean profile at a given time; the mean
profile is horizontal and is defined by the means of all shore-normal records.
By plotting the profiles in a vertical time sequence, the shore-normal move-
ments of areas of erosion and accretion are made visible. Clearly, the plot
shows the crest of an oscillation that progressed landward across the swash
zone. The migration rate slowed in a landward direction; the mean rate of
migration was about 0.8 m min-'. These landward migrating features were
associated with the 10- t o 15-min oscillations of Fig.2; the higher frequency
oscillations lower in the swash zone were not well represented in Fig.4.
Figure 6 shows how a sediment-level oscillation migrates across the swash
zone. At 11 min, a distinct accretional area had developed at the seaward
end of the swash zone (Fig.6A). Within 4 min, a crest of an oscillation was
apparent in the midst of our shore-normal array. Excluding the initial land-
ward migration (between 15 and 18 min), the crest progressed landward
mostly by erosion of its seaward flank (Fig.6B). Note that the eroded
sediment was apparently transported seaward. Thus, the migration involved
more of a form migration landward than a net landward sediment transport.
161
. I
0 3 6 9 12 15 18
DISTANCE SEAWARD FROM
LANDWARD STAKE (m)
Fig.5. Time sequence of shore-normal profiles of sediment-level elevations. An individual

profile shows the difference in bed elevation from a mean profile at a given time; the
mean profile is horizontal and is defined by the means of all shore-normal records. Linear
trends were removed from records prior to constructing these profiles.
T0 6 -- A
Y
.’-
LEGEND
z - _/c-
_ / - - - -11 min
20- , I ---13 min
I- 15 m i n
a - \
> \
,
w -
1
W-6-
LEGEND
15 m i n
_-- 18 m i n
-----21 min
2 4 min
.. . . . ., .. 27 m i n
2
w-61
0 3
I
6 9
, I
12
I
15 18
I
D I S T A N C E S E A W A R D F R O M L A N D W A R D S T A K E (m)
Fig.6. A. Shore-normal profiles of bed elevations showing how the crest of a sediment-
level oscillation develops. B. Shore-normal profiles of bed elevations showing how a crest
of a sediment-level oscillation migrates.
162
DISCUSSION
The type of sediment-level oscillation described above appears to be a

feature common to swash zones. We have repeated our experiment at the
same beach during flooding and ebbing tides and when the foreshore was
undergoing net erosion and net accretion. In all of these experiments the
results were similar, the features were always present and were always mi-
grating landward. We have also conducted two smaller scale experiments on a
different Monterey Bay beach. The beach was composed of much finer
sand (mean -0.3 mm) and was of gentler slope (-3.8') than the beach dis-
cussed above. The same type of feature appeared to be present in one of
the experiments. In the other experiment, conducted on a day of low
incident energy, the sediment level did not change appreciably. The reason
that the sediment-level oscillations are not better known is probably because
of their very low steepness making the oscillations difficult to detect visually.
The cross-shore length for an oscillation is order 10 m (Figs.6 and 6) and
height is about 6 cm (Fig.3) giving a steepness of 0.006.
The sediment-level oscillations described here are similar to those des-
cribed by Waddell (1973, 1976) although there are important differences.
Phase relationships between records at Waddell's two measurement locations
indicated that his oscillations were migrating offshore rather than the on-
shore migrations observed in this study. Oscillations measured by Waddell
had periods on the order of a minute whereas the oscillations observed in
this study had much longer periods, in the range 6-15 min. With only two
measurement locations, Waddell was unable to show, as we did in the present
study, whether the migration of his oscillations involved a net sediment
transport in the direction of migration.
The sediment-level oscillations were not sand waves, as suggested by
Waddell, or any of the various types of swash zone morphologic features
described previously. Sand waves generally have net transport in the direc-
tion of migration whereas the features observed here migrated landward
while sediment was being transported seaward (Fig.6). Since the oscillations
were landward progressive and had shore-normal length scales on the order
of the swash-zone width, they resembled small-scale swash bars. Swash bars
(called by some ridge and runnel topography) have been described by Davis
et al. (1972), Owens and Frobel (1977), and others. However, swash bar
migration involves a net landward sediment transport which is different from
the mode of migration of oscillations described here. Backwash ripples,
whose origin is discussed by Broome and Komar (1979), differ from the
sediment-level oscillations described here in that they do not migrate exten-
sively and have much shorter shore-normal length scales (50-70 cm).
When we first observed the net accretion that occurred in the upper
swash zone during our experiment (Fig.3), we thought that the landward
progressive oscillations may have contributed to the accretion. However, as
discussed above, the migration of the sediment-level oscillations does not
involve a net landward sediment transport. The observed accretion may have
163
been related to tidal cycle sedimentation; with a rising tide the upper swash
zone should accrete (see Duncan, 1964).
The origin of the sediment-level oscillations is not at all clear. As discussed
in the Introduction, Waddell (1976) hypothesized that the oscillations were
driven by low-frequency oscillations in ground water. Ground-water oscil-
lations were supposedly caused by waves of the same frequency in the surf
zone. However, this hypothesis cannot readily explain the observed change
in frequency of sediment-level oscillations across the swash zone. Low-
frequency waves may, however, be important. Using time-lapse photography,
we measured time series of runup. In low-pass filtered runup records, there
was some evidence of waves with periods of order 10 min. Such very low-
frequency waves may be shelf waves or, since we were working in Monterey
Bay, bay seiches. In a low-passed record, runup oscillations were apparent
of the same frequency and 180" out of phase with sediment-level oscilla-
tions at stake 12. However, the relationship was not conclusive. Low-
frequency waves may be important, but the exact processes are not clear.
CONCLUSIONS
Sediment-level oscillations with heights of 6 cm and shore-normal lengths

roughly the swash zone width were measured in the swash zone of a high-
energy, coarse-sand beach. The period of oscillation decreased across the
swash zone ranging from 6 min below the mean swash level to 10-15 min
above the mean swash level. The sediment-level oscillations were progressive
landward; the oscillations in the middle t o upper swash zone migrated at an
average rate of roughly 0.8 m min-'. The sediment-level oscillations differed
from sand waves and the other swash-zone bedforms previously described
in that only the form of an oscillation migrated landward. These results show
that in a high-energy environment, the swash-zone profile changed its con-
figuration in a surprisingly rapid and well-organized manner.
ACKNOWLEDGEMENTS
We thank Beth Laband, Jeff List, Bruce Jaffe, and Geof Caras for help in
the field and Beth Laband and Jeff List for help in data reduction.
REFERENCES
Broome, R. and Komar, P.D., 1979. Undular hydraulic jumps and the formation of back-
wash ripples on beaches. Sedimentology, 26: 543-559.
Davis, R.A., Fox, W.T., Hayes, M.O. and Boothroyd, J.C., 1972. Comparison of ridge
and runnel systems in tidal and non-tidal environments, J. Sediment. Petrol., 42:
4 1 3-421.
Duncan, J.R., 1964. The effects of water table and tide cycle on swash-backwash, sedi-
ment distribution, and beach profile development. Mar. Geol., 2: 186-197.
Owens, E.H. and Frobel, D.H., 1977. Ridge and runnel systems in the Magdalen Islands,
Quebec. J. Sediment. Petrol., 4 7 : 191-198.
164
Waddell, E., 1973. Dynamics of swash and implication to beach response. Coastal Studies
Institute, Louisiana State University, Baton Rouge, La., Tech. Rep. 139, 49 pp.
Waddell, E., 1976. Swash-groundwaterbeach profile interactions. In: R.L. Davis and
R.L. Ethington (Editors), Beach and Nearshore Sedimentation. Soc. Econ. Paleontol.
Mineral., Spec. Publ., 2 4 : 115-125.
Marine G e o l o g y , 6 0 ( 1 9 8 4 ) 165-198 165
Elsevier Science Publishers B.V.. Amsterdam - Printed in The Netherlands
WAVE-FORMED STRUCTURES AND PALEOENVIRONMENTAL

RECONSTRUCTION
H. EDWARD CLIFTON and JOHN R. DINGLER

U.S. Geological Survey, 345 Middlefield R o a d , Menlo Park, C A 94025 ( U . S . A . )
(Received September 15, 1983; revised and accepted January 14, 1984)
ABSTRACT
Clifton, H.E. and Dingler, J.R., 1984. Wave-formed structures and paleoenvironmental
reconstruction. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and
Sedimentation in Wave-Dominated Coastal Environments. Mar. Geol., 6 0 : 165-198.
Wave-formed sedimentary structures can be powerful interpretive tools because they

reflect n o t only the velocity and direction of the oscillatory currents, but also the length
of the horizontal component of orbital motion and the presence of velocity asymmetry
within the flow. Several of these aspects can be related through standard wave theories to
combinations of wave dimensions and water depth that have definable natural limits. For
a particular grain size, threshold of particle movement and that of conversion from a
rippled t o flat bed indicate flow-velocity limits. The ratio of ripple spacing t o grain size
provides an estimate of the length of the near-bottom orbital motion. The degree of
velocity asymmetry is related t o the asymmetry of the bedforms, though it presently
cannot be estimated with confidence. A plot of water depth versus wave height (h--H
diagram) provides a convenient approach for showing the combination of wave parameters
and water depths capable of generating any particular structure in sand of a given grain
size. Natural limits o n wave height and inferences o r assumptions regarding either water
depth o r wave period based o n geologic evidence allow refinement of the paleoenviron-
mental reconstruction. The assumptions and the degree of approximation involved in the
different techniques impose significant constraints. Inferences based o n wave-formed
structures are most reliable when they are drawn in the context of other evidence such as
the association of sedimentary features o r progradational sequences.
INTRODUCTION
Sedimentary geologists have long sought to use depositional structures for

interpreting ancient depositional environments. Quantitative analysis of
paleo-processes based on such structures has proved at best only partly
successful. Even where the structures can be related with reasonable precision
to ancient processes, those processes commonly cannot be meaningfully incor-
porated into a broader environmental interpretation. For example, the flow-
regime concept provides a comprehensive model for interpreting structures
produced by unidirectional flow (Harms et al., 1982). Nonetheless, even
where application of the concept generates specific data on such parameters
as flow velocity and water depth, it commonly is unclear how these param-
eters contribute significantly to a paleoenvironmental reconstruction.

166
In contrast, parameters interpreted from wave-generated structures

commonly can be linked clearly to important aspects of the depositional
setting. Water depth, for example, which is of somewhat uncertain influence
on the development of structures produced by unidirectional flow, bears in a
direct, calculable way on the origin and nature of wave-formed structures.
Until recently, wave-generated sedimentary structures were poorly under-
stood, and, consequently, they could not be used t o successfully interpret
depositional environments. Considerable data based on field and laboratory
experiments and observations have accumulated in recent years (for example,
Lofquist, 1978; Miller and Komar, 1980a, b; Dingler and Clifton, 1984, this
volume), and several interpretative models have been proposed based on
wave-generated structures (e.g., Allen, 1970; Tanner, 1971; Komar, 1974;
Clifton, 1976; Allen, 1979, 1980, 1982; Harms et al., 1982). Although the
relations between wave-generated structures and the associated fluid dyna-
mics are still not fully understood, useful interpretations are possible. This
paper summarizes the published research on wave-formed sedimentary struc-
tures and outlines a procedure whereby wave-formed structures can be used
to interpret ancient depositional environments. The procedure is presented
step by step, noting the physical basis for the parameters employed, and
assessing the validity of the various approaches and techniques. In conclusion
the procedure is applied to specific geological problems.
THE INTERPRETIVE PROCEDURE
The procedure of interpreting paleoenvironments from wave-generated

structures requires three discrete steps. The first involves inferring flow param-
eters from specific aspects of the wave-generated structures using results
from empirical investigations or experimental studies. The second step
employs wave theory to determine the combinations of water depth and
wave size and shape that could produce the inferred flow parameters. The
third step utilizes the natural limits that exist for waves, geologic reasoning,
or wave-hindcasting techniques to constrain the range of possible combina-
tions of water depth and wave size and to relate those that are feasible to the
paleoenvironment.
STEP 1. INFERRING FLOW PARAMETERS FROM WAVE-FORMED FEATURES
Wave-formed structures
Under wave action, the character of the flow and the composition of the
bed (texture and mineralogy) combine to determine the general configuration
of the bed (flat, hummocky, or rippled) and the size and shape of the bed-
forms themselves. Accordingly, aspects of size and shape of the bedforms
can be used to infer previously existing flow parameters, which can in turn
be applied to the interpretation of depositional environments.
Oscillation ripples are the predominant wave-generated bedforms. In profile,
spacing A, height and symmetry P / A characterize these ripples (Fig. 1).In
167
D i r e c t i o n of W a v e P r o p a g a t i o n I,
Fig.1. Geologically important parameters of waves, water motion and wave-formed ripples.
Wave length ( L ) is t h e horizontal distance between successive wave crests; wave height (H)
is t h e vertical distance between wave crest and trough; water depth ( h ) is t h e vertical dis-
tance from still water level t o the seafloor; orbital diameter ( d o )is the maximum horizon-
tal distance of excursion of water particles as a wave passes ( a circular motion in deep
water, an elliptical motion in shallow water); ripple spacing ( k ) is t h e average horizontal
distance between ripple crests; ripple height ( q ) is t h e average vertical distance between
ripple crests and troughs; ripple asymmetry ( p / h )is the ratio between t h e average distance
from ripple crest to leading trough ( p ) and the average ripple spacing (k). Not shown: the
wave period (T), the time required for successive wave crests to pass a given point; and
the maximum orbital velocity (urn), the maximum horizontal velocity in t h e direction of
wave passage.
plan, crest length relative to spacing and crest sinuosity are primary character-
istics; Inman (1957) called ripples short-crested, intermediate-crested, or
long-crested if their crest-length to spacing ratio was less than 3, 3-8, or
greater than 8, respectively. Crest pattern ranges from straight to sinuous; in
the extreme they can take on a crescentic shape, such as the lunate mega-
ripples of Clifton et al. (1971). Most oscillation ripples are transverse to the
forming current, but a few types such as the cross ripples described by Clifton
et al. (1971) are oblique to the flow.
The ratio of ripple height to wavelength v/X is the ripple steepness; it and
its inverse, the ripple index (Reineck and Singh, 1973) or vertical form index
(Bucher, 1919), have been used to describe ripples (e.g., Dingler, 1974; Allen,
1980). Allen (1980) indicates that a wide range of ripple indices (steepnesses)
is possible. Dingler and Inman (1977) showed that for fine sand near La Jolla,
California, ripple steepness remained at a value of about 0.15 with increasing
wave energy until, as sheet flow conditions were approached, the steepness
decreased systematically t o zero.
Symmetric ripples have a symmetry factor that approaches 0.5, or a
ripple symmetry index [(A - /3)/p; Reineck and Singh, 19731 that approaches
1.0. The steeper side of most asymmetric ripples faces in the direction of
ripple migration, making the symmetry factor less than 0.5 and the ripple
168
symmetry index greater than 1.0. The maximum value of the ripple symmetry
index acquired by asymmetric wave ripples is reportedly 3.8 (Reineck and
Singh, 1973, p.27), equivalent t o a symmetry factor of about 0.25.
The nature of the sediment that composes the bed is an important and
potentially troublesome factor. Several different aspects of texture or com-
position such as mean grain size (D), sorting, shape, and particle density can
bear on bedform development. The influence of mean grain size is fairly well
known (Clifton, 1976), butvery little is known about the effects of the other
three factors. It would seem likely, however, that a bed of coquina would
respond t o a given type of flow differently than would a bed of quartz sand
of similar mean grain size.
The recognition of structures as formed by waves is obviously critical to
their use as interpretive tools. Symmetric ripples are generally accepted a
priori as produced by wave activity, although the common presence of sym-
metrical ripples in deep-sea photographs (Heezen and Hollister, 1971, p.348)
suggests other possible mechanisms. Asymmetric bedforms generated by
waves may be difficult to distinguish from those formed by unidirectional
currents. Tanner (1967), Reineck and Wunderlich (1968), Boersma (1970),
and Reineck and Singh (1973) present criteria for recognizing wave-produced
bedforms.
The problem of identifying effects of waves is further complicated in expo-
sures where the bedforms themselves are poorly expressed. In such a case,
the influence of waves must be inferred, often with difficulty, from the
internal structure produced by migrating bedforms. Boersma (1970) and
Allen (1982) offer a number of criteria for recognizing wave-formed ripples
on the basis of internal structure. The orientation of the ripples can in some
cases suggest their origin. Because of the general absence of shoreward-flowing
currents, Clifton (1981) inferred that ripples that faced or migrated in a
shoreward direction were solely the product of waves.
Most of the expressed relationships between wave-formed structures, flow
parameters, and waves assume an absence of superimposed unidirectional
current (Clifton, 1976; Allen, 1981a). Yet in natural environments, combined
oscillatory and unidirectional flow, in the form of tidal, rip or longshore
currents is fairly common. A few studies have described combined flow
ripples (Reineck and Wunderlich, 1968; Harms, 1969; Bliven et al., 1977),
but presently they cannot be used with confidence in environmental inter-
pretation (Harms et al., 1982, pp.2-42). Distinguishing between purely
oscillatory and combined-flow ripples may be very difficult.
The identification of flat bedding produced by oscillatory sheet flow can
be particularly difficult. First, it may be impossible t o recognize the contri-
bution of superimposed unidirectional flow t o the development of sheet
flow conditions. Second, ripples that migrate across the seafloor may produce
a very similar, nearly flat stratification that is analogous to the climbing
translatent strata observed in eolian deposits (Hunter, 1977). Clifton (1976)
suggests several criteria (mostly based on lithologic association) that may
prove useful for distinguishing between sheet-flow laminae and subaqueous
climbing translatent strata.
169
Flow parameters
The flow parameters that can be inferred from wave-generated structures

are relatively straightforward. As a wave travels along the surface of the
water, it sets the water particles in motion (Fig.1). If the water depth ( h )is
large relative t o the length ( L ) of wave (that is, h > L/4), the wave form is
sinusoidal and the induced water motion is essentially circular (Fig.1). The
diameter of the circle ( d o ) diminishes exponentially with depth, reaching
zero above the bottom. In shallower water, where the wave interacts with
the bottom, the wave may retain its nearly sinusoidal shape, but the water
particles move in ellipses that become progressively flatter and smaller with
depth (Fig.1). Just above the sea floor the elliptical motion becomes a hori-
zontal oscillation, the length of which still is referred to as “orbital diameter”.
In very shallow water, just before breaking, the wave may lose its sinusoidal
shape, and the water motion is nearly horizontal throughout the water
column (Komar, 1976).
The velocity of the water particles, which is a critical parameter in the
threshold of grain movement and in the shaping of bedforms, depends both
on the magnitude of the orbital diameter and on the wave period ( T ) . For
deep-water waves (where water motion is circular), the maximum orbital
velocity (u,) equals the average orbital velocity, the circumference of the
orbital motion ( n d o ) divided by the time required t o complete an orbit (7’).
In shallower water, the maximum orbital velocity above the bottom differs
from the average velocity, but the relation, u, = r d o / T ,remains valid.
As a wave approaches the shore, its form changes (Fig.2) such that the
crest becomes increasingly narrow and peaked and the trough broad and flat.
As the wave begins t o break, it also becomes asymmetric about a vertical
plane through and parallel to the crest, because its landward face steepens
relative to its seaward face. These changes impart an asymmetry to the orbital
motion.
Part of the physical basis for this asymmetry can be seen in Fig.2. If mass
transport is assumed to be nil, the volume of water that moves forward under
the crest of a wave must equal that which moves in the opposite direction
under the trough. Because the crest of the wave is foreshortened relative to
DIRECTION OF W A V E PROPAGATION
c-
I .
* t +
Fig 2. Typical form of a wave as it passes into shallow water. Note that the time available
for movement of water in the direction of wave propagation under the crest of the wave
is substantially less than that available for movement in the opposite direction under the
wave trough. The result is a velocity-magnitude and velocity-time asymmetry whereby
the forward motion of the water under the crest is strong but of short duration relative to
the reverse motion under the trough.
170
the trough, water under the crest must move more rapidly to maintain mass
balance. This condition causes the bottom orbital flow under the crest to be
abrupt and strong relative to that under the trough. This onshore flow would
be further reinforced by any shoreward mass transport.
The resulting orbital velocity asymmetry can be viewed as consisting of
both a velocity-magnitude and a velocity-time component (Kemp, 1975).
Velocity-magnitude asymmetry, as used here, refers to any difference be-
tween the peak or maximum velocity under the crest and trough of the wave.
Velocity-time asymmetry refers to any difference between the duration of
flow in the direction of wave propagation and that in the opposite direction.
Figure 3 illustrates some of the conceivable velocity profiles that could be
generated by shoaling waves. It should be noted that, in nature, asymmetry
of flow is almost always due to a combination of velocity-magnitude and
velocity-time asymmetry and is therefore complex.
The amount of water mass transport generated by asymmetric orbital
motions seems variable and, under certain conditions, may be of minimal
importance. The character of flow was qualitatively examined over a field of
shoreward-facing lunate megaripples located seaward of the surf zone on the
southern Oregon coast, using neutrally buoyant drifters, vertical streaks of
dye, and clouds of sand thrown into suspension on the leeward side of the
lunate megaripples (Clifton et al., 1971). In no case evidence was seen for
shoreward water mass transport, even as the lunate megaripples migrated
towards the shore.
In summary, orbital velocity asymmetry derives from differences in magni-
tude and duration of the back-and-forth components of oscillatory flow. Both
aspects are important to the movement of sediment. Velocity-magnitude
asymmetry is particularly important where only the stronger component
exceeds the threshold velocity for movement of a given grain size (Kemp,
1975). Moreover, since bedload transport is thought to vary approximately
with the third or fourth power of velocity (Inman, 1963; Wells, 1967),
velocity-magnitude asymmetry may significantly influence onshore/offshore
sediment transport. Net water transport is an additional factor that may be
most important for the movement of suspended fine sand (Kemp, 1975).
Because of the complexities involved, an acceptable measure of velocity
asymmetry is yet to be defined. Clifton’s (1976) parameter Au, is the
absolute difference in the peak orbital velocity under the crest and the trough
of a wave; Kemp’s parameter v, is the ratio between the two. Neither mea-
sure takes into account the duration of the opposing flows which must be
accounted as important. Other authors (Dingler, 1974; Allen, 1979, 1980)
measure orbital asymmetry in terms of the associated net drift of the water
or the ratio of this drift t o maximum orbital velocity. This approach does
not accommodate the important influence of the velocity-magnitude asym-
metry. Kemp (1975) suggests using the time-velocity curves to estimate the
potential transport of a grain of a particular size. Such a process is laborious
but should give the most reliable measure of the effects of orbital velocity
asymmetry.
171
TI2 3T14
I
\
Fig.3. Possible profiles of velocity ( u ) over time ( t ) imparted by a wave of period T. Ver-
tical axis = velocity (positive in the direction of wave approach); horizontal axis = time.
(a) Neither velocity-magnitude nor velocity-time asymmetry; no net transport (typical
symmetrical velocity profile under a sinusoidal wave; common in nature). (b) Velocity-
magnitude asymmetry without velocity-time asymmetry; probable net transport in direc-
tion of stronger flow (not likely to occur in nature). (c) Velocity-time asymmetry with-
out velocity-magnitude asymmetry; probable net flow in the direction of flow with
longer duration (not likely to occur in nature). (d) Both velocity-magnitude and velocity-
time asymmetry, balanced such that net transport is negligible (observed qualitatively
over active highly asymmetric bedforms just seaward from the southern Oregon surf
zone). (e) Both velocity-magnitude and velocity-time asymmetry, some net transport in
direction of wave propagation. Note subtle difference from profile d (may be common
under natural shoaling waves). Note that a superimposed unidirectional flow can impart
both velocity-magnitude and velocity-time asymmetry to any of the profiles including
profile a, where the effect can be visualized by adding a constant velocity to the curve
shown (i.e., moving the curve up o r down relative to the zero-velocity axis.
172
The velocity asymmetry induced by shoaling waves is extremely important

in the sorting and transport of sand and in the development of sedimentary
structures. It is, however, complex, involving both differences in the velocity
components of oscillatory currents and a net transport of water. The issue
can be further complicated by the presence of unidirectional flows such as
rip currents or wind-driven flow, that are unrelated to the currents induced
by the passing wave, but can further modify their character (e.g., Inman and
Bowen, 1963).
Basis f o r inference o f f l o w parameters
The basis for inferring the foregoing flow parameters from wave-generated
structures lies largely in either empirical analysis of field data or experimental
studies in the laboratory. Neither of these approaches produces completely
satisfactory results relative to geological applications. Field studies encompass
such a large number of variables that it is difficult to ascertain the critical
relationships, and the spectre of metastability haunts the results. Laboratory
experiments can reduce the number of variables and can generate equilibrium
conditions; commonly, however, such experiments cannot satisfactorily
duplicate natural conditions.
One approach to dealing with the variables encountered in field studies is
to make simplifying assumptions regarding the viscosity and density of the
water and the size, shape, and density of the sediment (Clifton, 1976). Some
of these assumptions, unfortunately, have little basis. Although wave-
winnowed sand typically is well-sorted, it is not uniform in texture and
composition.
Metastability can be an important problem attending empirical studies in
that the observed bedforms may not be completely in equilibrium with the
processes active at the time of observation (Harms et al., 1982). Disequilibrium
may result from two different factors. First, the observed bedformsmay have
developed under more energetic conditions than those at the time of observa-
tion and retained their initial form under the influence of less intense pro-
cesses. Second, the bedforms themselves may influence the flow in such a
way as to retain their original character. Commonly it is impossible to establ-
ish in the field if the observed structures are responding completely to on-
going processes without prior influence.
Laboratory studies circumvent many of these problems because sand of
uniform size can be used as bed material and the waves, or flow, carefully
controlled. Although these experimental studies can do much to define the
influence of specific flow parameters, they can duplicate only a small range
of natural environmental conditions. Specifically, laboratory studies have yet
to replicate conditions imposed by large, long-period oceanic waves. More-
over, certain types of experimental techniques (specifically the use of an
oscillating bed) may produce misleading results (Miller and Komar, 1980a;
Harms et al., 1982).
Empirical and experimental studies of wave-generated bedforms provide
the basis for interpreting maximum orbital velocity, orbital diameter, wave
173
period, and questionably, orbital velocity asymmetry. Estimates of orbital

velocity are based on threshold criteria for grain movement or for sheet flow.
Estimates of orbital diameter are predicated on the relation of ripple spacing
or steepness to grain size. Estimates of orbital velocity asymmetry derive
from the degree of asymmetry of the depositional structures.
Threshold velocities
Two threshold velocities can be defined for oscillatory flow: that required
to initiate grain movement and that required t o produce sheet flow. Under
the oscillatory currents produced by surface gravity waves, ripples form
quickly upon the initiation of grain movement (Dingler, 1974); the lower
flat-bed regime that occurs in unidirectional flow appears to be largely
suppressed. Bagnold (1946), Komar and Miller (1973, 1975), and Dingler
(1979) are among several investigators who have studied the threshold of
grain motion in oscillatory flow. The relation for the onset of grain motion
under oscillatory flow resembles the Shields (1936)-Bagnold (1966) relation-
ship for onset under unidirectional flow (Madsen and Grant, 1976; Dingler,
1979).
Threshold criteria are most accurately presented in terms of shear stress, r ,
which is related to the mean velocity by the equation T = fpu2,/2 (Jonsson,
1967) where p is the fluid density and f is an empirically obtained friction
factor. Because the friction factor is hard to determine, most investigators
present threshold curves using the calculated near-bottom maximum orbital
velocity. Komar and Miller (1973) defined the threshold for movement of
grains smaller than 0.5 mm with the dimensionless equation:
where p s is sediment density and g is the gravitational constant. For quartz

'ndo
sand in water, the relationship urn = - gives:
T
u, = 0.337(gz (2)
which, in units of centimeters and seconds is equivalent to 33.3 (TD)'I3cm s-I
(Clifton, 1976). For movement of grains coarser than 0.5 mm, Komar and
Miller define threshold conditions by the dimensionless equation:
which for quartz sand in water reduces to:

urn = 1.395($TD3)'/' (4)
In units of seconds and centimeters, this is equivalent t o 71.4(TD3)'17cm s-'
(Clifton, 1976). The threshold curves of Komar and Miller (1975) for motion
of sediment equivalent in density to quartz are based on eqs. 1 and 3 and
shown in Fig.4.
174
'/
0.0 1 0.1 1 .o 10 100

D (rnrn)
Fig.4. Velocity thresholds for grain movement and sheet flow of quartz sand in water.
Solid lines are threshold curves of Komar and Miller ( 1 9 7 5 ) for grain movement. Dashed
lines connected by dots are threshold curves of Dingler ( 1 9 7 9 ) in the range of experi-
mental evidence; dots are absent where curves are extrapolated. Threshold curve for
sheetflow, from Dingler and Inman (1977), is solid in the size range of experimental evid-
ence and dashed where extrapolated. Note effect of differing wave period o n threshold of
motion.
Dingler, in a separate experimental study (1979), found that motion for

grain sizes between 0.18 and 1.454 mm commenced when the dimensionless
criterion:
is satisfied, where p is the fluid viscosity and the other terms as defined
above. In terms of threshold velocity, eq. 5 reduces to:
which for quartz sand in water, in units of grams, centimeters, and seconds,
is equivalent to 52.4 (TD)'" cm s-l (Clifton, 1976). Threshold curves based
on this equation are also shown in Fig.4.
It should be noted that all of the above threshold equations show a depen-
dence on wave period. For any particular grain size a longer period wave
requires a higher velocity to initiate grain movement than does a shorter
period wave. The basis for this relationship is unresolved. Possibly it derives
from the more abrupt onset of flow that occurs under a shorter period wave
and the gradient of stress that is associated with greater acceleration.
175
The threshold curves for onset of grain motion are plotted in Fig.4. In the
range of fine sand, the sets of curves are fairly consistent, but they deviate
markedly when extrapolated to coarser bed material. Unfortunately, it is the
larger size ranges where threshold values are most useful for defining minimum
possible wave size and water depth combinations.
Dingler and Inman (1977) determined that sheet flow occurs in fine sand
under a relationship whereby p u ; / ( p , - p)gD = 240. For quartz sand in
water this equation reduces to:
u , = 19.9(gD)”2 (7)
which, in units of centimeters and seconds is equivalent to 623 D1”cm s-’.
The threshold curve for sheetflow derived therefrom is shown on Fig.4. It
should be noted that this curve was observed only in a narrow range of grain
size (0.0128-0.0158 cm) and extrapolation beyond this range must be done
with caution.
All the above threshold equations assume spherical grains of uniform size
on a flat bed. Bagnold (1963) suggested that sand grains on a previously
rippled bed would move at somewhat lower flow velocities, and Southard
and Dingler (1971) showed that ripples under unidirectional flow could
propagate downstream of a disturbance under subthreshold conditions.
Hallermeier (1981) notes that the velocity required to initiate movement
over a rippled bed may be half of that required for the same material on a
flat bed. In the absence of a definitive study on this problem, reliance must
be placed on the flat-bed thresholds noted in the foregoing.
When a range of grain sizes occurs, as is the case outside the laboratory,
most people use the median or arithmetic mean diameters, which are easily
calculated. Some evidence exists that the effective size for consideration of
movement of poorly sorted sediment is less than the median diameter
(Hallermeier, 1981). Bagnold (1966) recommended that the geometric mean
diameter be used because it more realistically weights the size distribution.
Inasmuch as wave-worked sands are typically well-sorted, the median diam-
eter probably suffices.
Threshold values commonly have their greatest use in calculating the forces
required t o move the coarsest clasts available. In many cases, the size of
these largest clasts substantially exceeds that of the bulk of the bed material.
The assumption of uniformly sized particles in the foregoing equations casts
doubt on their applicability t o isolated large clasts on a smaller bed.
Fahnestock and Haushild (1962) suggest that isolated cobbles would move
under unidirectional flow as easily on a plane sand bed as on a bed of gravel.
But would the threshold velocity thereby be significantly reduced? Prelimi-
nary experiments in a water-tunnel indicate that the threshold for movement
of subspherical quartz grains about 1 cm in diameter on a bed of about
0.05 cm sand is not greatly less than that predicted by Komar and Miller
(1975). Under the highest oscillatory velocity possible with the apparatus
(85 cm s-l at T = 16 s), the clasts remained immobile (R. August, pers.
commun., 1983). According to the curves of Komar and Miller (1975),
176
threshold velocity of such clasts under 16 s waves is slightly more than

100 cm s-l (Fig.4).
The effect of grain shape on the threshold curve has not been determined.
However, using the equivalent sphere diameter is satisfactory in most situa-
tions involving terrigenous detrital material. Shelly or other non-spherical
biogenic material would almost certainly require different threshold equa-
tions.
Relations between ripple spacing and orbital diameter

Two disparate views exist regarding the relation of ripple spacing and
orbital diameter. Inman (1957) suggested that for a given grain size, ripple
spacing is directly proportional to orbital diameter until some critical maxi-
mum orbital diameter is reached, whereupon spacing becomes inversely pro-
portional to orbital diameter, diminishes and ultimately reaches a constant
intermediate value. Dingler (1974), plotting both Inman’s (1957) field data
and original field and laboratory measurements found a similar relationship.
In contrast, Allen (1979), after plotting a large amount of existing data
(mostly laboratory), found no well-defined relation between orbital diameter
and ripple spacing for a given size of sand. He therefore concluded that
Inman’s bell-shaped curve was spurious. Miller and Komar (1980a), after
analysing much of the same data, concluded that there were differences in
the data sets that could be attributed to the type of laboratory device used
to generate the oscillatory motions. In particular, the results of oscillating
bed experiments, which dominate Allen’s data, are different from water-
tunnel, wave-channel, and presumably, field results. Oscillating bed experi-
ments indicate that, for a given grain size, ripple spacing increases with orbital
diameter until it reaches a maximum and then remains constant. Plots of
water tunnel and wave channel experiments, in contrast, show a tendency
toward the bell-shaped curve (Miller and Komar, 1980a).
Figure 5a is a dimensionless plot of h / D against d o / D for a number of field
and laboratory studies. Oscillating bed experiments are specifically omitted.
The field data tend to dominate the right side of the diagram (high do/D
values), whereas the laboratory experiments dominate the plot at low d o / D
values. As might be expected the field data are more broadly scattered, but
both sets of data show the bell-shaped relationship.
Using a similar plot that incorporated Inman’s (1957) and Dingler’s (1974)
data, Clifton (1976) subdivided symmetric ripples into three types based on
the relationship between ripple spacing and orbital diameter. This subdivision
(Fig.5b) appears to be valid for the larger data set presented here in Fig.5a.
Orbital ripples are those on the left side of Fig.5b where ripple spacing is
proportional to orbital diameter in the approximate relationship (Miller and
Komar, 1980a):
h = 0.65 do (8)
Such ripples can form under conditions where the d o / D ratio lies in the
range of 100-3000 or more (Fig.5b). Their spacing-to-grain-size ratio ( h / D )
177
1o3
x /D
102
a. 103 do/D 104 105
t Orbital r i p p l e s
x /D
b. 104 1o5
'03 do/D
Fig.5. a. Plot of ratio of ripple spacing to grain size against ratio of orbital diameter t o
grain size. Data include field observations (crosses) and experimental (wave channel,
water tunnel) data (circles). Field data from Inman (1957), Dingler (1974), Miller and
Komar (1980b) and Dingler and Clifton (this volume). Experimental data from Carstens
et al. (1969), Mogridge and Kamphuis (1972), and Miller and Komar (1980a). b. Classi-
fication of ripples based on the distribution shown in a.
ranges from less than 100 to more than 2000. Because of the requirement
for short oscillatory motion, orbital ripples occur most commonly in very
shallow water under short-period waves. Long-period waves can generate
similarly short orbital flow at the bottom in deeper water, but because of the
relationship u, = n d o / T ,thevelocity will be reduced and threshold conditions
less likely to be reached. The spacing of orbital ripples tends to increase in a
shoreward direction, paralleling the increase in do as a wave shoals (Komar,
1974). The spacing appears to be independent of grain size.
Ripple spacing remains proportional t o orbital diameter until the do/Dratio
reaches the range of 1000-3000 (Clifton, 1976, and calculations from
178
Miller and Komar, 1980a). Under such conditions (Fig.5b), the ripple spacing
decreases as orbital diameter increases. Ripples formed under these conditions
were accordingly termed “suborbital” (Clifton, 1976). The ripple spacing
appears to depend both on orbital diameter and grain size in some undefined
relationship.
A t d o / D values in excess of 5000, ripple spacing stabilizes at a value that is
independent of orbital diameter (Fig.5b). Termed “anorbital ripples” by
Clifton (1976), such ripples are most commonly observed in fine sand where
they have a spacing of 5-10 cm. Typically their AID ratio lies in the range of
400-600 under conditions of a single train of waves. Recent field studies
indicate that, under a polymodal wave spectrum (more than one wave train
present), the spacing-to-grain size ratio of anorbital ripples may be on the
order of 1200 (Miller and Komar, 1980b). Anorbital ripples are probably the
only type to form in fine sand under very long period (>12 s) waves (assum-
ing a threshold velocity of 1 5 cm s-’ for sand 0.125 mm in median diameter,
a 1 2 s wave will induce a threshold orbital diameter of nearly 60 cm, and a
d,/D ratio of 4800). Anorbital ripples include the “reversing” ripples desc-
ribed by Inman (1957), which alternate their direction of asymmetry with
each reversal of the oscillatory flow.
An intriguing relation exists between anorbital ripples and the “maximum
ripples” produced by an oscillating bed. Both have been described using the
dimensional parameter h/D”2 = 60 cm’/2 (Clifton, 1976, for anorbital ripples;
Bagnold, 1946, for “maximum” or “natural pitch” ripples). If this relation-
ship is not entirely coincidental, it may provide insight into a fundamental
difference between ripples formed on an oscillating bed and those formed
under oscillating fluid. Both ultimately generate ripples for which the spacing
is independent of orbital diameter and can be defined as h = 60 D1/? if both
h and D are in cm. Ripples on oscillating beds reach this spacing by contin-
uously increasing their size; ripples formed by fluid motion such as those
occurring in nature seem to have the potential to grow as orbital ripples
beyond the size of the maximum ripple of the oscillating bed. Miller and
Komar (1980a) suggest that this growth ceases at the point whereby h =
14.7 X lo3 (both h and D measured in cm). Further increase in orbital
diameter causes the spacing t o shrink (suborbital ripples) until the “natural
pitch” is achieved (anorbital ripples).
Except for reversing ripples, the relationships between h and do described
in the foregoing paragraphs appear to be valid only for symmetrical ripples.
The spacing of asymmetric ripples appears to follow a different pattern,
which remains t o be resolved (Clifton, 1976). The continuum that appears to
exist between small and large symmetric bedforms is lacking for asymmetric
bedforms. The marked difference in size between wave-formed lunate mega-
ripples and associated long-crested asymmetric ripples near the high-energy
surf zone (Clifton et al., 1971) suggests a discontinuity in the scale of asym-
metric wave-formed bedforms similar t o that within the lower regime of
unidirectional flow.
Nielsen (1981) relates ripple spacing t o a “mobility number” $ (Brebner,
1981) that is equivalent to the relative stress of Komar and Miller (1972)
179
( n d o / T ) 2 p / ( p , - p)gD. Nielsen proposes that the spacing of naturally occur-

ring ripples is best described by the equation:
693 - 0.37 ln8$
+
0.75 ln7$ (9)
This expression is complicated by the presence of three variables (do,D,T ) ,
but can be solved as a set of curves relating h and d o for a specific grain size
(Fig.6) or a specific wave period. As shown in Fig.6, Nielsen's equation pre-
dicts a A/do relation similar to that described by Inman (1957) and Dingler
(1974) in which orbital, suborbital and anorbital ripples can be readily iden-
tified. It should be noted that on this diagram, the ripple type is no longer
simply a function of d,/D.Nielsen (1981) suggested that the spacing of
ripples formed in the laboratory follows a somewhat different pattern (Fig.7)
indicated by the equation:
100
10
x
(cm)
10
D . 0 0 1 cm
01
10 10 100
do (cml
Fig.6. Plot of spacing of naturally occurring quartz sand ( D = 0.01 cm) ripples against
orbital diameter for waves of different period based on eq. 9 (Nielsen, 1981). The curve
representing the indicated wave period is terminated on the left side of the diagram at the
threshold orbital diameter (using eq. 19) and on the right side of the diagram at the
maximum possible orbital diameter (using a combination of Fig.9 and eq. 16). Fields for
orbital, suborbital and anorbital ripples are indicated as a function of the nature of the
relation between spacing and orbital diameter. Lines of equal ripple steepness ( X / q ) are
drawn based on eq. 11 (Nielsen, 1981). Data set includes no waves with periods less than
6 s.
180
d
X = 2 (2.2 - 0.345 9°..”4) (10)
2
This equation is of limited validity at larger values of 9 , where the corre-
sponding values of X become negative (see Fig.7). The difference between
the spacing relations of naturally occurring and laboratory ripples is attrib-
uted without elaboration to the irregularity of natural waves (Nielsen, 1981).
The question of the influence of wave period on ripple spacing remains
unresolved. Bagnold’s (1949) statement that, in all of his experimental
studies, the pitch (spacing) was independent of the speed of oscillation
implies that period was not a factor. The spacing of anorbital ripples described
by Miller and Komar (1980b) does not change significantly under unimodal
waves of periods that ranged from 8 to 1 6 s. Nonetheless, as noted before,
they found that the spacing of anorbital ripples under a polymodal wave
spectrum was more than twice that of ripples formed in sand of the same
size under unimodal waves of similar periods. On the basis of experimental
evidence, Nielsen (1981)suggests that spacing depends on wave period, partic-
ularly at the shorter periods (in the range of 1-2 s). Nielsen’s expression for
the spacing of naturally occurring ripples (eq. 9) contains wave period as a
variable, the effect of which can be seen in Fig. 6. It should be noted, how-
ever, that the data base from which Nielsen derives his expression contains no
\
\
‘T = l
- Laboratory ripples
--- N a t u r a l l y occurring ripples
D = 0.01 c m
0.1
1.o 10 100
do ( c m l
Fig.7. Comparison of spacing-orbital diameter relations f o r laboratory and naturally
occurring ripples in quartz sand (D= 0.01 cm) under waves with periods of 1 and 2 s using
eqs. 9 and 10 (Nielsen, 1981). Curves drawn for naturally occurring waves are suspect
because of the absence of data for waves of 1 and 2 s. Curve for laboratory ripples formed
under 2-s waves is suspect at the higher values of d o where it begins a precipitous decline.
181
waves with periods of less than 6 s. For waves of shorter period, the relation
based on laboratory ripples (Fig. 7) may be of greater validity, particularly at
the less extreme values of d o for waves of a given period.
In summary, field observations and laboratory experiments using wave

channels and water tunnels suggest that orbital diameter can be estimated
from the ratio of ripple spacing to grain size, both readily measured param-
eters. At h / D values less than 400, ripple spacing seems to depend on
orbital diameter in the approximate relationship X = 0.65 d o (orbital ripples).
A ratio in the range of AID = 400-600 indicates either orbital ripples or
anorbital ripples ( d o > 5000 D).Spacing-to-grain size ratios in excess of 600
suggest either orbital ripples or suborbital ripples ( d o = 1000-5000 D).Where
the type of ripple is ambiguous (orbital or anorbital, orbital or suborbital),
the complete process of interpretation may indicate which is more likely.
Orbital diameter cannot presently be estimated from the spacing-to-grain
size relationship of asymmetric wave-formed ripples. Variation in wave
period may further complicate the interpretive process in ways that are not
yet fully understood.
Relations between ripple steepness and orbital diameter

Another approach to determining orbital diameter is based on the ratio of
ripple height to wavelength (77 /A ) or ripple steepness. Several investigators
have described three types of asymmetrical ripples based on steepness:
rolling-grain ripples, vortex ripples, and post-vortex ripples. Bagnold (1946)
gave the name rolling-grain ripples to low-amplitude ripples that form on flat
beds under oscillatory flows just above the threshold for grain motion. Sleath
(1976) and Allen (1979) apply this name t o all low-amplitude, wave-generated
ripples. The steepness ( ~ / hof) rolling-grain ripples ranges from zero t o about
0.12 (a VFI or h / of ~ about 8; Allen, 1979); they are too low in amplitude
for a vortex to form in the ripple' troughs. A t least in the lower part of the
ripple regime these ripples are metastable; they readily convert to vortex
ripples if there are many large disturbances on the bed or if the flow velocity
increases. The discussion by Miller and Komar (1980a) suggests that rolling-
grain ripples are stable bedforms only on oscillating beds.
Vortex ripples occupy much of the ripple regime under oscillating fluid,
extending from near the threshold of grain motion to near the onset of sheet
flow (Miller and Komar, 1980). The large, sediment-laden vortex, which
forms in the lee of each crest, gives these ripples their name (Bagnold, 1946).
Ripple steepness is essentially constant throughout the ripple regime, having
a typical value of about 0.15 (Dingler and Inman, 1977) and a range of
about 0.12-0.22 (or VFI between 4.5 and 8; Bagnold, 1946).
As flow velocity increases over the vortex ripples, a point is reached where
sand is stripped from Me rfpple cresds. Or&&Zdirne&mze zepylage when
this velocity is attained, and the ripple wavelength is unchanged by increasing
flow. The net result is post-vortex ripples (Dingler and Inman, 1977) or
rolling-grain ripples (Allen, 1979) that show a systematic decrease in ripple
steepness from 0.15 to 0 as sheet flow conditions are approached (Dingler
and Inman, 1977).
182
The relationship between ripple steepness and ripple spacing for given
values of orbital diameter and sediment grain size is not completely clear.
Several workers (Allen, 1979; Allen, 1981a, b) equate vortex ripples with
orbital ripples. By this interpretation, ripples that have steepness values in
the range of 0.12-0.20 can be used t o calculate orbital diameter from eq. 8.
This approach may be overly simplistic, however. A plot of field and wave
tank data (Fig.8) indicates that vortex ripples ( q / h > 0.1) exist for d o / D
values of less than 5000. This plot confirms that orbital ripples (do/D< 1000)
are vortex ripples, but demonstrates that the converse may not be true.
Vortex ripples also form at do/Dvalues of 1000-5000, where ripple spacing
may be inversely related to orbital diameter. Post-vortex ripples appear to be
stable only under conditions where anorbital ripples form, and therefore
indicative of d o / D values >5000.
Nielsen (1981) proposes that ripple steepness, like spacing, is a function of
the mobility number I). Using the same data set as incorporated into Fig.7,
he suggests that for naturally occurring ripples,
q/X = 0.342 - 0.34(1/2 f , ~ ) ) " ~ (11)
where f, is a friction factor equivalent to exp [ 5.213 (5D/d0)'.lg4 - 5.9771.
Curves of equal steepness for ripples in quartz sand 0.01 cm in diameter on
Fig.6 also indicate that the transition from vortex to post-vortex ripples (in
the range of q / h = 0.12) accompanies the transition t o anorbital ripples.
It should be noted that the steepness of vortex ripples may be reduced by
faunal activity, compaction, or other post-depositional processes (Reineck
and Wunderlich, 1968; Boersma, 1970; Allen, 1981a). Therefore, low values
of ripple steepness may not necessarily reflect anorbital conditions. For
1.01 I I ' , 1 1 1 1 1 I I I 1 ! 1 1 l 1 I I 1 1 1 I J ' I I 1 1 ""'j

I I
VORTEX RIPPLES
-
I VFI
rl/A -
- - - --I-
ORBITAL xfg &xw&x ANORBITAL - ~ s
0.1
RIPPLES
I
+?&
P
z
ls
:*
I
3::
- --20
P 0 ST-VORT EX R I P PL E S -
0.0 1 I I I 1 1 1 1 1 1 I I I , / 1 , 1 1
I I 1 I I ( , , I , , , , , ,,,;
Fig.8. Plot of ripple steepness ( q / h ) against the orbital diameter to grain size ratio ( d , / D ) ,
showing the relationship between vortex and post-vortex ripples (defined by ripple steep-
ness) and orbital and anorbital ripples (defined by the d,/D ratio). Vertical form index
(VFI) scale on right side of plot. Ripples with d,/D < 1000 are assumed to be orbital;
those with d,/D ratios >5000 are assumed to be anorbital. Ripples with d,/D ratios
between 1000 and 5000 are considered t o be transitional between orbital and anorbital
ripples (suborbital ripples). Data from Inman (1957) and Dingler (1974).
183
these reasons, we feel that ripple steepness by itself presently is not a reliable
indicator of orbital diameter in ancient deposits.
Asymmetry of bedforms
Little is known about the degree of orbital velocity asymmetry that is

required to generate asymmetric bedforms. Clifton (1976) used several lines
of reasoning to suggest that a velocity-magnitude difference A u , of a few
centimeters per second could produce asymmetry in small ripples, and Allen
(1979) concluded that the degree of ripple asymmetry is proportional to the
ratio of the wave-drift velocity to the near-bottom velocity maximum. This
conclusion is highly tentative and virtually no data exist regarding the nature
of flow that causes asymmetry of larger ripples or wave-formed lunate mega-
ripples (which, as noted in the foregoing, have been observed to migrate in
the absence of an observable wave-drift current). Tietze (1978) produced
asymmetric ripples in a small experimental wave tank under measured A u ,
values between 1 and 11cm s-’. He demonstrates a relationship between the
degree of ripple asymmetry and the ratio between drift velocity and maxi-
mum orbital velocity, but does not indicate where in the water column the
net drift was measured. His observation that ripple asymmetry is increased
by adding coarse sand to the bed strongly suggests an influence by velocity-
magnitude asymmetry rather than by net drift (Kemp, 1975). Field observa-
tions show that shoreward-facing, wave-induced asymmetry of bedforms is
most common in shallow water near the breaker zone (Davidson-Arnott and
Greenwood, 1974; Clifton, 1976). Without a documented relationship, how-
ever, between ripple shape and some measure of flow asymmetry, quantitative
estimates of the paleo-asymmetry of oscillatory flow are presently question-
able.
STEP 2. ESTIMATING POSSIBLE COMBINATIONS OF WAVE PARAMETERS AND

WATER DEPTHS FROM FLOW PARAMETERS
Once flow parameters, such as orbital diameter or maximum velocity,

have been derived from the sedimentary structures, a range of wave condi-
tions and water depths can be estimated. The normal complexity of the sea
surface forces an investigator to undertake a great deal of simplification;
commonly a “typical wave form” is identified that can be represented by
mathematical equations from an appropriate wave theory. Such simple waves
can be approximated in experimental studies. Spectral analysis provides a
more accurate approach to describing natural waves (Dingler, 1974; Miller
and Komar, 1980b; Dingler and Clifton, 1984, this volume). It is worth
noting that a complicated wave field, composed of several different trains of
waves can profoundly influence the development of bedforms (Clifton et al.,
1971; Miller and Komar, 1980b).
Various wave theories can be used to relate flow parameters to basic wave
parameters. The four most commonly cited are the: (1)Airy; (2) Stokes; (3)
184
cnoidal; and (4) solitary wave theories. Each is most applicable under a
specific set of conditions of wave height, wave period and water depth (Fig. 9).
Each has associated disadvantages and each should be considered only an
approximation. A brief description of these theories is presented here; for
further information, the reader is directed to the useful summary provided
by Komar (1976).
Airy theory, which treats waves as sinusoidal forms, is the simplest in
application. It is applicable to small amplitude waves in a wide range of
water depths (relative to wave length) and provides for reasonable approxi-
mation of measured orbital diameter and near-bottom maximum velocities
for real waves in shallow water (LeMehaute et al., 1969). It does not, how-
ever, provide for asymmetric flows.
The other theories noted above apply to waves with peaked crests and
flattened troughs, a shoaling transformation of the sinusoidal wave. All
predict asymmetric oscillatory motion. The Stokes wave theory is relatively
simple, but, according to Komar (1976), becomes inaccurate for large waves
when extended into shallow water. Figure 10 indicates the combinations of
water depth and wave height under which the Airy and Stokes theories apply
for waves of different periods using the criterion employed by Komar (1976)
whereby the expression HL2/h3 = 32 defines the boundary between
cnoidal and Airy or Stokes waves, and H / L = 0.0625 tanh (27rh/L) defines
the boundary between Airy and Stokes waves. This figure shows, for example,
that the Stokes wave theory can describe wave form and water motions for
a 1 0 s wave in 3 m of water, provided the wave height does not exceed 1 m.
It should be noted that in water depths greater than about 7 m, cnoidal wave
theory does not apply regardless of wave height. A t greater depths and/or
Fig.9. Conditions of wave length ( L ) , wave height ( H ) and water depth ( h ) for which
different wave theories are most applicable (from Komar, 1976). Approximate theoretical
waveform shown within each field.
185
t 1
.................T = 2 sec
I
Fig.10. Areas of applicability (in terms of water depth and wave height) for Airy, Stokes,
and cnoidal wave theories for waves of different period. Limiting relations shown for 5-s
wave are similar for the waves of other periods. Limits between Stokes and Airy theory
defined by H / L = 0.0625 tanh ( 2 n h / L ) and between Stokes-Airy and cnoidal theory by
HL21h3= 3 2 n2" (Komar, 1976).
smaller waves, the less cumbersome Airy theory provides an equally valid
approximation.
Cnoidal wave theory may be more accurate than Stokes theory for large
waves in shallow waters (Wiegel, 1960), but its complexity severely limits its
use. In many cases the Stokes or Airy theory may sufficiently approximate
the water motion under conditions where cnoidal wave theory is otherwise
indicated (Komar, 1976, p.62).
Solitary waves are individual progressive waves composed of a single crest.
Waves very close to shore commonly resemble the solitary wave shape
although such waves can be described by their wavelength and period, terms
that are not appropriate for true solitary waves. Although solitary wave
theory is relatively accessible, it does not truly describe periodic oscillatory
motion of real wind-generated waves. This and the deviation of predicted
results from measured parameters casts doubt on the use of solitary wave
theory for nearshore studies (Komar, 1976, p.59).
Wave length and period
The flow parameters described in Step One (urn, d o , etc.) result from
different combinations of wave size, shape, and water depth. The size of a
186
wave is most readily perceived in terms of its height H and length L (Fig.1).
The significant wave height (average of the highest 1/3 of the waves) is often
used to represent a wave field (Dingler, 1974), even though the root-mean-
square wave height is naturally associated with spectral analysis. The length
of a wave is a variable parameter inasmuch as it changes significantly as a
wave shoals. The wave period T , the time required for one oscillation, is
related to the wave length by T = L/C, where C is the phasevelocity (velocity
of propagation) of the wave in question. Both phase velocity and wave length
decrease progressively at the same rate as a wave shoals. The wave period
remains unchanged and is therefore a more useful description of a wave than
either length or velocity of propagation.
Airy wave theory relates the length of a given wave to its period and to
the water depth h by the equation:
Equation 12 can be simplified in deep water (h/Lo > 1/4) where the hyper-
bolic term approaches unity:
Lo=-gT2
2n
or, in mks units, Lo = 1.56 T 2 m. This relationship implies that, for waves of
any period, there is an associated, easily calculated deep-water wave length.
The deep-water wave length can be introduced into eq. 12 whereby:
L = L otanh (hh) (14)
where h , the wave number, equals 2 n/L. Expressed this way, the hyperbolic
term can be viewed as a shoaling factor that is applied to the deep-water
wave length to give the wave length at any water depth. Equation 14, how-
ever, remains complicated by the presence of the unknown (L) on both sides
of the equation.
This problem can be resolved by dividing both sides of eq. 1 4 by the water
depth h and rearranging such that:
La-z
- h tanh(hh)
This expression has been solved by Wiegel (1954) and presented in tabular
form in the Shore Protection Manual (US. Army Coastal Engineering
Research Center, 1973). Using these tables the solution of many of the wave
equations is greatly simplified. For any ratio of water depth to deep-water
wave length (or, by inference, any combination of water depth and wave
period), the tables indicate the corresponding values of hlL, kh, the different
hyperbolic and trigonometric functions of (hh), and other parameters. Such
relations allow the wave size at any water depth to be readily expressed in
terms of wave height and wave period.
187
Wave height
Some natural limits t o wave height provide useful constraints to the com-
bination of solutions possible from the wave equations. One such limit occurs
in shallow water whereby waves become unstable and break at some critical
water depth. The ratio of wave height to water depth at which breaking
occurs depends on the beach slope and the initial wave steepness (Iverson,
1952). A value of 0.78 has been most widely accepted for this ratio (Komar,
1976). In deeper water, waves will break if their height exceeds the value L / 7
(Miche, 1944). Although the largest possible wind wave could theoretically
exceed 6 5 m in height (Bascom, 1980, p.58), the largest recorded remains
the 34-m wave observed from the U.S.S. “Ramapo” in 1933. Accordingly,
40 m seems a reasonable maximum height for a set of wind-generated waves.
Figure 11illustrates the maximum stable height for waves of different periods.
Orbital diameter and maximum velocity
In Airy wave theory the orbital diameter at the sea floor do is:
H
d o = sinh ( k h )
0 1 I 1 I I 1 ,,,I I I I I 1 1 1 1 1 I , I 1 1 8 1
0 1 10 10
H (rn)
Fig.11. Maximum wave height ( H ) for waves of different period as a function of water
depth, Curves for waves of different period terminate at approximately h = L o (= 1.56 T2m).
188
and the maximum orbital velocity at the seafloor urn under the crest of the
wave is:
The velocity under the trough of an Airy wave is of similar magnitude but
opposite in direction. Although Airy theory is specifically applicable to con-
ditions of relatively small waves in deep water, LeMehaute et al. (1969)
showed that it provides a reasonable approximation to measured orbital
diameters and near-bottom maximum velocities for finite amplitude waves in
shallow water. Equations 1 6 and 17 are valuable for calculating the combina-
tions of wave height, period, and water depth that will generate a particular
wave-formed bedform or internal structure. One can recast these equations
in terms of wave height
u,T sinh ( h h )
H = do sinh ( h h ) =
71
and then, using the structurally indicated value of do or u,, solve the equation
for a series of selected water depths under waves of several different periods.
In these calculations (shown in more detail in the first example, following),
the chosen combination of wave period and water depth determines the value
of h/Lo (equal to h/1.56 T 2in meters), which then can be used to enter the
wave tables to compute the appropriate value of h/L. The results can then be
plotted as curves of equal wave period on an “H-h” (wave height vs. water
depth) diagram. Figure 1 2 is an example of an H-h diagram that shows the
combinations of wave height and water depth required to produce sheet flow
in medium-grained sand under waves of several different periods.
One can also combine estimates of orbital diameter with those of threshold
velocity t o determine the maximum wave period that can produce the given
combination. Longer period waves are capable of generating the same orbital
diameter in deeper water, but the orbital velocities will be reduced below
threshold level. Dingler (1979) combines his threshold equation for grain
movement of quartz sand with the relation u, = ndo/T (all measurements in
cm and s) to derive the expression for maximum or threshold wave period:
T = 0.17(dg/D)’/3s (19)
The threshold equations of Komar and Miller (1975) provide for similar
e)””
expressions of threshold wave period (in units of cm and s):
T = 0.17 - s (for D < 0.05 cm)

T = 0.065 ( d;:
p)s (for D > 0.05 cm)
189
0.V
0.1 1.o 10
H (m)
Fig.12. Combinations of wave height and water depth that will generate sheetflow of
0.250 mm quartz sand (u, = 100 em s-l) under waves of different period. Curves for
waves terminated at maximum stable wave height.
Velocity asymmetry
As noted in the foregoing section, the Airy theory cannot provide esti-
mates of velocity asymmetry, and a higher order theory must be invoked.
Stokes second-order theory provides the easiest calculable estimate of velocity
asymmetry in shallow-water. Asymmetries can be calculated from the other
shallow-water wave equations, but their inherent complexity generally limits
their application in paleoenvironmental interpretations.
Stokes’ (1847) solution for waves of finite height results in water-particle
motion that is asymmetric with respect to both maximum velocity and time.
The velocity of the water moving forward under the wave crests exceeds the
reverse velocity under the wave trough. The duration of forward flow is less
than that of the reverse flow, but the net result is an onshore migration of
the water particles.
Stokes second-order wave theory provided equations both for velocity-
magnitude asymmetry and for the net drift velocity of the water particles.
The complete equation for near-bottom orbital velocity under the crest of a
Stokes wave is:
190
l7H
u; =
T sinh ( k h )
whereas that under the trough of the wave is:
The first term in eqs. 22 and 23 is the expression for maximum orbital velocity
under an Airy wave; the second term can be viewed as a correction factor
imposed by the Stokes wave. Clifton’s (1976) expression for velocity magni-
tudes asymmetry is the sum of eqs. 22 and 23 (in effect, twice the correction
factor):
3(7m - 14.8 H 2
Au, =
2 LT sinh4 ( k h ) - LT sinh4 ( k h )
The net drift velocity generated by a Stokes wave is derived by integrat-
ing (over a wave period) the Stokes second-order equation for water particle
displacement. The result involves assumptions (infinite channel length, con-
stant depth, absence of viscosity) that are inappropriate for most natural
situations. Accordingly, Longuet-Higgins (1953) developed a wave drift
relationship for the case of a Stokes wave in a channel of finite length with a
real viscous fluid. The resulting equation:
*- 5(rrH)’
= 4 LT sinh2 ( k h )
describes a slow onshore mass transport of water due to-wave passage.

Unfortunately, as noted previously, neither A u, nor Uas defined in eqs. 24
and 25 is completely satisfactory for expressing the velocity asymmetry
imparted by shoaling waves. Compounding this problem is the fact that the
Stokes second-order equation can be of uncertain validity at the shallow,
near-breaker-zone depths where asymmetry of flow is most important. Unfor-
tunately, cnoidal wave theory is generally too complicated for geologic
application and solitary wave theory has been shown to be unreliable under
real conditions (Komar, 1976, p.59). It thus appears that, despite its obvious
importance in determining bedform shape and sediment transport, there
presently is no satisfactory means of quantitatively estimating velocity
asymmetry.
STEP 3. DERIVING PALEOENVIRONMENT INTERPRETATIONS FROM INFERRED

COMBINATIONS OF WAVE CHARACTER AND WATER DEPTH
Typically, the results of an analysis of ancient, wave-generated structures

take the form of an H--h diagram on which are plotted the various combina-
tions of wave height, period and water depth that will generate a particular
flow parameter. Commonly, these combinations span a broad range of water
depth and wave size. To apply these data t o a paleoenvironmental problem
191
requires that the combinations be restricted to a smaller range. The natural

limits to wave height described in the foregoing section provide one such re-
striction. Further constraints can be placed by an application of wave hind-
casting techniques, geological reasoning, or a combination of the two. Wave
hindcasting is a technique for estimating the height and period of waves on
the basis of past speeds and duration of winds, fetch length and water depth.
(Shore Protection Manual, U.S. Army Corps of Engineers Coastal Research
Center, 1973). The inferred period and height of ancient waves can be thus
related to basin size and wind speed aad direction. Paleogeographic recon-
struction may thereby impose limits on the waves, or conversely, the recon-
structed waves may indicate size and shape of the depositional basin.
Geological evidence can also place constraints on water depth. Evidence
for water depth may be drawn from paleoecologic analyses, the nature of
associated facies, or the vertical distance t o the inferred base-of-beach in a
prograding shoreline deposit (see Duprk, 1984, this volume). Directional
features may indicate the direction of wave approach reIative to the shoreline
and the presence of wave-driven currents near the shoreline. The use of geo-
logical reasoning is limited only by the availability of critical data and the
resourcefulness of the investigator.
EXAMPLES
The following examples illustrate different ways in which wave-formed

structures can be used to interpret aspects of the paleoenvironment. In each
case the procedure is outlined in detail, noting particularly each of the three
steps involved.
Example 1 . Estimating wave size from threshold velocity
This example has been published (Hunter and Clifton, 1982) but is
included here for analysis and to illustrate the procedure. The problem is
determining whether or not storm waves were involved in the formation of
hummocky cross-stratification in sandstone of Late Cretaceous age exposed
at Cape Sebastian on the coast of southern Oregon. The hummocky cross-
stratification occurs in the lower part of sediment cycles typically tens of
centimeters thick. The hummocky cross-stratified sandstone overlies an
erosional surface and grades upward into horizontally stratified sandstone
that in turn grades upward into thoroughly bioturbated sandstone. The cycles
are composed of slightly graded fine sand. Small pebbles are scattered in the
hummocky cross-stratified sandstone and a few lie within the overlying hori-
zontally stratified sand. Exposures of bedding surfaces of the horizontally
stratified sandstone commonly show straight-crested symmetrical ripple
marks.
The largest pebble, about 5 cm in diameter, found in the horizontally
stratified sandstone forms the basis for the calculations presented here. The
common presence of symmetrical ripple marks in this facies implies the
192
presence of waves, which suggests the possibility of using the threshold for
movement t o calculate maximum orbital velocity.
Step 1. The threshold velocity for moving a 5 cm pebble can be estimated
from Fig.4. The curves of Komar and Miller (1975) provide a more conserva-
tive value at wave periods larger than 5 s and are therefore employed. These
curves indicate an orbital velocity on the order of 200 cm s-'. It should be
noted that the fact that the pebble moved on a bed of fine sand rather than on
a bed of similarly sized pebbles (assumed for the threshold curves) introduces
a measure of uncertainty.
Step 2. To determine the combination of wave size and water depth that
would produce the orbital velocity derived in Step 1requires selection of the
appropriate wave theory. Since asymmetry of flow is not involved, Airy
theory should provide for reasonable calculations based on the indicated
maximum orbital velocity using eq. 18. A table can subsequently be con-
structed to determine the wave height that at given water depths will generate
an orbital velocity of 200 cm s-'. Table I, for example, indicates the wave
heights that will generate this orbital velocity a t a variety of water depths for
a 1 2 4 wave. Similar tables can be constructed for waves of other periods and
the results plotted on an H-h diagram (Fig.13) that indicates the combina-
tion of water depth and wave height at which waves of several specific periods
will generate a near-bottom maximum orbital velocity of 200 cm s-'. Note
that the curves for the waves of smaller period are terminated a t their maxi-
mum stable wave height.
Step 3. Figure 13 indicates that the specified orbital velocity of 200 cm s-l
can be generated by waves of widely varying size. Constraints, however, can
be imposed by geologic reasoning. The absence of unidirectional crossbedding
and shoreline progradational sequences from the cyclic part of the Upper
Cretaceous sandstone at Cape Sebastian led Hunter and Clifton (1982) to
conclude that deposition did not occur in very shallow water close to a
TABLE I
Computation of wave heights (H) required to generate u, = 2.0 m s-' at various water
depths ( h ) under a 12 s wave, using Airy wave theory (eq. 18). Lo = 1.56 T1m = 225 m ;
Hm = 0.142 L,' = 32 m
h1 hlLO2 sinh (y ) H4
5 0.022 0.3 90 3.0

10 0.044 0.579 4.4
20 0.089 0.922 7.0
30 0.133 1.27 9.7
50 0.222 2.21 17.0
70 0.311 3.72 28.0
100 0.444 8.27 63.0
Iselected arbitrarily; 2calculated from indicated values; 3computed from Wave Data tables
(US.Army Coastal Engineering Research Center, 1973); 4calculated from eq. 18.
193
I , I I I I
CONDITIONS T O GENERATE Urn= 200 C M l S E C

700 -
-
-
-
-
(THRESHOLD FOR M O V E M E N T
- OF 5 CY O U A R T Z P E B B L E 1
70 --- - --
-
h i
-
(m)
1.0- I 1 I ' f 1 ' 1 1 I I 1 1 1 1 1 1 1 I I , I 1 1 4 1

0.1 1 .o 10
shoreline. Accordingly, they felt that a depth of 10 m was a conservative

minimum for deposition of this part of the unit. Applying this limit to Fig.13
eliminates the shorter period waves and implies wave heights of at least 4 m.
Since these conditions presumably prevailed during the waning phase of the
event that created the sedimentary cycle, even larger waves attended the
development of the hummocky cross-stratified sandstone.
Example 2. Determining depositional environment o n the basis of grain size

and ripple spacing
The Berea Sandstone is a paralic deposit of Mississippian Age that crops

out in anorth-south belt across the state of Ohio (Pepper et al., 1954; Coogan
et al., 1982). In exposures southeast of the city of Cleveland the unit consists
of fine sandstone in which sets of trough crossbedding 1-3 m thick alternate
with thin (typically less than 1 m thick) sequences of flat-bedded or wavy-
bedded sandstone. Bedding surfaces of the flat- or wavy-bedded sandstone
commonly exhibit small symmetrical ripple marks. Thp drossbeds dip to the
west and northwest, presumably an onshore direction, and channelling is
conspicuously uncommon (Coogan et al., 1982). The deposit has beenvari-
ously attributed to accumulation in a subaerial delta, river system, barrier,
bar, lagoon, wind-tidal flat or tidal channel system (Coogan et al., 1982).
Step 1. The symmetrical ripples prove particularly useful in resolving the
depositional environment of this unit. The smallest ripples, with a spacing of
1.5 cm, were composed of sand of about 0.01 cm diameter. The h / D ratio of
150 indicates that the ripples are orbital ripples (A = 0.65 d o ) formed under
194
an orbital diameter of 2.3 cm. This interpretation is consistent with h-do

relations predicted by Nielsen's equation for laboratory ripples (Fig.'7), which
may be more applicable than that for naturally occurring ripples at low values
of A . Equations 19 and 20 indicate maximum wave periods of 1.3 and 1.0 s,
respectively, for forming these ripples.
Step 2. The combinations of water depth and wave height whereby a wave
with a period of 1.3 s (the more conservative at the indicated values) will
generate a near-bottom orbital diameter of 2.3 cm can be estimated using
Airy wave theory (eq. 18, this paper). Following the procedure outlined in
the previous example, an H-h diagram can be constructed (Fig.l4), which
indicates that the observed ripples formed in very shallow water (less than
2 m deep).
Step 3. The inference that the ripple-marked sandstone was deposited at
water depths smaller than the thickness of the intercalated cross-bed sets
imposes special constraints. The nature of the contacts between the two
lithologies becomes critical t o the interpretation. In every observed case, the
planar- or wavy-bedded sandstone sharply overlies the sandstone below and
extends gradationally into overlying cross-bedding foresets above. This rela-
tion implies that the bedforms that produced the crossbedding migrated over
the planar- or wavy-bedded sandstone with the small ripples. The truncated
top to the cross-beds indicates that the bedform relief exceeded the thick-
ness of the cross-bedding unit, which in turn generally is greater than the
water depth implied by the small symmetric ripples. The implication that the
height of the bedforms exceeded (perhaps substantially) the depth of water
into which they migrated and the scarcity of current ripples relative t o sym-
metric ripples effectively eliminate a subaqueous origin for the crossbedding.
The bedforms are best interpreted as aeolian dunes that migrated over a
surface that at least part of the time was the site of interdune ponds.
1
0.10.01
H
0.1
(rn)
3 .O
Fig.14. Combinations of water depth and wave height whereby a wave with a period of
1.3 s will generate a near-bottom orbital diameter of 2.3 cm (threshold for forming orbital
ripples with a spacing of 1.5 cm in 0.100 mm quartz sand). Wave curve terminated at
maximum stable wave height.
195
Further examination of the deposit supports this interpretation. The wide

variability of the trend of the ripple marks (sets on the same bedding surface
may diverge by nearly 90") suggests local winds blowing over very shallow
water. Mud cracks on a few of the ripple surfaces are consistent with a pond
environment. The crossbedded sand locally shows subtle bedding features
(climbing translatent strata, sandflow tongues) attributed t o aeolian deposi-
tion (Hunter, 1977, 1981).
Other examples. Several recent papers use wave-generated structures to
reconstruct paleoenvironments. Allen (1981) uses an approach similar to
that analysed in Example 2 t o estimate the size and depth of a Devonian lake
on southeast Shetland. Clifton (1981) draws heavily on the orientation of
wave-formed structures t o develop a detailed reconstruction of a Miocene
shoreline in the southern Coast Range of California. Papers by Allen and
Dupr6 in this volume provide additional examples.
CONCLUSIONS
Wave-formed sedimentary structures can be a valuable tool for interpreting

paleo-environments. A number of caveats should be noted relative to the
implementation of these tools:
(1) The results are approximations. The use of wave equations typically
gives precise solutions, but the equations themselves are approximations in a
real environment. The indicated values should be considered as reasonable
estimates only.
(2) Much uncertainty exists regarding the nature of the structures and the
hydrodynamic processes involved in their formation. Threshold equations
should be extrapolated with care, particularly to the coarser grain sizes. The
threshold of movement of large isolated clasts on a bed of smaller material
requires further study. The relations between ripple spacing, grain size and
orbital diameter need t o be defined more closely particularly in terms of the
influence of wave period. The basis for asymmetry of wave-formed structures
is controversial and awaits further study before being applicable in any
hydrodynamic sense.
(3) The interpretation of wave-formed structures is best done in conjunc-
tion with other geologic evidence, Inferences of water depth or wave size
based on wave-formed features are most credible when supported by other
observations.
Much further field observation and experimental study must be done
before wave-formed structures reach their full potential for paleoenviron-
mental interpretation. Presently, they can provide powerful, but often
tantalizing, clues regarding ancient environmental conditions.
REFERENCES
Allen, J.R.L., 1970. Physical Processes of Sedimentation. Allen and Unwin, London,
248 pp.
196
Allen, J.R.L., 1979. A model for the interpretation of wave ripple-marks using their wave-
length, textural composition, and shape. J. Geol. SOC.London, 136: 6 7 3 4 8 2 .
Allen, J.R.L., 1980. Sand waves: a model of origin and internal structure. Sediment.
Geol., 26: 281-328.
Allen, J.R.L., 1982. Sedimentary Structures - Their Character and Physical Basis, Vol. 1.
(Developments in Sedimentology, 30A). Elsevier, Amsterdam, 593 pp.
Allen, P.A., 1981a. Some guidelines in reconstructing ancient sea conditions from wave
ripples. Mar. Geol., 43: M59-M67.
Allen, P.A., 1981b. Wave-generated structures in the Devonian lacustrine sediments of
south-east Shetland and ancient wave conditions. Sedimentology, 28: 369-379.
Allen, P.A., 1984. Miocene waves and tides from the Swiss Molasse. In: B. Greenwood
Coastal Environments. Mar. Geol., 60: 455-473 (this volume).
Bagnold, R.A., 1946. Motion of waves in shallow water. Interaction of waves and sand
bottoms. Proc. R. SOC.London, Ser. A, 187: 1-16.
Bagnold, R.A., 1963. Mechanics of marine sedimentation. In: M.N. Hill (Editor), The Sea,
Vol. 3. Wiley-Interscience, New York, N.Y., pp. 507-528.
Bagnold, R.A., 1966. An approach to the sediment transport problem from general
physics. U.S. Geol. Suw., Prof. Pap., 422-1.
Bascom, W.A., 1980. Waves and Beaches - Revised and updated. Anchor Press/Doubleday,
Garden City, N.Y., 366 pp.
Bliven, L., Huang, N.E. and Janowitz, G.S., 1977. An experimental investigation of some
combined flow sediment transport phenomena. North Carolina State Univ., Center for
Marine and Coastal Studies, Rept. No. 77-3.
Boersma, J.R., 1970. Distinguishing features of wave-ripple cross-stratification and mor-
phology. Thesis, University of Utrecht, Utrecht, 6 5 pp.
Brebner, A . , 1981. Sand bedform lengths under oscillatory motion. Proc. 7th Conf.
Coastal Engineering, Sydney, N.S.W., pp.1340-1341.
Bucher, W.H., 1919. On ripples and related sediment surface forms and their paleogeo-
graphical interpretations. Am. J. Sci., 47 : 149-210, 241-269.
Clifton, H.E., 1976. Wave-generated structures - a conceptual model. In: R.A. Davis and
R.L. Ethington (Editors), Beach and Nearshore Processes. SOC.Econ. Paleontol.
Mineral., Spec. Publ., 24: 126-148.
Clifton, H.E., 1981. Progradational sequences in Miocene shoreline deposits, southeastern
Caliente Range, California. J. Sediment. Petrol., 51: 165-184.
in the high-energy nonbarred nearshore. J. Sediment. Petrol., 41: 651-670.
Coogan, A.H., Heimlich, R.A., Malcuit, R.J., Bork, K.B. and Lewis, T.L., 1982. Early
Mississippian deltaic sedimentation in central and northeastern Ohio. In: T.G. Roberts
(Editor), G.S.A. Cincinnati '81 Field Trip Guidebooks, Vol. 1:Stratigraphy, Sedimen-
tology. Am. Geol. Inst., Falls Church, Va., pp.113-152.
with bar topography in the shallow-water wave environment. J. Sediment. Petrol., 44:
6 98-7 04.
Dingler, J.R., 1974. Wave-formed ripples in nearshore sands. Thesis, Univ. Calif. at San
Diego, San Diego, Calif., 136 pp.
Dingler, J.R., 1979. The threshold of grain motion under oscillatory flow in a laboratory
wave channel. J. Sediment. Petrol., 49: 287-294.
Dingler, J.R. and Clifton, H.E., 1984. Tidal-cycle changes in oscillation ripples on the
inner part of an estuarine sand flat. In: B. Greenwood and R.A. Davis, Jr. (Editors),
Geol., 60: 219-233 (this volume).
Dingler, J.R. and Inman, D.L., 1977. Wave-formed ripples in nearshore sands. Proc. 15th
Conf. on Coastal Engineering, Honolulu, Hawaii, pp.2109-2126.
197
Dupr6, W.R., 1984. Reconstruction of paleo-wave conditions from Pleistocene marine

terrace deposits, Monterey Bay, California. In: B. Greenwood and R.A. Davis, Jr.
Fahnestock, R.K. and Haushild, W.L., 1962. Flume studies of the transport of pebbles
and cobbles on a sand bed. Geol. SOC.Am., 73: 1431-1436.
Hallermeier, R.J., 1981. Critical wave conditions for sand motion initiation. Coastal Eng.
Tech. Aid No. 81-10, U.S. Army Corps Engineers Coastal Eng. Res. Center, 16 pp.
Harms, J.C., 1969. Hydraulic significance of some sand ripples. Geol. SOC.Am. Bull., 80:
363-396.
Harms, J.C., Southard, J.B. and Walker, R.G., 1982. Structures and sequences in clastic
rocks. S.E.P.M. Short Course No. 9, SOC.Econ. Paleontol. Mineral., Tulsa, Okla.,
249 pp.
Heezen, B.C. and Hollister, C.D., 1971. The Face of the Deep. Oxford Univ. Press, New
York, N.Y., 659 pp.
Hunter, R.E., 1977. Basic types of stratification in small aeolian dunes. Sedimentology,
24: 361-387.
Hunter, R.E., 1981. Stratification styles in eolian sandstones: some Pennsylvanian to
Jurassic examples from the western interior, U.S.A. In: F.G. Ethridge and R.M. Flores
(Editors), Recent and Ancient Nonmarine Depositional Environments: Models for
Exploration. SOC.Econ. Paleontol. Mineral., Spec. Publ., 31: 315-329.
Hunter, R.E. and Clifton, H.E., 1982. Cyclic deposits and hummocky cross-stratification
of probable storm origin in Upper Cretaceous area, southwestern Oregon. J. Sediment.
Petrol., 52: 127-146.
Inman, D.L., 1957. Wave generated ripples in nearshore sands. Tech. Memo. Beach Ero-
sion. Bd. U.S. Army Corps of Engineers, 100.
Inman, D.L., 1963. Ocean waves and associated currents. In: F.P. Shepard (Editor), Sub-
marine Geology (2nd ed.). Harper and Row, New York, N.Y., pp.49-81.
Inman, D.L. and Bowen, A.J., 1963. Flume experiments on sand transport by waves and
currents. Proc. 8th Conf. Coastal Engineering, Berkeley, Calif., pp.137-150.
Iverson, H.W., 1952. Studies of wave transformation in shoaling water, including breaking.
Natl. Bur. Stand., Circ., 521: 9-32.
Jonsson, I.G., 1967. Wave boundary layers and friction factors. Proc. 10th Conf. Coastal
Engineering, Tokyo, pp.127-148.
Kemp, P.H., 1975. Wave asymmetry in the nearshore zone and breaker area. In: J. Hails
and A. Carr (Editors), Nearshore Sediment Dynamics and Sedimentation. Wiley, New
York, N.Y., pp.47-5.
Komar, P.D., 1974. Oscillatory ripple marks and the evaluation of ancient wave conditions
and environments. J. Sediment. Petrol., 44: 169-180.
N.J., 417 pp.
Komar, P.D. and Miller, M.C., 1973. The threshold of movement under oscillatory water
waves. J. Sediment. Petrol., 43: 1101-1110.
Komar, P.D. and Miller, M.C., 1975. The initiation of oscillatory ripple marks and the
development of plane-bed at high shear stresses under waves. J. Sediment. Petrol., 45:
697-703.
LeMehaute, B., Divoky, D. and Lin, A., 1969. Shallow water waves: a comparison of
theories and experiments. Proc. 1l t h Conf. Coastal Engineering, London, pp.86-96.
Lofquist, K.E.G., 1978. Sand ripple growth in an oscillatory-flow water tunnel. Tech.
Pap. Coastal Eng. Res. Center, US.Army Corps Eng., 75-8.
Longuet-Higgins, M.S., 1953. Mass transport in water waves. Philos. Trans. R. SOC.London,
Ser. A, 245: 535-581.
Madsen, O.S. and Grant, W.D., 1976. Sediment transport in the coastal environment.
Rept. no. 209, Ralph M. Parsons Lab., Mass. Inst. Tech., 1 0 5 pp.
Miche, R., 1944. Undulatory movements of the sea in constant and decreasing depth.
Ann de Ponts e t ChaussGes, May-June, July-August, pp.25-78, 131-164, 369-406.
198
Miller, M.C. and Komar, P.D., 1980a. Oscillation sand ripples generated by laboratory
apparatus. J. Sediment. Petrol., 5 0 : 173-182.
Miller, M.C. and Komar, P.D., 1980b. A field investigation of the relationship between
oscillation ripples spacing and the near-bottom water orbital motions. J. Sediment.
Petrol., 50: 183-191.
Nielsen, P., 1 9 8 1 . Dynamics and geometry of wave-generated ripples. J. Geophys. Res.,
8 6 : 6467-472.
Pepper, J.F., DeWitt, W. and Desmarest, D.F., 1 9 5 4 . Geology of the Bedford shale and
Berea sandstone in the Appalachian Basin. U.S. Geol. Surv., Prof. Pap., 259, 1 0 6 pp.
Reineck, H.E. and Singh, I.B., 1973. Depositional Sedimentary Environments. Springer,
New York, N.Y., 4 3 9 pp.
Reineck, H.E. and Wunderlich, F., 1968. Zur Unterscheidung von asymmetrischen Oszil-
lationrippeln und Stromungsrippeln. Senkenbergiana Lethaea, 4 9 : 321-345.
Shields, A., 1936. Anwendung der Ahnlichkeitsmechanik und der Turbulenzforschung
auf die Geschiebebewegung. Mitt. Preuss. Versuchsanst. Wasserbau Schiffbau, 2 6 , 2 0 pp.
Sleath, J.F.A., 1975. A contribution to the study of vortex ripples. J. Hydraul. Res., 1 3 :
3 15-3 28.
Sleath, J.F.A., 1976. On rolling-grain ripples. J. Hydraul. Res., 1 4 : 69-81.
Southard, J.B. and Dingler, J.R., 1971. Flume study of ripple propagation behind mounds
on flat sand beds. Sedimentology, 1 6 : 251-263.
Stokes, G.G., 1847. On the theory of oscillatory waves. Trans. Cambridge Philos. SOC.,8:
441.
Tanner, W.F., 1967. Ripple mark indices and their uses. Sedimentology, 9 : 89-104.
Tanner, W.F., 1971. Numerical estimates of ancient waves, water depth and fetch. Sedi-
mentology, 1 6 : 71-88.
Tietze, K.W., 1978. Zur Geometrie von Wellenrippeln in Sanden unterschiedlicher Korn-
grosse. Geol. Rundsch., 6 7 : 1016-1033.
U S . Army Coastal Engineering Research Center, 1 9 7 3 . Shore Protection Manual, 3 vols.
Wells, D.R., 1967. Beach equilibrium and second-order wave theory. J. Geophys. Res.,
7 2 : 497-504.
Wiegel, R.L., 1954. Gravity Wave, Tables of Functions. Council on Wave Research, Eng.
Found. Univ. California, Berkeley, Calif., 3 0 pp.
Wiegel, R.L., 1960. A presentation of cnoidal wave theory for practical application. J.
Fluid Mech., 7 : 273-286.
Marine Geology, 60 (1984) 199-218 199
BOUNDARY ROUGHNESS AND BEDFORMS IN THE SURF ZONE
DOUGLAS J. SHERMAN* and BRIAN GREENWOOD

Department of Geography, Scarborough Campus, University of Toronto, Scarborough,
Ont. M l C 1A4 (Canada)
Departments of Geography and Geology, Scarborough Campus, University o f Toronto,
Scarborough, Ont. M l C l A 4 (Canada)
(Received April 16, 1983;revised and accepted August 31, 1983)
ABSTRACT
Sherman, D.J. and Greenwood, B., 1984. Boundary roughness and bedforms in the surf
zone. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and Sedimenta-
tion in Wave-Dominated Coastal Environments. Mar. Geol., 60: 199-218.
Hydrodynamical models of the nearshore system frequently assume that a single fric-
tion coefficient is sufficient to represent flow conditions at a point in the surf zone.
Furthermore, models attempting to relate bed configuration to surf zone flows have relied
primarily upon the wave orbital velocity as an indicator of potential bedforms, and thus
as the control on boundary roughness. The data presented here point out potential errors
arising from either of these approaches. The results of a field experiment conducted a t
Wendake Beach, Ontario, show that at a single location in an active surf zone, the Darcy-
Weisbach friction coefficient, f, varied by approximately 250% (in this case between
0.016 and 0.041).
I t is also shown that existing bedform models, based upon primary wave motions alone,
do not accurately predict conditions at this study site. For a relatively constant wave
orbital velocity and velocity asymmetry, it is found that changes in bed roughness, as a
result of bedform development, are reflected mainly in the vertical profile of the long-
shore current velocity. A sequence of bedforms, from oscillatory ripples through flat bed,
is inferred from the data, and found to be supported by diver observations and preserved
primary sedimentary structures.
INTRODUCTION
Considerable time and effort has been expended on research into the
nature of the complex interactions at the fluidsediment interface, especially
concerning the development of bedforms. These interactions are important
to both the accurate modeling of modern prototype fluid systems, and the
analysis of bedding genesis and interpretation of sedimentary environments.
However, in nearshore systems dominated by wave-generated oscillatory
flows and wave-induced, quasi-steady currents, our knowledge of fluid-
*Present address: Department of Ocean Engineering, Woods Hole Oceanographic Institu-

tion, Woods Hole, MA 02543, U.S.A.

200
sediment interaction remains rudimentary. According to Allen (1982, Vol. I,

p.444), “Little attempt has been made to define the existance fields of bed-
forms in oscillatory flows”, and about wave-current ripples he goes on to say
(1982, Vol. I, p.448), “These forms are less well known and understood than
their more symmetrical relatives”.
To a large extent, this lack of understanding can be attributed to the
paucity of prototype data describing these bedforms. Although considerable
advances have been made in the last decade or so in formulating conceptual
models of bedform sequences in oscillatory flows (e.g., Clifton, 1976;
Davidson-Arnott and Greenwood, 1974, 1976; Allen, 1982; Clifton and
Dingler, 1984) there is still relatively little empirical evidence to support
these models (see, for example, Clifton et al., 1971; Davidson-Arnott and
Greenwood, 1976; or Dabrio and Polo, 1981, for some prototype results).
Thus, a primary motivation for this research was to improve the data base,
attempt to determine the sequential development of bedforms through a
storm, and recognize the key fluidsediment relationships. Toward this end,
a field experiment was conducted at Wendake Beach, Ontario, in May and
June, 1980.
STUDY SITE AND EXPERIMENTAL DESIGN
The Wendake Beach study site was located on the southeastern shore of
Nottawasaga Bay, in Lake Huron (Fig.1). The nearshore slope is gentle
(2 0.015), with three low-amplitude, nearshore bars. The mean sediment
diameter by weight frequency distribution, D ,is 0.21 mm. Additional details
of this site are presented in Randall (1982), Sherman (1982) and Greenwood
and Sherman (1983, and 1984, this volume). For this field experiment, six
bi-directional, fast-response electromagnetic water current meters (Marsh-
McBirney models 511 and 512) were deployed in two vertical arrays (of
three current meters each) along the edges of the outer nearshore trough. In
vertical array 1 (VAl), about 50 m from the still water line, the current
meter elevations were 0.10, 0.60, and 1.00 m above the bed, and in vertical
array 2 (VA2), about 105 m offshore, the installation elevations were 0.10,
1.00, and 1.45 m (Fig.2). Water depths at the two arrays were 1.60 and
1.70 m, respectively. The purpose of the vertical arrays was to measure an
assumed logarithmic velocity profile in the longshore current. I t has long
been recognized that the structure of these vertical profiles should reflect
the magnitude of boundary roughness (e.g., Nikuradse, 1933;Wooding et al.,
1973), although in the nearshore there are complications arising primarily
from the presence of the wave boundary layer (Grant and Madsen, 1979)
and near-bed sediment transport (Smith and McLean, 1977; Grant and
Madsen, 1982). Stratification due to temperature or suspended sediments
was not considered to be a factor a t this site. There was occasional direct
observation of bed configuration during the storm by divers, and indirect
monitoring through post-storm box core samples taken at the vertical arrays.
The evidence from these measurements and observations allow a t least a
qualitative description of the nearshore flow regime.
201
Wymbol wood
)
Beach
Wendake
Beach
Fig.1. Location map of Wendake Beach study site.
Nine sets of measurements were .obtained from the current meters during
a storm that occurred on June 8, 1980, with specific record lengths of either
nine or eighteen minutes, and with sampling at approximately 0.5 s intervals,
or less. Thus each file for each current meter comprises about either 1000 or
3000 smpJees.
FLOW STRUCTURE AND BOUNDARY ROUGHNESS
Mean flow conditions through the storm were determined in the first stage
of data analysis. From each of the vertical arrays, all of the longshore current
velocity measurements from the set of three current meters were averaged to
obtain a depth-integrated, mean velocity at that location. The velocities at
each elevation were then averaged and compared to the depth-integrated
mean. These results are illustrated in Fig.3. The forms of these profiles indi-
cate the presence of the type of near bed deformation expected, although
the anomalous surface decrease in velocity is not readily explainable. Also
note that the relatively small amount of vertical variability (plus or minus
202
I - CURRENT METER LOCATIONS: LINE 0

E
1
WENDAKE BEACH:1980: 06:08
a
2 -4-
0
te -5-
1 1 1 1
W I 1 1 1 1 1 1
0 20 40 60 80 100 120 140 160 180 200
DISTANCE FROM BASELINE (in)
Fig.2. Current meter locations in the vertical arrays, 1980:06:08.SWL is still water level.
RELATIVE VELOCITY PROFILES WENOAKE BEACH 1980:06:08
VA1 VA2
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.10 0.15 0.20 0.25 0.30 0.35
v (ms-’)
Fig.3. Relative longshore current profiles through a storm, 1980:06:08. Vertical dashed
lines at V A l and VA2 represent depth integrated mean current velocities. Points are mean
velocities relative to the array mean at different elevations above the bed and the associ-
ated numbers are the ratios of point velocities to the depth-integrated average.
about 10% of the overall average) supports the general longshore current
modeling assumption of velocity homogeneity through the water column
(see discussion in Basco, 1982). Nevertheless, the two bottom current meters
in VA2 consistently produced results that were consistent with the presence
of a logarithmic boundary layer. The data from V A 1 do not fit the logarith-
mic model because the velocity gradient between the two elevations is too
203
steep to yield physically plausible boundary roughness lengths. This is be-

lieved to reflect, in part, the relative insensitivities of the two lower current
meters in VA1 (model 511). These instruments have a calibration precision
of only 3 cm s-', whereas the other four current meters (model 512) cali-
brate at k0.3 cm s-'. These terms do not include the &lo% measurement
errors of the instruments, Thus these data are suspect and not considered
further.
Using the 0.10 and 1.00 m current meters (CM6 and CM4, respectively) in
VA2, the apparent roughness length zo,as felt by the longshore current, may
be derived using analytic geometry:
(Vb.lnz,) - ( V a - l n z , )
lnz, =
Vb - va
where V and z refer to the longshore current velocity and elevation above
the bed, respectively, and the subscripts a and b refer to values for the lower
and upper current meters, respectively.
The first analytic complication arises in the determination of the elevation
for the current meter sensors. Although they were originally installed at 0.10
and 1.00 m above the bed, it is known, through data obtained from locally
emplaced, depth-of-disturbance rods (Greenwood et al., 1980; Greenwood
and Hale, 1980) and box core data, that the bed elevation decreased a maxi-
mum of 0.16 m through the storm. Thus, at some stage of the storm, pre-
sumably at its peak, the elevations of the lower current meters, CM6 and
CM4, were 0.26 and 1.26 m above the bed, respectively. These changes in
the values of z have important effects in the derivation of z,. For example,
the natural log of 10 cm is 2.30, whereas In 26 cm is 3.26. This change in
In z can result in apparent changes in the roughness length of several orders of
magnitude. It was therefore decided to attempt t o model bed elevation change
through the storm based upon several assumptions.
First it must be assumed that the maximum amount of bed depression is
coincident with the largest longshore current velocities. For the June 8
storm, this velocity was 0.55 m s-' a t 1130 h (Eastern Daylight Saving
Time). Thus, at this time it is assumed that the elevation of CM4 is 1.26 m
above the bed. I t is also, somewhat arbitrarily, assumed that bed depression
begins when the longshore current velocity exceeds 0.10 m 6'. Thus, when
the current velocity is less than 0.10 m s-l the bed elevation is a constant.
Above 0.10 m s-', the bed elevation is assumed to vary as a function of the
square of the velocity. This assumption is based upon bed shear stress in-
creasing solely with Vz and ignoring potential changes associated with wave
effects.
Given the above qualifications, changes in the value for z for each instru-
ment were calculated through the storm. These changes are shown in Fig.4.
Of the ten values shown, three are measured; the two end points and the
maximum depression. The result of the predictions for the other values is
a reasonable sequence of elevations through the storm. Thus values for both
the mean longshore current velocities (averaged over a record length) and the
204
PREDICTED BED ELEVATION CHANGES 1980.06:08
Time (h)
04:OO 08:OO moo 16:OO 2o:oo 24:OO
O I = -3 I I I I I I I I I
i
c
B 10
3
zW 12
V
o 14
m
l6
18 t
Fig.4. Predicted sequence of bed elevation changes through the storm.
projected height above the bed are available for each of the current meters.
These values are substituted, in turn, into eq.1 to calculate the apparent
roughness lengths at VA2 through the storm.
An independent estimate of the minimum physical roughness of the bed
may be obtained by using grain size information to calculate the Nikuradse
(1933) equivalent sand grain roughness, k,, and thus derive a value for z, for
plane bed conditions. It is assumed here that h, = 2Dso, where DSo is the
mean grain size by weight (0.21 mm at Wendake Beach), and the constant
of proportionality represents the effect of an uneven surface packing of the
sediment on an otherwise plane bed (Yalin, 1972). Further assuming that:
k, = 30 z, (2)
(Schlichting, 1968), the minimum value of z, for the Wendake Beach surf
zone should be about 0.014 mm (minimum k , = 0.42 mm).
Table I presents a summary of the VA2 data used in these calculations and
the results. Note that the difference between the k, estimates of 6.73 X 10”
m at 0910 and 5.26 X 10” m at 0935 may not be significant because the
measurements used for these calculations are at or near the limits of the
current meter accuracy. For example, the only difference in velocities for
the file pair WAV2O and VERT5 is the 0.005 m S-’ velocity at CM6. Because
of the instrument calibration, however, the smallest increment of velocity
that can be measured is 0.003 m s”. Thus these values may be virtually
identical. Note that even this small difference in velocity results in a change
of about 25% in the estimate of k,. Therefore all values of k, presented here
must be considered as approximations only. Figure 5 illustrates the velocity
profiles and the resulting estimates of z,, including the minimum estimate
based upon the grain size procedure. Note also the dashed lines labeled 10
and 100 cm. These indicate the unadjusted instrument elevations. I t can be
seen that the failure to account for bed elevation changes could greatly
reduce the derived estimates of z,. Indeed, most estimates would then fall
below the minimum physical limit set by 2D,0/30.
205
TABLE I
Summary of data used for the derivation of bed roughness estimates, Wendake Beach surf
zone, 1980:06:08. The velocities from CM6 and CM4 are denoted by subscripts a and b,
respectively. The predicted changes in bed elevation are Az. VERT files are of 18 min
duration, WAV files are 9 min long
Filename Time va ub Az 20 ks
(h) (m s') (m s-l) (m) (m x i O - 3 ) (m x lo-*)
VERT2 0420 0.137 0.180 -0.01 0.09 0.28
WAV19 0530 0.313 0.405 -0.07 0.33 0.98
VERT3 0600 0.359 0.450 -0.10 0.24 0.72
WAV20 0910 0.381 0.512 -0.13 2.24 6.73
VERT 5 0935 0.386 0.512 -0.13 1.75 5.26
WAV2l 1130 0.460 0.551 -0.16 0.14 0.42
WAV22 1400 0.267 0.346 -0.09 0.52 1.56
VERT8 1515 0.136 0.181 -0.07 0.65 1.96
WAV24 1900 0.199 0.288 4.08 3.28 9.83
W18 W24 W22 W19 V 3 W21
100 cm
10 cm
198006:08
-7
-6 1 -
_-_-_-_-_-_-_---------------------
2 40
30
Fig.5. Graphical estimation of boundary roughness length from measurements of average

longshore current velocity and estimated elevations above the bed. V and W are file desig-
nators here, representing record lengths of 18 and 9 min, respectively. 20,,/30 is a sand
grain roughness estimate of the minimum physical roughness length.
206
The values of z, shown here range from 0.09 to 3.28 mm. Using the
already mentioned relationship between z, and h,, the range of equivalent
sand grain roughness is thus from 2.8 to 98 mm. These values of h, can be
used to obtain graphic estimates of the Darcy-Weisbach f from the familiar
Moody diagram, where specific values of f are found as a function of the
Reynolds number and relative roughness. Plotting the Wendake Beach data
against this relationship, we obtain the results shown in Fig.6. The arrows
and numbers are to indicate the sequence of findings. The data show a mini-
mum f of 0.016 and a maximum estimate of 0.041, representing a variabil-
ity of about 250%. Most previous attempts at quantifying bed friction in
the surf zone have assumed either a constant value for the entire nearshore
or constant friction coefficient at a given location with spatial variability
according to changes in local slope and relative roughnesses (e.g., Wright et
al., 1982). This is clearly not the case here, where a large variability is found
at a point. These changes may arise from several sources: a change in the
turbulent structure of the flow; a change in the wave boundary layer thick-
ness; a change in the thickness of the near bed sediment transport layer,
changes in bed configuration; or combinations of the above. It is believed
that changes in bed configuration are primarily responsible for the range
of f found in these data, for the reasons presented below.
RESULTS AND DISCUSSION
For the Wendake Beach data, most values of the Reynolds number are
shown to be near or beyond the limit for fully developed turbulent flow
(this limit is indicated by the dashed line curving from upper left to lower
right in Fig.6). For fully rough flow, the Darcy-Weisbach friction factor
depends solely upon the relative roughness (e.g., Vennard, 1961). Therefore
it is not likely that the variability in f is attributable to changes in Reynolds
number.
Grant and Madsen (1979) have theoretically proposed that the presence of
the wave boundary layer within the current boundary layer will result in an
increase in apparent boundary roughness (that felt by the current) from the
physical boundary roughness and enhanced stress. Thus changes in the thick-
ness of the wave boundary layer would be reflected in changes in the appar-
ent bottom roughness length. According to Grant and Madsen (1979), the
thickness of the wave boundary layer ( 6 ) may be approximated by:
6 =2K lU*,,l/O (3)
where K is the Von Karman constant (0.4),w is the radian frequency, 2n/T,
and I U * ~ , I is the shear velocity due to wave and current interaction:
207
0.05 0.02
0.04
10-2
$ 0.03
“
Y
0.025
c
0 4P
u
‘L
Y
0.02 I 0-3
0.01 5
0.0I 2 I0-4
104 2 105 107
Fig.6. Wendake Beach storm data plotted o n a Moody Diagram. Arrows and numbers indi-
cate the sequence off. Values o f f range from about 0.016 to about 0.041.
where f,, is the combined wave and current friction coefficient, and for
trans-directional flow:
a!= 1 + (u/u,)2 (5)
where u is the steady current velocity and urn is the wave orbital velocity:
urn = y(gh)”2 (6)
where g is the gravitational constant, h is water depth, and y is the wave
(amplitude) breaking criterion (0.4). Grant and Madsen (1979) also propose
that f,, is a function of u/u, and k,/Ab, where A b is the horizontal orbital
amplitude, A b = u,/o. From these relationships it can be seen that, for acon-
stant water depth and bottom roughness, 6 will depend only upon T and u.
Changes in u affect the ratio of the steady current to the oscillating cur-
rent, u/urn.This ratio is one control on f,, (Grant and Madsen, 1979). How-
ever, for trans-directional flow, relatively large changes in u/u, change f,,
only slightly. For the range of u/u, measured at Wendake Beach (Table 11),
estimates of f,, change by only about 10% and the ratio is almost constant
through the peak of the storm. It is therefore presumed that for these re-
sults, the wave boundary layer thickness is dependent primarily upon wave
period. As wave period increases, so should the boundary layer thickness.
For the Wendake Beach data, through the storm, the peak wave period
varied from a minimum d 2.3 s to a maximum of 6.4 s, a variability of
about 69%. However, through the middle of the storm, when there were
large changes in apparent roughness, the mean period varied only about 17%,
208
TABLE I1
Predicted and measured orbital velocities and relative velocities at VA2, 1980:06 :08,
Wendake Beach
File urnax* urn, Val urns

( m s-:) ( m s-')
VERT2 1.24 1.09 0.13

WAV19 1.03 1.08 0.29
VERT3 1.53 1.12 0.32
WAV2O 1.26 1.07 0.34
VERT5 1.40 1.12 0.35
WAV21 1.37 1.18 0.39
WAV22 1.37 1.12 0.24
VERT8 1.15 0.96 0.15
WAV24 1.18 1.05 0.19
~
*Ford = 2.00 m, Urn (predicted) = 1.73.
between 5.3 s and 6.4 s. Further, T increased from 5.3 s at 0910 to 6.4 s at
1130 while k , dropped from 6.73 to 0.42 cm. Thus, while there is no doubt
that the presence of the wave boundary layer is reflected in the measure-
ments leading to the derivation of z, (or k s ) , it alone cannot account for
either the direction or magnitude of variability in the boundary roughness.
Smith and McLean (1977), and Grant and Madsen (1982), have shown
how apparent roughness is also affected by the thickness of the near-bed
sediment transport layer. According to Grant and Madsen (1982) the thick-
ness of this layer, as reflected in its contribution to the total roughness, is
approximated by:
hbs = 160 (s + c,)Dl//,
'''),)I/'$([ - 0.71' (7)
where kbs is the sediment transport induced roughness, s = p s / p , where ps is
the density of the sediment and p is the fluid density, C , is an added mass
coefficient (taken as 1 / 2 for spheres), $, is the critical value of the Shields
parameter for the initiation of motion and l//' is the maximum value. Grant
and Madsen (1982, p.471) define:
b!" = T k / P (S - I)@ (8)
where T L is~ the maximum value of the skin friction under the wave:
where fk is based upon k , and T. Therefore, for constant sediment character-

istics, and with an otherwise unchanging bed roughness, the thickness of the
near bed sediment transport layer depends upon urn and T only, and as
noted above (eq.4), urn is a function of water depth in the surf zone, and
T is relatively stable through the storm. Thus, in an otherwise constant
209
environment, changes in h, or z, cannot be solely accounted for by changes

in kbs.
It should be noted that even under optimum conditions, i.e., with several
high-resolution current meters in an array and the bottom precisely defined,
large errors in z, (hence h,) are to be expected. Grant et al. (1983) working
on the continental shelf, used four acoustic current meters to measure water
velocities and an echo sounder to sense the bottom. Regression lines fit to
the data with R Z > 0.99 showed estimates of z, to vary on the order of
2100% a t the 95% confidence level. Uncertainty as to the bed elevation at
Wendake Beach, and the overall precision of measurement required to
accurately locate z,, imply that values of bed roughness reported here be
used only as indicators, not absolutes. However, the order of magnitude
variability in roughness estimates (Table I) is much greater than the potential
error terms alone can account for. Therefore, although specific values of h,
(or z,) may be inaccurate, the temporal variability is real.
As expected, attention may therefore be focused upon changes in the
surf zone bed configuration as being the primary agent affecting measure-
ments of z,. It is assumed that the major influence on bed deformation is
the wave orbital velocity, with the current adding a secondary, but impor-
tant contribution to the total stress. Indeed most attempts to formulate
models of bed form development due to periodic waves (Clifton, 1976;
Davidson-Arnott and Greenwood, 1976; Allen, 1982; Clifton and Dingler,
1983) rely, at least in part, on the use of orbital velocity and mean grain
diameter as the main variables. Using Clifton’s (1976) model, the bed form
thresholds are found t o be functions of urn, D ,and, in some cases, T. For
the threshold of sheet flow and flat bed, Clifton has used Dingler’s (1974)
data to set this limit based upon grain size. The critical orbital velocity for
transition to sheet flow is determined from the relationship:
k c - (3.880 x 1 0 y (10)
where u, is the threshold velocity. Using this relationship and the Wendake
Beach mean grain diameter of 0.021 mm, the critical orbital velocity for plane
bed configuration is about 0.90 m s-’.
Estimates of the Wendake Beach orbital velocities are obtained from linear
wave theory, using eq.4, and from the current meter data that provide a
maximum measured velocity (urn=) and a predicted velocity, u,, = 2.8 s,
where s is the standard deviation of the current meter record and 2.8 is an
H,,, analogy (after C.E.R.C., 1977, pp.3--12). This yields a conservative
estimate for the orbital velocity. From eq.4, the predicted value of u, is
1.73 m s-’. The results obtained from the current meter record through the
storm are presented in Table 11. For all data records summarized in Table 11,
the values of urn exceed the threshold velocity given by eq.8. Thus, accord-
ing to Clifton’s (1976) model, the bed configuration throughout the storm
is predicted as flat, and therefore could not account for the measured vari-
ability in 2,.
Clifton’s (1976) model also attempts t o define the effects of wave orbital
210
asymmetry, A u , , on bedform development. His work suggests that for a

given wave period and sediment grain diameter, the nature of bed deforma-
tion is a function of the relationship between u, and nu,. Again, however,
the current meter data indicate that there was little variability in Au,,,
through the storm. Indeed this factor was almost constant when changes in
z, were greatest, near the peak of the storm.
Of the surf zone velocities measured, only the longshore current showed a
substantial variation through the monitoring period. Figure 7 illustrates the
measurements of V, u,, and A u , ( A u , is plotted as the mean flow asym-
metry, after Dingler, 1974) through the storm. Given the relative uniformi-
t y of the latter two variables, they are considered constants against which
changes associated with current can be evaluated.
Diver observations and box core data, taken across the surf zone, clearly
indicate that plane bed conditions were not present at all times throughout
the storm. There was instead a sequential development of bedforms, and
this sequence seems t o qualitatively correspond with changes in longshore
current velocities (implicitly superimposed on a constant urn). Figure 8
VARIABILITY OF SURF ZONE VELOCITIES WENDAKE BEACH 1980:06:08
-
0.9 -
0.8 -
- 0.7 -
E
-
- 0.6 -
._2.
r
m
0.4
0.3
0.2
0.1
0.0
04:OO 08:OO 12:oo 16:OO 2o:oo
Time [h)
Fig.7. Sequence of surf zone velocity measurements through the storm, 1980:06:08.
211
LONGSHORE CURRENT VELOCITY AND ROUGHNESS LENGTH WENDAKE BEACH 1980:06:08

60 0.0
50 -1.0
40 -2.0
-- -
-$ 30 -3.0 2E,
> -c
20 -4.0
i
10 -5.0
I 1 I I I I I I I I I -6.0
0
0o:oo 04:OO 08:OO 12:oo 16:OO 20:oo 24:OO
Time (h)
Fig.8. Concurrent changes in longshore current velocity and apparent roughness length,
1980:06:08.
shows the concurrent changes in the current velocity, V, and lnz,, through
the storm. There is a general, coincident increase in V and In z, up to about
50 cm s-I. Then the incremental increase in velocity (at 1130 h) is accom-
panied by a large decrease in the apparent roughness length. Thereafter, the
lower velocities are associated with increasing roughness. These changes in
bed roughness are primarily attributed to changes in bed configuration, a
finding in contradiction with nearshore bedform development models
employing U , only.
A tentative means of identifying the bedforms is offered in Fig.9. Here,
the natural log of the equivalent sand grain roughness is plotted through the
storm. The upper pair of lines represent predicted roughness values obtained
using relationships between ripple height and k,. From a limited data set
obtained by direct measurement and from box core data, a typical ripple
height, 77, for Wendake Beach varied between 1.5 and 2.5 cm. According to
Stefanick (as cited in Jonsson, 1980) a reasonable approximation of the
bedform induced boundary roughness is h, = 2.5 77. This relationship is
plotted for the pair of estimated 77. Working backwards, these lines, there-
fore, serve as indicators of the expected bed configurations, given a mea-
sured estimate of k,. There are pairs of points at the beginning and end of
the series to represent the variability in k, that arises from a 2 cm error in the
estimate of z use3 to obtain z,. These represent maximum errors for the
212
EQUIVALENT SAND GRAIN ROUGHNESS WENDAKE BEACH 1980:06:08

I
I
2.0 I
1.0
-
5
I
Y
= o
-1.0 I- I
I
-2.0 I I I I I 1 I I I I I I
0o:oo 04:OO 08:OO 12:oo 16:OO 2o:oo 24:OO
Time (hi
Fig.9. Equivalent sand grain roughness estimates and roughness values attributed to spe-
cific bedforms. The upper dashed lines are ripple values.
series because the relative change in In z is greatest for the smaller values of
z. The value of 2 cm was selected as being approximately the potential ripple
height migration effect on bed elevation.
The information presented in Fig.9 suggests that for most observations
ripples were present. This is not surprising, as the ripples occur at the be-
ginning and end of the storm sequence. The lower and upper end points
of the first and last values of h,, respectively, are considered errors based
upon physical considerations of what was occurring at those times. A larger
than rippled bed is present at 0910 h, with a large decrease in roughness at
1130 h. This drop in roughness, associated with increased velocities, is
assumed to indicate a change to a flat bed configuration. The difference
between this roughness derived from the velocity profiles and the grain
roughness (4.2 and 0.42 mm, respectively) is attributed to nearbed sedi-
ment transport. These results ignore any possible effects of bedform orien-
tation (relative to the flow) on boundary roughness.
Based upon this analysis, and field observations, inferred bed form types
~ 1/2 p f V 2 ;with f found from Fig.6) and
are plotted with shear stress, T~ ( T =
V (Fig.10). The numbers at each point are to identify the sequence of the
data. This figure shows two clusters of supposed ripples, two megaripple
predictions and one value at plane bed. The change of position between the
213
SHEAR STRESS, CURRENT VELOCITY
AND INFERRED BED FORMS
4
100 - 45
9 2 0 3
-
7
0 7 0
E
E
hio
010 - 80
1.
0 , 0 =RIPPLES
A =MEGARIPPLES
:FLAT BED
0 01
0 00
u 0 20 0 40 0 60
V (ms-’)
Fig.10. Boundary shear stress, longshore current velocity and inferred bedforms. Numbers
indicate the sequence o f data.
ripple clusters is attributed to changes in the thickness of the near-bed sedi-

ment transport layer due to changing bed roughness and velocity. Note also
that among the six ripple observations, there is a distinct segregation be-
tween points from the waxing and waning limbs of the storm (as indicated
by the point numbers). Points 1, 2 and 3 lie almost in a straight line, and
points 7, 8 and 9 are all above that line. This corresponds well with what is
found in studies of river flow regime, where the rating loop effect is well
documented (Allen, 1982). This effect is a function of the response time of
the bed to changes in flow. With decreasing current velocities, the bedforms
change more slowly, response time is increased, and thus relict roughnesses
may be present during velocity measurements. The rating loop effect also
seems to be indicated in the Fig.8 data.
Figure 11 is Southard’s (from Middleton and Southard, 1977) conceptual
model of the relationships between bed form, velocity ( U ) , and boundary
shear stress ( T ~ ) In
. this model it is shown that a given current velocity is asso-
ciated with a unique bedform. For increases in velocity within a specific bed-
form class, there is a relatively uniform change in bed shear stress [e.g.,
increasing in the ranges of ripples (R) and sand dunes (SO)]. There are also
relatively abrupt changes in T~ associated with small changes in U in the
transition between bedform regimes. Of particular interest is the rapid
decrease in T,, despite an increase in U , with the change from dunes t o flat
bed (3’). Note that in general, the data shown in Fig.10 follow a similar
form of curve. Figure 1 2 is the Wendake Beach data plotted in a manner
214
RELATIONSHIP BETWEEN MEAN FLOW VELOCITY u

AND BOUNDARY SHEAR STRESS to
U
Fig. 11. Southard’s (Middleton and Southard, 1 9 7 7) conceptual model of the relationship
between mean flow velocity, boundary shear stress and bedforms. R is ripple, SD is sand
dunes, F is flat bed, and A is antidunes.
PREDICTED BED FORM TRANSITIONS
WENDAKE BEACH 1980:06:08
I
1.00-
N
-
E
z
v
0.10-
0 =RIPPLES
A =MEGARIPPLES
+ =FLAT BED
I I 1
0.0lI
0.00 0.20 0.40 0.60
v (ms-‘)
Fig.12. Predicted bedform transitions through a storm sequence, as interpreted from the
Wendake Beach data and Southard’s conceptual model (Fig.ll). All gradients and cut-off
points are estimates.
corresponding with that used by Southard. These data show changes in 7,

with changes in V that are gradual within bedform groups and with steep
gradients between the inferred groups. The cross-bars indicate the estimated
breaks between groups. This analysis is speculative, of course, because of the
215
small size of the data set and the uncertainties with some of the measure-
ments. Nevertheless, the concurrence of this progression with that suggested
by Southard indicates that these findings are not unreasonable.
More concrete evidence of the propriety of the relationship suggested in
Fig.12 is obtained from analysis of an epoxy relief peel of a box core taken
at VA2 after the storm. This peel (Fig.13) clearly shows a quasi-planar
structure at -0.16 m. This is at the maximum depth of disturbance, as
indicated above. The presence of the plane lamination at -0.16 m (equiva-
lent to a flat bed configuration) reinforces the validity of the method used
for measuring bed roughness, as this is the configuration predicted from the
data. Further, if bedforms produced during the falling limb of the storm at
least partially reflect the rising limit sequence, the supposition of the sequen-
tial development of ripples through flat bed is also supported. Although the
presence of megaripples is not clearly indicated in the peel (see Davidson-
Arnott and Greenwood, 1976, figs.7e and 8b for some examples), some form
of large-scale, rhythmic roughness is apparent. Allen (1981) describes similar
structures, undulatory laminations, that he attributes to the existence of
low-amplitude, rolling grain ripples, occurring subsequent to flat bedding
and before vortex ripples appear (in a waning sequence). This was perhaps
the case at Wendake Beach. Small scale ripple cross-lamination is present
in the upper 3 or 4 cm of the peel. These last forms are the product of
oscillatory ripple formation toward the end of the storm.
CONCLUSIONS
Although wave orbital velocity is the primary agent responsible for gener-
ating the shear stress that causes bed deformation in most wave-dominated
nearshore zones, the magnitude and variability in wave parameters are not
sufficient to correlate directly with changing bedforms. Indeed, as Grant and
Madsen ( 1979) suggest, wave-current interaction can substantially enhance
the stress over that predicted for waves alone. However, present models
relating nearshore bedforms t o flow conditions consider only the direct
wave effects. For example, Clifton and Dingler (1984, this volume) note that
their model is designed t o operate only in the absence of secondary, unidirec-
tional flows, and is thus inappropriate for the description of conditions in near-
shore zones where such currents are present. The formulation of a comprehen-
sive wave and current model may be possible using a format similar t o that
shown in Figs. 11and 1 2 where a total contribution t o T,, and V by the waves
and current can be determined. This is, however, not attempted here, because
it is beyond the intent of the present study. Where wave and current charac-
teristics are unknown, approximations using existing theories will have to
be employed, but they must also predict the magnitude of the longshore
current. Further, a great deal of additional field data is required to accu-
rately fix the relationships and limits shown in Fig.12 (indeed, to see if they
are real). Finally, it has been shown that apparent bed roughness, as reflected
by the Darcy-Weisbach friction coefficient, can vary at least 250% a t a given
216
Fig.13. Epoxy peel from box core taken a t VA2 after the 1980:06:08 storm. Near-plane
bed configuration is present at maximum depth of disturbance as predicted from data.
Large-scale undulatory laminations may be associated with post-vortex, rolling grain
ripples (Allen, 1981). Small-scale ripple cross-stratification is present near the t o p of the
peel. Slight landward slope at -0.16 m is due t o core position o n landward slope of the
outer bar. Some distortion due t o sampling appears along the edges of t h e peel. Area
shown is approximately 3 5 x 25 cm.
217
location in the surf zone ( a t Wendake Beach, f values ranged from 0.016 t o
0.041 in 2 m of water). This result implies that care must be taken in apply-
ing models that assume a uniform friction coefficient t o the solution of surf
zone flow problems. Ideally, future models will not consider this parameter
to be a constant.
ACKNOWLEDGEMENTS
This study forms a part of a continuing research programme at Scarborough

College in Nearshore Hydrodynamics and Sedimentation supported by both
capital (E39218, E6614) and operating grants (A7956) from the Natural
Sciences and Engineering Research Council of Canada awarded t o B.G.
University of Toronto Open Fellowships and an Ontario Graduate Scholar-
ship are acknowledged by D.J.S. Assistance with computer costs was given
by both Scarborough College and the Department of Geography, University
of Toronto. The Academic Workshops at Scarborough College assisted with
instrument design and construction, while the Graphics and Photography
Division of the College produced the illustrations. We gratefully acknowl-
edge field and laboratory assistance from M. Rollingson, P. Christilaw, J.
McDonnell and R. Sutherland, Scarborough College and R.G.D. Davidson-
Arnott and D. Randall, Department of Geography, University of Guelph.
Gretchen McManamin, Woods Hole Oceanographic Institution, typed the
manuscript. J.R. Dingler (USGS) and W.D. Grant (WHOI) are thanked for
their constructive criticism. The results and interpretations presented in this
paper are, however, the sole responsibility of the authors.
REFERENCES
Allen, J.R.L., 1982. Sedimentary Structures: Their Character and Physical Basis, Vols. I
and 11. Elsevier, Amsterdam, Vol. I, 593 pp., Vol. 11, 663 pp.
Allen, P.A., 1981. Wave-generated structures in the Devonian lacustrine sediments of
Basco, D.R., 1982. S u r f Zone Currents, Vol. I : State of Knowledge. US. Army Corps of
Engineers, Coastal Engineering Research Center, Miscell. Rep. 82-7 (I), 243 pp.
C.E.R.C., 1977. Shore Protection Manual. US. Army Corps of Engineers, Coastal Engi-
neering Research Center, Ft. Bevoir, Va.
Clifton, H.E., 1976. Wave formed sedimentary structures - a conceptual model. In: R.A.
Clifton, H.E. and Dingler, J.R., 1984. Wave-formed structures as interpretive tools. In:
Wave-Dominated Coastal Environments, Mar. Geol., 60: 165-198 (this volume).
Clifton, H.E., Hunter, R.E. and Phillips, R.L., 197 1. Depositional structures and processes
in the non-barred high energy nearshore. J. Sediment. Petrol., 41 : 651-670.
Dabrio, C.J. and Polo, M.D., 1981. Flow regime and bedforms in a ridge and runnel
system, S.E. Spain. Sediment. Geol., 28: 97-110.
218
Davidson-Arnott, R.G.D. and Greenwood, B., 1976. Facies relationships on a barred coast,
Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davis and R.L. Ethington
Publ., 24: 149-168.
Dingler, J.R., 1974. Wave-formed ripples in nearshore sands. Ph.D. Dissertation, Univer-
sity of California, San Diego, Calif., 136 pp.
Grant, W.D. and Madsen, O.S., 1979. Combined wave and current interaction with a
row& bottom. J. Geophys. Res.,84 <c4\.. 1797-18Q8.
Grant, W.D. and Madsen, O.S., 1982. Moveable bed roughness in unsteady oscillatory
flow. J. Geophys. Res., 87 ( C l ) : 469-481.
Grant, W.D., Williams 111, A.J., Glenn, S.M., Cacchione, D.A. and Drake, D.E., 1983. High
frequency bottom stress variability and its prediction in the CODE Region. Woods
Hole Oceanographic Institution, Tech. Rep. WHOI-83-19, 72 pp.
Canada: Littoral Processes and Shore Morphology. Geol. Surv. Can. Pap., 80-10: 89-
109.
Greenwood, B. and Sherman, D.J., 1983. Shore-parallel flows in a barred nearshore. Proc.
18th Coastal Engineering Conf., Cape Town, A.S.C.E., pp.1677-1696.
Greenwood, B. and Sherman, D.J., 1984. Waves, currents, sediment flux and morpho-
logical response in a barred nearshore system. In: B. Greenwood and R.A. Davis, Jr.
ments. Mar. Geol., 60 : 3 1-6 1 (this volume).
nearshore zone: prototype measurements. In: Workshop on Instrumentation for Cur-
rents and Sediments in the Nearshore Zone, National Research Council of Canada,
Associate Committee for Research on Shoreline Erosion and Sedimentation, Ottawa,
Ont., pp. 99-120.
Jonsson, I.G., 1980. A new approach to oscillatory rough turbulent boundary layers.
Ocean Eng., 7 : 109-152.
Middleton, G.V. and Southard, J.B., 1977. Mechanics of Sediment Transport. SOC.Econ.
Paleontol. Mineral., Short Course Notes, 3.
Nikuradse, J., 1933. Laws of Flow in Rough Pipes. National Advisory Committee for
Aeronautics Tech. Memo. 1292 ( 1 950 translation).
Randall, D.C., 1982. Changes in the form and energy spectra of storm waves across a
nearshore bar system, Wendake Beach, Nottawasaga Bay, Ontario. M.Sc. Diss., Uni-
versity of Guelph, Guelph, Ont., 133 pp.
Schlichting, H., 1968. Boundary Layer Theory. McGraw-Hill, New York, N.Y., 748 pp.
Sherman, D.J., 1982. Longshore currents: A stress balance approach. Ph.D. Diss., Univer-
sity of Toronto, Scarborough, Ont., 208 pp.
Smith, J.D. and McLean, S.R., 1977. Spatially averaged flow over a wavy surface. J. Geo-
phys. Res., 8 2 : 1735-1746.
Vennard, J.K., 1961. One-dimensional flow. In: V.L. Streeter (Editor), Handbook of
Fluid Dynamics. McGraw-Hill, New York, N.Y., pp.3-1-3-30.
Wooding, R.A., Bradley, E.F. and Marshall, J.K., 1973. Drag due to regular arrays of
roughness elements of varying geometry. Boundary Layer Meteorol., 5: 285-308.
Wright, L.D., Nielsen, P., Short, A.D. and Green, M.O., 1982. Morphodynamics of a
macrotidal beach. Mar. Geol., 50: 97-128.
Yalin, M.S., 1972. Mechanics of Sediment Transport. Pergamon, New York, N.Y., 290 pp.
Marine Geology, 60 ( 1 9 8 4 ) 219-233
219
TIDAL-CYCLE CHANGES IN OSCILLATION RIPPLES ON THE INNER

PART OF AN ESTUARINE SAND FLAT
J.R. DINGLER and H.E. CLIFTON

U.S. Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025 (U.S.A.)
(Received March 14, 1983; revised and accepted December 17, 1983)
ABSTRACT
Dingler, J.R. and Clifton, H.E., 1984. Tidal-cycle changes in oscillation ripples on the
inner part of an estuarine sand flat. In: B. Greenwood and R.A. Davis, Jr. (Editors),
Geol., 60: 219-233.
Oscillation ripples form on subaqueous sand beds when wave-generated, near-bottom

water motions are strong enough t o move sand grains. The threshold of grain motion is
the lower bound of the regime of oscillation ripples and the onset of sheet flow is the
upper bound. Based on the relation between ripple spacing and orbital diameter, three
types of symmetrical ripples occur within the ripple regime. In the lower part of the
ripple regime (orbital ripples), spacing is proportional to orbital diameter; in the upper
part (anorbital ripples) spacing is independent of orbital diameter. Between these regions
occurs a transitional region (suborbital ripples).
Oscillation ripples develop on a sandy tidal flat in Willapa Bay, Washington, as a result
of waves traversing the area when it is submerged. Because wave energy is usually low
within the bay, the ripples are primarily orbital in type. This means that their spacing
should respond in a systematic way to changes in wave conditions. During the high-water
parts of some tidal cycles, ripples near. the beach decrease in spacing during the latter
stage of the ebb tide while ripples farther offshore d o not change. Observations made
over several tidal cycles show that the zone of active ripples shifts on- or offshore in
response t o different wave conditions.
Detailed bed profiles and current measurements taken during the high-water part of
spring tides show the manner in which the oscillation ripples change with changes in
orbital diameter. Changes in ripple spacing at the study site could be correlated with
changes in orbital diameter in the manner suggested by the criterion for orbital ripples.
However, there appeared t o be a lag time between a decrease in orbital diameter and the
corresponding decrease in ripple spacing. Absence of change during a tidal cycle could
be attributed to orbital velocities below the threshold for grain motion that negated the
effects of changes in orbital diameter.
Because changes in sand-flat ripples depend both upon changes in orbital diameter
and upon the magnitude of the orbital velocity, exposed ripples were not necessarily
produced during the preceding high tide. In fact, some ripples may have been just pro-
duced, while others, farther offshore, may have been produced an unknown number of
tides earlier. Therefore, when interpreting past wave conditions over tidal flats from low-
tide ripples, one must remember that wave periods have to be short enough to produce
velocities greater than the threshold velocity for the orbital diameters calculated from
the observed ripple spacings.

220
INTRODUCTION
Wave size, wave shape and water depth, in concert with grain size, deter-
mine the size and shape of wave-formed ripples (Inman, 1957; Dingler,
1974; Clifton, 1976; Miller and Komar, 1980). Little is known, however,
about the way ripple-form responds in the field to changing wave condi-
tions. For example, do ripples continuously adjust and thereby remain in
constant equilibrium with changing wave conditions, or do they retain a
metastable form until some threshold for change is exceeded? In the process
of changing t o meet new conditions, does the spacing continuously expand
or contract or do the old ripples influence the size of the new ones? Such
questions are highly relevant to the interpretation of wave-formed ripples
in paleoenvironmental analysis (Clifton and Dingler, 1984, this volume).
The specific relation between flow parameters and ripple size and shape is
also important t o studies of the movement of nearshore sand.
This paper chronicles the change in oscillation-ripple size that attends the
rise and fall of the tide on a tidal flat of an estuary in the United States
Pacific Northwest. We document the character of the change and relate it
to the variation of wave character during the tidal excursion.
BACKGROUND
Active oscillation ripples exist for a definable range of wave conditions,

called the ripple regime (Dingler and Inman, 1977). With respect to sedi-
ment transport, the threshold of grain motion is the lower boundary of
the ripple regime, and the onset of sheet flow the upper boundary. Bagnold
(1946) first quantified the threshold of grain motion for an oscillating flow,
and several investigators (e.g., Komar and Miller, 1973; Dingler, 1974) have
carried that work to the point where the threshold can be accurately calcu-
lated for a wide range of grain sizes and wave conditions. Dingler and Inman
(1977) presented a dimensionless criterion for the onset of sheet flow
based on field work in fine sand.
Three parameters - spacing A , height 17, and asymmetry /3/A - define the
size and shape of an oscillation ripple for sand of median grain size.D, and
density p,. Wave period T and either orbital diameter do or maximum orbital
velocity u, define the near-bottom flow conditions. Based on the work of
Dingler (1979) the equation
do = 14.3 (T3D,)'.' (1)
defines the minimum orbital diameter required to move quartz sand in sea
water.
Inman (1957) empirically defined the ripple regime in a plot of ripple
spacing versus orbital diameter for various grain sizes. Clifton (1976) sec-
tioned a form of that plot into orbital, suborbital, and anorbital ripples.
Orbital ripples occur in the lower part of the ripple regime, and their spacing
is directly proportional t o orbital diameter. Anorbital ripples occur in the
221
upper part of the ripple regime, and their spacing is independent of orbital
diameter. Suborbital ripples are a transitional form between orbital and
anorbital ripples; some investigators, though, question their existence (e.g.,
Allen, 1979).
The spacing of orbital ripples is given by
X = Cdo (2)
where C is a constant between 0.65 (after Miller and Komar, 1980) and 0.8
(Komar, 1974). The maximum ripple spacing, which defines the upper limit
for orbital ripples, can be calculated from the equation
x = 14.7 x 1030,y* (3)
where and D, are both measured in centimeters (Miller and Komar,
1980). Dingler and Inman (1977) stated that the ratio of ripple height to
spacing for orbital ripples was a constant such that
q / h = 0.16 (4)
LOCATION
The study area is part of a sandy tidal flat on the west side of the Bay
Center Peninsula in Willapa Bay, Washington (Fig.1). The bay, which is
about 40 km long, is about 6.5 km wide at the study area. Because sediment
moves northward along this section of coast, the entrance to Willapa Bay lies
near the north end of the bay, west northwest of the study area.
A bluff composed of Pleistocene terrace deposits runs along the west side
of the Bay Center Peninsula, and, in many places, a narrow beach sits at the
base of the bluff. A well-defined break in slope and a concentration of
coarse debris marks the toe of the beach. From this toe a sandy tidal flat
extends bayward for about 1 km with a gradient of 10-20 cm per 100 m.
The median diameter of the sand on the inner part of the flat is 0.0165 cm.
Figure 2 shows a profile across the beach from the bluff to a distance of
270 m offshore from the toe of the beach.
Our experiments took place during spring tides in the month of August
1976 and 1977, a time when the local winds were variable. Tides were semi-
diurnal with a maximum range of 2.7 m; at high water on observational days,
one to one-and-a-half meters of water covered the study site. Under fair-
weather conditions, when our experiments took place, the wind was from
the north and northwest with wind speeds of 5 m s-' or less recorded about
50 km south at Astoria, Oregon.
Most of the waves traversing the sand flat are generated within Willapa
Bay because the energy of oceanic waves dissipates over the shoals at the
mouth of the bay. Under particularly calm conditions, waves with periods
of 8-10 s, breaking on the flats or beach, reflect the remnants of oceanic
swell. These long-period waves, however, are typically of very small ampli-
tude and generate smaller near-bottom water velocities than those produced
within the bay by wind blowing over fetches of 5-10 km. The waves observed
at the study site were mostly of a period of a few seconds and no more than
a few tens of centimeters high.
9
2
.-a
cn
N
w
U
0
3
c
A
n
(n
e r a 0
EVATION
,
I
Fig.2. Profile of beach and tidal flat at site of ripple study. EMCM means electromagnetic
current meter. Elevations are relative to a marker at the base of the cliff.
223
PRELIMINARY OBSERVATIONS
From preliminary observations of sedimentary processes on the sand flat,

we discovered that the ripples present when the innermost flats were sub-
merged often differed in size from those present when the flat was next
exposed. We confirmed this observation by comparing ripples encased at
high water with those exposed in the same area during the ensuing low water
(Fig.3A, B). Also, at high water the positions of successive ripple crests were
marked with short segments of copper wire; when the sand flat was next
exposed the spacings between wires were greater than those of the exposed
ripples.
These observations verified that ripple spacing commonly decreased sig-
nificantly on the tidal flats near the toe of the beach during the fall of the
tide (Fig.3A). Rather surprisingly, the spacing of ripples a few tens of
meters farther onto the flat did not diminish (Fig.3B). The ripples exposed
there at low tide were flatter and more rounded, but typically neither
migrated nor decreased in spacing during the fall of the tide.
FIELD METHODS
To establish the basis for and the nature of the change in ripple spacing,
and to understand why spacing decreased selectively near the toe of the
beach, we carefully monitored the ripples and the wave-generated, near-
bottom flow parameters during several tidal cycles. The measuring equip-
ment consisted of a high-resolution sonar to make detailed profiles of the
bottom and an electromagnetic current meter to measure near-bottom flow
velocities. The interpretations hereinafter rely on the sonar and current-
meter data in conjunction with visual observations.
The current meter, which is a commercially available, two-axis unit with
a 2.5 cm diameter sensor mounted at the end of a rod, was mounted to the
sonar frame. The center of the sphere was 20 cm above the bed, which is
above the vortex layer for ripples of the size found on the sand flat (Dingler,
1974).
Dingler et al. (1977) detailed the high-resolution sonar used in this study.
The unit, which spanned two meters of the sand bed, consists of a trans-
ducer mounted in a frame and a small electronics package. The skeletal
nature of the frame (Fig.4) permits easy underwater transport and minimizes
flow disturbance. The sonar head, which contains the transducer mounted
in a plastic housing, moves in a channel by means of a manually operated
chain drive. During this study, the transducer sat 28 cm above the bed, and
the frame was oriented with its long dimension in a shore-normal direction.
Resolution is of the order of a millimeter, which is necessary because ripples
in the study area typically had spacings of less than 10 cm and heights of
less than 2 em.
224
Fig.3. “High-tide’’ ripples compared with ripples normally exposed a t low tide. A box
placed upside-down o n the bottom (outline of box is in center of photographs) a t high
water preserved t h e ripples through t h e falling tide. Beach is t o the left of b o t h p h o t o -
graphs. A . Ripples close t o the t o e of the beach. Note dramatic change in spacing. Scale
is 60 cm long, including t h e case. B . Ripples a few tens of meters bayward from A . Note
t h a t neither ripple spacing nor position changed during t h e fall of the tide. Scale is 16 cm
long.
225
Fig.4. High-resolution sonic profiler t h a t was used to measure ripples during the study.
Beach is t o the right. T h e frame is two-meters long; it has mounted upon it an electro-
magnetic current meter ( A ) . T h e transducer, which rides in the aluminum channel, is
driven b y the hand crank ( B ) .
Electrical cables connected the sensors t o a pulse-code-modulated encoder

onshore. Wave data were digitized at either 16 (1976) or 8 samples per s
(1977) and sonar profiles at 64 samples per s. Data collection took place
aperiodically over an approximately four-hour span when the sensors were
underwater; each set of measurements consisted of a minimum of 1 0 sonar
scans and/or a 5-min (minimum) time series of the wave field. All of the
encoded data were recorded on quarter-inch magnetic tape for subsequent
analysis.
DATA ANALYSIS
Spectral analysis of the current-meter time series produced estimates of

the dominant wave period and near-bottom velocity. Reformatting the
data occurred, when necessary, before spectral analysis; each 1976 current-
meter record was subsampled so that all time series were 2048 points long at
8 samples per s.
Each of the U (shore-normal) and V (shore-parallel) spectra were derived
by ensemble averaging four complete spectral functions, each computed
from a different quarter of the appropriate time series. This procedure
produced spectra with 256 spectral coefficients; these were then smoothed
with a five-point, running binomial filter (Otnes and Enochson, 1978) and
plotted as in Fig.5. The dominant wave period occurs a t the highest spectral
226
VELOCITY SPECTRA
Willapa B a y , Washington
Fig.5. Wave-velocity spectra (shore-normal and shore-parallel components of velocity)

taken during the period of high tide at observation site on 25 August, 1976 beginning at
1510 h.
peak, and orbital velocity is proportional to the area under the spectral curve
in the region of the dominant period. Because the current meter may have
been oriented at an angle to the wave orthogonals, the real velocity was
taken to be the sum of the shore-normal and alongshore velocity com-
ponents, especially because spectral analysis removes the mean current.
Because the ripple data had been recorded on two encoder channels-
elevation on one and distance along the bed on the other- the profiles
were plotted onto graph paper for analysis (Fig.6). Counting the peaks on
a ripple profile and dividing that number into the distance between the
first and last peak gave an average ripple spacing. Then, the ripple traces
were redigitized to produce 512 points spaced 0.371 cm apart, and wave-
number spectra were calculated using a similar procedure to the one used
with the current-meter data. This produced a plot (Fig.7) on which the
peaks represented the ripple spacings that dominated the profile.
221
I time 2 8 August 1 9 7 6 I
2
1
0
cm
Fig.6. Ripple profiles from the study site a t four times on 28 August 1976. High water
occurred a t 1621 h.
A 1405
____ 1513
5 1405
-__- 1721
Fig.7. Spectral analysis of ripple spacing at the same site during a four-hour period on
28 August 1976. First five spectra appear in A; the last four spectra along with the first
one appear in B. High water occurred at 1621 h.
228
RIPPLE CHANGES
We observed that sand-flat ripples only change under certain wave condi-
tions; twice, during 10 spring tidal cycles in August 1976 and 1977, de-
creases in ripple spacing were documented. The measurements of 28 August
1976 (Tables Ib and IIb) show the response of the sand bed to changes in
the wave field; those of 25 August 1976 (Tables la and IIa) show no
significant changes. Figure 6 shows selected bed profiles and Fig.7A, B
compare ripple spectra from various times during the tidal cycle on
28 August.
The bed was profiled 9 times on 28 August, thus obtaining a picture of
the bedforms from just after the water rose above the sonar head (1405 h)
t o just before the water dropped below the sonar (1815 h). The nine profiles
were irregularly spaced relative to the 1621 h high water of about 1.5 m -
three occurred before high water and six afterwards. Before the tide reached
the area, a measurement of 26 ripples under the sonar frame yielded an
average spacing of 7.3 cm. Scaled photographs of the bed after the tide
receded show an average spacing of about 6 cm.
On 28 August the average ripple spacing was 9.3 cm at the time of the
first profile (1405 h). The spacing increased slightly to 10 cm and then
dropped t o 8.8 cm at 1644 h. The average spacing was still 8.8 cm at 1721 h,
but then it decreased t o 6.6 cm where it remained. This change, which
Fig.8A shows, paralleled the change in orbital diameter calculated from the
velocity spectra.
The ripple spectra for 28 August generally showed multiple, broad peaks.
When the profiles were only frequency averaged (no ensemble averaging)
as in Fig.7, each spectral peak encompassed a range of spacings of the order
TABLE I
Spectrally dominant wave period T,near-bottom orbital diameter d o , and near-bottom

maximum velocity u, for 8.53 min time series starting at the indicated times
Time T do urn
th) (s) (em) (cm s-)
a. 25 August 1976
1210 2.4 8.3 10.8
1300 3.4 13.4 12.4
1346 3.2 11.6 11.4
1425 3.6 11.0 9.6
1510 4.0 10.2 8.0
1546 3.8 10.1 8.3
b. 28 August 1976
1442 4.0 16.8 13.2
1615 3.2 14.9 14.6
1715 2.2 8.5 12.2
1733 1.6 5.7 11.3
1804 1.6 4.1 8.0
229
TABLE I1
Average h and spectral A s ripple spacings for profiles taken at the indicated times. Spectral
spacings A,, and A, are ranked relative t o the height of the spectral peak with the higher
one first (see text for the discussion of this ranking)
a. 25 August 1976
1320 6.9 9.1 6.4
1400 6.5 9.6 7.4
1445 6.8 6.6 9.6
1501 6.6 6.6 9.6
1527 6.8 6.6 8 .O
1545 6.3 6.6 8.0
1601 5.7 10.1 6.6
b. 28 August 1976
1405 9.3 9.1 13.7
1513 10.0 12.8 7.7
1608 10.1 12.0 9.6
1644 8.8 12.0 8.7
1709 8.5 9.1 7.4
1721 8.8 9.1 7.4
1745 6.6 6.2 8.0
1800 6.5 6.2 8.4
1815 6.4 6.2 9.1
of 1-2 cm. Over time the trend, as with the average spacing, was for the
peaks representing longer spacings to predominate early in the tide and
for the peaks a t shorter spacings to emerge during the falling tide. None of
the spectra were unipeaked, as would be expected if the ripples were in com-
plete equilibrium with the waves at the measurement time.
The ripple data for 25 August were much more static (Fig.8A); the aver-
age spacing remained constant during the experiment except for a slight de-
crease at the end of the day. Early in the tidal cycle, the ripple spectra
developed a dominant peak at 6.6 cm that existed throughout the experiment.
DISCUSSION
Because the ripple height for orbital ripples is related to its spacing (eq.4),
only one equilibrium ripple height exists for a given spacing. Therefore, a
switch from one ripple spacing t o a shorter one produces a decrease in the
spectral variance, which is proportional to ripple height. The dependence of
the maximum height of a spectral peak on spacing makes the interpretation
of ripple spectra difficult. Thus, the presence of peaks of equal size, but at
two different spacings, means that the shorter ripples are closer to their
equilibrium size than are the longer ones. Such appears to be the case in the
ripple spectra from 28 August; the late emergence of the spectral peak at
6.2 cm suggests that the ripples are slowly changing to that shorter spacing.
M e a s u r e d Urn= o 8
Threshold Um' x
---a. .
- 25 August 1976
- _ - - 28 August 1976
1 I , I I I
1200 1300 1400 1500 1600 1700 1800 1900

TIME (hours1
Fig.8. Comparison of observed and calculated wave and ripple parameters from
25 August 1976 and 28 August 1976. A . Observed ripple spacing ( A ) and orbital diameter
( d o). B . Observed maximum velocity and calculated threshold velocity calculated using
eq.1.
The existence of multiple peaks in the ripple spectra indicates that the
transition from a longer spacing to a smaller one is a gradual process, and,
under a gentle wave climate, that remnants of the larger ripples often exist
up t o the time the flat is exposed.
The average spacing, which is calculated directly from the profile, does
not differentiate the various spacings making up a bed. However, as the
ripples change, the average spacing also changes. Figure 8A is a plot of the
average spacing and orbital diameter against time. Not only d o the changes
in ripple size appear t o be discontinuous, Fig.8A suggests that there is a
time lag between the time when the orbital diameter decreases to when the
spacing reaches a smaller value. This time lag is indicative of the difficulty
that a less energetic oscillatory flow has in destroying the older bedforms.
Both the magnitude of the orbital diametergrain-size ratio (Clifton and
Dingler, 1984, this volume) and the parallelism between the spacing and
orbital diameter curves, demonstrate that these are orbital ripples. Further-
more, the spacing t o orbital diameter ratio falls between 0.65 and 0.8 values
of eq.2 during much of the tidal cycle with the points that fall above the
0.8 d o curve representing the time lag between the decrease in do and that
in A. Rather than the exact value of that ratio, the parallelism of the curves
is the important factor here.
231
Thus, as expected, ripple spacing follows orbital diameter for orbital

ripples. On 28 August 1976 this meant a slight increase in spacing early in
the incoming tide, and a greater decrease during the latter part of the falling
tide. This does not occur at the same depth every tidal cycle, or even at
different depths during the same tidal cycle because there is a second, often
overlooked, control t o ripple formation - the threshold of grain motion.
The threshold of grain motion depends on both wave period and orbital
diameter for sand of a given grain size (eq.1). In Fig.9, arrows mark the
minimum orbital diameters required to move 0.165 mm sand (the size found
on the sand flat) for various wave periods; these arrows are superimposed on
a schematic of the X vs. d o relationship. Thus, longer periods require larger
orbital diameters to initiate grain motion, and periods greater than 5 s do
not even fall within the range of orbital ripples. Figure 8 compares observed
orbital velocity with the calculated threshold velocity, which is based on the
dominant wave period for each record. Figure 8A and B together show that
ripple spacing follows the orbital diameter when the observed velocity
exceeds the threshold velocity. When, as occurs near the end of the ebb tide
or during days of longer dominant periods (e.g., 25 August 1976), the actual
velocity approximates or is less than the threshold velocity, the ripples
do not change even though the orbital diameter continues to fluctuate.
Therefore, ripples on the sand flatt will respond t o changes in orbital diam-
eter only when the near-bottom oscillatory flow exceeds the threshold
velocity.
This control of ripple change by threshold velocity may explain why
some ripples remained unchanged through a falling tide while the spacing of
ripples further inshore decreased noticeably (Fig.3A, B). Ripples at both
/
/
/
1 10 100
ORBITAL DIAMETER (cm)
Fig.9. Threshold orbital diameters (arrows) superimposed on a plot of ripple spacing ( h )

versus orbital diameter ( d o )for 0.165-mm-diameter sand.
232
locations experienced a similar decrease in water depth as the tide level fell.
Apparently, however, when the water depth over the outer ripples reached
the level at which the spacing of ripples at shallower depths diminished a
short time earlier, the orbital velocity was subthreshold. This situation could
be caused by an increase in wave period during the fall of the tide, a
phenomenon that was observed, but which did not seem to occur consis-
tently (Table I).
An alternate possibility relates to the tidal-flat profile and the changes in
wave energy that attend the falling tide. Within about 60 m of the toe of
the beach, the slope of the profile increases. The ripples that decreased in
spacing occur on the upper part of this slope, and the unchanged ripples on
the lower part. By the time these latter ripples were exposed to the water
depths at which the inshore ripples changed, the general depth over the flats
would be significantly reduced. The loss of wave energy due to frictional
interaction with the bottom would be greater, resulting in reduced wave
heights. Accordingly, the associated orbital velocities would have more
quickly become subthreshold.
The ripple response process outlined here has important implications for
estimating past wave conditions from low-tide ripples. In many cases inter-
pretations will be inaccurate because ripples exposed at low tide could have
been produced anytime during the previous tidal cycle or even during an
earlier one.
CONCLUSIONS
In low-energy wave environments both wave period and orbital diameter

control ripple size. Ripple spacing and, therefore, ripple height, are propor-
tional to orbital diameter only when the near-bottom orbital velocity
exceeds the threshold velocity; threshold velocity is proportional to the ratio
of orbital diameter t o wave period. Ripples exposed at low tide may have
formed at different times during the previous high tide, or even during
earlier tidal cycles. The decrease in ripple spacing sometimes observed
during the falling tide comes about when decreasing orbital diameters occur
with short wave periods above the threshold. The decrease in ripple spacing
with the decrease in orbital diameter is discontinuous with a time lag caused
by the existence of larger ripples.
ACKNOWLEDGMENTS
We thank the many field assistants who participated in this study. Special
thanks go to Jeff Hedenquist and Curt Peterson, who spent hours in the
water operating the sonar. The junior author thanks his family, who spent
several nights with him on the exposed, wind-swept tide flats of Willapa Bay,
searching with flashlights for ripples marked with little bits of copper wire.
The senior author started this project while a National Research Council
Postdoctoral Fellow with the U.S. Geological Survey, working with
E.D. McKee.
233
REFERENCES
Allen, J.R.L., 1979. A model for the interpretation of wave ripple-marks using their
wavelength, textural composition, and shape. J. Geol. Soc. London, 136: 673-682.
Bagnold, R.A., 1946. Motion of waves in shallow water. Interaction between waves and
sand bottoms. Proc. R. SOC.London, Ser. A, 187: 1-18.
Clifton, H.E., 1976. Wave-formed sedimentary structures - a conceptual model. In:
R.A. Davis, Jr. and R.L. Ethington (Editors), Beach and Nearshore Sedimentation.
SOC.Econ. Paleontol. Mineral., Spec. Publ., 24: 126-148.
Clifton, H.E. and Dingler, J.R., 1984. Wave-generated structures and paleoenvironmental
reconstruction. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics
and Sedimentation in Wave-Dominated Coastal Environments. Mar. Geol., 60:
165-198 (this volume).
Dingler, J.R., 1974. Wave-formed ripples in nearshore sands. Unpubl. Ph.D. thesis, Uni-
versity of California, San Diego, Calif., 1 3 6 pp.
Dingler, J.R. and Inman, D.L., 1977. Wave-formed ripples in nearshore sands. Am. SOC.
Civ. Eng., Proc. 15th Coastal Engineering Conf., 210+2126.
Dingler, J.R., Boylls, J.C. and Lowe, R.L., 1977. A high-frequency sonar for profiling
small-scale subaqueous bedforms. Mar. Geol., 24: 279-288.
Inman, D.L., 1957. Wave-generated ripples in nearshore sands. Dept. of the Army, Corps
of Engineers, Tech. Memo, 100, 65 pp.
Komar, P.D., 1974. Oscillatory ripple marks and the evaluation of ancient wave condi-
tions and environments. J. Sediment. Petrol., 4 : 169-180.
N.J., 429 pp.
Komar, P.D. and Miller, M.C., 1973. The threshold of sediment movement under oscil-
latory water waves. J. Sediment Petrol., 43: 1101-1110.
Miller, M.C. and Komar, P.D., 1980. Oscillation sand ripples generated by laboratory
apparatus. J. Sediment. Petrol., 50: 173-182.
Otnes, R.K. and Enochson, E., 1978. Applied time series analysis, Vol. 1. Wiley, New
York, N.Y., 449 pp.
Marine Geology, 60 (1984) 235-259 235
Elsevier Science Publishers B.V., Amsterdam -Printed in The Netherlands
BEDFORMS AND DEPOSITIONAL SEDIMENTARY STRUCTURES OF A

BARRED NEARSHORE SYSTEM, EASTERN LONG ISLAND,
NEW YORK
R. CRAIG SHIPP*
Marine Systems Laboratory, Smithsonian Institution, Washington, DC 20560 (U.S.A.)
(Received July 15, 1983; revised and accepted October 18, 1983)
ABSTRACT
Shipp, R.C.,1984. Bedforms and depositional sedimentary structures of a barred near-

shore system, eastern Long Island, New York. In: B. Greenwood and R.A. Davis, Jr.
ments. Mar. Geol., 60: 235-259.
The depositional sedimentary structures and textures of a single-barred nearshore

system o n the Atlantic coast of eastern Long Island, New York, were studied along seven
shore-normal transects. Data along these transects consisted of textural analysis of 160
sediment samples, temporal bedform observations, and 42 can cores for the analysis of
sedimentary structures.
Six sedimentary subenvironments were observed, based on distinct combinations of
sediment color and texture, bedforms, physical, and biogenic sedimentary structures, and
benthic infaunal communities. The shoreface environment is divided into the upper shore-
face, the longshore trough, and the longshore bar. The divisions of the inner shelf environ-
ment are the shoreface-inner shelf transition, the offshore, and the coarse-grained deposit.
The first five subenvironments are arranged in bands parallel to the shoreline, whereas the
coarse-grained deposit occurs in patches across the inner shelf.
The location of fakweather wave base, coinciding with a reduction in slope (3.0-0.3”)
from the shoreface to the inner shelf, is characterized by the cessation of debris surge in
the troughs of ripples, the formation of a “rust layer” of microorganismsover the bedform
surface, and a sediment color change caused by an increase in organic detritus. The se-
quence of bedforms and physical sedimentary structures observed in this system fits well
with existing wave-generated (oscillatory) flow regime models. These models explain the
observed sequences as a response to the degree of asymmetric flow created by shoaling
waves. Distribution of biogenic structures and assemblages of infaunal organisms is influ-
enced by the distance landward or seaward of fair-weather wave base.
The overall relationships of this nearshore system can then be summarized as a hypo-
thetical prograding stratigraphic sequence. The entire sequence is underlain by organic-
rich, bioturbated, offshore deposits. Overlying the offshore is the planar-laminated
sediments of the transition. Grading upward from the transition are the cleaner, planar-
laminated, seaward slope deposits of the longshore bar. Above this, is a distinct erosional
surface indicating the base of the massive to cross-laminated coarse sediments of the
*Present address: Oceanography Program, 14 Coburn Hall, University of Maine, Orono,

ME 04469, U.S.A.

236
longshore trough. Capping the sequence are the cross- to planar-laminated, clean sands of
the upper shoreface and foreshore.
INTRODUCTION
The study of the nearshore zone has been long neglected due t o poor
accessibility. The historical concentration of research efforts on the beach
and offshore has created a large gap in the understanding of nearshore pro-
cesses and sedimentation. Concomitant with difficult access, investigation of
the nearshore zone has presented a broad array of logistical problems includ-
ing sampling strategy, equipment deployment, and weather dependence.
Only in the last two decades have a handful of detailed geologic nearshore
studies emerged. Initially, these studies defined nearshore subenvironments
by examining sediment texture, physical sedimentary structures, and bio-
genic features (Reineck, 1963; Reineck and Singh, 1971; Howard and
Reineck, 1972a, b, 1981). Later, these initial ideas were developed further
by relating nearshore subenvironments to “an oscillatory flow regime con-
cept” (Clifton et al., 1971; Davidson-Arnott and Greenwood, 1974, 1976;
Clifton, 1976). More recently, investigations have dealt either with the
morphology or sedimentology of prominent nearshore features such as
intertidal and submarine bars (Greenwood and Mittler, 1979; Hunter et al.,
1979; Greenwood and Hale, 1980; Goldsmith et al., 1982), or with the
sedimentary and biogenic trends from the supratidal zone across the
intertidal and through the nearshore zone (Kent, 1976; Hill and Hunter,
1976). Few studies have explored the characteristics of the nearshore zone
from mean low water (MLW) across the shoreface to the inner shelf.
The present study describes the morphologic variation and sedimentary
subenvironments of a single-barred nearshore system off eastern Long Island,
New York. Even though most data were collected during fair-weather con-
ditions, the overall sedimentological setting (fair-weather and storm condi-
tions) of this nearshore system are suggested. Finally, the depositional
relationships are summarized as a hypothetical prograding stratigraphic
sequence.
Of all the published studies t o date, the conceptual model of Clifton
(1976) seems the most useful in explaining nearshore dynamics and the
sequence of bedforms generated by shoaling waves. This model integrates
Airy (linear) and solitary wave theory and empirical measurements t o
explain sedimentologic and stratigraphic observations. An important element
in Clifton’s model is the change from symmetric to asymmetric oscillatory
flow as the wave shoals. This change in flow symmetry of a shoaling wave, in
turn, alters the symmetry of the bedforms. The depth a t which the asym-
metry develops is a function of wave height, wave period, grain size, slope,
and bottom roughness (Newton, 1968). Clifton applies this principle by
suggesting a flow sequence of wave-generated bedforms (Fig.lA). In addi-
tion, the work of Davidson-Arnott and Greenwood (1976) have modified the
Clifton sequence for a barred nearshore system. They suggest that the Clifton
237
A.
-LAND (Direction of Wave Travel)
- ASYMMETRIC SYMMETRIC +
ACTIVE INACTIVE
LUNATE CROSS
' R ~ l ~ ~ ~ ~LINEAR
[ R RIPPLES
F L A T BEDS MEGARIPPLES RIPPLES
7 6 5 4 3 2b
-
O N MAJORITY OF SHOREFACE
No M o v e m e n t Rippler +
Lunote
Megarippler
la+ ~ e d s
1 ASYMMETRIC
OSCILLATORY
FLOW
O N INNER SHELF 8, SEAWARD BAR SLOPE

SYMMETRIC
NO Movement -* Rippler A Flat Beds OSClLLATORY
FLOW
Fig.1. Bedform sequences in the nearshore zone. A. Hierarchical sequence of bedforms
produced by shoaling waves (modified from Clifton, 1976). Numbers along the bottom
correspond to the bedforms named above. B. Suggested flow regime sequences for barred
nearshore systems (modified from Davidson-Arnott and Greenwood, 1976).
sequence is valid over the majority of the shoreface for strong asymmetric
oscillatory flow during fair-weather swell conditions, while a distinct sym-
metric oscillatory regime would be produced during periods of local storms
on the inner shelf and the seaward slope of the bar (Fig.lB).
Because of its peculiar geometry, a particularly interesting form in the
sequence is the cross ripple (5 on Fig.lA). This bed configuration consists
of two sets of ripples oriented obliquely to the oscillatory current (wave
approach observed directly from the bed). One set is longcrested, while the
second set is composed of short ripples occupying the troughs of the longer
set. Cross ripples seem to represent a structural transition between asym-
238
metric ripples and lunate megaripples (Clifton e t al., 1971) and are shown in
the present study to be an easily recognized field indicator of asymmetric
oscillatory flow.
STUDY AREA
The study was conducted off southeastern Long Island in Suffolk County,
New York (Fig.2A). The study area, adjacent t o the villages of Beach
Hampton, Amagansett, and East Hampton, extends for a total shore-parallel
distance of 13 km. The limit of the investigation was 3.25 km seaward of the
shoreline, approximately at the 2 2 m isobath (Fig.2B).
The beach and nearshore deposits of Long Island consist of reworked
outwash deposits and sediments transported from the updrift Montauk Till
Member (located 25 km to the east a t Montauk Point) of the late Wisconsin
Ronkonkoma Moraine (Taney, 1961). The entire eastern Long Island near-
shore system is dominated by a linear to slightly irregular subtidal bar. The
bar crest reaches within 3.5 m of the water surface and varies from 100 to
300 m seaward of MLW. The gross morphology of the bar varies from a
1l0
10"
9"
239
LONG %
' <
ISLAND /
Fig.2. A . Location of the study area within the New York Bight. The box on the southern
fork of eastern Long Island is the approximate geographic limits of B. B. Location of the
seven transects off the south shore of eastern Long Island. The six short transects estab-
lished along the shoreface-inner shelf contact vary from 100 to 300 m and are not drawn
to scale. 3A and 3B refer to location of photographs in Fig.3. Isobaths in feet.
straight and narrow, landward asymmetric form (Fig.3A) to a slightly irregu-

lar and wider, symmetric form (Fig.3B).
Prevailing winds in the study area are southwest in the summer, with a
3.6-5.2 m s-' average hourly velocity range, and northwest in the winter,
with a 4.7-6.8 m s-' average range (Lettau et al., 1976).The dominant wind
generally blows from the northeasterly to easterly sectors during storms that
are either tropical (predominantly hurricanes) or extratropical (north-
eastern). Tropical storms are most prevalent from August through October,
whereas extratropical storms usually occur most frequently between October
and April (U.S. Army Corps of Engineers District, 1977).
Visual and remotely measured data from eastern Long Island show a mean
annual breaker height of 0.7 m and wave period of 8.0 s (DeWall, 1979).
According to the wave hindcasting calculations of Saville (1954),72% of all
240
Fig.3. Oblique aerial photographs of the longshore bar in the study area. A . Straight, land-
ward asymmetric form. B. Slightly irregular, symmetric form. Locations of photographs
o n Fig.2B.
deep-water waves in western Long Island approach from the east-northeast

through south-southeast. The largest predicted waves are 7.6-9.1 m. The
wave dominance from the easterly quadrants continues in the study area,
w h c h is located 170 km t o the east. In addition, a secondary dominance
from the southwest is caused by t,he increased fetch between the New Jersey
shoreline and the eastern Long Island study site (Fig.2A).
241
The semidiurnal spring tide in eastern Long Island is 0.7 m at Montauk

Point and increases t o 1.1 m a t Shinnecock Inlet (US.Department of
Commerce, 1978). An estimated spring range for the study area is therefore
0.85 m. This range places the study area well within the microtidal classifi-
cation of Davies (1964).
METHODS
Transect lines
In the study area, one long intermittent transect was established from
MLW to a distance of 3.25 km offshore. Six shorter transects (100-300 m
in length) were surveyed over the change in slope at the seaward base of the
longshore bar (Fig.2B). With the aid of Scuba, polypropylene-line transects
were placed shore-normally across the bottom, anchored every 3.0 m with
a 0.5 m section of hooked reinforcing bar (rebar). Transect orientation was
maintained by compass. With this method, 200 m of line could be deployed
a t a time. All transects were surveyed once and were usually removed the
same day, except for the main transect at Beach Hampton (Fig.2B). This
transect was buoyed and maintained for the duration of the study (two
months).
Along the transects, sampling intervals were chosen wherever bedform or
textural changes occurred. The interval of sampling varied from 3 t o 6 m, on
the landward slopes of the longshore bar, t o over 1km on the inner shelf. At
each sample site, bedforms were examined for planimetric geometry, height,
spacing, degree of asymmetry, presence of wave surge, and amount and type
of biological activity.
Sediment textural a m lysis
Sediment samples were analyzed for size (mean) and sorting (standard
deviation) utilizing the Hydraulic Equivalent Sediment Analyzer (HESA)
described by Anan (1972) that was later modified by replacement of the
original pressure transducer system with an electrobalance. The modified
HESA is interfaced with a HP 9825A calculator and a HP 9872A graphic
plotter (Hewlett-Packard Calculator Products Division). The system is
programmed t o tabulate grain-size distribution and graphically depict size-
frequency and cumulative probability curves for individual samples. Deter-
mination of mud content ( > 4 @)was made by wet sieving the sediment
samples. A complete dispersion of silts and clays was assured by agitating the
sample for 15 min in an ultrasonic disruptor after the methods of Kravitz
(1966).
Sediment cores
A total of 42 can cores were taken a t selected sites on four of the seven
transects. The can coring technique is a modification of the procedure used
242
by Howard and Frey (1975). The corer consists of a 7.6 1 metal gas can
(measuring 26.9 X 21.7 X 14.7 cm) with the bottom removed. A slurry of
sand, water, and red lead (NL Industries) was mixed in a plastic bag and
distributed over the area to be cored. The incorporation of the red lead
into the sediment revealed structures that are caused by the coring process
(Howard and Reineck, 1972a) and also aided in the preservation of the
surface bedforms (Davidson-Arnott and Greenwood, 1974).
The cores were extruded in 2 cm thick slabs into plastic trays for trans-
port t o the laboratory. The leftover sand from the extruded cores was
elutriated and the infaunal invertebrates collected and preserved. In the
laboratory, the slabbed cores were first X-rayed using routine radiographic
techniques for unconsolidated sediments (e.g., Howard, 1968; Bouma, 1969;
Hamblin, 1971). Next, two different methods were used to make relief peels
of the core slices. For coarse sand and gravel the epoxy-resin technique of
Howard and Frey (1975) was employed. A second technique utilizing a
lacquer cement (Bouma, 1969) was found to be more suitable for finer
sediments.
Some structural information in the relief peels was lost due to post-core
amphipod bioturbation. This generally occurred only in the upper 10 cm of
the core. The problem was eliminated in later cores by saturating the sedi-
ment with a 10%solution of MgCI2.
RESULTS
In the Long Island nearshore system, six sedimentary subenvironments

can be defined. These areas are described by distinct combinations of sedi-
ment color and texture, physical sedimentary structures, and, occasionally,
infaunal assemblages and/or biogenic structures. Beginning at MLW, the
subenvironments defined are: (1) upper shoreface; (2) longshore trough;
(3) longshore bar including the subdivisions: (a) landward bar slope, (b) bar
crest, and (c) seaward bar slope; (4)shoreface-inner shelf transition (here-
after referred to as transition); ( 5 ) offshore; and (6) inner shelf coarse-
grained deposit. The first five subenvironments occur in continuous bands
parallel t o the shoreline. The coarse-grained deposit is unique in that it
occurs in small patches scattered across the inner shelf. Additionally, the
sedimentary subenvironments of the Long Island nearshore zone are grouped
into two major environments dependent on their location relative t o fair-
weather wave base (Fig.4). The first three subenvironments are landward of
fair-weather wave base, thus comprising the shoreface. The last three sub-
environments are seaward of wave base and, therefore, are on the inner shelf.
Fair-weather wave base
A change in the offshore slope from approximately 3 . 0 4 . 3 "occurs at the

seaward base of the longshore bar at a depth of approximately 10 m (Fig.4).
This slope change marks the seaward extent of the shoreface and the land-
243
INNER SHELF
\\\\-\\\
Fig.4. Location of the sedimentary environments in the Long Island nearshore system.
Coarse-grained deposits are located in scattered patches on the inner shelf and are not
necessarily drawn to scale.
ward boundary of the inner shelf. Observations of this area during fair
weather indicate that this contact is the limit of wave influence, and, there-
fore, the greatest depth of active sand transport during these conditions. This
transitional region has been defined as the effective or fair-weather wave base
(Curray and Moore, 1964; Dietz and Fairbridge, 1968; Elliott, 1978). I t is
characterized in the study area by:
(1) The point at which debris in the ripple troughs ceases to surge and
becomes motionless.
(2) A distinct color change from a clean, yellow-tan sand on the shoreface
to a gray-brown sand on the inner shelf. This color change is caused by an
increase in organic detritus seaward of the contact.
(3) A layer of rust colored microbenthos covering the bedforms seaward
of the fair-weather wave base. The first observation of a rusty bottom
(hereafter referred to as the “rust layer”) was documented by Kumar and
Sanders, 1976, and may be composed of diatoms, algae, and other micro-
organisms.
244
A series of observations taken over a four-week period (August, 1978)

during fair-weather conditions indicated a nearly exact correlation between
the location of wave-surge cessation and the initiation of the rust layer. This
change from debris surging in the ripple troughs to the formation of the rust
layer vaned across a 50 m band. A series of observations over three consecu-
tive days shows that a thin rust layer can form in 24 h or less. The transition
from a yellow-tan to gray-brown colored sediment is also gradational across
the 50 m band. However, the color shift may require several months or a
season to respond to changes in wave climate as opposed t o the rapid re-
sponse of the rust layer. During low to moderate wave-energy conditions,
this color shift is an excellent indicator of fair-weather wave base.
Sedimentary subenvironments of the shoreface
Upper shoreface
The landwardmost shoreface environment is characterized by: (1)a cover
of medium sand; (2) a shore-normal sequence of bedforms; (3) three differ-
ent types of physical structures; and (4) a few biogenic structures. This area
extends seaward from MLW t o approximately 200 m offshore in water
depths ranging from 0 to 6.0 m. The average slope is approximately 2.0"
(Fig.4).
The upper shoreface consists of well to moderately sorted medium sand
(1.6-2.1 4, 0.45-0.55 u) that fines seaward and has a white-tan color due to
a lack of fine-grained material. The landward progression of bedforms on the
upper shoreface exhibits an asymmetric set complete t o flat beds as pro-
posed by the Clifton sequence (2b-7 Fig.lA). A typical can core taken from
the linear-rippled region on the upper shoreface displays subhorizontal
planar lamination overlain by coarser landward-dipping, medium-scale fore-
set bedding, and capped by small-scale ripple cross-lamination (Fig.5A). Only
a few biogenic structures are present despite the great abundance of Hau-
storiid amphipods. The few short wispy bioturbate structures that are pre-
served are usually in the top 10 cm of sediment, where amphipods are most
commonly found (Howard, 1968).
Longshore trough
This subenvironment seaward of the upper shoreface is distinguished by:
(1)a coarse sand size; (2) a sediment surface dominated by large oscillatory
ripples; ( 3 ) a lack of internal depositional structures;and (4)no biogenic struc-
tures. The longshore trough varies from 5.5 to 6.0 m in depth and averages
Fig.5. Relief peels of can cores taken from the shoreface: A . upper shoreface (depth -
4.6 m). B. longshore trough (depth - 6.0 m). C. landward slope of the longshore bar
(depth - 5.5 m). D. crest of the longshore bar (depth - 3.5 m ) . E. seaward slope of the
longshore bar (depth - 7.5 m). Some of the ripple cross-lamination at the top of the
relief peels have been altered by postcore amphipod bioturbation. Black arrows at top
left point landward. Scales at bottom right are 3 cm long.
245
246
50 m in width. The trough exhibits a concave-upward bathymetric profile with

approximately 0.5 m in relief between the center and the landward and
seaward boundaries (Fig.4).
The longshore trough is composed of subangular, moderately well-sorted
coarse sand (0.5-0.7 q5, 0.57-0.65 a) and at least a 2% component of sub-
rounded gravel. The large oscillatory ripples are continuous in a shore-
parallel direction for several hundred meters and have an average 60 cm
wavelength and a 1 5 cm amplitude (2a on Fig.lA). A typical can core taken
from the longshore trough displays massively bedded sand and gravel with
no discernable physical or biogenic structures (Fig. 5B). The reason for the
massive nature of these sediments seems related to the presence of an in-
faunal fish. During fair weather any disturbance, such as the passage of a
large wave overhead, causes the American sandlaunce Ammodytes ameri-
canus to dart across to an adjacent oscillatory ripple. This in turn is followed
by rapid burrowing into the ripple crest, thus disrupting any bedding.
Longshore bar
The third subenvironment is the longshore bar that extends from 250 to
530 m seaward of MLW. It is characterized by: (1)moderately sorted fine to
medium sand; (2) several sets of bedforms that follow the Clifton sequence;
(3) three distinct vertical sequences of physical structures; and (4) no bio-
genic structures even though two species of Haustoriid amphipods and the
isopod Cirolana concharum are in moderate abundance. The three subdivi-
sions associated with the longshore bar are the landward slope, the bar crest,
and the seaward slope.
Landward slope. The landward slope of the longshore bar consists of a

50 m shore-parallel band of well- t o moderately well-sorted medium sand
(1.7-2.1 q5, 0.41-0.58 u) dipping landward at an approximate slope of 3"
(Fig.4). The series of bedforms found on this slope follows the asymmetric
ripple set from the Clifton sequence except it is in the reverse order ( 2 a - 6
on Fig.lA). The sequence from the top to the bottom of the landward
slope includes cross ripples (Fig.6), irregular ripples, and linear ripples. A
typical can core taken on the middle of the slope displays landward-dipping,
medium-scale foreset bedding overlain by small-scale ripple cross-lamination
(Fig.5C).
Bar crest. The bar crest is characterized by a narrow band of moderately

well-sorted medium sand (1.9-2.0 q5, 0.53-0.55 u). Here, the type of bed-
form is most sensitive t o the incident wave climate. During low wave-energy
conditions the bar crest is covered by irregular ripples to cross ripples (3-5
on Fig.lA). Observations during higher wave-energy conditions (1.25 m
height and 6.0 s period) revealed a 5 0 m band centered on of the bar apex
that consisted of lunate megaripples having a 60-80 cm wavelength and
14-18 cm amplitude (6 on Fig.lA). A typical can core in this region shows
a narrow band of small-scale ripple cross-lamination overlying cosets of
247
Fig.6. Cross ripples. Arrow points to a secondary ripple crest in a trough between two
long-crested ripples. Both sets of ripples are oriented oblique to the oscillatory current.
Knife handle is 15 cm long.
landward dipping, truncated, low-angle foreset bedding of megaripple origin

(Fig.5D).
Seaward slope. The seaward bar slope is characterized by a 200 m wide

band of moderately sorted fine sand (1.8-2.6 @, 0.41-0.67 u) dipping sea-
ward a t approximately 3.0" (Fig.4). During low waveenergy conditions,
bedforms of this region vary from symmetric linear ripples at the seaward
base of the longshore bar to asymmetric cross ripples (Fig.6) just seaward of
the bar crest (2a-5 on Fig.lA). At the time of higher wave-energy7the upper
portion of this area shifts to lunate megaripples grading to cross ripples fur-
ther down the seaward slope ( 5 4 on Fig.lA). In fair weather, the boundary
between symmetric and asymmetric bedforms may be as far as halfway up
the seaward slope and then shift downslope in response to increased wave
activity. A typical can core in this region reveals a sequence of physical struc-
ture quite different from the landward slope. Instead of small-scale ripple
cross-laminations overlying medium-scale foreset bedding, the vertical se-
quence consists of small-scale ripple cross-laminations overlying horizontal
to subhorizontal planar lamination (Fig.5E). A similar sequence on the sea-
ward slope of the longshore bar was also reported by Davidson-Arnott and
Greenwood (1976) and Greenwood and Mittler (1979).
218
Sedimentary subenvironments of the inner shelf
Transition
The transition is the landwardmost subenvironment of the inner shelf and
is characterized by: (1)moderately well-sorted fine sand; ( 2 ) intermittently
active, symmetric, linear ripples; ( 3 ) a sequence of physical structures similar
t o the seaward bar slope; and ( 4 ) a few biogenic structures. The depth range
of this environment varies from 10 m t o a depth of 14-16 m. Across the
profile, the transition extends from 530 m to a distance of approximately
1.5 kin seaward of MLW (Fig.4).
The transition consists of moderately well sorted fine sand (2.4-2.7 @,
0 . 5 3 4 . 6 7 u) that is light gray-brown in color. The darker color is due to an
increase in the organic constituent of the sediment. The mobility of the
linear ripples is a function of the ambient wave conditions and could be
precisely determined by observing the onset of the rust layer. A typical can
core in this region exhibits the same sequence as the seaward bar slope -
small-scale ripple cross-lamination overlying horizontal planar lamination
- except the units of the small-scale lamination are thicker (Fig.7A). In
addition, at the same level in the core, organic detritus drapes the preserved
ripple troughs indicating periods of quiescence when fair-weather wave base
was further landward. I t is during these quiet periods that the rust layer
forms over the sediment (see p. 242). The limited bioturbation (white wispy
structures on X-ray radiographs) present in this subenvironment is caused by
Lysianassid amphipods.
0ffsh o re
The offshore subenvironment is characterized by: (1) organic-rich, fine
sand; ( 2 ) inactive symmetric linear to irregular ripples; ( 3 ) a lack of physical
sedimentary structures; and (4) extensive bioturbation. The landward boun-
dary of this environment begins approximately 1 . 5 km seaward of MLW in
an approximate 1 5 m depth and continues seaward beyond the 3.25 km
distance and the 22 m depth limits of this study area.
The offshore consists of well-sorted, fine sand (2.2-2.5 @, 0.47-0.62 u)
containing a large amount of organic detritus and shell hash. The surface is
covered with inactive, symmetric, linear ripples stabilized by the rust layer
(1 on Fig. 1A). In addition, a high density of the sand dollar Eclzinarachnius
parma covers the surface (Fig.8). Virtually no primary depositional struc-
tures are preserved due to the extensive bioturbation by a high density of
infaunal organisms. A typical can core for this region exhibits extensive
Fig.7. Relief peels (left) and X-ray radiographs (right) of can cores taken from the inner
shelf. A . Transition (depth - 11.0 m ) . OD o n radiograph marks areas draped with organic
detritus in the troughs of small-scale ripple cross-lamination. B. Offshore ( d e p t h - 20.5
m). IB marks an infilled burrow and S D indicates a n in place sand dollar Echinarachnius
parma. C. Coarse-grained deposit ( d e p t h - 14.5 m), Black arrows a t t o p left point land-
ward. Scales a t bottom right are 3 cm long.
219
250
Fig.8. Inactive symmetric linear ripples in the offshore ( d e p t h - 20.5 m). S D indicates
the sand dollar Echinarachnius p a r m a and ST marks a surface trail of E. parma. Scale is
1 5 cm long.
reworking by E. parma. The only bioturbate structures preserved are wispy

structures of amphipods and infilled burrows (Fig.7B).
Coarse-gra ined deposit

The last subenvironnient observed in the Long Island nearshore system
consists of patchy units of coarse material. This coarse-grained deposit is
characterized by: (1)a coarse to very coarse sand texture; (2) large, sym-
metric, linear ripples; ( 3 ) physical structures consisting of medium-scale
foreset bedding; and (4) moderate bioturbation. Surficially, the coarse-
grained deposit appears similar to the longshore trough. The significant
difference between the two areas is the occurrence of cross-bedding and
abundance of gravel in the coarse-grained deposit. This subenvironment was
observed at numerous locations varying in depth from 10 to 16 m and seems
to occur commonly as exposed lenses throughout the transition.
The coarse-grained deposit consists of yellow-tan, moderately sorted coarse
sand (0.7-0.9 4;0.73-0.84 u ) mixed with subrounded gravel. Similar to the
longshore trough, the surface is dominated by large symmetric oscillatory
ripples having a 60-80 cm wavelength and a 10-20 cm amplitude which
contain large amounts of organic and inorganic debris in the ripple troughs.
251
Unlike the longshore trough, depositional structures are quite abundant. A

typical can core reveals steep landward-dipping foreset bedding (Fig.7C). The
preservation of bedding in this region is attributed to the absence of A . ameri-
canus. Wispy bioturbate structures are observed along the bedding planes in
the coarse-grained deposit, caused by the presence of Haustoriid amphipods.
In addition, several species of bivalves are found in this subenvironment.
DISCUSSION
Bedform sequences
The bedform sequence in the Long Island nearshore system follows the
structural model of wave-formed features suggested by Clifton (1976). Even
though simultaneous measurements of wave period, wave height, and bed-
form dimensions were not available, the observations of bedform symmetry
and the presence of cross ripples allow qualitative determination of sym-
metric and asymmetric oscillatory flow for each environment (Fig.9). Uni-
directional longshore or rip currents on the shoreface, noted by Clifton
(1976) and discussed thoroughly by Davidson-Arnott and Greenwood
(1976), were not observed during the present study. Recently, however
observations made bi-weekly by aerial reconnaissance have confirmed the
existence of seaward flowing rip currents along other parts of the Long
Island coast t o the west of the study area (G.A. Zarillo, pers. commun.,
1983).
SHOREFACE INNER SHELF

UPPER
SHOREFACE
LONGSHORE
TROUGH
LONGSHORE
BAR I TRANSITION OFFSHORE
Ii
I I
I I
I I
2 I
In
a I
: 4 I
W
I
5
f 8
+
n
g 10
12
18
CONDITIONS
I
Fig.9. Summary o f bedforms and physical sedimentary structures. The numbers above the
typical can cores refer to the bedform types in Fig.lA. The arrows associated with the
numbers indicate the direction in which the bedform sequence progresses.
252
The sequence of bedforms across the six subenvironments can be ex-

plained as a response to shoaling waves (Fig.lA). The symmetric, linear, to
occasionally slightly irregular bedforms of the transition, the coarse-grained
deposit, and offshore subenvironments are formed by infrequent large waves
during storms or by long period swell. The bedforms in these environments
maintain their symmetry, because waves in deeper water depth ( > l o m )
restricts orbital flow to symmetric oscillations. Waves with periods of 5-8 s
on coasts that are affected infrequently by longer period swell (such as Long
Island) may generate a symmetrically-rippled bottom to depths of 100 m
(Clifton, 1976). The sequence of near-symmetric linear ripples at the sea-
ward base of the longshore bar (depth - 10 m) to lunate megaripple at the
bar crest (depth - 4 m) is controlled by wave shoaling and adheres t o the
order of bedforms proposed by the Clifton sequence. The increased size of
the lunate megaripples on the bar crest during higher wave-energy conditions
is an expression of increased velocity of asymmetric flows caused by larger
waves. A reverse of the Clifton sequence of cross ripples at the bar crest to
linear ripples at the bar trough contact (depth - 6 m) is a response t o de-
creased asymmetric flow velocity caused by a rapid increase in depth on the
landward slope of the bar. The large symmetric oscillatory ripples in the
coarse sand of the longshore trough (depth - 5.5 t o 6.0 m) are controlled by
two factors - increased depth and larger grain size. The sequence of asym-
metric linear ripples (depth - 5.5 m) t o flat beds (depth - 0 m) on the
upper shoreface is again an expression of an increased asymmetric flow
velocity caused by reformed shoaling waves.
In addition to the above qualitative determination of the transition from
symmetric to asymmetric bedforms, interpolation of data from Clifton
(1976, table 2) provides further support for the empirical model of asym-
metric wave-formed structures. Assuming an 8 s wave period and a deep-
water wave height of 0.7-0.8 m (which is representative for the Atlantic
coast of Long Island), the transition should occur between 5 and 10 m for
“typical” fair-weather conditions (Table I). The two locations, where bed-
forms grade from symmetric to asymmetric features, are at the seaward base
of the bar (depth - 10 m) and at the bar trough-upper shoreface contact
(depth - 5.5 t o 6.0 m). Both of these locations are included in the transition
envelope predicted from the interpolation of Clifton’s data. The difference
in transitional depth between these two locations is caused by the longshore
bar. The effect of the bar on the incoming waves is t o dissipate wave energy
by breaking. Therefore, the reformed waves are smaller and thus orbital
velocities near the bed are reduced (Wood, 1971).
Physical and biogenic structures
The sedimentary structures preserved in the upper portion of t h e can

cores are different from those preserved in the lower half (Fig.9). This
pattern has been interpreted as low-wave energy reworking of underlying
structures deposited during higher energy conditions (Howard and Reineck,
253
TABLE I
Depth of transition during fair-weather conditions from symmetric to asymmetric ripples

in the nearshore zone. T is the wave period in seconds. H , is the deep-water wave height
in meters. Eastern Long Island transition depth is an interpolation of data presented in
Clifton (1976) which is verified by observations in the present study
TYPICAL
FAIR-WEATHER TRANSITION
__ .~ WAVE
~~~~~
DEPTH
T(sec) Holm)
~~ - ~.. ~~~
WILLAPA BAY, WASHINGTON 2 0.2 lm
SOUTHEASTERN SPAIN 5 0.5 3-5m
IEASTERN
LONG I S L A N D 8 0.7-0.8 5-10m I
SOUTHERN OREGON 10
-- ~ ~ ~ _ _ _ _ _
1 .o 10-201-11
1972a; Greenwood and Hale, 1980) and provides each core with a record of
two phenomena. The top of the cores are the fair-weather phase, whereas
the bottom shows the most recent storm event. With only one exception,
the Clifton sequence can be used to explain relative flow conditions of the
higher wave-energy (deeper) sedimentary structures.
The deeper bedding structures of the upper shoreface consists of planar
lamination overlain by landward-dipping medium-scale foreset bedding. The
longshore trough would most likely be characterized by landward-dipping,
medium-scale foreset bedding similar to the stratification found in the
coarse-grained deposit, if it were not for the burrowing activities of A . ameri-
canus.
The three subdivisions of the longshore bar display separate sequences of
physical structures. The landward-dipping, medium-scale foreset bedding on
the landward slope indicates a moderate intensity of asymmetric flow, which
may be caused by wave energy dissipation on the bar crest. The bar crest is
characterized by larger scale foreset bedding (up t o 3 X ) signifying larger
bedforms and more intense asymmetric flow. The planar lamination of the
seaward slope seems inconsistent with the Clifton sequence (i.e., planar
lamination found in shallower water depth than foreset bedding). In the
study of the barred coast of Kouchibouguac Bay, New Brunswick, Canada,
an explanation has been suggested for the absence of foreset bedding on the
seaward bar slope. The planar lamination have been interpreted as upper
flow-regime structures of a symmetric oscillatory flow (Davidson-Arnott and
Greenwood, 1976). This sequence progresses from ripples directly t o flat
beds with no intermediate formation of megaripples (Fig.lB).
The structures at depth in the cores of the transition subenvironment are
the same as those of the seaward bar slope. Since the only difference be-
tween both environments is the distance landward or seaward of fair-weather
wave base, the major distinction between these two environments is textural.
254
In the transition the degree of bioturbation increases seaward. This trend is

similar to the “parallel laminated-to-burrow sets” observed in cores from the
nearshore system off Sapelo Island, Georgia, and viewed in outcrop from the
late Cretaceous of east-central Utah (Howard, 1971). The offshore sub-
environment is characterized by a lack of physical structures at depth due
to the high degree of bioturbation and the apparent infrequency of compe-
tent flows. Occasionally, biogenic structures are infilled with organic detritus
and preserved. The dominant physical structures throughout the inner shelf
coarse-grained deposit are landward-dipping, medium-scale foreset bedding
caused by the occasional migration of the steep slipfaces of the oscillatory
ripples.
The shoreface-inner shelf contact at the base of the longshore bar (depth
- 1 0 m) marks the most landward indication of major biogenic reworking of
nearshore system. Landward of this point, physical sedimentary structures
are dominant. The transition subenvironment is characterized by a seaward
increase in bioturbation. A t the point where few physical structures are
present due to bioturbation, the offshore subenvironment begins. In eastern
Long Island, the change from the transition to the offshore is approximately
at a 14-16 m depth between 1.4-1.6 km seaward of MLW. This study
clearly documents the transition as a distinct subenvironment identifiable
by discrete sedimentological characteristics. This view challenges the existing
convention of a “transition zone” which has been considered as a simple
“mixing” of shoreface and offshore components (e.g., Emery, 1960; Reineck
and Singh, 1975).
Origin of the coarse-grained deposit
The presence of fields of large, linear bedforms partially composed of

coarse-grained sediment has been documented for the inner shelf environment
off Atlantic Beach, western Long Island (Swift and Freeland, 1978; Swift et al.,
1979). With the aid of side-scane sonar, well-defined sets of features, interpret-
ed as low relief sandwaves, have been observed in water depths between 9 and
14 m. These features exhibit a 10-100 m spacing, 1m height, and crestlines
oriented perpendicular to the shoreline. The sandwaves consist of a smooth,
low-angle stoss slope (east-facing), a less smooth higher angle lee slope
(west-facing), and a rippled flat trough. The bed roughness of the sandwave
troughs is due to large oscillatory ripples generated by infrequent storm
waves, and, therefore, are oriented crest-parallel to the shoreline.
Swift and Freeland (1978) and Swift et al. (1979) attributed the origin of
the inner shelf sandwaves to unidirectional flow caused by extratropical
storms. Measurements by Scott and Csanady (1976) and Lavelle et al.
(1978a, b) revealed storm-induced westward bottom currents flowing
parallel to the isobaths of the Long Island shelf. The large ripples in the
troughs of sandwaves described for the western Long Island shelf seem
analogous t o the coarse-grained deposits of eastern Long Island observed
in the present study.
255
Stratigraphic implicatw ns
A stratigraphic model for progradation of the Long Island nearshore

system would likely consists of a t least a portion of all six sedimentary
subenvironments (Fig.10). The entire sequence would be underlain by
organic-rich bioturbated offshore deposits. Conformably overlying the
offshore would be the planar-laminated, slightly bioturbated, transition
sediments. The contact between the offshore and the transition would be
highly gradational and is placed a t 15 m in this section. Dispersed through-
out the offshore and transition subenvironments would be lenses of inner
shelf coarse-grained deposit. Grading upward from the transition are the
cleaner, planar-laminated, lower seaward slope sediments of the longshore
bar. Above this is a distinct erosional surface indicating the base of the
coarse, massive to possibly cross-bedded deposits of the longshore trough.
The 2 m unit of bar sands below this erosional surface would be the only
LANDWARD
FLOW
FORESHORE
5
I
I
0
K
U
I
I-
n
W
n
LONGSHORE
LT
W BAR
5 10
Fig. 10. Hypothetical prograding stratigraphic sequence for the eastern Long Island near
shore system. Solid flow arrows indicate observed flow conditions, while dashed arrows
indicate postulated flow conditions. The structures in the foreshore were not observed in
the present study and are assumed from such sources as Elliott (1978) and Hunter et al.
(1979). The lensoidal structures in the transition are the coarse-grained deposits.
256
portion of the longshore bar preserved. This would be due to erosion of

shallow bar crest deposits by the progradation of the longshore trough.
Capping the sequence is the planar- to cross-laminated clean sands of the
upper shoreface and the proposed planar laminations of the foreshore.
One application of this study to interpretation of ancient environments
may be the development of firmer criteria for determining the relative
position of fair-weather wave base (or at least position landward or seaward
of it). Some of these points have been discussed by earlier workers (e.g.,
Howard, 1971; Howard and Reineck, 1972a), but consideration of all
criteria together may assist future investigations. The three criteria are:
(1)The amount and type of physical and biogenic structures. Preservation
of biogenic structures would indicate a position seaward of wave base.
Generally, the more bioturbation, the farther seaward the environment.
(2) The presence of drapes (mud or organics) in small-scale ripple cross-
lamination (flaser bedding). This criterion is a very sensitive indicator of
distance seaward of wave base. The greater the seaward distance, the more
opportunity for the finer material to settle out.
(3) An increased darkening of overall sediment color may indicate approx-
imate distance seaward of wave base. This criterion would be subtle and/or
poorly preserved, but may be useful when considered with other indicators.
CONCLUSIONS
Four conclusions can be drawn concerning the barred nearshore sedimen-

tary system from this study of Long Island:
(1) S i x subenvironments were observed during fair-weather conditions.
The shoreface environment is divided into the upper shoreface, the longshore
trough, and the longshore bar. The subenvironments of the inner shelf are
the transition, the offshore, and the coarse-grained deposits. These sub-
environments were defined during fair-weather conditions as distinct com-
binations of sediment texture and color, bedforms, physical and biogenic
sedimentary structures, and infaunal assemblages.
(2) The order of magnitude reduction in slope (3.0-0.3") a t the shoreface-
inner shelf boundary coincides with fair-weather wave base. The location of
wave base can be characterized at any moment in time by: (a) the cessation
of debris surge in ripple troughs; ( b ) the onset of the rust layer formation
over the bedform surface; (c) the darkening of sediments caused by the
increased presence of organic detritus; and (4)subtle changes in sediment
texture.
(3) Sedimentary structures produced by shoaling waves follow a consistent
shoreward progression which is controlled by variations in oscillatory velo-
city and velocity asymmetry, This oscillatory flow regime concept as pro-
posed by Clifton (1976) and modified by Davidson-Arnott and Greenwood
(1976) can be used to qualitatively predict oscillatory flow asymmetry.
Specifically, cross ripples are a particularly good indicator of the asymmetric
transition between asymmetric ripples and lunate megaripples.
257
(4)Criteria useful for distinguishing position relative t o fair-weather wave

base in ancient nearshore environments are: (a) the amount and type of
physical and biogenic structures; (b) the occurrence of mud or organic
drapes in the small-scale ripple cross-laminations; and (c) the overall possible
darkening of sediment color due to the increase of fine-grained material.
ACKNOWLEDGMENTS
This study was carried out as part of the research program of the Coastal
Research Division, Department of Geology, University of South Carolina.
The project was made possible by a grant from the Morris Companies of
Secaucus, New Jersey. The final preparation of the manuscript was under-
taken with the support of the facilities a t the National Museum of Natural
History, Smithsonian Institution. Miles 0. Hayes, John C. Horne, Robert
Ehrlich, Larry G. Ward, and Gary A. Zarillo read earlier copies of the manu-
script. Robin G.D. Davidson-Arnott, Brian Greenwood, and Gary A. Zarillo
critically reviewed the final copy. Charlotte Johnson assisted in the drafting
of the figures. Appreciation is also extended t o Stephanie A. Staples in the
preparation of the final manuscript. A special thanks is offered t o John H.
Banvis and Robert Ehrlich for their initial suggestions and enthusiasm for
the undertaking of this study. Their continued support throughout the en-
tire study is also acknowledged.
REFERENCES
Anan, F.S., 1972. Hydraulic equivalent sediment analyzer (HESA). Coastal Research
Center, Univ. of Massachusetts, Techn., Rep. 3-CRC, 37 pp.
Bouma, A.H., 1969. Methods for the Study of Sedimentary Structures. Wiley, New York,
N.Y., 458 pp.
Clifton, H.E., 1976. Wave-formed sedimentary structures - a conceptual model. In: R.A.
in the non-barred high energy nearshore. J. Sediment. Petrol., 41: 651-670.
Curray, J.R. and Moore, D.G., 1964. Pleistocene deltaic progradation of continental ter-
race, Costa de Nayarit, Mexico. In: T.H. van Andel and G.G. Shor (Editors), Marine
Geology of the Gulf of California. Am, Assoc. Pet. Geol., Tulsa, Okla., pp.193-215.
New Brunswick, Canada. J. Sediment. Petrol., 44 : 698-704.
coast, Kouchibouguac Bay, New Brunswick, Canada. In : R.A. Davis and R.L. Ethington
(Editors), Beach and Nearshore Sedimentation, SOC.Econ. Paleontol. Mineral., Spec.
Publ., 24: 149-168.
Davies, J.L., 1964. A morphogenic approach to world shorelines. Ann. Geomorphol., 8 :
127-1 42.
DeWall, A.E., 1979. Beach Changes at Westhampton Beach, New York, 1962-1973.
Coastal Engineering Research Center, Misc. Rep., 79-5, Corps of Engineers, Fort
Belvoir, Va., 1 2 9 pp.
258
Dietz, R.S. and Fairbridge, R.W., 1968. Wave base. In: R.W. Fairbridge (Editor), Dictio-
nary of Geomorphology (Vol. 111). Reinhold, New York, N.Y.
Elliott, T., 1978. Clastic shorelines. In : H.G. Reading (Editor), Sedimentary Environment
and Facies. Elsevier, New York, N.Y., pp.143-175.
Emery, K.O., 1960. The Sea off Southern California. Wiley, New York, N.Y., 366 pp.
Goldsmith, V., Bowman, D. and Krinsley, K., 1982. Sequential stage development of
crescentic bars: Hahotherm Beach, southeastern Mediterranean. J. Sediment. Petrol.,
52: 233-249.
Greenwood, B. and Hale, P.B., 1980. Depth of activity, sediment flux, and morphological
Canada: Littoral Processes and Shoreline Morphology. Geol. Surv. Can., Pap. 80-10 :
89-109.
Greenwood, B. and Mittler, P.R., 1979. Structural indices of sediment transport in a
straight, wave-formed nearshore bar. Mar. Geol., 32: 191-204.
Hamblin, W.K., 1971. X-ray photography. I n : R.E. Carver (Editor), Procedures in Sedi-
mentary Petrology. Wiley-Interscience, New York, N.Y., pp.251-284.
Hill, G.W. and Hunter, R.E., 1976. Interaction of biological and geological processes in
the beach and nearshore environments, northern Padre Island, Texas. In : R.A. Davis
and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SOC.Econ. Pale-
ontol. Mineral., Spec. Publ., 24: 169-187.
Howard, J.D., 1968. X-ray radiography for examination of burrowing in sediments by
marine invertebrate organisms. Sedimentology, 11: 249-258.
Howard, J.D., 1971. Amphipod bioturbate textures in recent and Pleistocene beach sedi-
ments. In: J.D. Howard, J.W. Valentine and J.E. Warme (Editors), Recent Advances in
Paleoecology and Ichnology. Am. Geol. Inst., Short Course Lecture Notes, pp.213-
233.
Howard, J.D. and Frey, R.W., 1975. Estuaries of the Georgia coast, U.S.A. - sediment-
ology and biology, I : Introduction. Senckenbergiana Mar., 7: 1-31.
Howard, J.D. and Reineck, H-E., 1972a. Georgia coastal region, Sapelo Island, U.S.A. -
sedimentology and biology, IV: physical and biogenic sedimentary structures of the
nearshore shelf. Senckenbergiana Marit., 4:81-123.
Howard, J.D. and Reineck, H-E., 197213. Georgia coastal region, Sapelo Island, U.S.A. -
sedimentology and biology, VIII: conclusions. Senckenbergiana Marit., 4 : 217-222.
Howard, J.D. and Reineck, H-E., 1981. Depositional facies of high-energy beach-to-off-
shore sequence: comparison with low-energy sequence. Bull. Am. Assoc. Pet. Geol.,
65: 807-830.
Kent, H.C., 1976. Modern Coastal Sedimentary Environments - Alabama and Northwest
Florida. Geological Exploration Associates, Golden, Colo., 96 pp.
Kravitz, J.H., 1966. Using an ultrasonic disruption as an aid to wet sieving. J. Sediment.
Petrol., 36: 811-812.
Lavelle, J.W., Swift, D.J.P., Gadd, P.E., Stubblefield, W.L., Case, F.N., Brashear, H.R. and
Haff, K.W., 1 9 7 8 a Fair weather and storm sand transport on the Long Island, New
York, inner shelf. Sedimentology, 25: 823-842.
Lavelle, J.W., Young, R.A., Swift, D.J.P. and Clarke, T.L., 1978b. Near-bottom sediment
concentration and fluid velocity measurements on the inner continental shelf. J. Geo-
phys. Res., 83: 6052-6062.
Lettau, B., Brower, W.A. and Quayle, R.G., 1976. Marine Climatology. MESA New York
Bight Atlas Monograph 7, New York Sea Grant Institute, Albany, New York, 239 pp.
Newton, R.S., 1968. Internal structure of wave-formed ripple marks in the nearshore
zone. Sedimentology, 11: 275-292.
259
Reineck, H-E., 1963. Sedimentgefuge im Bereich der Sudlichen Nordsee. Abh. Senckenb.
Naturforsch. Ges., 505, 138 pp.
Reineck, H-E. and Singh, I.B., 1971. Der Golf von Gaeta (Tyrrhenisches Meer) - 111. Die
Gefuge von Vorstrand- und Schelfsedimenten. Senckenbergiana Marit., 3 : 185-201.
Reineck, H-E. and Singh, I.B., 197 5. Depositional Sedimentary Environments. Springer,
New York, N.Y., 439 pp.
Saville, T., 1954. North Atlantic coast wave statistics hindcast by Bretschneider-revised
Sverdrup-Munk method. Beach Erosion Board, Techn. Mem. 55, Corps of Engineers,
Washington, D.C., 76 pp.
Scott, J.T. and Csanady, G.T., 1976. Nearshore currents off Long Island, J. Geophys.
Res., 81: 5403-5407.
Swift, D.J.P. and Freeland, G.L., 1978. Current laminations and sandwaves on the inner
shelf, middle Atlantic bight of North America. J. Sediment. Petrol., 48: 1257-1266.
Swift, D.J.P., Freeland, G.L. and Young, R.A., 1979. Time and space distribution of
megaripples and associated bedforms. Sedimentology, 26 : 389-406.
Taney, N.E., 1961. Geomorphology of the south shore of Long Island, New York. Beach
Erosion Board, Techn. Mem. 128, Corps of Engineers, Washington, D.C., 50 pp.
U.S. Army Corps of Engineers District, New York, 1977. Final Environmental Impact
Statement for Fire Island Inlet to Montauk Point, New York - Beach Erosion Control
and Hurricane Protection Project. Vols. I and 11, 375 pp.
U.S. Dept. of Commerce, 1978. Tide Tables - East Coast of North and South America.
Natl. Ocean. Atmos. Admin., National Ocean Survey, Rockville, Md., 288 pp.
Wood, W.L., 1971. Transformation of breaking wave parameters over a submarine bar.
Unpubl. Ph.D. Diss., Michigan State University, East Lansing, Mich., 224 pp.
Marine Geology, 60 (1984) 261-282 26 1
BEACH AND NEARSHORE FACIES: SOUTHEAST AUSTRALIA
A.D. SHORT
Coastal Studies Unit, Department of Geogmphy, University of Sydney, Sydney,
N . S . W. 2006 (Australia)
(Received December 31, 1982; revised and accepted August 29,1983)
ABSTRACT
Short, A.D., 1984. Beach and nearshore facies: southeast Australia. In: B. Greenwood
Coastal Environments. Mar. Geol., 60: 261-282.
The morphology, texture and facies sequence on seven sand beaches, located in low,
moderate and high wave energy, microtidal environments in southern Australia were
investigated using box coring and Scuba observations. Systematic variation in facies occur
both within and between the beaches. Low-energy reflective beaches are limited in
lateral and vertical extent and in facies to beach laminations separated by coarse step
deposits from finer nearshore cross-lamination facies. Moderate-energy intermediate
beaches characterised by rip circulation possess increasingly wider surfzones with ridge
and runnel and bar-trough facies separating the beach and step facies from the more
extensive nearshore sequence. High-energy dissipative beaches may have 600 m wide surf-
zones containing multiple bar-trough topography. Fine beach laminations with backwash
structures grade into 4-5 m thick bar-trough sequences then the extensive nearshore
facies. As wave energy increases from low (Hb< 1 m) to high (Hb > 2.6 m) the vertical
extent of the beach to nearshore sequence increases from < 1 0 m to approximately 30 m,
and the width from 100 m to several kilometres. Consequently one would expect higher-
energy paleo-beach sequences to be represented more by diagonal than vertical facies
sequences.
INTRODUCTION
The beach and nearshore zone extends from the upper swash limit across
the surfzone t o modal wave base, the limit to which modal waves actively
entrain sediments. The entire zone consists of depositional facies formed by
wave-current dynamics and the associated boundary layer flows at the bed.
The abundance of paleo-beaches attests to the preservation potential of such
deposits during progradational episodes. Understanding the relationship
between beach facies and the environmental conditions that produce them is
important for several reasons. First, it enables a more complete definition of
the morpho-stratigraphic characteristics of a beach; second, in so far as the
facies reflect processes at the time of deposition they can indicate present
day wave and bed dynamics (e.g., Clifton et al., 1971; Davidson-Arnott and
Greenwood, 1976; Davidson-Arnott and Pember, 1980); and third, given the
0025-3227/84/$03.00 0 1984 Eloevier Science Publishers B.V.

262
above, if the structures are preserved in the rock record they can equally be
used to identify paleo-beaches and t o interpret the prevailing environmental
conditions at the time of deposition (e.g., Clifton et al., 1971; Reineck
and Singh, 1973; McCubbin, 1982; Allen, 1982; Duprk, 1984, this volume).
The pioneering, though recent, investigations in this field recognised the
potential range of beach environments in response to varying levels of
wave energy and beach configuration. High- and low-energy systems were
described by Clifton (1976), and Howard and Reineck (1981). Specific
beach configurations such as barred nearshores (Davidson-Arnott and
Greenwood, 1974, 1976; Hunter et al., 1979; Greenwood and Mittler, in
press) illustrate alternatives to the classic planar nearshore of Clifton et al.
(1971). The recently developed models of morphodynamic beach response
to low (<1m) or high (>2.5 m) waves (Fig. 1)by Wright and Short (1983),
and Short and Wright (1983), provides a basis for a systematic study of
beach and nearshore facies across a range of wave environments. The aim of
this paper is to describe the bedforms, textures, structures and resulting
facies from a range of beaches, exposed to modally low, moderate and high
wave activity. More specifically results are presented from seven contrasting
coastal environments in southeast Australia. The characteristic facies se-
quence of each is presented together with the extent of the various facies
within each system.
FIELD SITES AND METHODS
Seven beaches in southeast Australia (Fig.2) that are representive of

Wright and Short’s (1983) six beach states (Fig. 1)were investigated. The
morphodynamic characteristics of each field site are given in Table I.
East coast sites are located in a micro-tidal (spring range 1.6 m), east coast
swell environment, with highly variable wave regime. The modal deep
water wave is 1.5 m in height with a period of 10 s. Goolwa on the south
coast has a 1.0 m spring tide range and a modal 3 m, 1 2 s wave, typical of
this west coast swell environment. Sands predominantly consist of quartz
with variable percentages of carbonates (shell fragments). All systems were
surveyed using the Emery method in shallow water and echosounder in deep
water. All systems were box-cored subaerially and subaqueously the later by
Scuba divers who also measured bedforms (Short and Wright, in press).
Using a modified method of Burger et al. (1969), box-cores were impreg-
nated with Ciba-Geigy araldite (K79 kit) t o give a 30 X 20 cm cast of near
surface structures. In addition samples of sediment were dry sieved at 0.25
phi intervals t o determine grain-size statistics.
BEACH TEXTURES AND STRUCTURES
The results from the seven beach systems are presented briefly to give an
indication of the nature, extent and relationship of the facies within each
representative beach type. These results are then combined into a more
general classification.
263
T
(a) 200
-d
v)
v)
100
LONGSHORE BAR TROUGH

300 m
200
RHYTHMIC BAR AND BEACH
u 200 m ~ A'
(c) 100
0 -B'
DISSIPATIVE
200 m A'
(d) 100 E DOMAIN V A R l P.BLE

P A T I V E DOMAIN
0 B'
H I G H T I D E REFLECTIVE
LOW T I D E D I S S I P A T I V E
200 m A'
100 LOW T I D E D I S S I P A T I V E
0 B'
REFLECTIVE DOMAIN
/(HIGH t LOW TIDE1
H
2 200 m I
E
0 (f) 100
u
$P 0
100 200
Id L 300 m
Fig.1. Plan and profile configuration and basic surfzone circulation pattern of the six
beach states (from Wright and Short, 1983).
264
I
50" E
I
151'
I
152' ' 153°C
PORT MACQUARIE
32'
33"
PEARL BEACH
r
NARRABEEN BEACH
FISHERMANS BEACH
+ 34
0
m
-
SCALE
- 50 Km
S E V E N MICE BEACH
35'5
Fig.2. Location of the seven beach sites (arrowed) selected for sediment sampling and
box-coring. The wave-sediment characteristics of each are listed in Table I.
LOW-ENERGY REFLECTIVE
Fishermans Beach is a modally low wave energy reflective beach (see

Figs. If, 3a and 4) composed of medium to very coarse sand. Modal breaker
height is 30 cm, with breakers rarely exceeding 1 m (Table I). The beach
consists of a moderately steep (10") beach face capped by an incipient
foredune. A very coarse-grained step lies at about mean low water separating
the beach face from the nearshore. The nearshore has a lower gradient (3"),
and is approximately 50 m wide terminating at a depth of 3 m (modal wave
base).
265
TABLE I
Wave-sediment characteristics of beach sites
I LOCATION I FISHERMANS P E A R L \ H A W K S NEST NARRABEEN GRANTS M.SEVEN GOOLWA I

BEACH T Y P E REFLECTIVE
I INTERMEDIATE DISSIPATIVE
10 17 14 35 21 25 12
Hb(m) .3 .5 1 1.5 1.6 1.6 3
T 10 10 10 10 10 10 12
Gd(mm1 .35 .5 .26 .3 .3 .27 .2
ws .05 ,076 .035 .04 .04 ,036 ,026
n .60 .65 2.85 3.75 4.0 4.4 9.6
Gradient 1.9 1:12 1 13 1.15 1 29 1.37 1.33
8-5 Reflective Reflective Ridge Transverse Rhythmic Bar Dissipative

Runnel Bar-Rip Bar-Beach Trough
Facies (Figs.5 and 6a)
The beach face is composed of thinly bedded sub-parallel, laminations

of varying thickness with variable lateral continuity, that dip seaward at
-5" (Fig.5a). Sediment varies from very coarse to medium sands. A high
degree of variability in texture and laminae thickness reflects the immaturity
of the beach face sediments in this low-energy environment.
The step consists of medium-scale sets (- 10 cm) of cross-stratified gravels
(Fig.5b). These gravels are poorly sorted, and rich in carbonate. The step is
located at the toe of the beach face at the point of wave surging. It is usually
a few decimetres in thickness. It overlies and abruptly grades into nearshore
sediments. The nearshore has fine-to-medium, well-sorted, low-carbonate
sands (Fig.6); sets of medium-scale, landward-dipping cross-laminations with
0-20" dip angles are predominant (Fig.5~).The cross-lamination is pro-
duced by asymmetrical oscillation ripples with straight parallel to slightly
sinuous crests [length ( L ) = 40 cm, height ( H ) = 7 cm] .
HIGHER ENERGY REFLECTIVE
Pearl Beach is a more energetic beach than Fishermans, though still reflec-
tive and, apart from its greater extent (Fig. 3b) it has several other distin-
guishing characteristics resulting from the higher wave energy. The beach is
wider consisting of a foredune fronted by a berm runnel and series of beach
cusps. The 10" beach face grades into a coarse step. The nearshore zone
slopes seaward at 5-6O from the step, t o a depth of 6-8 m where it levels
out into a low gradient (0.5') offshore zone (Fig.4).
266
(a) (b)
Ws 005 WS 0076
4
n1 - 0 6 0 fl ~ 0 6 5
0 100rn 0 100 rn 0 100 rn 0 100 200 m
1 - ;(f) , I
G R A N T S BEACH M I D - SEVEN MILE B E A C H

RHYHMIC BAR AND BEACH BAR - TROUGH
MAY I82 I2 c l3.11k81
2l \ Beach Face
Bar
-4 -4
0 100 200 m 0 100 200 300 rn
4 ,f&T7--T-, , , , , , , , , , , , , , ,
GOOLWA BEACH
DISSIPATIVE
MLw kOuter Breaker Zone-+
-2 -
Hb= 30
W, = 0026
-4 - A = 96
-6 ' l l ' ' l l ' ~ ' l ~ l l ~ ' l ' ~ l ~ i l l

100 200 300 400 500111
Fig. 3. Beach-surfzone profiles across the field sites showing cross-sectional morphology
and location of box cores, indicated by squares and solid squares (illustrated in Figs.5
and 7-12). See Table I for definition of H,, w, and a .
Facies (Figs.6a and 7 )
The berm crest consists of 2-10 cm thick units of planar continuous par-
allel laminae (see Fig.7a). The units alternate between coarse (runnel depos-
its) and finer sand. The beach face contains seaward dipping (<5"), continu-
ous, planar to curved parallel beds, 10-15 cm thick, and composed of either
coarse or finer sand. The lower beach face has steeper (10-20") more dis-
continuous planar to curved, non-parallel beds of coarse (shell rich) sand.
These grade into the 30 cm thick, partially cross-laminated (seaward dipping
Z O O ) , coarse sand beds of the step (Fig.7b). Immediately seaward of the step,
267
PRnFll FS i
- 3 @ L A - L - - L l ~ 1- J I I L L I1 I 1 1 - I l i
I 2 3 4
DISTANCE SEAWARD ( Km)
Fig.4. Nearshore profiles across the field sites showing location of nearshore box cores
(squares).
Fig.5. Fisherman's beach box cores. Shore to left. ( a ) beach face; ( b ) step; ( c ) nearshore,
depth 2 m. Scale in centimetres.
sediments fine rapidly and sinuous oscillation ripples produce predominantly

landward dipping (10") 5-10 cm thick cross laminations ( F i g . 7 ~ ) .These
extend offshore t o modal wave base at 8 m where bioturbation affected
10--20% of the 30 cm deep core.
RIDGE AND RUNNEL
Hawks Nest Beach is located toward the southern end of the 16 km long
Fens embayment. Fens grades from a moderate t o high energy rhythmic
bar and beach system ( F i g . 1 ~at
) the northern end t o a low-energy, reflective
268
( 0 ) SIZE o SORTING C A R B O N A T E % (b) SIZE8 5 {TING CARBONATE

-1 !i I 2 3 0s I 0 I0 23 30 -1 0 I 2 3
l-i 1 ' \ I
II'
---- FISHERMA NS
PEARL -(a
--- HAWKS NEST 1
NARRAEEEN ~1
GRANTS
MID SEVEN MILE
.................. GOOL WA
\
Fig.6. Vertical sequence of sediment characteristics (size, sorting, and percent carbonate)
across the beach, surfzone and nearshore zones of the seven field sites. Non-carbonate
sediments are predominantly quartz grains.
system at the more protected southern end. Hawks Nest beach experiences
a range of wave levels (Hb up t o 3 m) and beach types, but modally is a
ridge and runnel type (Fig.le) as it was during the field investigation
. has a low foredune fronted by a 40 m wide berm-runnel and
( F i g . 3 ~ ) It
berm, with a 6" beach face which terminates at a low tide step. The surfzone
consists of a shallow narrow runnel and flat ridge. Past the breakpoint, slope
increases t o 6" until reaching modal wave base and a 0.2" gradient offshore
zone at 10 m depth (Fig.4).
269
Fig.7. Pearl beach box cores. Shore to left. (a) berm crest; ( b ) step; ( c ) nearshore, depth
1 m. Scale in centimetres.
Facies (Figs.6a and 8)
Sediments are predominantly medium t o fine sand (Fig.Ga), finer than the
two reflective beaches described above. The berm consists of continuous
parallel laminae, arranged in thin beds, grain size is in the medium t o fine
sand range. The upper (high tide) beach face contains >30 cm thick beds of
seaward dipping (5") continuous, parallel laminae (up t o 3 cm in thickness)
of medium-grained sand with occasional coarser grained, shelly laminae
(Fig.8a). The lower (low-tide) beach face had shallow scour depressions on
the surface with structures alternating between those similar t o the upper
beach face and 1 0 cm thick beds of coarse t o very coarse seaward dipping
(15") shelly material. The latter represent the step deposits formed at high
tide when the beach is more reflective (Fig.8b).
The runnel contained long-crested wave ripples ( L = 100 cm, H = 10 cm),
which produced steeply dipping (20") cross-stratification consisting of tabular
t o trough parallel laminae arranged in 20 cm beds of medium-grained sand.
These deposits are overlain by ridge sediments which follow the sequence
of Davis e t al. (1972). The onshore part of the ridge contained landward
dipping (10-30") cross strata while the crest contained sub-horizontal
strata (Fig.&).
Cross ripples occurred seaward of the break point and were best developed
at 4 m depth ( L = 20-40 cm, H = 5-8 cm). They produced predominantly
landward dipping cross strata (10-30", Fig.8d) similar t o that described by
Clifton (1976). Between depths of 4 and 5 m were large megaripples ( L =
400 cm, H = 50 cm; which could have been produced by 3 m waves three
days previously). The megaripples had tangential, predominantly landward-
dipping (10-20") cross strata on the crest, consisting of fine t o medium
sand, with shelly cross strata (10-15 cm thick) in the trough. This overlay
seaward dipping (10") slightly bioturbated parallel laminae (pre-high waves?)
(Fig. 8e). The zone of megaripples graded into parallel sharp crested, wave
Fig.8. Hawks Nest box cores. Shore t o left. ( a ) upper beach face; ( b ) lower beach face
step; ( c ) bar crest; ( d ) nearshore (cross ripples), depth 4 m ; ( e ) nearshore (megaripple
trough), depth 5 m ; ( f ) nearshore, depth 6 m. Scale in centimetres.
ripples ( L = 30 cm, N = 5 cm) which became sinuous at 8-10 m depth.

Internal structures were planar t o ripple cross laminations with tangential
landward dipping (lCF2O”) cross beds, of fine t o medium sand, and shell
rich (Banhavia) (20-40”) landward-dipping layers that were bioturbated
in the deepest part of this zone (10-15 cm thick). Banhavia both living and
dead were abundant on the surface and in cores taken between 5 and 9 m
water depth (Fig.8fj.
TRANSVERSE BAR AND RIP
Narrabeen Beach ranges in morphology from reflective t o intermediate

type in response t o a highly variable wave climate with breakers frequently
> 3 m and occasionally > 5 m ; modal wave height is 1.5 m (Short and Wright,
1981). The beach was sampled and cored on three occasions when well-
271
developed transverse bar and rip systems (Fig. I d ) dominated the morphology
(Fig.3d). The berm and beach face varied in width from 20 to 50 m depend-
ing on location relative to the megacusp horns and embayments (Short,
1979). The longshore spacing between megacusps averaged 150 m. The
megacusp embayments were fronted by rip feeder and rip channels (0.5-
1.5 m deep) which run normal to the shoreline across the bar (see Wright and
Short, 1983). The megacusp horns were attached to the transverse bars
which continued 50-80 m seaward to the break point. The nearshore zone
slopes at 1.5" from the break point to beyond the modal wave base at 18 m
(Fig.4). The system is morphologically analogous to the connected inner bar
systems described by Greenwood and Davidson-Arnott (1975).
Fucies (Figs.6b and 9)
The berm and upper beach face deposits are similar to the previous
two beaches. On the megacusp horns the beach face has a 5" gradient with
seaward dipping (1-2") parallel laminae which grade into coarser grained,
10 cm thick cross strata (10-15") at the junction with the bar. A step is
absent. In the embayment coarser grained, 2-5" seaward dipping tangential
laminae, grade into a zone of cross strata before a very coarse grained, shelly
step with steeply (20") seaward dipping beds is reached (Fig.9a). The bar
facies is similar in sequence to the Hawks Nest ridge deposits; they are how-
ever more extensive, slightly coarser and higher in skeletal carbonates. The
rip feeder channels, analogous to Hunter et al.'s (1979) 'longshore trough
facies' and Davidson-Arnott and Greenwood's ( 1976) 'trough' facies, con-
tained both wave and current ripple structures. They consisted of very
coarse, shell-rich, predominantly seaward dipping (5-40") cross strata with
laminae arranged in alternating 5-10 cm thick sets (Fig.9b). The rip channel
contained seaward migrating megaripples ( L = 150 cm, H = 25-30 cm).
The medium- to coarse-grained megaripple crests produce seaward dipping
(10-25") tangential laminae (Fig.9c). The troughs consist of coarse sand
arranged in medium-scale cross strata (10-20" dip). These structures are
basically identical to the rip-channel facies described by Davidson-Arnott
and Greenwood (1976) and Hunter et al. (1979).
Immediately seaward of the break point and to a depth of 5 m cross
ripples dominated ( L = 50-100 cm, H = 5-15 cm). They contained
5-10 cm sets of landward dipping (5") tangential laminations, overlain by
sets of steeper (10-40") predominantly landward dipping cross laminations
(Fig.9d). Beyond 6 m depth, sinuous, sharp crested wave ripples ( L =
40-80 cm, H = 7-10 cm), pass laterally into parallel-crested forms ( L =
40-60 cm, H = 1 0 cm) which extend to modal wave base a t 18 m depth.
Cross strata dominate with 5" landward dipping laminae truncating (20")
seaward dipping strata arranged in 5-10 cm thick sets (Fig.9e). Beyond
modal wave base sediments rapidly coarsen, becoming shellier with poorer
sorting (Fig.6 b). However, periodic high waves produce well-developed
sharp-crested, parallel wave ripples ( L = 50 cm, H = 10 cm) out to a depth of
272
Fig.9. Narrabeen beach box cores. Shore to left. (a) step; ( b ) longshore trough; (c) rip
channel (megaripple crest); ( d ) nearshore (cross ripples), depth 3 m; (e) nearshore, depth
8 m; (f) nearshoreoffshore, depth 19 m ; ( g ) offshore, depth 24 m. Scale in centimetres.
273
at least 30 m. Crossbedding structures were apparent in the cores. Relatively

thick (1-2 cm) landward dipping (5") laminae overly 10-20" seaward dip-
ping beds (Fig.9f and g). Bioturbation increased markedly seaward of 18 m
depth.
RHYTHMIC BAR AND BEACH
Grants Beach is more exposed t o deep-water waves and has slightly finer
sediments than Narrabeen (Fig.6b). Consequently it is modally more
energetic (Hb= 1.6 m) and more often has a rhythmic bar and beach morph-
ology (Fig.le). It is rarely reflective and under high waves can become
dissipative. It was investigated when a well-developed crescentic bar system
was present, the bars were not attached t o the shoreline. The sampling line
crossed the 80 m wide moderate gradient beach, a 60 m wide 3 m deep
trough with a 1.5 m deep bar crest lying 7 F 9 0 in seaward of the shoreline
(Fig.3e). Waves were low (0.5 m ) at the time of sampling. The nearshore
zone is convex in shape (Fig.4) and perhaps bedrock controlled, though none
was observed.
Facies (Figs.6b and 10)
Berm structures consisted of horizontal, parallel, thin laminae composed

of medium- and fine-grained sands. The upper beach face contained seaward
dipping (5") parallel and tangential beds of alternating fine and coarse
grained laminae (Fig.lOa). These graded into predominantly seaward dipping
(10-15") cross strata on the lower beach face (Fig.lOb) and a coarse-grained
shelly 1 F 2 0 cm thick step sequence at low water. The beach face sequence
was similar in gross form to Hawks Nest and the Narrabeen rip embayment.
The distinctive characteristic of this beach type is however the extensiw
bar and trough sequence. The deep rip trough contained straight to sinuous
parallel wave ripples ( L = 50-60 cm, H = 10-20 cm) in coarse-grained
shelly sediments. Structures were similar to the Narrabeen rip channel with
seaward dipping (20") laminae, arranged in 10-20 cm thick sets of alter-
nately coarse and fine sand (Fig.lOc). The trough facies contrasts with the
finer-grained better sorted bar sediments which contained predominantly
landward dipping (5-20") parallel laminae overlain with ripple cross strata
on the crest (Fig.lOd), produced by sinuous wave ripples ( L = 50 cm, H =
1 5 cm) present a t the time of coring.
Seaward of the breaker zone lunate megaripples ( L = 5+60 cm, H =
1 0 cm) produced 10-15 cm thick sets of predominantly landward dipping
( 1 F 3 0 " ) parallel laminae. Between 4 and 1 2 m depth, sharp-crested parallel,
wave ripples ( L = 6-8 cm, H = 1-2 cm) were encountered. They produced
horizontal t o slightly landward dipping wavy laminae (Fig.lOe). These ap-
pear analogous to the 'inner offshore' facies of Hunter et al. (1979). At
Grants Beach the small scale structures and increasing bioturbation, which
occurred seaward of 8 m depth (Fig.lOf and g), reflected the prevailing
274
Fig.10. Grants Beach box cores. Shore to left. (a) mid beach face; ( b ) lower beach face,
above step; ( c ) longshore trough; (d) bar crest; ( e ) nearshore, depth 8 m; ( f ) nearshore,
depth 1 2 m; ( 9 ) nearshore, depth 16 m. Scale in centimetres.
275
low swell conditions. Under more normal wave conditions a sequence such
as the Narrabeen nearshore would be expected. The small wave ripples
became sinuous beyond 16 m depth. Banhavia were prominent both on the
surface and in the shallow cores between 1 2 and 21 m depth.
BAR TROUGH
Mid-Seven Mile beach is fully exposed t o the regional deep-water wave

regime (H,, = 1.6 m) which, combined with predominantly fine sand, results
in a low gradient beach and a double bar-trough surfzone (Figs.lb and 3f).
The inner bar varies from ridge and runnel t o bar-trough in response to
varying wave conditions, while the outer bar-trough, apart from on-offshore
movement of the bar crest maintains its form year round (Short and Wright,
1983). The outer bar commonly lies over 100 m seaward of low water
(Fig.3f). The beach, inner bar-trough and nearshore (Fig.4) were cored, while
breakers prevented coring of the outer bar-trough system.
Fucies (Figs.6b and 11)
The wide, low gradient (2") generally featureless beach face, exhibits
characteristics of the high-energy dissipative beach face. On the upper beach-
face parallel, horizontal t o slightly seaward dipping laminae are arranged in
uniformly fine-grained sets in thick to very thick beds. On the lower beach
face low-frequency backwash associated with surfbeat setdown (see Wright
et al., 1982) produces a slowly seaward-migrating, hydraulic jump and leaves
antidunes ( L = 70-100 cm, H = 1.5 cm) on the surface. They result in land-
ward dipping truncation of horizontal beds and landward dipping (10") even
laminae (Fig.1la). Similar structures are described by Panin and Panin
(1967) and Reineck and Singh (1973, p. 303).
The step characteristic of lower-energy beach systems and common in
coarser-grained bar-trough systems is usually absent on the fine-grained
more dissipative beach faces. The coarser-grained lower beach face sediments
grade into the inner trough sequence. The troughs and associated rip
channels are dominated by both wave and current ripples and in gross form
follow the Narrabeen sequence. Generally coarser-grained sediments and
medium-scale, cross bedding underly seaward migrating lunate mega-
ripples, with 10-15 cm thick co-sets of cross lamination dipping both sea-
ward (2CF30") and landward (10-20") (Fig.llb). The fine-grained, rela-
tively thin (<1cm) laminae and absence of coarse-grained beds distinguishes
these trough deposits from those in the previous beach systems (i.e. Hawks
Nest, Narrabeen, Grants). The inner bar deposits followed the usual bar
pattern with steeply landward dipping (30") parallel laminae on the inner
section, and lower angle (10") laminae on the crest (Fig.llc). The outer
bar-trough was not cored. However, it would be expected to follow sequence
described by Davidson-Arnott and Greenwood (1976) and Greenwood and
Mittler (in press) with cross bedding in the trough and a mixture of cross
276
Fig.11. Seven Mile Beach box cores. Shore t o left. (a) lower beach face; ( b ) longshore
trough; ( c ) bar crest; ( d ) nearshore, depth 5 m ; (e) nearshore, depth 1 2 m ; ( f ) nearshore,
d e p t h 18 m. Scale in centimetres.
bedding or horizontal laminae on the crest, possibly with larger scale bed-
forms owing t o the more energetic wave conditions. The nearshore sequence
a t greater than 5 m depth consisted of sinuous-crested wave ripples, initially
of large scale ( L = 150 cm, H = 20 cm) which produced medium beds of
landward dipping (10-30") tangential laminae (Fig. l l d ) . Ripple size de-
creased ( L = 20-50 cm, H = 5-10 cm) between 1 2 and 21 m depth. Medium
scale cross-bedding and increasing bioturbation by Banhavia dominated
( F i g . l l e ) with few structures apparent below a depth of 20 m (Fig.llf).
Beyond the modal wave base at 25 m, large well-developed, parallel sharp-
crested, wave ripples ( L = 100 cm, H = 25 cm) were present. These forms,
produced by 2 m waves five days previously, were composed of very coarse,
shelly, poorly sorted offshore sediments, which combined with biotwbation
masked any structures.
277
DISSIPATIVE
Goolwa Beach in South Australia is a modally high energy, (Hb = 3 m,

with H , > 2 m 70% of the year), fine-grained dissipative beach, with a wide,
low gradient beach face (1.5”), and 500 m wide surfzone. The surfzone con-
sists of an inner and outer breaker zone separated by a 4 rn deep trough
region (Fig.3g). Sediment and bedform observations were made across the
region during a period of low (-1 m) waves, however cores could only be
obtained from the beach face and depths greater than 6 m. The sediment
characteristics are given in Fig.6b.
Facies (Figs.6b and 1 2 )
The 100 m wide beach face was essentially similar t o the previous beach
(Mid-Seven Mile). However the upper beach face did contain coarser sedi-
ments arranged in parallel laminations (Fig.12a). The coarseness
perhaps represents swash limit deposition of coarser particles and is
equivalent t o the coarser berm and cusp deposits of lower-energy beaches.
The lower beach face contained thin parallel-to-tangential horizontal-to-low-
angle seaward and landward dips (Fig.12b); the latter are due t o the back-
wash processes described for Seven Mile beach.
In the 4 m deep trough and over the 3 m deep bar crest wave oscillations
maintained a plane bed. Small ephemeral parallel wave ripples ( L = 5 cm,
H = 1 cm) began at 4 m depth on the seaward slope and dominated from 6
to 10 m depth. These produced slightly landward dipping thin parallel
laminations, in 20 cm sets over an erosional contact (Fig.12~).The erosion
was probably due t o 4-5 m high, 1 2 s waves two days previously. At 14 m
depth low parallel ripples ( L = 30-40 cm, H = 2-3 cm) were underlain by
similar structures and a shell rich erosion contact, (Fig.12e). Given the
previous high waves these cores (Fig.l2c, d and e) resemble the “shoreface
storm layers” with upper laminated tempesites over an erosion contact,
described by Aigner and Reineck (1982). If so the lower convex curved
laminations observed in the 10 m depth core may represent hummocky cross
stratification.
A coarsening in grain size below 18 m (due t o inner shelf lag deposits)
produced large ( L = 150 cm, H = 30 cm), sharp crested, parallel wave oscil-
lations ripples, with predominantly steeply landward dipping cross strata
(Fig.12f). Smaller active ripples ( L = 40 cm, H = 1 0 cm) with more sinuous
crests were observed at 22 and 25 m depth.
DISCUSSION
The foregoing description of beach morphology, texture, bedforms and

structures provide a ‘representative’ facies sequence for each of the six beach
states presented in Fig.1. In Fig.13 the observed facies are ranked according
to their beach type. The figure illustrates two important features of the
nature and extent of facies relative t o wave energy.
2 78
Fig.12. Goolwa beach box cores. Shore t o left. ( a ) upper beach face; (b) lower beach face;
( c ) nearshore, depth 6 m ; ( d ) nearshore, depth 1 0 m ; ( e ) nearshore, depth 14 m ; ( f )
nearshore, depth 18 m. Scale in centimetres.
First, the thickness or depth (beach t o nearshore) of each facies sequence

increases with increasing wave energy from 5 t o 1 0 m in low-energy reflec-
tive beaches, t o 10 t o 30 m in intermediate beaches and 30 m or more in
highenergy dissipative beaches (Fig.13). At the same time the horizontal
extent of the active sequence increases from 100 m t o several kilometres
(Fig. 4). On a prograding shoreline this means that a complete vertical se-
quence of a reflective beach may be preserved after 100 m of shoreline
progradation, whereas several kilometres of progradation could be required
t o produce a similar vertical sequence for higher-energy intermediate and
dissipative sequences. In other words the higher-energy sequences will be
spread over a greater horizontal distance, as indicated by Fig.4. On
stationary or regressive shorelines a diagonal sequence would at best be
preserved.
279
REFL EC TI VE ~ INTERMEDIA T f ~ ~ ~ ~
w, DISSIPA TI VE
Hb WsT < I 1 - 6 > S
Fisherman's Pearl Nest Narrabeen Grants S e v e n M l l e Goolwa
b . upper beach face

c - lower beach f a c e
,~
d - step r
e - inner trough c r o s s lamination
f - inner bar c r e s t ~
_- ~~ medium s c a l e 5 - 10 c m
9 - rip channel .~~~...
=~ large scale
~ 10 Crn
h - o u t e r trough
r - outer bar crest ,,_130_
i - inner n e a r s h o r e :.',, shells

k - o u t e r nearshore I I ' ( I bioturbarion
Fig.13. Idealised vertical sequence of all possible facies for each beach system. Bracketed
facies (f and i) have a low preservation potential. Higher-energy intermediate and dissi-
pative sequences are more likely to occur in diagonal sequences due t o massive prograda-
tion required t o produce vertical sequences.
Second, the occurrence of individual facies (a-k in Fig.13) is depen-

dent of the prerequisite morphodynamic coupling. Consequently, higher-
energy beaches will have features and structures not found on lower energy
beaches and vice versa. The lowest-energy reflective beaches consist solely
of a beach face, step and nearshore (b, c, d, k sequence, Fig.13). Higher-
energy reflective beaches have a berm crest, upper and lower beach face,
step, and deeper nearshore (a, b, c, d, j , k ) . The berm persists through the
intermediate beaches becoming wider and lower in gradient. The berm is
usually absent on dissipative beach faces which widen and have more exten-
sive upper and lower beach-face deposits. Steps at first prominent become
discontinuous longshore on rhythmic beach shorelines (absent on horns)
and absent on finer-grained, high-energy intermediate and dissipative beaches.
Mackaness (1981) used discriminant analysis of beach face textures and
structures t o statistically distinguish reflective, intermediate and dissipative
beaches. The bar-trough facies is initiated on a small scale in the low- t o
280
moderate-energy ridge and runnel beach state; it is prominent in the bar and
rip state, with the trough increasing in depth to 3 m below MLW in the bar-
trough state. Davidson-Arnott and Greenwood (1976) and Hunter et al.
(1979) suggest that in the progradation of such systems only the swash
(beach face), swash-trough transition, longshore trough (rip feeder channel),
rip channel and nearshore facies would be preserved with little preservation
of the bar sequence.
The inner nearshore regions of intermediate and dissipative beaches are
dominated by what Clifton et al. (1971) termed the ‘outer rough facies’
containing megaripples. The outer nearshore to modal wave base, and
nearshore of reflective beaches, is dominated by Clifton et al. (1972) ‘asym-
metric ripple facies’. This sequence has also been described by Shipp (1984,
this volume).
Moderate-energy intermediate beaches will therefore have an a-b-d-e-(f)-
g-j-k sequence, with higher-energy , intermediate and dissipative beaches a-c-e-
(f)-g-h-(i)-j-ksequence. The bars (f and i) have a low preservation potential,
and the nature of the outer nearshore facies (k) is highly dependent on
grain size.
The overall preservation potential of beach-nearshore systems has been
well documented in the literature (Clifton et al., 1971; McCubbin, 1982).
In southeast Australia, Thom et al. (1981) have completed extensive
augering of numerous Holocene and Pleistocene barrier systems. Using grain
size, colour and percent carbonate they have been able to discriminate be-
tween dune, beach-nearshore, and shelly nearshore (offshore) facies. The size
and extent of these systems, which included intact buried Pleistocene
barriers, suggest an overall high preservation potential. In southern and
western Australia formation of calcrete and consequent partial lithification
of the barriers increases preservation potential enabling them to survive
sealevel transgressions (Short and Hesp, in press).
The gradation in shoreface facies between low- and high-energy systems,
first proposed by Clifton et al. (1971) and elaborated by Davidson-Arnott
and Greenwood (1976) has been both confirmed and extended. This study
of seven beaches located in low, moderate and high wave environments has
provided additional information on the beach morphodynamics and as-
sociated texture, bedforms and structures. While Figs.6 and 13 illustrate the
vertical sequence of the systems, the increasing width of the higher energy
systems would dispose them to a more diagonal sequence of preservation.
The high energy dissipative Goolwa system would require several kilometres
of shoreline progradation to produce a straight vertical sequence of all
beach-nearshore-offshore facies.
The occurrence, sequence and extent of individual facies (a-k) in Fig.13
may assist identification of paleo-beach type and thereby levels of wave
energy. The arrangement of the facies sequence, vertical to diagonal, will
be an indication of degree of shoreline stability, with vertical high-energy
sequences indicative of massive shoreline progradation, and diagonal
sequences of stable and/or regressive shorelines.
281
ACKNOWLEDGEMENTS
This study was supported in part by the Australian Research Grants

Committee and Australian Marine Science and Technologies Committee. In
the field G. Lloyd was essential, assisting in all SCUBA operations dong with
the excellent assistance of P. Cowell, J.M. Short, N.L. Trenaman, and L.D.
Wright. Reviews by J.R. Dingler and B. Greenwood greatly assisted the
revision of this manuscript. Figures were drafted by J. de Roder, cores
photographed by A. Pritchard and manuscript typed by J.M. Martin.
REFERENCES
Aigner, T. and Reineck, H-E., 1982. Proximality trends in modern storm sands from the
Helgoland Bight (North Sea) and their implications for basin analysis. Senckenbergiana
Marit., 14: 183-215.
Allen, J.R.L., 1982. Sedimentary Structures: Their Character and Physical Basis. (Dev.
Sedimentol., 30A, 593 pp; 30B, 663 pp) Elsevier, Amsterdam.
Burger, J.A., Klein, G. deV. and Sanders, J.E., 1969. A field technique for making epoxy
relief-peels in sandy sediments saturated with saltwater. J. Sediment. Petrol., 39:
3 38- 34 1.
Clifton, H.E., 1976. Wave-formed sedimentary structures - a conceptual model. In:
Clifton, H.E., Hunter, R.E. and Phillips, L., 1971. Depositional structures and processes
in the non-barred high-energy nearshore. J. Sediment Petrol., 41: 651-670.
coast, Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davies, Jr. and R.L.
Ethington (Editors), Beach and Nearshore Sedimentation. SOC. Econ. Paleontol.
Davidson-Arnott, R.G.D. and Pember, G.F., 1980. Morphology and sedimentology of
multiple parallel bar systems, southern Georgian Bay, Ontario. In: S.B. McCann
(Editor), The Coastline of Canada. Geol. Surv. Can., Pap. 80-10, pp. 417-428.
Davis Jr., R.A., Fox, W.T.,Hayes, M.O. and Boothroyd, J.C., 1972. Comparison of ridge
and runnel systems in tidal and non-tidal environments. J. Sediment. Petrol., 42:
4 13-42 1.
Dupr6, W.R., 1984. Reconstruction of paleo-wave conditions from Pleistocene marine
terrace deposits, Monterey Bay, California. In: B. Greenwood and R.J. Davis, Jr.,
tary processes Kouchibouguac Bay, New Brunswick. In: J. Hails and A. Carr.(Editors),
Nearshore Sediment Dynamics and Sedimentation. Wiley-Interscience, London,
pp. 123-150.
Greenwood, B. and Mittler, P.R., in press. Vertical sequence and lateral transitions in the
facies of a barred nearshore. J. Sediment. Petrol.
Howard, J.D. and Reineck, H.-E., 1981. Depositional facies of high energy beach to off-
shore sequence: comparison with low-energy sequence. Am. Assoc. Pet. Geol., 65:
807-830.
Hunter, R.E., Clifton, H.E. and Phillips, R.L., 1979. Depositional processes, sedimen-
tary structures, and predicted vertical sequences in barred nearshore systems, southern
282
Mackaness, J., 1981. Microsedimentary structures of intertidal sand beaches. B.A. Hons
Thesis, Department of Geography, University of Sydney, Sydney, 111 pp.
McCubbin, D.G., 1982. Barrier-island and strand-flat facies. In: P.A. Schoole and
D. Spearing (Editors), Sandstone Depositional Environments. Am. Assoc. Pet. Geol.,
Tulsa, Okla., pp. 247-279.
Panin, N. and Panin, St., 1967. Regressive sand waves on the Black Sea Shore. Mar. Geol.,
5 : 221-226.
Reineck, H-E. and Singh, I.B., 1973. Depositional Sedimentary Environments. Springer,
Berlin, 439 pp.
Shipp, R.C., 1982. Nearshore depositional facies of Long Island, New York, U.S.A. In:
B. Greenwood and R.J. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in
Wave-Dominated Coastal Environments. Mar. Geol., 60: 235-259 (this volume).
Short, A.D., 1979. Three dimensional beach stage model. J. Geol., 87: 553-571.
Short, A.D. and Wright, L.D., 1981. Beach Systems of the Sydney Region. Aust. Geogr.,
1 5 : 8-16.
Short, A.D. and Hesp, P.A., in press. Coastal morphodynamics of the South East Coast of
South Australia. Coastal Studies Unit Technical Report, 8 4 / 1 , Coastal Studies Unit,
Department of Geography, University of Sydney, Sydney, N.S.W.
Short, A.D. and Wright, L.D., 1983. Physical variability of sandy beaches. In: A. McLach-
lan and H. Erasmus (Editors), Sandy Beaches as Ecosystems. Junk, The Hague, pp.
133-1 4 4 .
Short, A.D. and Wright, L.D., in press. Field methods in wave dominated surfzone and
nearshore environments. Occasional Papers, Department of Biology, Memorial Uni-
versity of Newfoundland.
Thom, B.G., Bowman, G.M., Gillispie, R., Temple, R. and Barbetti, M., 1981. Radio-
carbon dating of Holocene beach-ridge sequences in south-east Australia. Monogr. 11,
Department of Geography, University of N.S.W., R.M.C., Duntroon, A.C.T., 36 pp.
Wright, L.D. and Short, A.D., 1983. Morphodynamics of beaches and surfzones in Austra-
lia. In: P.D. Komar (Editor), Handbook of Coastal Processes and Erosion. CRC Press,
pp. 3 5 - 6 4 .
Wright, L.D., Guza, R.T. and Short, A.D., 1982. Dynamics of a high energy dissipative
surfzone. Mar. Geol., 4 5 : 41-62.
Marine Geology, 60 (1984) 283-311 28 3
Elsevier Science Publishers B.V., Amsterdam - Printed in T h e Netherlands
STRUCTURES IN DEPOSITS FROM BEACH RECOVERY, AFTER ERO-

SION BY SWELL M'AVES AROUND THE SOUTHWESTERN COAST OF
ARUBA (NETHERLANDS ANTILLES)
J.H.J. TERWINDT', C.H. HULSBERGEN* and L.H.M. KOHSIEK' *

'Department o f Physical Geography, State University o f Utrecht, P.O. Box 80.1 15,
3508 TC Utrecht (The Netherlands)
*Delft Hydraulics Laboratory, P.O. Box 152, 8300 A D Emmeloord (The Netherlands)
(Received February 21, 1983; revised and accepted July 14, 1983)
ABSTRACT
Terwindt, J.H.J., Hulsbergen, C.H. and Kohsiek, L.H.M., 1984. Structures in deposits
from beach recovery, after erosion by swell waves around the southwestern coast of
Aruba (Netherlands Antilles). In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydro-
dynamics and Sedimentation in Wave-Dominated Coastal Environments. Mar. Geol.,
60: 283-311.
Hurricane- or storm-generated swell waves may cause erosion and deposition along
coasts which are situated thousands of kilometers outside the generating wind field.
Marked beach erosion, caused by such swell waves, was observed along the micro-tidal
west coast of Aruba. During the process of erosion a swash bar was formed, which moved
up-beach during t h e waxing part of t h e swell event. The swash bar welded t o the beach
during the waning part of the event. Rapid sedimentation occurred o n the upper beach.
Finally, recovery of the beach was observed. The formation of a swash bar was attributed
to a n erosive, dissipative interval of a normally accretionary reflective beach. The sedi-
mentary structures, although generally in line with observations o n other beaches, show
several peculiar characteristics: ( 1 ) t h e great thickness of the laminae in these calcareous
sands; (2) the succession of low-angle sigmoidal and tangential sets in the swash bar; ( 3 )
the relatively steep erosional lower set boundaries and the wedge-shaped lamination
in the successive stages of beach recovery; and (4) the several types of deformation
structures.
INTRODUCTION
The destructive effect of swell waves generated by hurricanes or severe

storms which pass or hit coastal zones has been long recognized in different
parts of the world (Howard, 1939; McKee, 1959; Hayes, 1967; Stoddart,
1971; Hopley, 1974; Kumar and Sanders, 1976). It has generally been obser-
ved that in most cases the recovery of the beaches takes place rather rapidly
after the swell has ceased (King and Williams, 1949; Reineck, 1963; Hayes,
1967; Hayes and Boothroyd, 1969; Davis et al., 1972; Owens and Frobel,
*Present address: Rijkswaterstaat, Delta Service, Van Alkemadelaan 400, 2597 AT The
Hague, The Netherlands.
0025-3227/84/$03.00 o 1984 Elsevier Science Publishers B.V

284
1977). However, the effects of hurricanes and storms are not restricted to
areas in the vicinity of the track of the eye. The generated swell waves may
influence coastal processes in areas, thousands of kilometers outside of the
wind field.
Along the west coast of the island of Aruba (Netherlands Antilles), the
existence of such swell waves, having travelled a considerable distance, was
demonstrated by Wilson (1968,1969) and Wilson et al. (1973). The low, but
long waves, reaching shallow water produce high breakers, which cause havoc
and damage t o the recreational resort areas (Kohsiek et al., in prep.).
The sequence of events in the surf zone during erosion and initial recovery
of the beach is difficult to establish because profiling and diving is almost
impossible during heavy surf. Another way to analyse these events is to
study the sedimentary succession. Three large, 2 m deep trenches were
excavated across the entire beach after an erosional event. This paper des-
cribes the observed characteristics of the exposed sediments.
GENERAL SETTING O F THE AREA O F INVESTIGATION
Aruba is situated in the Caribbean Sea in the zone swept by the east
trade winds (Fig.1). The mean wind velocity is about 7.7 m s-'. There are
minor seasonal variations in wind direction and speed. The diurnal tides are
low around Aruba: the spring tidal range is 0.43 m and the neap range is
0.13 m.
The wave climate is almost exclusively dominated by the trade winds. In
the Caribbean Sea the wind waves are heading to the west for 67% of the
time, for 18%to the southwest and for 11%t o the northwest. The average
wave height is about 1.5 m and the average period is 7 s.
The calcareous sandy beaches are situated at the leeward side of the island
between Malmok and Oranjestad (Fig.2) and are unprotected by barrier
reefs. The latter are present to the southeast of Oranjestad.
The westernmost part of the island is called Manshebu and the present
study is focused on Pelican Beach between Manshebu and Oranjestad and on
Eagle beach between Manshebu and Pos Chikitu (a bluff of an old lower
terrace barrier reef deposit).
HURRICANES AND TROPICAL STORMS AFFECTING THE AREA
The effect of tropical storms (8-11 B) or hurricanes (>11B) on the

Aruban beaches differs whether the track is within or without the Caribbean
Island Arch.
Within the Arch the storms are generated in the east and track toward
the west. As a result, important variations in wind velocity and wind direc-
tion may occur and this affects the direction of the swell waves. Around
Aruba the swell starts heading to a westerly direction merging in time
toward the south and finally toward the southeast. An example is presented
in Fig. 3, illustrating the track of the Hurricane David (Aug.-Sept. 1979)
285
I- U
BAHAMAS
0 200km
P P
0
PUERTORICO p
c 2 0
St CROlX +
%
b.
%+.
.LOUPE
0 0
BARBADOS
Fig.1. Location map of the eastern part of the Caribbean.
and the successive wave fields. Since 1970, a total of 19 tropical storms or
cyclones has passed the 70%’ meridian in the vicinity of Aruba.
Aruba may also be reached by waves generated by storms or hurricanes
tracking over the Atlantic, north of the Caribbean Island Arch. These storms
move from east t o west, and if powerful enough, may generate wind waves
of sufficient height t o become swell waves. Only swell waves travelling in
the direction of the corridors, the Mona and Anegada Passage, and some-
times even the Guadeloupe and Martinique Passage, can enter the Caribbean
Sea. This means that in the Caribbean, the travelling direction is more uni-
form, while, due t o the greater distance, the wave height is smaller and the
period longer in comparison with the Inner Arch swell waves under similar
boundary conditions. In Table I some recent data are gathered for events of
286
Fig.2. Island of Aruba. Sandy beaches are situated between Malmok and Oranjestad.
Dominant wave approach is from t h e east; wave refraction takes place around the north
and south capes of the island; the meeting area of refracted waves is near Manshebu.
severe erosion along the west coast of Aruba and the inferred tracks of
tropical cyclones on the Atlantic.
In conclusion, Aruba may be reached by different types of swell waves
either being multi-directional with variable travelling distances or uni-direc-
tiond with long travelling distances.
287
TABLE I
Recent data o n erosional events along t h e west coast of Aruba and the hindcast causes
and swell wave tracks
Date of severe erosion Meteorological cause; Inferred passage of the

o n Aruban coast L = low H = high Caribbean Island Arch
Nov. 2 4 , 1977 L, Cuba t o Puerto Rico Anagada

Dec. 1, 1977 H, Atlantic 50"W, 35% Guadeloupe or Martinique
Feb. 1 6 , 1 9 7 8 L, north of Cuba Mona
Aug. 26, 1 9 7 8 L , near Cuba Mona
Nov. 1 8 , 1 9 7 8 L , northeast of Puerto Rico Mona
Feb. 2 0 , 1 9 7 9 L, east of Florida Anagada or Guadeloupe
May 2 1 , 1 9 7 9 L , north of Hispaniola Mona
THE EFFECT OF WIND AND SWELL WAVES ON T H E SAND MOVEMENT
Wind and swell waves approaching the east coast of Aruba are refracted
around the north and south capes of the island (Fig.2). These refracted
waves meet each other at the west point of the island near Manshebu, gen-
erating a complicated cross pattern (Fig.4). The location of this meeting area
and the dominancy of one pattern over the other depends on the direction
of the incoming wave trains.
Under normal wind wave conditions the meeting area of the refracted
waves is situated just west of Manshebu. There is a net longshore drift
along Pelican as well as Eagle Beach towards Manshebu (Kohsiek et al.,
in prep.) resulting in an accumulation of sediment at the west point (Fig.5).
A different refraction pattern occurs during the presence of swell waves
which originate from outside the Caribbean Arch and come from a more
northerly direction. The southward travelling waves dominate along Eagle
and Pelican Beach, resulting in a southward littoral drift (Fig.5).
If the swell waves are derived from a storm within the Caribbean Arch,
the refracted waves approach the west point dominantly from the north-
west, gradually shifting toward the west. Thus, at first, there is an increasing
littoral drift and erosion along Eagle Beach and, at Manshebu, a decreasing
drift and even sedimentation along Pelican Beach. Swell, approaching from
the west causes erosion at Manshebu but ultimate accretion along Eagle and
Pelican Beach (Fig.5).
It appears that the littoral drift during a swell event is partly opposite to
that of the normal wind wave conditions. As an example Fig. 10 compares
the net sediment movement during normal conditions with that during the
passage of the hurricane David (Aug.-Sept. 1979). The sediment discharge
estimates are based on beach profiling and measurements of wave charac-
teristics, angle of incidence and littoral currents. The breaking swell waves
created high sediment transport during a short time as compared with the
wind wave transport. Furthermore, these high breakers caused a rapid
erosion of the foreshore and backshore. The eroded sediment is partly trans-
ported offshore below the breaker zone, but a larger part is transferred to
T-
‘ 7PW bLFW
aug 29,1979,
12 00 GMT aug 30,1979,
00 00 GMT
- cb 0 . 1 f fN
‘ 1VW 60W
00 GhTr
aug 30,1979,12
~ /%-4 A* . \
.........
..........
..........
..........
...........
:j;:;::;!::!;\
...........
..........
.- a
-
'
Y
0 **,
IOUW bww ' id w OWW
aug 31,1979, 12 00 GMT sept 01,1979, 00 00 GMT
ZPN
sept. 01, 1979, 12.00 GMT sept 02,1979,OO 00 GMT
b
4 .
2pNi 4 .
0 .
? U lu
Fig.3. Track of hurricane David, August 29-September 1, 1 9 7 9 and t h e calculated wave fields, based on meteorological data computa- ~
tions (courtesy Royal Dutch Meteorological Institute). Arrows indicate direction of wave travel. CD
a
290
adjacent stretches of the coast, as is evidenced by the changes in the beach

profiles. When the swell ceases the sediments gradually return by the action
of the smaller wind waves.
DESCRIPTION O F SECTIONS
The main purpose of this study was the ultimate fate of the beach sedi-
ments and the way in which erosion changes into accretion and the restora-
tion of the beach during recovery.
Three sections, 2 m deep, were excavated crossing the entire beach. TWO
were located on Eagle Beach and one on Pelican Beach. The sections showed
almost similar features and therefore only one (A on Fig.2) from Eagle
Beach will be treated here in detail.
All sections comprise the deposits of coastal recovery after aheavy erosion
by swell waves which occurred in the beginning of April 1979. The profile
of the ultimate erosion could be established by a deposit of rocks, broken
asphalt and coral rubble which was dumped by the local authorities in order
to stop or delay the retreat of the beach. It appeared that soon after the
dumping of the rocks, etc., the beach erosion stopped, the wave climate
calmed down and the beach recovery started. The sections were made on
9-10 May 1979, about 4 weeks after the erosion. During this time the beach
recovery was 20-25 m; a mean of about 0.8 m per day.
Eleven sedimentary units can be distinguished in the section under
consideration (Fig. 6). The boundaries between the units are ero-
sional surfaces. The units are numbered according t o their inferred order in
the succession of the recovery. The section shows a landward dipping part
(unit la-le), an almost horizontal part (unit 2, 4-6) lying on top and land-
ward of unit l c and several seaward dipping units ( 3 , 5-11). A similar dis-
tinction could be made in the other sections.
In the sections from Eagle Beach and Pelican Beach, dumped rock and
pieces of asphalt originating from the erosion prevention measures are found
on top of the landward dipping unit and are incorporated in the first seaward
dipping unit, comparable t o unit 3 (Fig. 6). The horizontal and seaward dip-
ping units lie above this rubble. This indicates that unit 1 was deposited
before the dumping of the fill. As it was impossible to perform beach pro-
filing during the period of heavy swell we cannot be absolutely sure that the
landward dipping unit was deposited during the swell period and that it is
not a relict. However, we think that unit 1 fits very well in the deduced
succession of events, and the accompanying deposits.
Furthermore, a similar unit is present in all sections and in a similar setting
(approximate distance from the shore, position and height in the profile,
the character of adjacent units). Thus, although not certain, we consider
unit 1 to belong to the sequence of swell erosion and recovery of the beach.
291
Fig.4. Aerial photograph of the cross-pattern of refracted waves near Manshebu.

292
SWELL FROM OUTSIDE CARRIBEAN SWELL FROM INSIDE CARRIBEAN
6) =N 330"
H=3,5 rn
T = 8 sec.
6) =N 20" = N 300-
H=2rn. H=2rn.
T = 7 sec. T = 7 sec.
- = sedimentation
- = erosion
)= equilibrium
Fig.5. Erosion and sedimentation on the west coast of Aruba (Manshebu) in relation to
swell and wind-wave conditions.
Fig. 6 . section showing sedimentary structures acrow Eagle Beach.
pp. 296298
ID
coarse toresets sigmoidal landward laridward rillconformity unconforinity

( 1 D) sets thinner laminae thicker laminap (1B-lC) ilA-IB)
I1 B) (1 C) ( 1 8)
Fig. 7 . Dclail uf section, s h o w i n g the landward dipping u n i t 1

I>P. 299-300
deformation
structures
hor. par. lam. tVpe 2 lense lense
t t t t
Fig. 8.Detail of section, showing units 1, 2, and 4-6.

pp. 301-302
deformation
structure
type 3
Fig. 9. Detail of section, showing units l a . 3. 5. and 8-9.

Fig. 10. Sand movement along the west coast of Aruba (Manshebu). a . Estimated sand transport during the passage of Hurricane David
(in m 3 per day). b. Estimated net yearly sand transport during normal wind wave conditions (in m 3 y r - ’ ) .
304
DESCRIPTION OF THE UNITS
Unit 1
This unit is made up of low t o moderate dipping landward, cross-strata

(Figs. 7 and 8). The unit is subdivided into five subunits (Fig. 6). The boun-
daries are unconformities, either erosion surfaces or sudden and marked
changes in textures and/or structural properties (foreset angles, nature of
toesets, specific changes in thickness of the laminae).
Unit l a is coarse at the base ( d S 0 = 1.0-1.5 mm), including small coral
rubble and fines upward to mediumlcoarse sand ( d S 0 = 0.4-0.8 mm). The
basal horizontal lamination merges upward and landward into landward
dipping (15-20") foresets and toesets.
Unit l b shows a much finer sediment (0.4-0.6 mm) and the structures
consist of very thinly laminated almost horizontal lamination, in landward
direction gradually merging into low-angle (10")foresets and sigmoidal toe-
sets. The laminae become progressively thicker going landward.
Unit l c differs from l b in that the foresets are better developed, are
somewhat steeper (15"), have thicker laminae, but the toesets are concave
and the laminae become progressively thinner.
Unit I d is a continuation of unit l c , but there are differences; the sub-
horizontal topsets are very thin and there is almost no vertical accretion.
There are thick, coarse-grained (0.8-1.2 mm) foreset laminae, which
become thinner toward the concave toesets. Two of such coarse-grained
foreset beds can be distinguished, separated by a thin intercalation of finer
material consisting of concave foresets and sigmoidal fore- toesets.
Unit l e shows concave topsets merging into increasingly thicker foresets
and sigmoidal toesets, in rather coarse material (dS0= 1.2-1.5 mm).
Unit 2
This unit consists primarily of thin, horizontal or slightly landward

dipping parallel lamination in medium-sand (dS0 = 0.3-0.5 mm; Fig. 8).
Most laminae can be traced over 2-5 m and landward they become
thicker. Some very flat lenses may be observed.
Units 3-1 1
Each unit is bounded at the base by an erosional surface, truncating the

underlying strata. The units show almost similar characteristics throughout.
Going from the beach toward the sea, the units consist of almost horizontal,
parallel lamination which merge into seaward-dipping downward-thickening
foresets and sigmoidal, almost horizontal, parallel-laminated toe and
bottomsets (Fig. 9).
The increasing thickness of the laminae results in a wedge-shaped set with
gradually flattening foreset angles. This would finally lead to very low
305
dipping strata. Occasionally, however, there were erosional conditions which

resulted in erosion surfaces having a steeper slope than the underlying strata.
In most cases coarser sediment gradually fines upward from these erosion
surfaces.
The almost horizontal nature of slightly landward-dipping topsets is
nearly concordant with unit 2. However, the material is much coarser and
the laminae become progressively thicker in landward direction.
The deposition of units 3-11 resulted in an overall aggradation.
INTERPRETATION
Initially the interpretation of the sedimentary units does not seem to be

very problematic.
Many authors have provided evidence that the landwaxd-dipping stratifi-
cation like that of unit 1 may be attributed to a shoreward migrating bar
(Reineck, 1963; Panin, 1967; Psuty, 1967; Hayes and Boothroyd, 1969;
Clifton et al., 1971; Davis et al., 1972; Davis and Fox, 1972; Wunderlich,
1972; Davidson-Arnott and Greenwood, 1976; Fraser and Hester, 1977;
Owens and Frobel, 1977; Van den Berg, 1977; Hine, 1979; Hunter et al.,
1979; Greenwood and Davidson-Arnott, 1979).
Unit 2 may be interpreted as an infilling of a depression landward of the
bar. This infilling finally results in a welding of the bar to the beach, anal-
ogous to the description by Sonu and Van Beek (1971), Davis et al. (1972),
Wunderlich (1972), Sonu (1973), Owens and Frobel (1977), Van den Berg
(1977), Hine (1979), Short (1979) and Wright et al. (1979).
Units 3-11 may be referred to as the seaward outbuilding (recovery) of
the beach showing similar deposits as described by Thompson (1937), Panin
(1967), Clifton et al. (1971), Davis et al. (1972), Van den Berg (1977),
Hine (1979) and Hunter et al. (1979).
Although these general interpretations are supported by previous observa-
tions of many investigators, several peculiar points may be recognized when
looking in greater detail.
First, a decision has to be made as to whether unit 1belongs to the swell
event or is a relict of a former beach deposit. During the heavy swell the
beach was strongly eroded and the strand line shifted landward. The bottom
sets of units 3-7 were not exposed and were below the base of the section.
The same holds for unit 2. This seems to indicate that the lowermost beach
profile, which developed during the April swell event lies below the present
section. Unfortunately this lowermost profile could not be excavated
because it is situated below the ground water table. Unit 1 in this concep-
tion lies above the lowermost erosion profile. Furthermore, if unit 1 was a
relict bar then an overall truncation plane should be present on top of the
unit. There are truncations on the upper side, but not on the landward side.
Here unit 1 gradually merges into unit 2, which may indicate a genetic
relationship. Thus, we think that it is justified t o incorporate unit 1 in the
features connected with the swell event of April 1979.
306
Under normal conditions of refracted wind waves, the Aruban beaches

have an almost straight, uniform beach profile, dipping at about 6-8" with
no bars and a small 0.1 m high step at the seaward side of the swash zone
( F i g J l ) , The step is composed of very coarse sand and coral rubble. Below
the step there is a linear, low-gradient nearshore profile consisting of fine
sand. There are no runnels, rips, circulation cells, etc.; however, beach cusps
of varying dimensions were frequently observed on the beach. These charac-
teristics have some resemblance to those described by Wright et al. (1979)
for reflective beaches. I t should be noted that the Australian beaches des-
cribed by Wright et al. are pocket beaches, while the Aruba beaches occur
along an open coast. Nevertheless, the overall characteristics are quite
similar.
According to Guza and Inman (1975) strong reflection of incident waves
may occur if the reflectivity parameter e < 2.0-2.5. The reflectivity para-
meter is defined as:
e = ai w 2 / g tan'o (1)
in which ai = incident wave amplitude near the breakpoint, w = 2 n/T,
where T = wave period, g = acceleration of gravity and p = beach slope.
Under the normal conditions along the western part of the Aruban coast
(ai = 0.5-1.0 m, 0 = 8-15', T = 5-8 s), E has values below 2.5, and the
beaches may be considered reflective. However, during a swell event the
beach gradient becomes much flatter. As tan p decreases and a l increases E
will be above 2.5 despite the fact that T increases and consequently w de-
creases somewhat. Then a dissipative type of beach may occur. It is char-
acterized by the presence of bars, rip cells, etc. (Wright et al., 1979).
Thus the swell event may be considered as an erosional dissipative inter-
val in the normal accretionary, reflective beach conditions although small
erosional fluctuations may be present. This interval is of a short duration (a
few days) and although sand movement is extensive, a complete adaptation
of the beach characteristics t o the dissipative condition is hardly to be expec-
ted in such a short time. Apparently, during this dissipative internal an initial
swash bar is formed. This is postulated from the fact that unit 1 is covered
L
0- -L-
-1 -
-2 -
-3 -
-4
301
by the rocks and asphalt, dumped during the erosion of the beach. Thus
the swash bar was already present in the swash zone, when the upper part
of the beach was still eroding. The conditions during the swell event are
comparable t o a very rapid transition from beach stage 1-2 and reverse to
1 as described by Short (1979) and a sequence of beach profile types put
forward by Sonu and Van Beek (1971) and Sonu (1973).
The continuous erosion of the upper beach during the dissipative interval
resulted in a landward migration of the swash zone and hence the swash bar.
This was accomplished by erosion of the seaward flank of the bar and
deposition at the leeward side where a slight depression exists. In the depres-
sion at the leeward side of the bar the suspension outfall of the coarse
material, produced the low-angle sigmoidal foresets and bottom sets of unit
l a , as described by Thompson (1937), Psuty (1967) and Panin (1967).
In some instances, perhaps during low water the orbital velocities
over the bar decreased somewhat. Then the steeper dipping tangential fore-
sets of unit l b may be deposited. This is analogous t o many observations
on ridge and runnel beaches where steep slip faces of the bar are encountered
during rather low orbital motion when water sweeps over the bar under
limited water depth (Wunderlich, 1972; Owens and Frobel, 1977). Finally,
an equilibrium developed between the flattened beach profile and the swash
and backwash produced by interacting swell and wind waves, thus stopping
migration of the bar.
Then the period of the gradual decrease of the swell waves started. This
had a direct effect on the intensity of the swash and backwash. The mean
wave run-up decreased. Still, occasional high swash ran far up the more or
less dry beach but the backwash was reduced due to percolation. As a result,
much material moved upslope and remained in the upper part of the profile,
especially in the depression landward of the swash bar. In this way unit 2
was formed which has a similar nearly horizontal, parallel, slightly out-
wedging stratification as described by Thompson (1937), Panin (1967),
Psuty (1967) and Hine (1979), although the laminae are remarkably thick.
This suggests a high deposition rate.
As the swell waves were further reduced the relative importance of the
wind waves increased. Low, wind-wave-generated swash did not overtop the
swash bar anymore and a seaward outbuilding occurred on the sea facing
slope of the bar. However, high (wind and swell wave) swash occasionally
still swept over the bar, but on the landward side there was a very flat beach.
As a result of percolation the backwash was very ineffective and almost all
material moved forward by the swash, remained in the uprush zone produc-
ing an almost horizontal lamination and a heightening of the upper beach
(units 3, 4).
Further decrease of the swell waves resulted in a gradual transition toward
the normal steeper reflective wind-wave profiles and a seaward outbuilding
of the beach (units 5-11). However, during this outbuilding, rather im-
portant erosional events took place as evidenced by truncated lamination.
Over the erosive lower boundaries of the units much coarser material is
308
found than is in the flatter upper parts of the units, The greater coarseness
may be attributed to stronger wave action (Thompson, 1937; Hine, 1979).
Apparently, increased wave energy results in erosion, steepening and selec-
tion of coarser grains in the swash zone. Waning wave energy produces
accretion, less steep profiles and deposition of finer material in upward
pointing wedge-shaped laminae. Such type of lamination was also reported
by Thompson (1937) and Van den Berg (1977) although a conclusive
explanation is still lacking.
DEFORMATION STRUCTURES
Three types of deformation structures were observed in the section:

Type 1 in unit l e and the lower part of unit 2; type 2 in unit 3, 4 and 6,
landward of the swash bar; and type 3 in several places in the seaward dip-
ping parts of units 5-11 (Figs.8, 9, and 10).
The deformations of type 1 in unit l e and lower part of unit 2 are
typically encountered in the trough of the swash bar and in the last (upper)
deposited sediments in this trough (Fig.9). Water escape structures and
convolutions (Lowe, 1975; Reineck and Singh, 1980) were observed at
several places. A t other places the lamination was completely absent indi-
cating post-depositional rearrangement of the fabric. Noteworthy are the
dimensions of the undulations in the trough which have amplitudes of some
decimeters. These deformation structures seem to point t o conditions of
“momentary failure” due to flow, induced within the porous bed (Madsen,
1974). As indicated by Madsen (1974) and by Sleath (1970), a horizontal
pressure gradient may be initiated of sufficient magnitude under breaking
waves to create a momentary bed failure. Structural indications of this
failure are generally observed in the trough of the swash bar and not at the
crest. This may indicate, that wave induced liquefaction, as described by
Dalrymple (1979, 1980), was not the generating process of the deformation,
because this type of liquefaction works out preferentially in the crestal zone
of major bedforms. The position of the deformation all over the trough
seems t o indicate that not only horizontal but also vertical pressure gradients
play an important role.
The deformations of type 2 in units 5 and 6 (Figs. 8 and 9) consist of
water escape structures (pillars), convolutions, concavities, and structureless
sand patches. At some places the stratification is completely disturbed. No
cavernous sand is observed in the deformed layers. This excludes the forma-
tion of the convolution as a result of entrapped air as described by De Boer
(1979). Furthermore, there is almost never a kind of stratification visible in
the lows between the updomed parts of the convolute layers, a feature con-
sidered representative by De Boer for an air-trapped origin of the convolu-
tions. The deformed layers much more resemble the water escape structures
(dish structures, pillars) as described by Lowe (1975, figs.4 and 5). These
structures are attributed t o liquified behaviour of the sediment. According
t o Lowe, after resedimentation liquified sands may show: (1)undeformed
309
primary structures; (2) deformed primary structures; (3) nearly complete

homogenization; and (4)water escape structures (pillar and dish structures).
The deformed sediments of units 5 and 6 do show these features. Loose
packing may be expected in these swash-produced layers in this part of the
units 4 and 5. The great extension of these deformed layers and the rather
sharp lower and upper boundary excludes a local origin. Generating condi-
tions appear to have been operative over the whole top part of the upper
beach. After the liquefaction the normal parallel beach lamination is res-
tored. The reason for this alternation of liquified and non-liquified sedimen-
tation is not clear.
The third type of deformation structure (Figs. 8 and 9) is especially
visible at some places in units 7-11. It consists of small concavities filled
in with concave-upward stratified laminae. This structure shows a resem-
blance of hoofprints or footprints (Van der Lingen and Andrews, 1969;
Lewis and Titheridge, 1978), made by horses or human beings, walking on
an almost water-saturated beach or moist dune sands. This type 3 deforma-
tion structure also resembles those shown by Lindstrom (1979, fig.4).
CONCLUSIONS
(1)Under normal trade wind-wave conditions the calcareous-sandy micro-

tidal beach along the west point of Aruba has a reflective character with
steep beach profiles and no bars. The beach is mostly accretionary, although
with small erosional fluctuations.
(2) Occasional swell events produce strong erosion and flatter beach pro-
files of a dissipative nature. During this condition a swash bar may be formed
moving shoreward and welding t o the upper beach, showing a succession of
steeper and flatter landward-dipping tangential and sigmoidal sets.
(3) The gradual transition from dissipative to the normal reflective condi-
tions creates rapid deposition in the upper part of the swash zone with
substantial sub-horizontal parallel-laminated sets. In the lower part of the
swash zone the beach recovery occurs by a succession of several units having
a steep erosive lower boundary over which is deposited coarse material.
Upward the material becomes finer and is incorporated in upward-pointing
wedge-shaped sets, resulting in a gradual and slight decrease in the steepness
of the beach profile.
(4)As to the preservation potential in a regressive sequence we may anti-
cipate deposits resembling units 3-11 in the lower part overlain by deposits
related to the swell wave events. Such events, although of rare occurrence,
may dominate the sedimentary sequence.
(5) Three types of deformation structures could be observed in the beach
section.
ACKNOWLEDGEMENTS
This study was a part of a larger investigation requested by the

Government of the Island of Aruba. We gratefully acknowledge the approval
310
by this Government to publish these results. Furthermore, we benefitted much

from the kind and effective cooperation of the Public Service Department
(D.O.W.). Our special gratitude goes to Mr. Camminga of D.O.W. We further
acknowledge the Royal Dutch Meteorological Institute for their computa-
tion of the wave fields of Hurricane David.
REFERENCES
in the non-barred high-energy nearshore. J. Sediment. Petrol., 41 : 651-670.
Dalrymple, R.W., 1979. Wave-induced liquefaction: a modern example from the Bay of
Fundy. Sedimentology, 26: 835-844.
Dalrymple, R.W., 1980. Wave-induced liquefaction: an addendum. Sedimentology, 27,
p. 461.
coast, Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davis and R.L.
Davis, R.A. and Fox, W.T., 1972. Coastal processes and nearshore sand bars. J. Sediment.
Petrol., 42: 401-412.
Davis, R.A., Fox, W.T., Hayes, M.O. and Boothroyd, J.C., 1972. Comparison of ridge and
runnel systems in tidal and non-tidal environments. J. Sediment. Petrol., 42:
4 13-4 2 1.
De Boer, P.L., 1979. Convolute lamination in modern sands of the estuary of the Ooster-
schelde, the Netherlands, formed as a result of entrapped air. Sedimentology, 26:
283-294.
Fraser, G.S. and Hester, N.C., 1977. Sediments and sedimentary structures of a beach-
ridge complex, southwest shore of Lake Michigan. J. Sediment. Petrol., 47: 1187-
1200.
wave formed bars: a review and case study. Can. J. Earth Sci., 16: 312-332.
Guza, R.T. and Inman, D.L., 1975. Edge waves and beach cusps. J. Geophys. Res., 80
(21): 2997-3012.
Hayes, M.O., 1967. Hurricanes as geological agents, south Texas coast. Bull. Am. Assoc.
Pet. Geol., 51 : 937-956.
Hayes, M.O. and Boothroyd, J.C., 1969. Storms as modifying agents in the coastal
environment. In: M.O. Hayes (Editor), Coastal Environments, N.E. Massachusetts and
New Hampshire. Coastal Res. Group, Contrib. 1: 245-265.
Hine, A.C., 1979. Mechanisms of berm development and resulting beach growth along a
barrier spit complex. Sedimentology, 26: 333-351.
Hopley, D., 1974. Coastal changes produced by cyclone Althea in Queensland, December
1971. Aust. Geogr., 1 2 : 446-456.
Howard, A.D., 1939. Hurricane modification of the offshore bar of Long Island, New
York. Geogr. Rev., 29: 400-415.
structures and predicted vertical sequences in barred nearshore systems, southern
Oregon Coast. J. Sediment. Petrol., 49: 711-726.
King, C.A.M. and Williams, W.W., 1949. The formation and movement of sand bars by
wave action. Geogr. J., 112: 70-85.
Kumar, N . and Sanders, J.E., 1976. Characteristics of shoreface storm deposits: modern
and ancient examples. J. Sediment. Petrol., 46: 145-162.
Lewis, D.W. and Titheridge, D.G., 1978. Small scale sedimentary structures resulting
from foot impressions in dune sands. J. Sediment. Petrol., 48: 835-838.
Lindstrom, M., 1979. Storm surge turbation. Sedimentology, 26: 115-124.
311
Lowe, D.R., 197 5. Water escape structures in coarse grained sediments. Sedimentology,
22: 157-204.
Madsen, O.S., 1974. Stability of a sand bed under breaking waves. Proc. 13th. Conf.
Coastal Engineering, 2: 776-794.
McKee, E.D., 1959. Storm sediments on a Pacific atoll. J. Sediment. Petrol., 29:
354-364.
Owens, E.H. and Frobel, D.H., 1977. Ridge and runnel systems in the Magdalen Islands
Quebec. J. Sediment. Petrol., 47: 191-198.
Panin, N., 1967. Structure de d6pbts de plage sur la cbte de la mer noire. Mar. Geol.,
5: 207-219.
Psuty, N.P., 1967. The geomorphology of beach ridges in Tabasco, Mexico. Louisiana
State Univ. Studies, Coastal Studies Ser., 18, 51 pp.
Reineck, H.E., 1963. Sedimentgefiige im Bereich der sudlichen Nordsee. Abh. Sencken-
bergiana Naturforsch. Ges., 505, 1 3 8 pp.
Berlin, 549 pp.
Short, A.D., 1979. Three dimensional beach-stage model. J. Geol., 87: 653-571.
Sleath, J.F.A., 1970. Wave-induced pressures in beds of sand. J. Hydraul. Div. ASCE,
96: 367-378.
Sonu, C.J., 1973. Three dimensional beach changes. J. Geol., 81: 42-64.
Sonu, C.J. and Van Beek, J.L., 1971. Systematic beach changes on the outer banks,
North Carolina. J. Geol., 79(4): 4 1 6 4 2 5 .
Stoddart, D.R., 1971. Coral reefs and Islands and Catastrophic Storms. In: J.A. Steers
(Editor), Applied Coastal Geomorphology. MacMillan, London, pp. 155-197.
Thompson, W.O., 1937. Original structures of beaches, bars and dunes. Geol. Soc. Am.
Bull., 48: 723-752.
Van den Berg, J.H., 1977. Morphodynamic development and preservation of physical
sedimentary structures in two prograding recent ridge and runnel beaches along the
Dutch coast. Geol. Mijnbouw, 56(3): 185-202.
Van der Lingen, G.J. and Andrews, P.B., 1969. Hoof-print structures in beach sand. J.
Wilson, W.S., 1968. On the origin of certain breakers off the Island off Aruba. Techn.
Rep. Chesapeake Bay Inst. No. 43, 27 pp.
Wilson, W.S., 1969. Field measurements of swell off the Island Aruba. Techn. Rep.
Chesapeake Bay Inst. No. 56, 6 4 pp.
Wilson, W.S., Wilson, D.G. and Michael, J.A., 1973. Analysis of swell near the Island of
Aruba. J. Geophys. Res., 78(33): 7834-7844.
dynamics of reflective and dissipative beach and inshore systems: south-eastern
Wunderlich, F., 1972. Beach dynamics and beach development. Senckenbergiana Marit.,
4: 47-79.
Marine Geology, 6 0 (1984) 313-329 313
WHAT IS A WAVE-DOMINATED COAST?
RICHARD A. DAVIS, Jr. and MILES 0. HAYES

Department o f Geology, University of South Florida, Tampa, F L 33620 (U.S.A.)
Research Planning Institute, 925 Gervais S t . , Columbia, SC 29201 ( U . S . A . )
(Received February 19,1983; revised and accepted July 29, 1983)
ABSTRACT
Davis, Jr., R.A. and Hayes, M.O., 1984. What is a wave-dominated coast? In: B. Greenwood
and R. A. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated
Coastal Environments. Mar.Geol., 60: 313-329.
During the past decade or so, various coasts have been designated as wave-dominated
or tide-dominated. Typically there is an association made between coastal morphology
and the dominant process that operates on the coast in question. Most authors consider
long, smooth, barrier coasts with few inlets and poorly developed ebb deltas as “wave-
dominated”. These coasts are associated with microtidal ranges. Conversely, mesotidal
coasts tend to develop short, drumstick-shaped barriers with well-developed ebb deltas.
They are considered as tide-dominated barriers. Such generalizations may be restricted
to coasts with moderate wave energy although this is commonly not stated.
Exceptions to these stated generalizations are so numerous that wave energy and tidal
prism must also be included in characterizing coasts. The relative effects of waves and
tides are of extreme importance. It is possible to have wave-dominated coasts with vir-
tually any tidal range and it is likewise possible to have tide-dominated coasts even with
very small ranges. The overprint of tidal prism will also produce tide-dominated morph-
ology on coasts with microtidal ranges.
INTRODUCTION
The title of this volume and the symposium from which it originated
implies that there are some coasts where the physical processes t o which
they are subjected are dominated by waves. There is the implication that
some coasts are dominated by tides, the other major physical process
acting upon coastal environments. Although a continuum is present bet-
ween domination by one process as contrasted t o the other, there is a
portion within this continuum where there is subequal impact by both
waves and tides. I t should be possible therefore to identify three types of
coasts from the standpoint of their influencing processes: (1)those domi-
nated by waves; (2) those dominated by tides; and (3)those with a balance
between waves and tides.
One would expect that the dominant physical process or processes would
leave an imprint on the coastal morphology in the form of the geometry
and type of sediment bodies that accumulate in the coastal zone. It is also
0025-3227/84/$03.00 @ 1984 Elsevier Science Publishers 33.V

314
likely that the combination of coastal morphology and sediment bodies

would produce a recognizeable stratigraphic sequence in the geologic record.
The recognition of sequences as wave-dominated or tide-dominated would
be of considerable value in reconstruction of the depositional history and
paleogeography of a coastal setting.
The following discussion will attempt to answer the question posed by
the title using modern coasts as examples. I t will be apparent to the reader
that the oversimplified and overgeneralized approaches to this problem
which have been stated in the literature are unwarranted (e.g. Price, 1955;
Hayes, 1975, 1979). It is now time t o look at the numerous exceptions to
these general “rules” and see if it is possible to refine and restate them in
light of a moie comprehensive look at the variables involved.
PREVIOUS APPROACHES T O T H E PROBLEM
A recent and comprehensive review of wave-dominated coastal environ-

ments by Heward (1981) states that “wave-dominated shorelines are those
where wave action causes significant sediment transport and predominates
over the effects of tides” (Heward, 1981, p. 223). This is certainly a reason-
able definition and will be used as the basis for this discussion. It is appro-
priate t o assume that the same definition can be used for tide-dominated
shorelines with appropriate substitutions in the wording.
Although coastal morphology has been examined and classified for many
years, it was not until the general classification of coasts according to tidal
range by Davies (1964, 1980) and by Hayes (1975, 1979) that the applica-
tion of tide-dominated and wave-dominated coasts became prominent.
Davies (1964) considered three major categories of coasts; microtidal (range
< 2 m), mesotidal (2-4 m) and macrotidal (>4m). Although the boundaries
are somewhat artificial, a general morphotype can be used to characterize
each tidal range category.
Hayes (1975) applied Davies’ terminology to general types of shoreline
morphology. The microtidal coast is characterized by long, narrow and
rather straight barriers with widely spaced inlets. Washover features are
prominent and flood tidal deltas are well developed but ebb deltas are small
or nonexistent (Fig.la). Waves are the dominant physical process (Hayes,
1975, 1979).
Mesotidal barrier coasts display short, rather wide barriers with closely
spaced inlets. The barriers have a drumstick configuration (Hayes, 1975)
with one end much wider than the other due t o shoreline progradation.
Inlets have well-developed ebb deltas (Fig.lb) testifying to a relatively pro-
nounced influence by tidal currents.
Macrotidal coasts do not develop barriers due to the predominance of
strong tidal currents in a shore-normal orientation. Funnel-shaped embay-
ments may be present containing linear sand bodies oriented parallel to tidal
currents (Hayes, 1975, 1979).
Hayes (1979) has further sub-divided these three categories into five as a
315
refinement of Davies’ (1964) original classification. In this classification,

the microtidal coast has a tidal range of <1 m with the same characteristics
as mentioned above. The low-mesotidal coast has a tidal range of 1-2 m and
is characterized by increasing numbers of tidal inlets with washovers dimin-
ishing. It is a coast of mixed tidal and wave energy but with waves dominant.
The New Jersey coast is an example.
The high-mesotidal coast has tidal ranges of 2-3.5 m with abundant tidal
inlets, large ebb-tidal deltas and drumstick barriers (Hayes, 1979). Plum
Island, Massachusetts and the Georgia-South Carolina coasts serve as excel-
lent examples. This is also a mixed-energy coast but tides are dominant. The
low-macrotidal coast has tidal ranges of 3.5-5 m. Some wave-built bars
may be present but barriers are not developed. The German Eight of the
North Sea is an example of this type of coast. The last category is the
macrotidal coast which has tidal ranges of > 5 m with numerous tidal-current
formed ridges and extensive tidal flats and marshes. Bristol Bay in Alaska
and the Bay of Fundy are examples (Hayes, 1979).
This approach t o coastal classification by Hayes (1979) uses tidal ranges
which do tend to develop similar morphologies through a spectrum of wave
climates. The classification is based on a broad range of coastal areas
throughout the world but generally is restricted to those having moderate
wave energy.
SOME NOTABLE EXCEPTIONS
The above described classifications and generalizations about the interac-

tions of waves and tides with the coasts and the resulting morphology are
based on numerous examples and they make good sense. Like most generali-
zations they are somewhat simplified and there are many exceptions.
High wave energy
As stated previously, Hayes’ (1975, 1979) classification is based largely on

coasts with low to moderate wave energy. It is also intended to be applied to
trailing edge, depositional coasts which would have high preservation poten-
tial. Along such coasts there is commonly a demonstrable relationship bet-
ween tidal range and morphology. The coast of Oregon and Washington is a
leading edge coast but it displays well developed beaches and contains a
rather large sediment prism along the coast and shoreface. This area is within
the zone of highest wave energy in Davies’ (1980) classification. Tidal ranges
along this coast reach 4 m thus placing it in Hayes’ (1975, 1979) low-
macrotidal type. The morphology of the coast is however, distinctly wave-
dominated with shore-parallel spits developed across estuary mouths and
essentially no ebb tidal deltas present (Fig.2). Bays in this area which have
large tidal prisms (e.g. Willapa Bay, Washington) may develop some ebb delta
morphology (H.E. Clifton, pers. commun., 1983).
316
OCEAN \$ -’-
317
Fig.1. Diagrams of Hayes’ coastal morphology types. ( a ) Microtidal, showing long narrow
barriers with numerous washovers and few inlets; (b) mesotidal, showing short, wide
barriers with numerous inlets; and (c) macrotidal, on which barriers are absent (after
Hayes, 1979).
Another good example is in the Bay of Fundy, an area well-known for its
extreme tidal ranges. The coast along the east side of Scot’s Bay, just west of
the entrance t o the Minas Basin experiences a tidal range of -10 m. This
coast is characterized by a smooth shoreline with a well-developed cobble
beach exhibiting a large storm berm (Fig.3). This coast, although only 9 km
long, is a classic example of a wave-dominated coast yet it is in an area of the
largest tidal ranges in the world. The Scot’s Bay coast is exposed to the
long fetch of the Bay of Fundy and the storm waves that are generated by
southwesterly winds.
A similar situation exists along the northwest coast of France although
sediment supply is limited in this area Such a coast would not fall under
Hayes’ (1975, 1979) classification because it is not a trailing edge, deposi-
318
Fig.2. Oblique aerial p h o t o at Siletz Spit, Oregon. N o ebb delta is present along this wave-
dominated coast although spring tides are 4 m ( p h o t o by W.T. F o x ) .
tional coast. I t is apparent however, that in coastal regions of high wave

energy, the tidal range is not a significant factor in determining coastal
morphology.
Low wave energy
Whereas the above section considers problems associated with high wave
energy and high tidal range, it is also important t o comment on exceptions
at the other end of the spectrum. Coasts that are characterized by low wave
energy and low tidal range also display numerous exceptions t o the afore-
stated generalizations.
Among the most notable of these is the west peninsular coast of Florida
along the Gulf of Mexico. This coast is characterized by barriers for a dis-
tance of more than 250 km. The barriers give way t o mangrove swamps to
the south and to coastal salt marshes to the north. Tidal range is less than
1 m throughout the barrier coast placing it in Hayes’ (1979) microtidal cate-
gory. Wave energy is low with mean annual wave height of 30 cm (Tanner,
1960; Hayes, 1979). Although this coast is within the belt of tropical storms,
the impact of these storms is infrequent; the most recent being in 1960.
The general morphology of the west-peninsular Florida barrier system is
very similar to that proposed by Hayes (1975, 1979) for the high-mesotidal
319
Fig.3. Cobble beach with washovers o n marsh a t Scot’s Bay, Bay of Fundy. The tidal
-
range o n this wave-dominated coast is 10 m.
coast of mixed energy but with tides-dominant. There are abundant tidal
inlets, large ebb deltas (Fig.4) and drumstick-shaped barriers (Fig.5). Down-
drift offset characterizes most of the inlets; another feature common to
mesotidal coasts. Although tidal range is low in this area, the wave energy
is also very low. This coupled with rather large tidal prism permits large ebb
deltas to develop. The infrequent occurrence of extreme storms allows
these ebb deltas to persist. Their presence causes the appropriate set of
process-response conditions that gives rise t o dmmstick barriers.
The coastal areas at both ends of the west-peninsular Florida barrier
system experience somewhat higher tidal ranges (1-1.2 m to the north and
1-1.4 m to the south) and lower wave energy (Tanner, 1960). As a con-
sequence the coast assumes a morphology that is tide-dominated (Price,
1955) with extensive salt marshes or mangrove swamps, tidal current ridges
and tidal flats, although the latter two features are not well-developed on
most of the Florida coast.
Another example of tide-dominated morphology in a region of low tidal
range is present on the west side of Andros Island in the Bahamas. This area
experiences tidal ranges of slightly less than 1m yet it exhibits a morphology
that is quite similar to the German Bight coast of the North Sea (Fig.6),
an example of the typical low-macrotidal coast (Hayes, 1979). The major
reason for the situation on Andros Island is the very low wave energy
320
Fig.4. Oblique aerial photo of Redfish Pass o n the Florida Gulf coast. This large ebb
delta develops with a spring tidal range of 0.7 m.
coupled with the extensive low relief coastal zone. I t is much like the “zero-
energy” coast of Florida (Tanner, 1960) described above.
ADDITIONAL IMPORTANT VARIABLES
This discussion has only considered wave energy and tidal range as the
variables which control the morphology and configuration of the coast.
Although these parameters are obviously quite important, consideration
must also be given t o other factors which may influence a coast and which
may determine if it is or is not wave-dominated. Among these are coastal
physiography, tidal prism, availability of sediment and influence of riverine
input; more or less in order of decreasing importance. The availability of
sediment and coastal physiography are commonly related. Coasts bounded
by resistant bedrock typically limit the amount of sediment especially on
high wave energy coasts where sediment is swept out of the shoreline area.
Physiography may be a limiting factor in that high relief coasts pro-
vide little space for sediment to accumulate either above or below mean
sea level. Much of the west coast of North and South America serves as an
example. These are high relief areas and are also high wave energy coasts.
As a result sediment collects only in small pockets or reentrants along the
coast. The shelf is narrow t o non-existant so that there is no place for
321
Fig.5. Oblique aerial photo of Caladesi Island, Florida, an excellent example of a drum-
stick barrier developed with a spring tidal range of 0.8 m.
large sediment accumulations t o develop. Such coasts are not likely to accu-
mulate sequences which would be preserved in the stratigraphic record.
Coastal plain areas provide abundant sediment as well as extensive sites
for sediment accumulation. Most of the world’s barrier systems are
developed along these physiographic provinces (Glaeser, 1978). It is this
type of coast that was the basis for Hayes’ (1975, 1979) original classifica-
tion.
Tidal prism represents one of the most important but commonly over-
looked factors in determining the morphology of barrier island type of
coasts. The amount of water that passes through any inlet is determined by
the product of the tidal range and the area of the bay landward of the barrier
which is serviced by this tidal inlet. Obviously in a situation where the bay
area is unchanged, a change in tidal range will effect the prism and the
reverse will also hold true.
Generally there will be little variation in the tidal range along a particular
reach of barrier coast although there are exceptions. For example, virtually
the entire Gulf of Mexico experiences tides of less than 1 m. When changes
occur they are of a regional nature such as along the south Atlantic coast of
the United States. From the Outer Banks there is a general and rather
pronounced increase in tidal range from 1 m toward the southern South
Carolina coast where it reaches 2.5 m, then a decrease t o the south into
322
Fig.6. a. Vertical aerial p h o t o of tide-dominated coast on the west side of Andros Island,
Bahamas. Tidal range is 0.9 m (photo courtesy of E.A. Shinn).
b . General map of coastal morphology in the German Bight area along the North Sea,
a macrotidal to high-mesotidal coast (after Hayes, 1979).
323
Florida where it is again about 1 m. There are many inlets along both of
these coasts and the tidal prisms of these vary greatly both from one to
the next and also from one area t o another. The reason for this is the variety
of sizes of estuaries, lagoons and bays which the inlets service. I t is possible
therefore t o have adjacent inlets with markedly different tidal prisms or to
have areas where tidal prisms are great even though the tidal range is small.
Data of this type are shown for several inlets on the Atlantic and Gulf
coasts and for one on the Pacific Coast (Fig.7). Only those inlets where
well-documented data on tidal range and prism are included. It can be seen
from this plot that there is no relationship between tidal range and tidal
prism. Note that the highest and the lowest prism values are on the Gulf
coast where all tidal ranges are less than 1 m.
The reason that this lengthy discussion of tidal prism is important t o the
question at hand is because large tidal prisms, especially in areas of low wave
energy, can explain large, well-developed ebb tidal deltas and development of
drumstick barriers.
Riverine input t o the coast in the form of both sediment and the physical
energy of the river itself has an important impact upon the coast. It is ob-
vious that much sediment is provided by rivers and that this sediment may
be dispersed in a variety of manners and in various amounts once it reaches
the coast. Great quantities of sediment may accumulate in the form of
deltas but the morphology of these deltas covers a broad spectrum depending
c ~ G U L FOF MEXICO
A-ATLANTIC
P - PACIFIC
TIDAL RANGE (M)
Fig.7. Plot of inlets along t h e coast of the United States showing relationship between
tidal range of tidal prism. Data from N O A A tables.
324
largely on the relative roles of the river, tides and waves. Such interactions
and their resulting morphology have been classified by Galloway (1975)
and by Scott (1969). In general, riverine dominated deltaic coasts are charac-
terized by digitate lobes on the delta with the Mississippi Delta being a good
example. Wave-dominated deltas achieve an accurate or cuspate outline with
a smooth outer boundary. Well-developed beaches and beach ridges are
common. The Sao Francisco River of Brazil is a good example (Wright,
1978). Tide-dominated deltas are similar in appearance to tide-dominated
estuaries. There is generally little protrusion beyond the regional shoreline
and sediment bodies are linear, paralleling the tidal currents. The Ord River
on the northern coast of Australia is an excellent example (Wright, 1978).
DISCUSSION
The data and examples described above and the numerous examples
described in the literature do enable some useful generalizations to be made
about the morphology of wave-dominated coasts and these in turn permit
related generalizations about stratigraphic sequences which accumulate on
wave-dominated coasts. The latter can serve as valuable data sets in recon-
structing paleoenvironmental conditions in the coastal regime.
Modern waue-dominated coasts
There can be no disagreement with the basic premise that rather straight
and smooth coasts, characterized by well-developed beaches are the result
of physical conditions dominated by waves and by wave-generated currents.
If barriers are present, they are typically long and smoothly accurate or
straight. Storm generated washovers may be present but are not a required
characteristic. Pocket beaches along bedrock coasts also fall into the wave-
dominated category.
The above characteristics are similar to those described by Hayes (1975,
1979). The only important deviation from Hayes’ classification is the associa-
tion of a particular tidal range (e.g. microtidal) with the wave-dominated
coast. Although this is a common association, there is no need to relate tidal
range to coastal morphotypes. The important relationship is that wave-
energy overwhelms tidal energy and in so doing, a characteristic morphology
is produced.
The relationships between tidal processes and wave-generated processes
have been shown by Hayes (1979) who presented a generalized diagram
based on many geographic areas (Fig.8). The five fields presented range from
tide-dominated to wave-dominated. An approximate limit of barrier island
development would be within the field labeled tide-dominated (low). Notice
that this field, as do all, covers a spectmm of tidal ranges and wave heights.
It is the relative effects of these processes which are important, not the
absolute values.
It is also important to be aware of the rather delicate balance between
325
0 100 200
MEAN WAVE H E I G H T (ern)
Fig.8. General relationships between tidal range and wave height as it relates to coastal
morphology. A particular coastal region may span several fields (modified after Hayes,
1979).
tide and wave processes as lower and lower values are approached. All of
the five fields converge at the low end of the spectrum (Fig.8). The conse-
quences of these relationships are that tide-dominated, wave-dominated
or mixed energy morphologies may develop with very little difference in
tide and wave parameters. Such relationships further emphasize the import-
ance of excluding absolute tidal range or wave height values from coastal
morphotypes.
The shoreline in wave-dominated environments is characterized by elon-
gate shore-parallel sediment bodies. These include longshore bars, beaches,
beach ridges, foredunes and even outer lobes of ebb-tidal deltas which are
wave-formed. Such features may be present along all types of coasts includ-
ing river deltas but are generally best developed along coastal plain shore
zones. They may be present on prograding coasts or on transgressive coasts.
The only sediment body type which commonly occurs on wave-
dominated coasts that is not shore-parallel is washover or blowover features.
These lobate or fan-shaped features are storm generated phenomena which
may be developed on all coastal types but are most common on wave-
dominated coasts.
Sediment which accumulates on wave-dominated coasts generally shows
moderate t o very good sorting due to continual reworking by waves but it
may be of virtually any grain size. Sand is by far the most common size as
it is in shore zones overall. Mud does occur, such as along the coast of
326
Surinam (Wells and Coleman, 1981) and on chenier coasts. Gravel and cobble
sediment is also common especially along many wave-dominated beaches in
moderate to high latitudes where glacial deposits are reworked along the
coasts and in some low latitudes where coral reef debris is abundant.
To summarize theref ore, modern wave-dominated coasts are characterized
by shore-parallel sediment bodies which may span the entire grain size spec-
trum. Tidal range itself is not a criterion in determining tidal- or wave-
dominance.
Stratigraphic sequences from wave-dominated coasts
A fundamental reason for considering the distinctions between wave-

dominated and tide-dominated coasts is the ability of making these distinc-
tions in the stratigraphic record. In order to consider this problem it is
necessary to consider both the individual stratigraphic column such as might
be seen in a core and also the geometry and interrelationships between
lith osomes.
Transgressive, wave-dominated sequences typically display a coarsening
upward texture. Considering a barrier island sequence, there would be a
sequence of estuarine, marsh or tidal flat, washover and beach and nearshore
deposits from bottom to top (Fig.9). Such a sequence could be resting com-
formably on almost any lithosome type. Many Holocene barrier systems on
the present Atlantic and Gulf coasts of the United States exhibit such
sequences (e.g. Fisk, 1959; Kraft, 1971).
A progradational (regressive) sequence of a wave-dominated system would
also display a generally coarsening upward texture with beach and dune
ridge lithosomes overlying nearshore and shoreface units. Thin units of
muddy and organic-rich sediment representing swales or cat’s eye ponds
would be expected t o be incorporated into the well-sorted, cross-stratified
sands of the prograding beach ridge lithosome (e.g. Barwis, 1976).
By way of contrast, the prograding tide-dominated coast would be expec-
ted to accumulate a fining-upward or at least a shoaling-upward sequence
with sandy subtidal shoal deposits, intertidal shoals, tidal flats and marsh
deposits from bottom to top. Good examples are those from the Wash in
England (Evans, 1965, 1975) and the Bay of Fundy (Knight and Dalrymple,
1975). A transgressive tide-dominated system would display a stratigraphic
sequence that is similar t o that of the transgressive wave-dominated system.
I t would contain a generally coarsening-upward textural trend with salt
marsh, tidal flat, and tidal sand bodies from the bottom to top. Subtle dif-
ferences would include nature and orientation of cross-stratification, exten-
sive reactivation surfaces in the tide-dominated sequence and the contained
fauna.
The most demonstrable method of distinguishing between these two types
of coastal sequences is by considering the overall geometry and orientation
of the sand bodies. The wave-dominated system displays shore-parallel sand
lithosomes whereas the tide-dominated system contains shore-normal sand
327
HOLOCENE TRANSGRESSIVE SEQUENCE

BAY MOUTH BARRIER AREA
DUNE F - M , WELL SORTED

EOLIAN CROSS BEDDING
BEACH- BERM M - V C . PEBBLY

WASHOVER LOW ANGLE LAMINATION
BACK BARRIER MARSH PEAT -CLAYEY SAND
- -
TIDAL DELTA M-PEBBLY. POORLY SORTED
ABUNDANT CROSS BEDDING
LAGOONAL SAND ..
a-.
RARE SILT STREAKS
0 .
- -
nu
LAGOON -- SOFT DARK GRAY CLAY- SILT
MARSH - PEAT
-
h
-
-
YL
-
PLEISTOCENE
COASTAL
ENVIRONMENTS
4
0-
.o
. TAN, ORANGE, GREENISH-GRAY
*a
Fig.9. Stratigraphic sequence of Holocene transgressive barrier system along the

Delaware coast (after Kraft, 1971).
lithosomes. Various directional and orientational structures such as cross-

strata, reactivation surfaces, parting lineation, scour marks and other indi-
cators of current direction must be used t o confirm these distinctions.
SUMMARY
A wave-dominated coast is simply a coast which is subjected t o physical

processes that are dominated by wave energy. This dominance is strictly
relative; it is not based on any absolute wave parameters. The same is true
for tide-dominated coasts. To relate the dominant process to a particular
tidal range or wave parameter is incorrect.
Modern coastal morphology confirms this by the wide range of wave and
tidal conditions which produce similar appearing coastal configurations.
Stratigraphic sequences in the ancient record may be recognized as tide- or
wave-dominated but the association with specific levels of tidal range is at
328
best, difficult. Paleotidal range determinations have been discussed by

Klein (1971; 1972) but such determinations are rarely attempted from the
stratigraphic record. Paleo-wave climate has been investigated by Komar
(1974) and Allen (1981) by relating bedform characteristics to wave para-
meters. Dupr6 (1984, this volume) has interpreted Pleistocene wave climate
on the basis of sediment and bedform features.
REFERENCES
Allen, P.A., 1981. Wave-dominated structures in the Devonian lacustrine sediments of

Barwis, J.H., 1976. Internal geometry of Kiawah Island beach ridges. In: M.O. Hayes and
T. W. Kana (Editors), Terrigenous Clastic Depositional Environments. Univ. South
Carolina, Coastal Res. Div., Columbia, S.C., pp.11-115-11-125.
Davies, J.L., 1964. A morphogenic approach t o world shorelines. Z. Geomorphol., 8:
27-4 2.
Davies, J.L., 1980. Geographical Variation in Coastal Development (2nd ed.). Longman,
Duprb, W.R., 1984. Reconstruction of paleo-wave conditions from Pleistocene marine
terrace deposits, Monterrey Bay, California. In: B. Greenwood and R.A. Davis, Jr.
Evans, G., 1965. Intertidal flat sediments and their environments of deposition in the
Wash. Q. J. Geol. SOC.London, 121: 209-245.
Evans, G., 1975. Intertidal flat deposits of the Wash, western margin of the North Sea.
In: R.N. Ginsburg (Editor), Tidal Deposits. Springer, New York, N.Y., pp.13-20.
Fisk, H.N., 1959. Padre Island and the Laguna Madre flats, coastal south Texas. Natl.
Acad. Sci., Natl. Res. Council, 2nd Geogr. Conf., pp.103--151.
Galloway, W.E., 1975. Process framework for describing the morphologic and strati-
graphic evolution of the deltaic depositional systems. In: M.L. Broussard (Editor),
Deltas, Models for Exploration. Houston Geol. SOC.,Houston, Texas, pp.87-98.
Glaeser, J.D., 1978. Global distribution of barrier islands in terms of tectonic setting. J.
Geol., 86: 283-297.
Hayes, M.O., 1975. Morphology of sand accumulations in estuaries: an introduction to
the symposium. In: L.E. Cronin (Editor), Estuarine Research, vol. 2. Academic Press,
Hayes, M.O., 1979. Barrier island morphology as a function of tidal and wave regime. In:
S.P. Leatherman (Editor), Barrier Islands. Academic Press, New York, N.Y., pp.1-27.
Heward, A.P., 1981. A review of wave-dominated clastic shoreline deposits. Earth Sci.
Rev., 1 7 : 223-276.
Klein, G. deV., 1971. A sedimentary model for determining paleotidal range. Geol. SOC.
Am. Bull., 82: 2585-2592.
Klein, G. deV., 1972. Determination of paleotidal range in clastic sedimentary rocks.
24th Inti. Geol. Congr. Proc., Sec. 6 : 397-405.
Knight, R.J. and Dalrymple, R.W., 1975. Intertidal sediments from the south shore of
Cobequid Bay, Bay of Fundy, Nova Scotia, Canada. In R.N. Ginsburg (Editors), Tidal
Deposits. Springer, New York, N.Y., pp.47-55.
Komar, P.D., 1974. Oscillatory ripple marks and the evaluation of ancient wave conditions
and environments. J. Sediment Petrol., 44: 169-180.
Kraft, J.C., 1971. Sedimentary facies patterns and geologic history of a Holocene marine
transgression. Geol. SOC.Am. Bull., 8 2 : 2131-2158.
Price, W.A., 1955. Development of shoreline and coasts. Dept. Oceanography, Texas
A & M Univ., Project 63, 9 pp.
329
Scott, A.J., 1969. Figure 7, p. 82. In: W.L. Fisher, L.F. Brown Jr., A.J. Scott and J.H.
McGowen, 1969. Delta Systems in the Exploration for Oil and Gas - A Research
Colloquium. Univ. Texas, Bur. Econ. Geol., Austin, Texas, 102 pp.
Tanner, W.F., 1960. Florida coastal classification. Trans. Gulf Coast Assoc. Geol. SOC.,
10: 259-266.
Wells, J.T. and Coleman, J.M., 1981. Physical processes and fine-grained sediment
dynamics, coast of Surinam, South America. J. Sediment Petrol., 51: 1053-1068.
Wright, L.D., 1978. River deltas. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary En-
vironments. Springer, New York, N.Y., pp.5-68.
Marine Geology, 60 (1984) 331-354 331
SHOREFACE MORPHODYNAMICS ON WAVE-DOMINATED COASTS
ALAN WILLIAM NIEDORODA' , DONALD J.P. SWIFT', THOMAS SAWYER HOPKINS3

and CHEN-MEAN MA4
'R.J. Brown and Assoc., Suite 200, 2010 North Loop West, Houston, TX 77018 ( U . S . A . )
'Exploration and Production Research, PGR-G130, A R C 0 Oil and Gas Company, P.O.
Box 281 9, Dallas, TX 75221 (U.S.A.)
30ceanographic Sciences Division, Department of Energy and Environment, Brookhaven
National Laboratory, Upton, N Y 11973 (U.S.A.)
'Woodward-Clyde-Oceaneering,7330 Westview Drive, Houston, T X 77055 ( U . S . A . )
(Received January 17, 1983; revised and accepted July 29, 1983)
ABSTRACT
Niedoroda, A.W., Swift, D.J.P., Hopkins, T.S.and Chen-Mean Ma, 1984. Shoreface mor-
phodynamics on wave-dominated coasts. In: B. Greenwood and R.A. Davis, Jr.
ments. Mar. Geol., 60: 331-354.
An open ocean shoreface typical of long, wave-dominated sandy coasts has been
examined through a combination of extensive field measurements of wave and current
patterns with computations of marine bedload transport and sedimentation. Sand
transport on the upper shoreface is dominantly controlled by waves with only secondary
transport by currents. Sand on the middle and lower shoreface, as well as the inner
continental shelf is entrained by storm waves and transported by a complex pattern of
bottom boundary layer currents.
Storm events have been studied and modeled for the shoreface off Tiana Beach, Long
Island. The dominant effect of coastal frontal storms is to cause significant shore-parallel
bedload transport with important shore-normal secondary components. These storms
tend t o result in net offshore transport of sand removed from the beach and surf zone
systems. The bedload transport during a storm is convergent on the shoreface leading to
accretion. Most accretion occurs on the upper shoreface with lesser deposits covering the
middle and lower shoreface as well as the inner continental shelf. Longer-term equilib-
rium can be maintained by slow return of sand up the shoreface during non-storm con-
ditions.
Annual and geologic time-scale budgets of shoreface sand transport and sedimentation
yield equilibrium, net accretion or net deposition. The annual balance results from an
integration of the event-scale bedload transport patterns and morphologic responses.
These processes and responses have feedback mechanisms which stabilize the system over
longer, but not geologic, time scales. Geologic time scale balances are controlled by rela-
tive sea level changes and relative availability of sediment supply with the event-scale
shoreface sand transporting processes providing the mechanism t o produce the changes
in long-term morphology and sedimentation patterns. In the area of study, the long-term
pattern is one of net shoreface erosion, and the permanent loss of sand t o the shelf floor.

332
INTRODUCTION
A major morphological feature of many continuous, sandy, wave-

dominated coasts is the concave-upward bottom slope lying between the
outer reaches of the surf zone and the inner continental shelf. This feature
is called the shoreface. Although the recognition of the prevalence of this
feature was demonstrated in the early portion of the 20th Century, there
have been remarkedly few studies conducted to understand the nature of
the sediment transport processes which occur in this zone and, in fact,
result in controlling its morphology. This zone is difficult to reach with
instruments supported from the beach and is generally considered too
close to the shore for classic oceanographic examinations. In addition, the
waves, currents, and sediment transport patterns in this zone are charac-
terized by strong spatial and temporal gradients. Unsteady flow conditions
make study of the region difficult without exhaustive theoretical or field
measurement programs.
In treating the overall morphodynamics of the shoreface, three time
scales must be considered. These are the event scale (order of 2-7 days),
the annual scale (order of a year), and the geologic time scale (order of
centuries or longer). The bulk of the following discussion will be directed at
event scale processes as these must be understood before longer time scale
effects can be defined. Only long, straight sandy coasts are considered.
Alterations in shoreface processes related to tidal inlets, estuary mouths, or
rocky headlands are not considered. Much of the experimental data presented
resulted from a series of field measurements off Tiana Beach, Long Island,
N.Y. These data are assumed t o be reasonably representative of typical
shoreface wave, current, and sediment transport conditions because the
shoreface in this area has a relatively simple shape and constant exposure to
open-ocean conditions.
DEFINITION O F THE SHOREFACE
From the days of classical geomorphology, the shoreface has been recog-
nized as a significant zone along much of the world’s coastline. Early
discussion of this zone can be found in Fenneman (1902), Barrel1 (1912),
and Johnson (1919). On long sandy wave-dominated coasts the shoreface is
the region lying immediately seaward of the surf zone and terminating at the
inner continental shelf. In general, it has a concave-upward shape with the
steepest slope near its top (Swift, 1976). The landward portion has steep-
nesses on the order of 1:20. This slope diminishes seaward until the shore-
face merges with the inner continental shelf where the slope is on the order
of 1:2000. The depths of the upper and lower boundaries of the shoreface
are variable depending upon local sediment supply and on wave and current
conditions. On the long sandy coasts of the mid-Atlantic and New York
bights, the upper shoreface joins the surf zone at a depth of approximately
4 m and the lower shoreface joins the inner continental shelf at a depth of
approximately 25 m.
333
Many authors have concluded that the shape and morphology of the
shoreface is an equilibrium response of an unconsolidated coast to the
typical local wave and current regime (Johnson, 1919; Bruun, 1962; Moody,
1964; Swift, 1976; Niedoroda and Swift, 1981). However, quantitative
relationships between process and response in this zone have been generally
lacking.
Limited information concerning the nature of dominant shoreface proces-
ses has resulted from geological interpretation of the character and distribu-
tion of shoreface sediments. Relatively coarse sand in the surf zone grades to
fine sand on the upper portion of the shoreface and is abruptly replaced by
coarser sand at the middle and base of the shoreface. Cook (1970), Langford-
Smith and Thom (1969), and Cook and Gorsline (1972) have concluded
that this pattern can be partially explained by the winnowing of fine sand
from deposits within the surf zone and deposition of this fine sand by rip
currents which diffuse over the upper shoreface. Variations in the slope and
curvature of shorefaces in widely separate regions have been qualitatively
attributed to differential sorting of sand deposits through the action of wave-
induced currents and to differences in the amount and type of sediment
available to local shoreface processes (Wells, 1967; Wright and Coleman,
1972). More recent work by Niedoroda (1980) and Niedoroda and Swift
( 1 9 8 l j has outlined many of the basic sediment dynamic processes on the
shoreface.
DYNAMIC ZONES O F THE COASTAL OCEAN
Niedoroda (1980) and Niedoroda and Swift (1981) have shown that the
marine bedload transport and morphodynamics of the shoreface are pri-
marily controlled by the net local wave and current environment. The dy-
namics controlling these two phenomena are independent of each other as
are their associated length scales, neither of which is necessarily coincident
with the width of the shoreface. All length scales are, however, ultimately
related to the mean bottom slope, the sediment supply, and the amount of
energy available to the system. Wave processes dominate the upper reaches
of the shoreface while current processes dominate the middle and lower
reaches of the shoreface.
For typical open ocean wave heights and periods non-linear wave behavior
becomes important in defining near bottom wave orbital kinematics in depth
ranges from about 8 to 18 m (approximately 1.5 km offshore). Thus, non-
linear wave behavior characterizes the upper portion of the shoreface zone.
The waves typically begin to break at depths of 3-5 m (approximately
0.5 km offshore). Under open-ocean storm conditions breaking waves may
extend offshore t o a depth of approximately 1 0 m. Thus, the inner edge of
the shoreface is occasionally occupied by the outer reaches of the surf zone.
The shore-normal or diabathic length scale for coastal ocean currents is
considerably wider than that for nonlinear behaviour of coastal ocean waves.
The coastal boundary layer represents the region of the ocean where the
334
presence of shoaling depths and the shoreline strongly affect the dynamics
of currents. Coastal ocean currents result from tidal forcing, wind forcing,
and horizontal pressure gradients related to slopes of the mean sea surface
and the isobars of the internal field of mass.
Non-tidal coastal ocean currents develop as a result of a complex relation-
ship between surface wind stresses, diabathic slopes of the mean sea surface
and/or the pycnocline, as well as flow in the upper and lower frictional
boundary layers. In general, the significant baroclinic diabathic length scale
is on the order of 5-10 km and the corresponding barotropic length scale
is on the order of 25-50 km for typical mid-Atlantic summer conditions
(Csanady, 1978; Winant and Beardsley, 1979; Hopkins, 1982; Hopkins and
Dieterle, 1983).
The characteristic time scales of coastal ocean waves and currents serve
to define the durations for which forcing mechanisms must remain nearly
constant to yield coherent and steady flow conditions. The steady coastal
ocean currents tend t o occur in distinguishable patterns, whereas unsteady
or transient currents tend t o develop in quite variable patterns. Waves occur
in a relatively narrow range of frequencies and adjust t o the wind within dura-
tions of hours t o one day. Coastal ocean currents have a wide range of
characteristic time scales which depend on the relative magnitude of the
forcing mechanisms and the nature of the local water column stratification.
Surface wind stresses are a principal forcing mechanism for coastal cur-
rents and their temporal dependency. Over deep water the wind must blow
with a nearly constant speed and direction for more than the inertial period
for steady currents t o result. Near coastal boundaries, wind transport causes
sea-level distortion and accompanying barotropic flow. Once initiated these
flows can remain coupled to the wind forcing, provided no sudden changes
in the wind occur. The greater the change the more uncoupled the flow
becomes until the lag is again at the inertial time scale (Hopkins, 1974). At
mid-latitudes this time scale falls between the semidiurnal and diurnal
frequencies (about 19 h). Over shallower water columns of the shoreface,
the response to wind forcing (including the barotrophic effects) is reduced
by the effects of increased turbulence and shallower depths. If the coastal
ocean is stratified, energy is diverted t o the baroclinic mode which opposes
and has longer time scales than the barotrophic mode. However, the baro-
clinic mode is reduced over the shoreface because stratification there is often
weak due to wave mixing. Well-organized coastal flows which are coupled to
the surface wind stresses require at least a day of nearly steady conditions to
develop. This often complicates the procedure of defining typical flow pat-
terns from measured data.
CURRENT PATTERNS IN THE COASTAL OCEAN
The purpose of this section is t o define a series of typical current patterns

over an open ocean shoreface. In subsequent subsections this information
will be used in combination with other measured wave and current data to
compute patterns of shoreface bedload transport.
335
The principal source of data concerning characteristic coastal ocean

currents over open ocean shorefaces is from a series of field experiments
conducted during 1976, 1977 and 1978. A complete description of these
experiments is given in Niedoroda (1980). Therefore, only a brief description
is provided here.
The field experiments were part of two interrelated projects. These were
project INSTEP (conducted jointly by the Coastal Research Center of the
University of Massachusetts and AOML/NOAA) and project COBOLT
(conducted jointly by Brookhaven National Laboratory and Woods Hole
Oceanographic Institute). A diabathic transect was established offshore of
Tiana Beach, Long Island (see Fig.1). Boat stations were established at 1km
intervals over a 1 2 km line across the coastal boundary layer in 1976. In
1977, these boat stations were established along a 3.5 km line across the
shoreface at approximately 500 m intervals. Nearly synoptic measurements
of currents, temperature, and salinity were made at 1and 2 m depth intervals
at these boat stations. The locations of these sampling points are given in
Figs.2 and 3. During 1976, two Shelton Spars were located on this transect
at 3 and 9 km offshore (see Fig.1). Each spar supported four 2-axes electro-
magnetic current meters and eight temperature and conductivity probes. A
Fig.1. Location map of the Tiana Beach, Long Island shoreface experiments (Projects
INSTEP and COBOLT). Solid line shows location of the boat station transect during
1976 and 1977. Dots are locations of the Project COBOLT Shelton Spars during 1976.
336
D I S T A N C E OFFSHORE (m)
0 500 1000 1500 2000 2500 30W 3sw 4000

I ' " ' I " ' ~ I ~ ' ~ ' J , I ~ , ' ~~1~~ I , I I I / 1 1 1 1 ,
Fig.2. Location of daily current meter, conductivity and temperature measurements

along the 1976 Tiana Beach Shoreface Transect.
DISTANCE OFFSHORE l m )
0 500 1000 I500 2000 2500 3000 35w 4000

r r ' I , 1 I ~ I r I I 7 I , , 1 - r
8 I
Fig.3. Location of daily current meter, conductivity and temperature measurements

along the 1977 Tiana Beach Shoreface Transect.
complete description of the project COBOLT Shelton Spars is given in

Dimmler et al. (1976), Scott et al. (1978) and Tucker (1974a, b, 1975). A
complete set of the data resulting from the boat station measurements is
given in a report by Niedoroda (1980).
This relatively comprehensive set of oceanographic data from a typical
open ocean shoreface has been processed to show characteristic flow pat-
terns. Measurement of the currents at the boat stations required a total of
approximately two hours. Therefore the data are nearly synoptic. In order
t o compare data taken on different times and days it is necessary to remove
the relatively large tidal components.
Time series of the currents measured at the 3 and 9 km Shelton Spars of
the COBOLT project were processed using rotary spectra and complex
demodulation (Mooers, 1973). The principal semidiurnal tidal current ampli-
tudes were determined. These are plotted on Fig.4. The regression lines
shown on Fig.4 indicate that the distribution of tidal current velocities can
331
100 1 I
I
I . / / .ool 10 /
10
a
/ .
t
I / /
crn/sec cm/sec
lo(
C
I
--
E
-E
-I
N N
/ 0.1
0.01
u 2 4 6 8 10 1214
crn/sec
Fig.4. Average semidiurnal tidal current components for various depths at the 3 and
9 km Project COBOLT Shelton Spars. Approximate semilogarithmic relationships are
shown.
be approximated by a semilogarithmic function whose equation is given on

the figure. Using these empirically derived semilogarithmic functions for the
nearshore tidal components a series of curves defining the maximum tidal
amplitude at each depth for each boat station of both the 1976 and 1977
measurements was determined. This is shown in Fig.5.
Data from Fig.5 were used to generate a series of masks representing esti-
mated tidal current components at the time that each data set was collected.
Maximum tidal amplitude was adjusted according t o a cosine function whose
argument was the phase of the semidiurnal tide. These values were then
subtracted from the measured current values. Representative results are
shown in Figs.6 and 7.
The occurrence of coastal upwelling or downwelling has long been recog-
nized but few data are available on the characteristic velocities and actual
338
0.01
10071
-
0 I
crn/sec
2 3
-i0t0.
0
0
1 '
0
2
7 '
3
'
4 5 6
crn/sec
--77-
7
-
1
8
,
9
I
1 0
,
0.01 -
0 I
crn/sec
2 3
O IO 2 l3 4 0 5' 6~
'
crn/sec
Fig. 5. Average tidal component curves used to generate synoptic tidal component masks.
These values were corrected for the instantaneous tidal phase via a cosine function.
r
' a' 9' 10
'
flow patterns in shoreface regions. Figure 6 shows a typical flow pattern off
Tiana Beach under a moderate southwest wind. Inside of the 1 2 m isobath
the shore-parallel current component dominates. The surface Ekman trans-
port, which has an offshore component, causes a horizontal divergence of
the upper layer over the shoreface. The divergence results in an upwelling of
bottom waters over the shoreface. This flow situation is represented diagram-
atically in the upper panel of Fig.8. The opposite situation is given by Fig.7
which shows flow over the Tiana Beach shoreface resulting from a mod-
erately strong easterly wind. A horizontal convergence in the upper layer
over the shoreface is caused by onshore surface Ekman transport. These
data are shown diagramatically in the lower panel of Fig.8.
339
DISTANCE, m e t e r s DISTANCE, m e t e r s
0. 500. . .EO?.. ! 5 e _ o 2 q 0 ! ~ _ 2 5 0 0 , . ? ~ ! E O O L I O . ?5?_0__?~!J?. ~.?5??. . !9oP. . !5??
~oo.~._’oo?.
N e a r H i g h T i d e (0.4)
Compooen t V D i s p l a y T i dv a l S u p p r e s s e d
component
Display Untreated
Wind 4 m/s S o u t h w e s t (245O) Wind 4 m / s S o u t h w e s t (245O)
DISTANCE, m e t e r s OISTANCE, m e t e r s
Y ... 500 --F!? 1 L?5!?-10!??L. ?ioo300_o_L?51~~ 0,. . . 2?OL,
.!?OP .25?0 . . 3900.
. _ ‘500_~.~?~00, .3??
STN, I S T N J STN 3 STN 4 STNA5 STN 6 STN I S T N 2 S T N 3 S T N , ~STN 5 STN 6 STN 7
12;
14:30 h r s 1 4 : 3 0 hrs
165
Near High T i d e N e a r H i g h T i d e (0.4)
component u Component U
D i 5 ; Iay U n t r e a t e d O i s p l a y T i d a l Suppressed
Wind 4 m/s S o u t h w e s t ( 2 4 5 0 ) Wind 4 m/- S o u t h w e s t ( 2 4 5 0 )
Fig.6. Shore parallel (parabathic, V ) and shore normal (diabathic, U ) components of

shoreface currents on 6 Aug. 1977. Contours are in cm/s. Dashed contours and positive
V-values show eastward parabathic flow. Dashed contours and positive U-values show
onshore flow. The left-hand panels are raw data and the right-hand panels show flow with
the tidal components removed. An eastward/upwelling flow event is shown.
DISTANCE, m e t e r s ‘DISTANCE, meters
0 . 50° 1000 1500 2000

~. I ~
2500
I _
3000
.~ 3500
. ~ . ~ .
0~ 500 1000 1500
-
2000
-.
2500
, -
. . 3000 . 3500 ~
STN I STN 2 STN 3 STN 3
Display Untreated Display T i d a l Suppressed

W i n d 5.5 m/s E a s t W i n d 5.5 m/s E a s t
DISTANCE, m e t e r s DISTANCE, m e t e r s
0
i
500
~I_~ ~
i
1000
_
1500 2000
.‘
2500
-, ~-
3000 3500
l
0 _
500
.~I_
1000
i..
1500
i
2000
. ~
2500 3000
. _. 3500
STN 1 STN 2 STN 3 STN 1 STN 2 STN 3
Display Untreated Display T i d a l Suppressed

W i n d 5.5 m / s E a s t W i n d 5.5 m/s E a s t
Fig.7. Longshore (parabathic, V ) and shore normal (diabathic, U ) components of shore-

face currents o n 2 4 August 1976. Contours are in cm/s. Solid contours and negative V-
values show westward parabathic flow. Solid contours and negative U-values show off-
shore flow. The left-hand panels are raw data and the right-hand panels show flow with
the tidal components removed. A downwelling westward flowing coastal jet is shown.
340
N - - - ? S
COAST W
PA M L L E L SURFACE
, STREAMING DIVERGENCE, M I D - S H E L F FLOW
I
I
- -
COAST
PARALLEL SURFACE
-- k -
- - c
- yc
Fig.8. Diagrams of two of the important classes of coastal ocean flows over the shoreface.
The upper panel shows an eastward upwelling event driven by a southwest wind along an
east-west coast. The lower panel shows a westward downwelling event driven by an east
wind along an east-west coast.
Similar characteristic flow patterns (not shown) for various wind stresses
were assembled from the original data. These patterns show representative
flows for westward downwelling, westward upwelling, eastward down-
welling, and eastward upwelling flow over the Tiana Beach shoreface. These
data were subsequently used to model flow and resulting marine bedload
transport on the shoreface during storm events.
341
SHOREFACE BEDLOAD SEDIMENT TRANSPORT
The subject of shoreface marine bedload transport has been discussed by

Niedoroda and Swift (1981). It was shown that during non-storm conditions
the sand residing on most of the shoreface was not transported even under
the combined effects of strong mean- and tidal-currents. Only at the upper
portions of the shoreface is there a tendency for sand to be driven upslope
by the asymmetrical near bottom wave orbital velocities. During storm
events, however, the larger waves cause large near-bottom orbital velocities
which entrain sand over the entire shoreface. This sand is then transported
both parallel and normal to the shore by the currents. Transport extends
across the shoreface onto the inner continental shelf. In this paper we ex-
tend the results of our earlier work to show typical shoreface bedload
transport patterns and morphological responses due to coastal storms.
Event scale bedload transport
The relative rates and patterns of shoreface sediment transport were

examined by combining data on waves and currents during storms to esti-
mate the spatial distribution of both high- and low-frequency fluid shear
stresses acting on the sea floor and computing the corresponding bedload.
Wave and current measurements from a profiling concentration velocity
probe (hereafter PCV-probe) were combined with current data from the
boat stations located across the shoreface t o produce a synthetic time series
of wave and current conditions across the shoreface during the storms.
The PCV-probe consists of an electromagnetic two-axes current meter
supported 1.1 m above the sea floor. This probe also has a pressure-type
wave gauge and a suspended sediment concentration transducer (Huff and
Friske, 1980; Young et al., 1982). This instrument was deployed at a 10 m
depth on the Tiana Beach shoreface during the fall of 1977 through the
spring of 1978.
Figure 9 shows a record of data measured by the PCV-probe during a
coastal storm which occurred between the 26th and the 29th of March
1978. As the storm developed, the waves built rapidly to a maximum during
the afternoon and evening of March 27 (Julian Day 86). As the storm center
crossed the area, the winds backed sharply and blew offshore. As a result,
the wave heights diminished rapidly during March 28 (Julian Day 87).
This storm was a classic Northeaster. The initial winds came from the east
to southeast and developed the heavy seas. During the first phase of the
storm the coastal flow was directed strongly to the west with the surface
current convergence over the shoreface producing a downwelling. The wind
veered slowly to the south and southwest before backing to the northwest
as the storm center passed. As the wind suddenly began to blow offshore,
the wave heights were rapidly reduced while the currents reversed t o an
easterly direction with first a downwelling and then an upwelling com-
ponent. These phases of the storm are noted on Fig.9.
34%
VOLUME FLUX (crn/cm.s)

I 2 3 4 5
0\\
10&
10 /
10/10 10
Fig. 9. Upper portion - Time series of total wave energy (upper panel), and current com-
ponent (middle panels) as measured by the NOAA PCV-probe at a depth of 1 0 m (current
meter 1.1 m above the sea floor) on the Tiana Beach shoreface during the storm of
26 March (day 8 5 ) t o 29 March (day 88) 1978. The symbols W/D, SW/D, W / U , E / D and
E / U mean westward/downwelling and eastward/upwelling, respectively. The letters a
through Q represent 6-h time intervals used to display shoreface bedload transport in
the lower panel of this figure.
Lower portion - The lower panel shows computed bedload transport at 500 m inter-
vals across the shoreface. Each subfigure ( a through q ) shows bedload fluxes for a par-
ticular time interval as defined by the corresponding letter's position in the time series of
the upper panel. The bedload flux vectors are on a logarithmic scale. The short horizontal
lines beneath the letters represent the shoreline. The vertical lines show distances off-
shore.
343
The data from the storm of 26-29 March 1978 shown on Fig.9 were
used in conjunction with the previously described catalog of characteristic
shoreface flow patterns to estimate bedload transport over the shoreface
during the event. The tidal oscillations were subjectively removed by
smoothing the shore-parallel and shore-normal current records. The smooth
lines shown in the middle and lower panel of Fig.9 represent the smooth
flow patterns used. To compute shoreface bedload transport the tidal
oscillations were added to the corresponding characteristic flow patterns to
yield estimates of the near bottom current over the duration of the storm.
The single-point measurement of current direction and intensity was thus
expanded to estimate bottom boundary layer flow over the whole shore-
face using the characteristic flow patterns for each of the different stages of
the storm (e.g., westward downwelling flow, westward upwelling flow, etc.).
A computer program based on the method for computing marine bedload
sediment transport due to the combined effects of waves and currents
(Madsen and Grant, 1976) was utilized. This method applies the Einstein-
Brown bedload transport relationship to marine conditions through the use
of a combined wave-current friction factor (Jonsson, 1966) and a modified
Shields parameter. Points at 500 m intervals across the Tiana Beach shore-
face were represented in the program. Wave and near-bottom current condi-
tions at each of these points were estimated from measured data at two-hour
time intervals over the duration of 26-29 March, 1978, storm. The time-aver-
aged bedload transport for each point and time interval was computed. An
“instantaneous” bedload volume flux was computed for time sub-intervals
equal t o one sixteenth the wave period. The sum of these values normalized
by the wave period and multiplied by the time interval was used t o repre-
sent the bedload transport of each two-hour interval. This method and
computer program has recently been calibrated with field data (Niedoroda
et al., 1982) and shown to yield calculated results within 20 percent of
measured transport values. The relative magnitudes of sediment transports
are less sensitive to calculation inaccuracies and are more important in
understanding the response of shoreface sediments to storm events.
Bedload transport was calculated for each two-hour interval of this
storm at seven locations at 500 m intervals across the shoreface.
Transport rates and directions for each of these points across the shore-
face were summed for 17 intervals (labeled a through q on Fig.9) over the
duration of the storm to compute the net transport during the storm.
The lower panel of Fig.9 shows vectors of the bedload sediment flux
whose magnitude is related to a logarithmitic scale given on the figure. In
each part of this figure (labeled a through q) the short horizontal line repre-
sents the shoreline. The vertical line represents a transect extending 3500 m
offshore. A t each 500 m interval, a bedload sediment transport flux vector
is shown indicating the magnitude and direction of the bedload transport
at each location during that period of the storm. The succession of figures
shows the time series of this transport.
These results indicate that during the early portion of the storm, sediment
344
was driven westward and offshore. As the storm intensified and the waves
became larger, more sediment was entrained and transported westward and
offshore. As the storm center passed through the area (between time period
k and time period l), the direction of shore parallel transport reversed to the
east but diabathic (shore-normal) transport remained offshore. Only during
the last portions of the storm did an upwelling flow develop. However, by
this time the offshore winds had significantly reduced the wave heights
resulting in only a minor pulse of upslope transport in the very final stages
of the storm.
Figure 10 shows the result of summing the diabathic component of shore-
face bedload transport during the entire storm. A t all points, the net volume
flux is offshore. The greatest volume flux occurs at the calculation point
highest on the shoreface (500 m). The net volume flux reduced sharply to a
distance of approximately 1500 m offshore and then fell off less rapidly to
the last calculation point near the base of the shoreface. The second curve
shown on this figure represents the difference in sediment transport between
each adjacent calculation point. Because of the convergence of the offshore
transport there is a deposition of sediment on the shoreface. The greatest
deposition occurs at the upper shoreface. The relative accretion of the
shoreface during the storm decreased over a distance of approximately
1700 m offshore. Seaward of this distance, the rate of deposition is approxi-
mately constant. Extrapolation beyond the last calculation point suggests
that this rate of deposition would continue out onto the inner continental
shelf.
DISTANCE OFFSHORE (m)

Fig.lO. The net bedload sediment flux (solid line) and net differential transport
or deposition on the shoreface (dashed line) as computed for the 26 through 29 March
1978 northeastern storm off Tiana Beach, Long Island. Note that the largest offshore
sand transport and largest spatial gradients of offshore sand transport (hence deposition)
occurred on the upper shoreface. However, offshore sand transport and deposition ex-
tends across the entire shoreface to the inner continental shelf.
345
The data shown on Fig.10 indicate that sediment removed from the surf
zone by northeast storms is deposited across the shoreface with the majority
coming to rest on the upper shoreface. Transport and deposition of this
sediment extends across the entire shoreface and onto the adjoining inner
continental shelf. That sediment which arrived at the lower shoreface or
inner portion of the continental shelf may, or may not, be returned to the
beach system. The sediment which was deposited during the storm on the
upper shoreface constitutes a reservoir which is then depleted by the slow
return of sand up the upper portions of the shoreface and into the surf zone
and beach during periods of non-storm conditions.
An obvious question is whether the one storm studied is indicative of
shoreface sediment transport and morphological response during other
coastal storms. This point is addressed with the following considerations. If
the storm follows an offshore track relative to say the New Jersey coastline,
then the first winds characterizing this storm will develop and strengthen
from the northeast. As the storm center migrates past the coastal point the
winds rapidly back and blow offshore from the northwest. With progressive
movement of the storm along the offshore track these northwesterly winds
diminish in intensity. If the storm follows an onshore track the winds will
develop from the southeast and strengthen as the storm center moves ad-
jacent t o the coastal point. As the storm center progresses past the coastal
point of interest the winds rapidly veer to the southwest and diminish in
intensity. These conditions are illustrated in Fig.11.
In either case, the occurrence of the low-pressure storm causes the initial
winds to have an onshore component leading t o the rapid generation of
heavy seas. The passage of the storm center results in a rapid shift of the
wind so that it has an offshore component which substantially reduces the
wave heights.
Figure 12 shows an idealization of this pattern. The upper stick diagram
in this figure indicates the time history of winds during the passage of a
northeastern storm for a typical location on the Long Island shoreline.
The second panel indicates rapid growth in local wave height during the early
phases of the storm. This corresponds t o the generation of an organized wind-
driven circulation over the shoreface. The onshore and southward surface
currents are characterized by a strong horizontal convergence over the shore-
face. This results in a downwelling. Combination of high waves and down-
welling current results in offshore sediment transport along the shoreface.
As the storm center passes and the winds back rapidly to the northwest, the
surface currents change and are characterized by a horizontal divergence
over the shoreface. This results in upwelling. However, the offshore winds
rapidly decrease the local wave heights so that the return of sediment up the
shoreface during the latter portion of the storm does not compensate for
the offshore transport in the early phase of the storm.
Figure 13 illustrates a similar pattern for a low-pressure storm following
an inland track. In this case the shore-parallel currents and resulting sedi-
ment transport during the initial downwelling stage are directed in the same
346
I,,
-1 D A Y I
Fig.11. Typical coastal storm wind patterns for the U.S. Northeast coast. The upper
left three panels show typical fronts and wind streamlines of an evolving low pressure
storm. Upper right panel shows a typical storm superimposed on a map of the eastern
U.S. Typical storm tracks are shown beneath. An inland track ( B ) and an offshore track
( A ) have been identified. The two wind stick diagrams show typical time histories of the
storm winds (track A, upper and track B, lower) for a particular point on the northeast
coast as the storm moves past. The beginning of the storm appears to the left and the end
appears to the right in the stick plots of the winds.
direction but are weaker than that of the previously discussed case. The
waves during the first stage of the storm will be larger than in the previous
example because of a more direct exposure of the coast to the wind. The
pattern of convergence of surface currents over the shoreface with the
resulting of downwelling during the initial period of the storm when the
waves are high, followed by a divergence of surface currents over the shore-
face and upwelling when the waves are low, is maintained.
347
TRACK A STORM
NORTHEASTER
Fig.12. Typical winds, wave heights and coastal ocean (shoreface) currents for a north-
eastern storm.
Although the above examples of storm induced shoreface currents show

only two of the complex set of strongly forced shoreface circulation pat-
terns, they do demonstrate a general effect. Storm winds with strong
onshore components cause high locally generated waves, convergence of
surface currents and downwelling over the shoreface. Most frontal storms
produce a rapid shift of the wind as the front or storm center propagates
past the coastal site. These later winds have offshore components which
rapidly decrease the height of the waves and cause surface current diver-
gence and upwelling over the shoreface. A hysteresis develops in the dia-
bathic bedload transport over the shoreface resulting in net offshore trans-
port during storms. More sediment is moved offshore in the first part of the
storm, when downwelling and high waves prevail, then is moved onshore
when lower waves and upwelling prevail. The processes causing the bedload
transport vary significantly over the shoreface. Most typically, this trans-
port decreases offshore leading t o a bedload convergence and accretion on
the shoreface.
Geographic generalization of this pattern of storm induced diabathic
shoreface bedload transport hysteresis is possible. Along the northern Gulf
Coast frontal storms occur in winter. They are called “Blue Northers” and
348
TRACK B STORM
SOUTHEASTER
23- W A V E HT
SECTION VIEW SECTION VIEW
Fig.13. Typical winds, wave heights and coastal ocean (shoreface) currents for a frontal
storm following track B o n Fig. 11.
result from cold fronts which sweep across the coast and propagate eastward.
As the front (which normally extends at a high angle or perpendicular to the
coastline) approaches a given coastal site the winds freshen from the south
or southeast. As the wind continues to increase they tend to swing to the
southwest and then veer suddenly t o the northwest as the front passes. The
southeast winds begin to build the waves while causing surface current con-
vergence and downwelling over the shoreface. As the wind strengthens and
veers to the south the waves increase in height while the convergence in
surface currents over the shoreface amplifies the downwelling. If the wind
continues to strengthen and veer to the southwest the shoreface shore-
parallel flow may reverse and the zone of surface current convergence will
move closer t o shore. Downwelling may be confined to the mid- and upper-
shoreface with a second upwelling circulation developing further offshore.
As the front passes the wind veers rapidly to the northwest. This strong
offshore wind quickly reduces the wave heights while causing divergence of
the surface currents and upwelling over the shoreface. The dominant near-
bottom current over the shoreface is offshore when the waves are high and
onshore when the waves are diminishing in height.
A somewhat similar pattern is exhibited for storms incident on U S .
Pacific northwest coast (i.e., north of Cape Mendocino). As winter low
pressure storms move eastward out of the Gulf of Alaska into Canada, the
349
early winds peak early out of the south or southwest and then veer and
become weaker northwesterlies as the low moves inland. Although the
winds remain onshore through the cycle the initial downwelling forcing is
stronger than the final upwelling forcing and thus the sense of the net shore-
face bedloah ‘manspohbysbmiis.>spreserued.
Many effects have not yet been considered. The prevalence of swell on the
Pacific coast must reduce the contrast in shoreface bedload transport between
storm and non-storm conditions. Hurricanes and tropical storms have not been
considered. The relatively small size and rapid propagation of these storms
can cause complex coastal currents which have marked longshore gradients
and may not be well coupled with the local winds.
Annual and geologic scale shoreface processes
The preceeding discussion of event-scale shoreface processes has shown

that the upper shoreface forms a reservoir for sand removed from the beach
and surf zone during coastal storms. I t has been shown in an earlier analysis
(Niedoroda, 1980; Niedoroda and Swift, 1981), that fair-weather wave
processes in the area of study will tend to return this material to the surf
zone and beach. Thus, this study, which extends further seaward than have
most studies of the coastal sediment budgets, leads to a somewhat more
complete and spatially comprehensive view of the annual cycle of
coastal sedimentation. Many earlier studies have described a cycle in
which sand tends to be stored in breakpoint bars during the stormy winter
months, but returns to the beach prism during the quiescent summer
months, as the bars migrate landward and weld to the berm. In this study,
the bar-berm cycle is seen to be part of a larger-scale cycle which extends
down the shoreface.
The study also carries important implications for coastal processes at
geologic time scales, since it demonstrates that a significant amount of
sand may be transported entirely across the shoreface, and on to the
adjacent inner shelf floor. Companion studies indicate that this sand is not
necessarily lost; the asymmetrical orbital currents of shoaling waves cease to
be effective in moving sand landward at about 15 m water depth (Niedoroda,
1980; Niedoroda and Swift, 1981), but wave-current interactions continue
to move sand landward on the lower shoreface below 15 m, and on the
adjacent inner shelf (Vincent et al., 19833. However, there is clearly the
potential for permanent loss, if downwelling storm currents transport sand
out onto the inner shelf faster than wave-current interactions can return it.
In fact, two basic coastal sand budgets can be defined in terms of the three
basic parameters of coastal evolution; the rate of sediment supply, the rate
of relative sea level change, and the rate of fluid power expenditure (Fig.14).
In an erosional regime, such as that of the Atlantic shelf of North America,
sea level is rising sufficiently rapidly with respect to the rate of river sedi-
ment input that the river mouths have become estuaries. They trap not only
350
PROGRADATIONAL EROSIONAL
SHOREFACE SHOREFACE
REGIME REGIME
n BEACH
RIVER - BEACH
RIVER
A? I
SHELF
L E G E N D
OFFSHORE STORM
DOWNWELLING TRANSPORl
SOURCE
0 SINK
4
I
I
ONSHORE WAVE ORBITAL
TRANSPORT
RESERVOIR
U WAVE-DRIVEN ALONGSHORE
TRANSPORT
Fig. 14. Schematic representation of general coastal sand budgets in prograding and
eroding shoreface regimes.
all fluvial sand, but also littoral drift (Meade, 1969). Storm downwelling
currents must move sand seaward faster than wave orbital currents can
return it, because the coast is retreating almost everywhere, and cores of
most shorefaces reveal a thin veneer of modern sand (several decimeters
thick) over older back barrier strata (Swift et al., 1985). In this kind of
coastal system, estuaries and the inner shelf floor become the ultimate sinks
for sand. The primary source is the eroding shoreface. Storms may strip off
the entire beach prism and back-barrier. Tree stumps, clays, and peats
become briefly exposed along hundreds of kilometers of beach before fair
weather waves return the sand (Harrison and Wagner, 1964). This denuda-
tion extends down the entire shoreface, during major events (Charlesworth,
1968), with ancient back-barrier deposits exposed at the sea floor to 10 or
15 m water depth. The Holocene sand sheet on the Atlantic shelf surface
appears to have been generated in this fashion by shoreface erosion, as the
shoreface retreated back across the shelf surface in response to post glacial
sea-level rise (Swift, 1976; Swift et al., 1985).
351
In a second coastal system described by Curray et al. (1969), the rate

of river sand input overwhelms the effects of sea-level rise, and river mouths
become deltas which inject sand directly into the surf zone. The beach, bar
and shoreface all receive sand more rapidly than they can exchange it with
adjacent environments. During neap tides, bars are captured as new berms
and the coast progrades seaward as a series of beach ridges, t o form a strand
plain.
CONCLUSIONS
The data, analyses, and results presented in this paper lead to the fol-
lowing conclusions.
The shoreface region of wave-dominated sandy coasts is a zone of active
bedload sediment transport. The processes which cause and control this
sediment transport are distinct from those which control sediment transport
in the surf zone. The rate of sediment transport on the shoreface is strongly
controlled by the height of waves which serve t o entrain the bottom sedi-
ments. Sediment on the upper shoreface is entrained by relatively small
waves and generally shifted landward during non-storm conditions as a result
of the asymmetrical wave orbital fluid stresses caused by non-linear near-
shore waves. This landward shift of sediment results in erosion of the shore-
face sand reservoir. The pattern of sediment transport on the shoreface
during storm events is controlled by the fluid shear stresses and transport of
water in the bottom boundary layer of the coastal ocean.
Frontal storms produce a relatively predictable pattern of waves, currents,
and bedload sediment transport on wave-dominated open-ocean shorefaces.
At the event-scale, the result of coastal storms tends t o be an offshore
transport of sediment over the entire shoreface. The offshore transport of
bedload during these storm events is convergent on the shoreface yielding
accretion. It has marked spatial gradients such that the majority of the
deposition is concentrated on the upper shoreface. A small percentage of
the sediment can be transported offshore, across the shoreface and deposited
on the inner continental shelf. Thus, the upper shoreface forms a reservoir
for sand removed from the surf zone during storm events. This reservoir
provides a source of the sand which generally returns to the surf zone and
beach during non-storm intervals.
A shoreface bedload transport hysteresis commonly results during storms.
On most of the U.S. coasts the initial stages of storms generally yield high
locally generated waves and bottom currents with an offshore component
due to downwelling over the shoreface. Strong offshore winds rapidly
reduce the wave heights and cause onshore bottom currents due to upwelling
over shoreface. The net bedload transport is offshore because the lower
waves associated with upwelling cannot transport as much bedload onshore
as was moved offshore by downwelling flow.
Annual scale sediment budgets and morphological balances result from the
integration of storm- and non-storm shoreface bedload transport patterns.
Such annual averaging tends to yield imbalances related to the average
352
number and intensity of storms during the year. Relatively sudden shifts in
shoreface position can result but its shape and morphology is preserved.
Over longer time spans the sediment transport processes on the shorefaces
provide a major avenue for the removal of sediment from coastal areas and
its consequent deposition at the base of the shoreface. Shorefaces can be
either erosional or depositional features depending upon the local supply of
coastal sediments from longshore processes and slow changes in relative sea
level.
ACKNOWLEDGEMENTS
Research which resulted in this publication was in part funded by con-

tracts from the National Oceanic and Atmospheric Administration
(Contract COM/NOAA 03-6-022-3511), the Office of Naval Research
(Contract N00014-76-C-0145), and the Department of Energy (DE-AC-2-
76CH00016) as well as a subcontract to Brookhaven National Laboratories
from NOAA/AMOL. Additional funding came from the University of
Massachusetts, Woodward-Clyde Consultants, and Dames & Moore Consul-
tants. Portions of the data were supplied by the Sediments Dynamics Group
at the Atlantic Oceanographic and Meteorologic Laboratories, NOAA,
Miami, from the INSTEP program which was funded by the Northeast
Office, Office of Marine Pollution Assessment, NOAA.
Many people assisted in this research. Particular appreciation is owed to
Mr. David Battisti for assistance with the field work and in the analysis of
tidal currents. Additional thanks is given for field assistance to H. Peg,
S. Wall, L. Wall, J. Collins, J. Lenard, G. Bowers, Dr. B. Brennickmeyer,
E. Divit and D. Carlson. Assistance in logistical support, data collection,
and data exchange obtained from Southampton College, Boston College,
and Woods Hole Oceanographic Institute is also acknowledged.
REFERENCES
Barrell, J., 1912. Criteria for the recognition of ancient delta deposits. Geol. SOC.Am.
Bull., 23: 377-446.
Bruun, P., 1962. Sea-level rise as a cause of shore erosion. J. Waterways Harbors Div.,
A.S.C.E., 88: 117-130.
Charlesworth, L.J., 1968. Bay, inlet and nearshore marine sedimentation: Beach Haven-
Little Egg inlet region, New Jersey. Doct. Diss., Univ. Michigan, Dept. Geol., 23 pp.
Cook, D.O., 1970. Occurrence and geologic work of rip-currents in Southern California.
Mar. Geol., 9: 173-186.
Cook, D.O. and Gorsline, D.S., 1972. Field observations of sand transport by shoaling
waves. Mar. Geol., 1 3 : 31-55.
Csanady, G.T., 1978. The arrested topographic wave. J. Phys. Oceanogr., 8: 47-62.
Curray, J.R., Emmel, F.J. and Crampton, P.J.S., 1969. Lagunas costercas, in simposio.
In: Mem. Simp. Int. Lagunas Costercas, UNAM-UNESCO, Nov. 28-30,1967, Mexico,
pp.6 3-1 00.
Dimmler, D.G., Rankowitz, S., Huszagh, D.W. and Scott, J.T., 1976. A controllable real-
time data collection system for coastal oceanography. Proc. Oceans '76, 256: 1-13.
Fenneman, N.M., 1902. Development of the profile of equilibrium of the subaqueous
shore terrace. J. Geol., 10: 9-32.
353
Harrison, W. and Wagner, K.A., 1964. Beach changes at Virginia Beach, Virginia Beach,
Virginia. U.S. Army Coastal Eng. Res. Center, Misc. Pap. No. 6-64, 25 pp.
Hopkins, T.S., 1974. On time dependent wind induced motions. Rapp. P.V. RBun. Cons.
Int. Explor. Mer, 1 6 : 721-736.
Hopkins, T.S., 1982. On the sea level forcing of the Mid-Atlantic Bight. J. Geophys.
Res., 87: 1997-2006.
Hopkins, T.S. and Dieterle, D.A., 1983. An Externally Forced Barotropic Circulation
Model for the New York Bight. Continental Shelf Res., 2: 49-73.
Huff, L.C. and Friske, D.A., 1980. Development of two sediment transport instrument
systems. Proc. 17th Int. Conf. Coastal Engineering, Sydney, N.S.W., pp.176--177.
Johnson, D., 1919. Shore Processes and Shoreline Development. Hafner, New York, N.Y.,
584 pp.
Jonsson, I.G., 1966. Wave boundary layers and friction factors. Proc. 10th Conf. Coastal
Engineering, ASCE, 1: 127-148.
Langford-Smith, T. and Thom, B.G., 1969. New South Wales coastal morphology. J.
Geol. SOC.Aust., 16: 572-580.
Madsen, O.S. and Grant, W.D., 1976. Sediment transport in the coastal environment.
Ralph M. Parsons Laboratory Rept. No. 209, 105 pp.
Meade, R.H., 1969. Landward transport of bottom sediments in estuaries of the Atlantic
Coastal Plain. J. Sediment Petrol., 34: 144-221.
Moody, D.W., 1964. Coastal morphology and processes in relation t o the development
of submarine sand ridges off Bethany Beach, Delaware. Thesis, Johns Hopkins Univ.,
167 pp. (unpublished).
Mooers, C.N.K., 1973. A technique for the cross spectrum analysis of pairs of complex-
valued time series, with emphasis on properties of polarized components and rota-
tional invariants. Deep-sea Res., 2 0 : 1129-1141.
Niedoroda, A.W., 1980. Shoreface-surf zone sediment exchange processes and shoreface
dynamics. NOAA Techn. Memo. OMPA-1, 8 9 pp.
Niedoroda, A.W. and Swift, D.J.P., 1981. Maintenance of the shoreface by wave orbital
currents and mean flow: observations from the Long Island coast. Geophys. Res.
Lett., 8: 337-340.
Niedoroda, A.W., Mean-Ma, C., Mangarella, P.A., Cross, R., Huntsman, S. and
Treadwell, D., 1982. Measured and computed coastal ocean bedload transport. Proc.
18th Int. Conf. Coastal Engineering. Am. SOC.Civ. Eng., 2: 1353-1368.
Scott, J.T., Hopkins, T.S., Pillsbury, R.D. and Divis, E.G., 1978. Velocity and tempera-
ture off Shinnecock Long Island in October-November 1976. Brookhaven Natl. Lab.
Rep. No. 50895, 3 5 pp.
Swift, D.J.P., 1976. Coastal sedimentation. In: D.J. Stanley and D.J.P. Swift (Editors),
Marine Sediment Transport and Environmental Management. Wiley, New York, N.Y.,
pp.255-310.
Swift, D.J.P., Niedoroda, A.W., Vincent, C.E. and Hopkins, T.S., 1985. Barrier island
evolution, middle Atlantic shelf, U.S.A. Part 1: Shoreface dynamics. In: G.F. Oertel
and S.P. Leatherman (Editors), Barrier Islands. Mar. Geol., 6 3 (in press).
Tucker, W.D., 1974a. Coastal shelf oceanography program. Progress Rept. No. 1,Brook-
haven Natl. Lab., 49 pp.
Tucker, W.D., 1974b. Coastal shelf oceanography program. Progress Rept. No. 2, Brook-
Tucker, W.D., 1975. Coastal shelf oceanography program. Progress Rept. No. 3, Brook-
Vincent, C.E., Young, R.A. and Swift, D.J.P., 1983. Sediment transport in the Long
Island shoreface, North American Atlantic Shelf: Role of waves and currents in
shoreface maintenance. Continental Shelf Res., 2: 163-181.
Wells, D.R., 1967. Beach equilibrium and second order wave theory. J. Geophys. Res.,
7 2 : 497-509.
Winant, C.D. and Beardsley, R.C., 1979. A comparison of some shallow wind-driven
currents. 3. Phys. Oceanogr., 9 : 218-220.
354
Wright, L.D. and Coleman, J.M., 1972. River delta morphology: wave climate and the
role of the subaqueous profile. Science, 176: 282-284.
Young, R.A., Merrill, J.T., Clark, T.L. and Proni, J.R., 1982. Acoustic profiling of sus-
pended sediments in the marine bottom boundary layer. Geophys. Res. Lett., 9 :
17 5-17 8.
Marine Geology, 60 (1984) 355-376 355
CONTROL OF BARRIER ISLAND SHAPE BY INLET SEDIMENT

BYPASSING: EAST FRISIAN ISLANDS, WEST GERMANY
DUNCAN M. FITZGERALD', SHEA PENLANDa and DAG NUMMEDAL3

'Department of Geology, Boston University, Boston, MA 0221 5 (U . S . A. )
'Louisiana Geological Survey, Louisiana State University, Baton Rouge, L A 70803
(U.S. A.)
)Department of Geology, Louisiana State University, Baton Rouge, L A 70803 (U.S.A.)
(Received January 17, 1983; revised and accepted September 11, 1983)
ABSTRACT
FitzGerald, D.M., Penland, S. and Nummedal, D., 1984. Control of barrier island shape
by inlet sediment bypassing: East Frisian Islands, West Germany. In: B. Greenwood
A study of the East Frisian Islands has shown that the plan form of these islands
can be explained by processes of inlet sediment bypassing. This island chain is located on
a high wave energy, high tide range shoreline where the average deep-water significant
wave height exceeds 1.0 m and the spring tidal range varies from 2.7 m at Juist to 2.9 m
at Wangerooge. An abundant sediment supply and a strong eastward component of wave
power (4.4 x lo3 W m - l ) have caused a persistent eastward growth of the barrier islands.
The eastward extension of the barriers has been accommodated more by inlet narrowing,
than by inlet migration.
It is estimated from morphological evidence that a minimum of 2.7 X lo5 m3 of sand
is delivered t o the inlets each year via the easterly longshore transport system. Much of
this sand ultimately bypasses the inlets in the form of large, migrating swash bars. The
location where the swash bars attach to the beach is controlled by the amount of overlap
of the ebb-tidal delta along the downdrift inlet shoreline. The configuration of the ebb-
tidal delta, in turn, is a function of inlet size and position of the main ebb channel. The
swash bar welding process has caused preferential beach nourishment and historical
shoreline progradation. Along the East Frisian Islands this process has produced barrier
islands with humpbacked, bulbous updrift and bulbous downdrift shapes. The model
of barrier island development presented in this paper not only explains well the con-
figuration of the German barriers but also the morphology of barriers along many other
mixed energy coasts.
INTRODUCTION
The morphology of barrier islands is influenced by a number of factors

including mode of formation, sediment supply, sea-level rise, and wave and
tidal processes. When the supply of sediment is sufficient to maintain the
barrier system the primary control of its general morphology is the hydro-
0025-3227/84/$03.00 o 1984 Elsevier Science Publishers B.V.

356
graphic regime of the region (Hayes, 1975). Hayes, using the tidal classifica-
tion of Davies (1964), showed that moderate wave energy barrier island
coasts (wave height = 60-150 cm) can be separated into two basic types:
microtidal and mesotidal coasts. In his microtidal model ( T R < 2.0 m),
barrier islands are long and continuous and have numerous washovers. Tidal
inlets occur infrequently and contain well-developed flood-tidal deltas but
poorly developed ebb-tidal deltas. Meso-tidal coasts (2.0 < TR < 4.0 m) have
short, stubby barrier islands and numerous tidal inlets. Tidal deltas are well
formed with ebb deltas more prominently developed. This classification
was later redefined by Hayes (1979) and Nummedal and Fischer (1978)
and based on a region’s tidal range and average wave height. Hayes’ meso-
tidal barrier island coast is now more appropriately termed a mixed energy
(tide-dominated) shoreline in which tidal and wave processes are equally
responsible for barrier island shape.
It has been observed by Hayes and Kana (1976) that the shape of barrier
islands along mixed energy shorelines is similar to that of adrumstick (Fig.1).
DOMINANT
LONGSHORE
PREDOMINANT
WAVE APPROACH
TRANSPORT
REVERSAL
Fig.1. Drumstick barrier island model after Hayes and Kana (1976). The accretionary
updrift portion of the barrier is formed from the longshore transport of sand toward
the inlet and from the landward migration of swash bars from the ebb-tidal delta. The
downdrift end of the barrier develops through spit accretion.
357
They have identified drumstick barrier islands along the coasts of Alaska,
South Carolina, Georgia and The Netherlands. Stephen (1981) has also
reported drumstick barriers along the coast of southwest Florida. Despite
these examples an examination of any mixed energy coast reveals many
other barrier island forms (Fig.2). One such barrier island chain which
exhibits an assortment of shapes is the East Frisian Islands off the North Sea
coast of West Germany. This coastline is strongly affected by a dominant
direction of wave energy flux, abundant sediment supply and tidal inlet
processes.
It is the intent of this paper t o demonstrate that the variability in shape of
the East Frisian barriers is primarily related to processes of inlet sediment
bypassing. Sand ultimately bypasses the inlets along this coast in the form
of large landward migrating swash bars. Depending upon where these bars
attach t o the downdrift inlet shoreline the barriers can be dmmstick-shaped,
humpbacked or even downdrift bulbous-shaped.
PHYSICAL SETTING
The East Frisian Islands consist of seven barrier islands located in the
southeast North Sea between the Ems River to the west and the Jade Bay to
the east (Fig.3). The barrier system is 90 km long and is separated from the
mainland by a 4-12 km wide tidal flat which has been incised by a network
of tidal channels.
The wind regime along this section of the North Sea is seasonal. A
recorder on the island of Norderney showed that over a 19-yr period (1947-
1966) prevailing winds during the fall and winter (SeptemberMarch)
generally blow from the southwest while during the rest of the year (April-
August) they blow from the northwest (Luck, 1976a). Dominant winds
come from the southwest with an average velocity of 1 0 m s-l.
Although there exists no reliable, long-term wave records for the near-
shore region, the Summary of Synoptic Meteorologic Observations (SSMO
data- U.S. Naval Weather Service Command, 1974) indicate that for the
Bremerhaven data square the resultant wave power is directed t o the east-
southeast (azimuth = 101"; Nummedal and Fischer, 1978). This resultant
vector yields a net eastward longshore power component of 4.4 X lo3 W m-'
(Fig.4).
Short-term wave gauge data off the island of Norderney indicate that at
the 1 0 m water depth the average significant wave height exceeds
1.6 m (Niemeyer, 1978). Another wave gauge in 8 m of water off the coast
of Sylt in the Northern Frisian Islands show that exceeds 1m daily and
4.8 m on an annual basis (Dette, 1977). For comparison, the average sig-
nificant wave height at Atlantic City, New Jersey, is 0.8 m (water depth =
5.2 m) and at Savannah, Georgia, Ifll3 = 0.9 m (water depth = 15.8 mi
(Thompson, 1977).
The mean tidal range along the East Frisian Islands increases in an easterly
direction from a low of 2.2 m at Borkum to a high at Wangerooge of 2.6 m
358
BARRIER I S L A N D M O R P H O L O G Y
w 16km -+I n l e i r r d Direction o f Dominant Longshore Sed!mrnl Transport
WASSAW ISLAND r-->
Georgia /'
/
BRIGANTINE ISLAND ci)

N e w Jersey
S T R A W B E R R Y REEF -
COPPER RIVER DELTA
SULLIVANS ISLAND
South Carolina
TFRSCHELLING ISLAND f'.
Fig.2. Barrier island morphology along various mixed energy coasts. The coastal classi-
fication is after Hayes (1979) and Nummedal and Fischer (1978).
359
Fig.3. Location map of East Frisian Islands along t h e West German North Sea coast.
(NOS, 1983). During spring tide conditions these values increase t o 2.5 and
2.9 m, respectively. Tides in the North Sea are semidiurnal.
SHORELINE PROCESSES
Historical changes
Morphological changes of the East Frisian Islands, tidal inlets and back-
barrier environment have been determined from sequential maps produced
by Homeier and Luck (1969) for the years 1660, 1750, 1860 and 1960. A
summary of these data is given in Tables I--111 and depicted in Figs. 5 and 6.
During the past 300 yrs there has been an 80% increase in the areal extent of
the barrier islands (Table 11).A large proportion of this increase, however, was
due to poldering behind the barriers, a process whereby marsh lands and
tidal flats are diked from the sea. Still, when the poldered region is sub-
tracted an 18.03 km2 or a 35% increase in total barrier island area is calcu-
lated. These data and reports of substantial dune growth during the past
century (Luck, 1975) indicate that there is a positive sand budget for these
islands despite rising sea level in the North Sea (Streif and Koster, 1978).
The supply of sand for these barriers is believed to come from a variety of
sources including: (1)the Rhine and Maas Rivers and erosion of old lobes of
the Rhine delta along the Dutch coast (Van Straaten, 1965); (2) sediment
discharged from the Ems River (Luck, 1976a); and (3) from offshore glacial
deposits (Luck, 1976a). The historical information demonstrates that except
for the erosion due t o inlet migration, the overall morphology of the East
Frisian barriers is a product of accretionary processes.
360
TABLE I
Historical changes in barrier island length and tidal inlet width (after Luck, 1975)
Barrier islands Length ( m

_-) Diff. between
1650-19 60
1650 1750 1860 1960
Juist 10,470 13,010 15,810 14,970 +4500

Norderney 8070 9440 13,470 13,870 +5800
Baltrum 8170 7560 5460 6050 -3120
Langeoog 9600 10,280 11,230 10,920 +1320
Spiekeroog 5230 5180 5910 9810 +4580
Wangerooge 7 300 7400 7950 8320 +lo20
(total length) 48,840 52,870 59,830 62,940 14,100
Width ( m ) Diff. between

__.__
Tidal inlets 1650 1750 1860 1960 1650-1960
Norderneyer Seegat 6800 4730 2600 2770 -4030

Wichter Ee 2170 2030 710 850 -1320
Accumer Ee 3150 2480 1360 1710 -1440
Otzumer Balje 2440 2400 2440 2410 -30
Harle 5800 5670 4810 2000 -3800
(total width) 20,360 17,310 11,920 9740 -10,620
Total barrier 69,200 70,180 71,750 72,680 +3480
island length
and tidal inlet width
TABLE I1
Areas of the barrier islands and poldered regions (in k m 2 )
[sland 1650 1750 1860 1960 Difference Poldered Difference

between region between
1650- 1650-1 960
1960 minus
poldered
region
Juist 11.15* 12.35 14.38 12.60 +1.35 2.6 -1.25

Norderney 11.38 13.25 20.45 24.68 +13.30 5.6 + 7.7
Baltrum 6.5 6.05 6.50 6.23 -0.27 2.1 -2.37
Langeoog 9.18 11.30 14.85 20.40 +11.22 7.3 + 3.9
Spiekeroog 6.43 7.38 10.10 21.30 +14.87 1.7 +13.17
Wangerooge 7.40 7.58 7.55 8.40 +1.00 4.1 -3.1
Total 52.14 57.91 73.83 93.61 +41.47 23.4 +18.07
*Areas of the barrier islands from Luck (1975).

36 1
TABLE I11
Drainage areas of the tidal inlets (km’)
1650 1750 1860 1960 Difference

between
1650-19 60
Norderneyer Seegat 125 103 98 97 -2 8

Wichter Ee 60 51 40 28 -3 2
Accumer E e 101 100 84 88 -13
Otzumer Balje 51 51 77 73 + 16
Harle 154 128 79 62 -92
Totals 497 433 372 348 -149
-
Longshore sediment transport
The strong easterly wave energy flux that is produced by dominant and
prevailing winds out of the westerly quadrant results in a strong eastward
movement of sand along the Frisian Island coast (Fig.4). A longshore sedi-
ment transport rate of 130,000 m3 yr-’ can be inferred from the average
volume of sand, in the form of large bars, that bypassed Norderney Seegat
over a 31-yr period of time (Homeier and Kramer, 1957). This is a mini-
mum rate because it does not account for the volume of sand that moved
past the inlet without significantly changing the bar forms. A volume of
270,000 m3 yr-’ is calculated from the increase in size of Norderney from
I 0
Fig.4. Distribution of deep-water wave power for t h e Bremerhaven data square as deter-
mined from the Summary of Synoptic Meteorologic Observations. The west-northwesterly
wave power vector resultant causes a strong easterly longshore transport direction along
the East Frisian Island shore.
WANGEROOGE
-.
....-.
SPIEKEROOG
,' ,
:', L
f, I
._. ,I
5
ANGEOOG % {
ec ~. 9
*
5,
s. I
5,
t*
2
...-..
I 1750
_ _ _ 1860
O e 5 M''er NORDERNEY
0 5Kmr
~ 1960 3
1 MEMMERT
Fig.5. Shoreline changes o f the islands and mainland area from 1650 to 1960. Constructed from maps by Homeier and Luck (1969).
The lines behind the barrier indicate changes in the position o f the drainage divides. Generally, the barriers have prograded eastward and
the drainage divides have migrated eastward as well.
36 3
1650 to 1960 (Table 11). However, during the same period of time, sand
must have been bypassing Wichter Ee, so even this value is most likely a
minimum estimate.
Over the past 300 yrs the easterly transport of sand has caused a lengthen-
ing of the barriers at the expense of inlet width. Although the western ends
of the barriers have eroded through inlet migration, this loss in shoreline was
more than compensated by the islands' eastward growth. As shown in Table I
and Fig.5 total barrier island length increased from 48,840 m in 1650 t o
62,940 m by 1960. During the same time, total inlet width decreased from
20,360 t o 9740 m. The 3480 m net increase in total length of the inlet-
barrier system was due t o the westward growth of Juist and the eastward
growth of Wangerooge (Fig.5).
The narrowing of the tidal inlets since 1650 can be explained by a de-
crease in tidal prism which was caused by a 30% decrease in backbarrier tidal
flats and channels (Table 111). The filling of the backbarrier region has been
brought about by: (1)poldering behind the barriers and along the mainland;
and (2) landward spit extension at the inlets. The width of the barriers pre-
cludes washovers from being deposited in the backbarrier except perhaps
along the very eastern ends of the island, where the barriers are young and
low in profile.
The eastern ends of the East Frisian barriers have been developed through
spit accretion at a relatively rapid rate (ave. = 30 m yr-') due t o an abundant
sediment supply and a high longshore sediment transport rate. However, it
should be noted that spit accretion has been accommodated more by inlet
narrowing than by inlet migration.
80 000
T o l a l B a r r i e r I s l a n d tenqili,
0 J - d 300
1650 1750 1860 1960
Fig.6. Frisian Island morphological changes from 1650 to 1960 (barrier island and inlet
width data from Luck, 1 9 7 5 ) . The 29% increase in barrier island length has been at the
expense of inlet width (52% decrease) and has resulted from a 30% decrease in drainage
area.
364
TIDAL INLET PROCESSES
Inlet sediment bypassing
The pattern of sand accretion along the East Frisian barriers is controlled
directly by the location where swash bars attach t o the shoreline during the
inlet sediment bypassing process. Mechanisms of inlet sediment bypassing
were first recognized and described by Bruun and Gerritsen (1959). They
documented that sand bypasses inlets by wave action along the terminal lobe
or by tidal currents coupled with channel and bar migrations. Detailed
studies of transport patterns at Norderneyer Seegat by Luck (1976b) and
Nummedal and Penland (1981) and at Harle by Hanisch (1981) have shown
that sand bypassing at the East Frisian inlets occurs by a combination of the
two processes. Their data consisted of historical morphological information,
grain-size analysis, bedform measurements, sand tracer studies and time-
series current velbcity readings. A general model of inlet sediment bypassing
for the East Frisian coast based upon these studies is shown in Fig.7 and
summarized below.
Sand moves eastward along the beach toward the inlet through wave
induced longshore sediment transport. Once in the vicinity of the inlet,
sand bypassing follows the following complex pathway t o the downdrift
inlet shoreline. Sand entering the inlet is transported by wave action and
by flood-tidal currents across the swash platform and through marginal flood
channels t o the main ebb channel. Some of this sand is then transported to
the backbarrier sand flats where it may remain for a period of time
(Nummedal and Penland, 1981) or it may be reworked back into the sand
transfer system. The main ebb channel at these inlets is dominated by ebb-
tidal currents (Kramer, 1961) and thus most of the sand dumped into the
inlet is moved in a net seaward direction t o the distal northern portion of the
ebb delta. At this location, sand is added to the reefbow (Luck, 1974) as a
series of swash bars that outline the terminal lobe portion of the ebb-tidal
delta. The fate of these bars is tied t o the second transport system and will
be discussed next.
The western, updrift side of the inlets’ ebb-tidal delta is composed of a
series of northwest-southeast trending, elongated bars that are separated
from one another by shallow ( < 3 m deep) channels. Some of these channels
are dominated by flood-tidal currents. The segregation of flow in these flood
and ebb channels (Nummedal and Penland, 1981; Hanisch, 1981) causes the
northeasterly, zig-zag transport of sand in a seaward direction. The strong
easterly component of wave energy also augments the eastward movement of
sediment. Once the sand reaches the outer portion of the ebb delta it be-
comes incorporated in swash bars that make up the reefbow. Some sand is
lost along this pathway when it is dumped into the main ebb channel where,
as described earlier, the dominant ebb-tidal currents transport it to the
outer portion of the ebb delta,
The inlet sediment bypassing process is completed as bars that comprise
365
Fig.7. Model of inlet sediment bypassing for the East Frisian Island coast based o n
historical and morphological data, bedform orientations, sand tracer studies, box cores
and tidal current measurements from the.investigations of Nummedal and Penland (1981)
and Hanisch ( 1 9 8 1 ) . Note that one of the end products of inlet sand bypassing is the
attachment of large swash bars to the downdrift beach.
the reefbow migrate in an arc along the periphery of the ebb delta to become
attached to the landward beach. Maps constructed by Homeier and Kramer
(1957) illustrate the movement of the center of gravity of bars at
Norderneyer Seegat over a 31-yr period of time (1926-1957) (Fig.8A) and
the shore-parallel elongation of the bars as they migrate onshore (Fig.8B).
Figure 8A and C demonstrate that the bars have a consistent west-east
migration and that their movement has been as great as 400 m yr-'. A
photographic time sequence of Wichter Ee and Accumer Ee illustrates the
migration of large bars toward the shoreline of Baltrum and Langeoog,
respectively (Fig.9). The morphology of a large swash bar welding to the
western end of Spiekeroog is shown in Fig.10. From Figs.8-100 it can be
seen that as they weld t o the beach, the bars are 1-3 km long. I t should be
noted that an unknown amount of sand is also continuously moved onshore
independent of the bar forms.
366
The Frisian tidal inlets exhibit many of the same mechanisms and path-
ways of sand bypassing as those of the U.S. East Coast. Mixed energy tidal in-
lets, including Price Inlet, South Carolina (FitzGerald et al., 1976), North In-
let, South Carolina (Finley, 1976) and Chatham Harbor Inlet, Massachusetts
(Hine, 1975), contain ebb- and flood-dominant channels that are similar in
hydraulics and morphology t o those of the western portion of Frisian ebb-
tidal deltas. These are the spillover channels noted by Oertel (1972) and
FitzGerald e t al. (1978) and the marginal flood channels described by Oertel
(1972) and Hayes et al. (1973).
7 7 7 14
53 44
r-xz--- BAR MIGRATION
b,
195711v
l95O/Vll = July
611
1950
367
I (C
Fig. 8. Bar migrational trends a t Norderneyer Seegat ( A and B). A . Map of bar movement
along the periphery of the ebb-tidal delta between 1 9 2 6 and 1 9 5 7 . Each d o t represents the
bar's center of mass o n successive years. Note the bar's movement is from west t o east and
t h a t the outer exposed bars move a t a faster rate than t h e inner bars (from Nummedal and
Penland, 1 9 8 1 ; after Homeier and Kramer, 1 9 5 7 ) . B . Morphologicalchangesof alargeswash
bar as it attached to the beach a t Norderiiey between 1 9 4 9 and 1 9 5 7 (from Nummedal and
Penland, 1 9 8 1 ; after Homeier and Kramer, 1 9 5 7 ) . C . Bar migrational trendsat Accumer Ee.
Note t h a t the swash bars attach t o t h e updrift end of Langeoog. T h e distance between n u m -
bers is o n e kilometer ( f r o m Homeier and Kramer, 1 9 5 7 ) .
Location of bar welding
In the inlet sediment bypassing process, the location where swash bars
attach t o the barrier island shoreline is controlled by the size and morph-
ology of the ebb-tidal delta. Generally, along the East Frisian Islands the
distance from the inlet that swash bars weld t o the downdrift shoreline
increases as: (1)inlet size increases; and (2) the downdrift skewness of the
ebb-tidal delta increases. The asymmetric configuration of the ebb deltas is
a product of the strong easterly longshore sediment transport direction, the
residual easterly tidal currents (Nummedal and Penland, 1981) and the
position of the main ebb channel with respect t o the inlet throat.
Using the reefbow as a relative measure of ebb-delta size, it is found that
the width of the deltas of the major inlets (Norderneyer Seegat, Accumer
Ee, Otzumer Balje and Harle) varies from a high of 8.3 km at Norderneyer
Seegat t o a low of 5.1 km at Accumer Ee (Table IV). The length of shore-
line that the ebb-tidal deltas overlap is also greatest at Norderneyer Seegat
(5 km) and least at Accumer Ee (1.9 km). However, this trend is not
entirely a function of size. As Table IV and Fig.11 illustrate, the percent
368
Fig.9. Sequential photographs of the bar welding process along the islands of Baltrum
and Langeoog during the period between May 1 9 6 9 and J u n e 1 9 7 3 (from Luck, 1974).
369
Fig.10. Oblique aerial photograph of t h e western end of Spiekeroog illustrating the mor-
phology of a large, landward migrating swash bar, approximately 1.5 km in length.
TABLE IV
Ebb-tidal delta configurations
Ebb-tidal delta Delta Downdrift % overlap

location width (km) barrier island
overlap (km)
Norderneyer Seegat 8.3 5.0 61

Accumer Ee 5.1 1.9 37
Otzumer Balje 6.0 3.4 57
Harle* 5.7 3.2 56
*Harle has been stabilized o n its east side b y 1.4 km length jetty that extends directly
southwest into t h e inlet.
overlap of the deltas is also controlled, in part, by the position of the main
ebb channel with respect t o the adjacent islands. At Norderneyer Seegat, the
main ebb channel abuts the downdrift island of Norderney which results in
the ebb-tidal delta being displaced to the east, hence, the large shoreline
overlap. The opposite inlet configuration occurs at Accumer Ee. Here the
main ebb channel flows against the updrift island of Baltrum resulting in
less eastward offset of the delta and thus a smaller length of shoreline over-
lap. The remaining two inlets, Otzumer Balje and Harle, have main ebb
channel positions in the middle of the inlet throat. The percentage of ebb
370
NORTH SEA
0G "..,
.t
-
Ebb Della
M a i n Inlet Channel
i
0
i~~
5 M.1,.
-2
Barrier Island 0 5 n,.,
Fig.11. A map illustrating the configurations of the ebb-tidal deltas. The strong easterly
component of wave energy flux causes a preferential overlap of the ebb-tidal delta along
the downdrift inlet shoreline. The amount of overlap is controlled by inlet size and
position of the main channel at the inlet throat.
delta-shoreline overlap of these two inlets is intermediate between that of

Norderneyer Seegat and Accumer Ee (Table IV).
BARRIER ISLAND MORPHOLOGY
The morphological variability of the East Frisian Islands is depicted in

Fig.12. The island of Juist, the westernmost barrier in the group, is long and
straight, and is narrower than the other islands. There has been little land
reclamation along the back side of this island compared to the rest. The con-
figurations of Norderney, Spiekeroog and Wangerooge are similar to one
another in that their seaward shorelines are humpbacked. The remaining two
islands, Baltrum and Langeoog, are bulbous at their downdrift and updrift
ends, respectively. With the exception of Baltrum, the narrowest portion of
the barriers coincides with their easterly downdrift ends, The landward
shorelines of the islands are largely a product of poldering.
Factors affecting barrier island shape
The development of the Frisian Island morphological diversity is ex-

plained by the factors summarized in Fig.13. As previously discussed, the
location where swash bars attach to the beach coincides with the bulbous
portion of the barrier and is controlled by the configuration of the ebb-
tidal delta. The symmetry of the ebb-tidal delta has been shown to be a func-
tion of the position of the main ebb channel at the inlet throat and the size
of the tidal inlet. The first factor is affected by: (1)the easterly longshore
sediment transport direction; and (2) the distribution, orientation and mag-
nitude of backbarrier tidal channels. Inlet size is primarily controlled by
371
u
.I I S7'
Location of Bar Welding Positive Sand Budget Negative Sand Budget
Neutral Sand Budget
Fig.12. Barrier morphology of the East Frisian Islands. T h e bulbous portion of the barrier
coincides with the position where swash bars attach to the beach. The location of bar
welding also correlates well with the erosional-depositional shoreline trends reported by
Luck (1975). Generally, t h e shoreline downdrift of the bar attachment site is stable to
depositional while updrift of the bar attachment site is erosional. An exception t o the
pattern occurs a t Langeoog where swash bars migrate into the inlet providing nourish-
m e n t t o t h e entire western end of the island.
drainage area, as tidal range does not change significantly along this section
of coast. The more easterly the location of the main ebb channel at the inlet
throat and the larger the size of the tidal inlet, the greater is the overlap of
the ebb-tidal delta along the downdrift island shoreline. A highly skewed ebb-
tidal delta configuration results in the attachment of swash bars to the shore-
line far from the inlet mouth. This condition produces humpbacked and
downdrift bulbous barrier islands. The opposite ebb-tidal delta arrangement
produces straight barrier islands and drumstick barriers.
Morphology of individual Frisian Islands
The long straight nature of Juist is explained by the fact that bar welding
occurs at the westernmost end of the island and not to the northward facing,
372
Drumstick
Pos,t,on of Barrier Island
Main Ebb Channel -.
at Inlet Throat
Amount of
Ebb-Tidol Delta
Overlap of
Downdrtft
Inlet Shoreline
- Location Where
Swash Bars
Attach to
the Beach
Barrier Island
Relative Size of
Inlet vs Island Downdrift Bulbous
- Barrier Island
x
,--
%,'
Fig.13. Model for the morphological development of barrier islands. Barriers may have a
variety of shapes depending upon the position along the downdrift inlet shoreline where
swash bars migrate onshore.
seaward shore (Fig.12). The northwest orientation of the channel between

Juist and Memmert controls this sedimentation pattern which has resulted
in a 3.0 km westward growth of the island since 1650 (Fig.5).
At Norderneyer Seegat the easterly position of the main ebb channel,
abutting the island of Norderney, results in a highly asymmetric configura-
tion of the ebb-tidal delta (Fig.11). Due to this morphology the reefbow
meets the downdrift shoreline 5.0 km east of the inlet mouth (Table IV).
This region coincides with the humpbacked portion of Norderney and some
of the highest dune fields on the island (Fig.12).
Quite a different pattern of sedimentation is found at Langeoog (Fig.12).
As discussed earlier, the small size of Accumer Ee, compared to that of
Norderneyer Seegat, combined with the westerly position of the main ebb
channel at the inlet throat (Fig.11) have produced an ebb-tidal delta that
fronts very little of the downdrift shoreline (Table IV). Due to this con-
figuration the onshore migration of swash bars from the reefbow occurs at
the western end of Langeoog along the inner inlet shoreline. This pathway
of sand movement has accounted for 2 km of shoreline progradation in the
southwest portion of the island (since 1860) (Fig.5). The stable to accre-
tionary trend of the inlet shoreline is evidenced by the lack of any groins in
this region.
Landward-migrating swash bars at Accumer Ee not only attach to the
western end of Langeoog but also migrate toward the inlet throat. This
process acts as a feedback mechanism which helps t o keep the main ebb
channel in a westerly position against Baltrum, thereby maintaining the loca-
tion of the eastern reefbow close to the inlet mouth. Similar movement
of swash bars toward the inlet throat has been recorded at central South
Carolina inlets by FitzGerald et al. (1978) and has been attributed to the
influence of flood-tidal currents on the wave-induced, bar migrational
process.
313
The mid-throat position of the main-ebb channel at Otzumer Balje and

Harle results in an ebb-tidal delta configuration that is intermediate between
that of Norderneyer Seegat and Accumer Ee (Fig.ll). The resulting hump-
backed barrier island morphology of Spiekeroog and Wangerooge is predic-
tably transitional between that of Norderney and Langeoog (Fig.12). The
bulbous downdrift shape of Baltrum is caused by the easterly location of
the main ebb channel at the inlet throat and the short length of the island
(Fig.11). Inlet sediment bypassing at Wichter Ee results in the welding of
swash bar t o the eastern end of Baltrum (Fig.12).
It should be recognized that the barrier island model explained above
(Fig.13) also works well for the East Frisian Islands during their construc-
tional history. For example, note in Fig.5 that as Harle migrated in an eas-
terly direction from 1650 t o 1960 so too did the humpbacked portion of
Wangerooge. Similar trends are also apparent at Langeoog and Spiekeroog.
DISCUSSION
Hayes and Kana (1976) have put forth the drumstick model (Fig.1) to
explain the morphology of barrier islands along mixed energy coasts. In their
model the bulbous updrift end of the barrier is attributed to a sediment
transport reversal caused by waves refracting around the ebb-tidal delta.
They suggest that this reduces the rate at which sand bypasses the inlet
resulting in a broad zone of accumulation. The downdrift end of the barrier
forms through spit accretion.
In clarification and amplification of Hayes and Kana’s (1976) model this
study has demonstrated two things. First, it is important t o note that depo-
sition on the downdrift inlet shoreline occurs primarily through the attach-
ment of swash bars from the ebb-tidal delta, regardless of whether or not
there exists a transport reversal., Also, the zone where bar welding takes
place is dependent on inlet size and ebb-tidal delta configuration and can
occur some distance from the inlet mouth. The barrier island model that
has been developed from this study (Fig.13) can account for barriers that are
drumstick-shaped, humpbacked and that have other forms. The East Frisian
Islands exemplify this concept well. The bulbous portion of these barriers is
a consequence of the location where swash bars move onshore and build up
the beach. In accordance with the Hayes and Kana model, the historical data
from this study has documented that the downdrift end of these barriers is
a product of spit accretion. However, it should be noted that spit growth in
the East Frisian Island is accommodated more through inlet narrowing than
inlet migration.
The use of our model to explain barrier island morphology along other
mixed energy coastlines requires that wave energy be sufficient to cause the
formation and landward migration of swash bars on the ebb-tidal delta. In
a recent paper (FitzGerald, 1982) it was documented that inlet sediment
bypassing occurs through landward bar migrations along the mixed energy
coasts of New England, New Jersey, Virginia, South Carolina and the Copper
374
River Delta region of Alaska. Thus, most of these barriers would be expected
to conform to our model.
For example, the shape of Sullivans Island, South Carolina (Fig.B), is a
product of an asymmetric ebb-tidal delta configuration and a southerly
downdrift location of the main ebb channel. Both of these factors contribute
t o swash bar attachment some distance from the inlet mouth (1km).
Although the Georgia coast has barrier islands of similar size compared to
the East Frisian Islands, swash bar development does not appear to be an
active process at the sounds along this coast. This is due to the large inlet
prisms (109-1011m3) (Jarrett, 1976), extensive ebb-tidal deltas, and small
wave energy = 90 cm) (Thompson, 1977) of the Georgia coast. Oertel
(1977) has shown that most of the transfer of sand from the delta to the
beach occurs very close to the inlet (200 m).
CONCLUSIONS
(1)An abundant sediment supply has led to a 35% increase in size of the
East Frisian Islands between 1650 and 1960. Some of this growth has been
contemporaneous with a narrowing of the tidal inlets. Poldering of the
back-barrier area coupled with natural sedimentation processes has reduced
the tidal prism resulting in smaller equilibrium inlet cross-sectional areas.
(2) The strong easterly component of wave energy flux delivers a mini-
mum of 270,000 m3 of sand to the inlets per year. This sediment bypasses
the inlet through tidal and wave-generated currents and is eventually added
t o the downdrift inlet shoreline in the form of large landward-migrating
swash bars (length = 1.0-1.5 km). Swash bar migration accounts for no
more than half of the volume of sand that is bypassed.
(3) Swash bars cause a preferential progradation of the beach at the site
of attachment. This process has produced a variety of barrier shapes along
the East Frisian Island chain. The morphology of individual barriers is con-
trolled by the location where the swash bars migrate onshore which, in turn,
is dependent on the amount of overlap of the ebb-tidal delta along the down-
drift inlet shoreline. The configuration of the ebb-tidal delta is a function of
inlet size and position of the main ebb channel at the inlet throat.
(4)The barrier island model presented in this paper differs from the model
proposed by Hayes and Kana (1976) in that barrier shape is dependent on
the position along the shoreline where swash bars weld to the beach. Our
model can account for not only drumstick-shaped barrier islands but bar-
riers that are humpbacked and other forms as well.
ACKNOWLEDGEMENTS
This investigation was supported by the Office of Naval Research, Coastal

Processes Program, through contract no. N00014-7842-0612 (Dag Nummedal,
Principal Investigator).
The authors would like to thank Dr. Gunter Luck, the director of the
375
Forschungsstelle Norderney, State of Lower Saxony, Germany, for the

logistical support that he provided during the field study. Gunter Luck,
Hanz Niemeyer and other research personnel at Forschungsstelle Norderney
are gratefully acknowledged for their scientific input and advice during our
investigation.
REFERENCES
Bruun, P. and Gerritsen, F . , 1959. Natural by-passing of sand a t coastal inlets. J. Water-
ways Harbors Div., Am. Assoc. Civ. Eng., 85: 75-107.
Davies, J.L., 1964. A morphogenetic approach to world shorelines. Z. Geomorphol.,
8 : 27-42.
Dette, H.H., 1977. Ein Vorschlag zur Analyse eines Wellenklimas. Die Kuste, 31:
1 66-1 80.
Finley, R.J., 1976. Hydraulics and dynamics of North Inlet, South Carolina, 1974-1975.
GITI Report, U.S. Army Coastal Eng. Research Center and Waterway Exp. Station,
Ft. Belvoir, V., 188 pp.
FitzGerald, D.M., 1982. Sediment bypassing of mixed energy tidal inlets. Am. SOC.Civ.
Eng., Proc. 18th Conf. Coastal Engineering, pp.1848--1861.
FitzGerald, D.M., Nummedal, D. and Kana, T.W., 1976. Sand circulation patterns at Price
Inlet, South Carolina. Am. SOC. Civ. Eng., Proc. 15th Conf. Coastal Engineering,
pp. 1868-1880.
FitzGerald, D.M., Hubbard, D.K. and Nummedal, D., 1978. Shoreline changes associated
with tidal inlets along the South Carolina Coast. Coastal Ocean Div., Am. SOC.Civ.
Eng., Coastal Zone 78, San Francisco, Calif., 1973-1994.
Hanisch, J., 1981. Sand transport in the tidal inlet between Wangerooge and Spiekeroog
(W. Germany). Spec. Publ. Int. Assoc. Sedimentol., 5: 176-185.
Hayes, M.O., 1975. Morphology of sand accumulation in estuaries: An introduction to
the Symposium. In: L.E. Cronin (Editor), Estuarine Research, VoI. 2. Academic Press,
S.P. Leatherman (Editor), Barrier Islands: From the Gulf of St. Lawrence to the Gulf
of Mexico. Academic Press, New York, N.Y., pp.1-28.
Hayes, M.O. and Kana, T.W., 1976. Terrigenous clastic depositional environments. Tech.
Report 11, CRD, Dept. of Geology, University of South Carolina, Columbia, S.C.,
364 pp.
Hayes, M.O., Owens, E.H., Hubbard, D.K. and Abele, R.W., 1973. The investigation of
forms and processes in the coastal zone. In: D.R. Coates (Editor), Coastal Geo-
morphology. State Univ. of New York, Binghamton, New York, pp.ll-41.
Hine, A.C., 1975. Bedform distribution and migration patterns on tidal deltas in the
Chatham Harbor Estuary, Cape Cod, Massachusetts. In: L.E. Cronin (Editor), Estuar-
ine Research, Vol. 2. Academic Press, New York, N.Y., pp.235-252.
Homeier, H. and Kramer, J., 1957. Verlagerung der Platen im Riffbogen vor Norderney
und ihre Anlandung an den Strand. Jahresber. 1956, Forschungsstelle Norderney,
8 : 37-60.
Homeier, H. and Luck, G., 1969. Das Historische Kartenwerk 1:50,000 der Niedersach-
sischen Wasserwirtschaftsvenaltung, Gottingen.
Jarrett, J.T., 1976. Tidal prism-inlet area relationships, General Investigations of Tidal
Inlets. Rept. no. 3., Coastal Eng. Research Center, Ft. Belvoir, Va., 32 pp.
Kramer, J., 1961. Strommessungen im Hafengebeit von Norderney: Jahresber. Forsch-
ungsstelle Norderney, 1 3 : 67-94.
Luck, G., 1974. Untersuchungen zur Sedimentbewegung mit Hilfe einer Unterwasser-
ferhnanlage. Jahresber. 1973, Forschungsstelle Norderney, 25: 55-78.
376
Luck, G., 1975. Der Einfluss der Schutzwerke der ostfriesischen Inseln auf die morpho-
logischen Vorgange im Bereich der Seegaten und ihre Einzugsgebeite. Mitt. Leicht-
weiss Inst., Braunschweig, 4 7 : 1-22.
Luck, G., 1976a. Inlet changes of t h e East Frisian Islands. Am. SOC.Civ. Eng., Proc. 1 5 t h
Conf. Coastal Engineering, New York, 11: 1938-1957.
Luck, G., 197613. Protection of the littoral and seabed against erosion. Fallstudie Nord-
erney, Jahresber. 1 9 7 5 . Forschungsstelle Norderney, 27 : 9-78.
National Ocean Survey, NOAA, 1983. Tide tables for Europe and West Coast of Africa
including t h e Mediterranean. 1 7 8 pp.
Niemeyer, H.D., 1978. Wave climate study in t h e region of the East Frisian Islands and
coast. Am. Soc. Civ. Eng., Proc. 1 6 t h Conf. Coastal Engineering, New York, 1:
134-151.
Nummedal, D. and Fischer, I., 1978. Process-response models for depositional shorelines:
T h e German and Georgia Bights. Am. Soc. Civ. Eng., Proc. 1 6 t h Conf. Coastal Engin-
eering, New York, 11: 1212-1231.
Nummedal, D. and Penland, S., 1981. Sediment dispersal in Norderneyer Seegat, West
Germany. Spec. Publ. Int. Assoc. Sedimentol., 5 : 187-210.
Oertel, G.F., 1972. Sediment transport of estuary entrance shoals and the formation of
swash platforms. J. Sediment. Petrol., 4 2 : 857-863.
Oertel, G.F., 1977. Geomorphic cycles in ebb deltas and related patterns of shore erosion
and accretion. J. Sediment. Petrol., 4 7 : 1121-1131.
Stephen, M.F., 1981. Effects of seawall construction o n beach and inlet morphology and
dynamics a t Caxambas Pass, Florida. Ph.D. Diss., University of South Carolina, Col-
umbia, S.C., 1 9 6 pp.
Streif, H. and Koster, R., 1978. T h e geology of t h e German North Sea Coast. Die Kuste,
32: 30-50.
Thompson, E.F., 1977. Wave climate a t selected locations along U.S. coasts. U.S. Army,
Coastal Engineering Research Center, Tech. Rept. No. 77-1, 3 6 4 pp.
U.S. Naval Weather Command, 1974. Summary of Synoptic Meteorological Observations,
Western European Coastal Marine Studies. Nat. Climatic Center, NOAA, Ashville,
N.C.
Van Straaten, L.M.J.U., 1965. Coastal barrier deposits in south and north Holland.
Meded. Rijks Geol. Dienst, 1 7 : 41-75.
Marine Geology, 60 ( 1 9 8 4 ) 377-389 377
COARSE CLASTIC BARRIER BEACHES: A DISCUSSION OF THE

DISTINCTIVE DYNAMIC AND MORPHOSEDIMENTARY
CHARACTERISTICS
R.W.G. CARTER and J.D. O R F O R D

School of Biological and Environmental Studies, T h e N e w University of Ulster, Coleraine,
Co. Londonderry B T 52 ISA (Northern Ireland)
Department of Geography, The Queen’s University of Belfast, Belfast B T 7 I N N
(Northern Ireland)
(Received March 1, 1983;revised and accepted August 1 2 , 1 9 8 3 )
ABSTRACT
Carter, R.W.G. and Orford, J.D., 1984. Coarse clastic barrier beaches: A discussion of the
distinctive dynamic and morphosedimentary characteristics. In: B. Greenwood and
R.A. Davis, Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated
Coarse clastic barriers are common on mid- and high-latitude coasts. They possess a
morphosedimentary and dynamic distinctiveness which sets them apart from sandy fine-
clastic barrier forms. The reflective nature of the seaward barrier favors the development
of zero mode, sub-harmonic edge waves particularly during long period swells (10-20 s),
manifest in the formation of high level cusps. In some circumstances the pattern of recent
overwashing of the barrier is related to cusps. Lack of distinct tidal passes, due partly to
the high seepage potential of coarse barriers, means that very little sediment is trans-
ported seaward. Thus the barriers roll steadily onshore, and sections show a variety of
washover facies, related t o the volume of. overwash surges.
INTRO DUCT10 N
Over the last decade there has been an upsurge of interest in barrier coast-
lines, both from viewpoints of shoreline stability (Leatherman, 1979a;
Kaufman and Pilkey, 1979) and as analogues for examining the reservoir
potential of ancient sedimentary bodies (McCubbin, 1982).
However, as Zenkovich’s (1967) classification shows, there is a number of
discrete barrier types, which Zenkovich divided into “free” and “fixed”
forms. An unfortunate trend has been to call all barriers “barrier islands”
(Glaesner, 1978; Hayes, 1979) or to assume that the barrier island sub-type
can be termed simply “barrier” (e.g. Godfrey et al., 1979; Leatherman
197913). There appear to have been no attempts to classify barriers on the
basis of grain size.
This paper concentrates on one type of barrier - the fixed, fringing gravel
barrier - with the aim of establishing its distinctiveness in terms of both
dynamics and morphosedimentary structure. These barriers are common on
0025-3227/84/$03.00 0 1 9 8 4 Elsevier Science Publishers B.V.

378
wave-dominated, mid- to high-latitude coasts, particularly where glacigenic

deposits are being reworked (Armon, 1974; Randall, 1977; Carter and
Orford, 1980; Wang and Piper, 19823. Similar structures have not been
recorded on low wave energy coasts, except in high latitudes as a result of
ice-push processes (Barnes, 1982). Conglomeratic barrier beaches also have
a presence in the geological column, albeit a minor one (Ali, 1976; Leckie
and Walker, 1982; Wright and Walker, 198lj. The paper is largely based on
the authors' research experience in southeast Ireland (Carter and Orford,
1980, 1981; Orford and Carter, 1982a, b, 1985) but reference is made t o
similar barrier structures elsewhere in Britain and Ireland (Fig.1).
Almost all these barrier beaches, around the coasts marginal to the south-
western approaches to the British Isles, are exposed to waves generated
within the North Atlantic. The barrier coasts are rarely influenced by storms
of exceptional severity, i.e. hurricanes. A more common picture is of
repeated lower-magnitude storms associated with the frequent southwest to
northeast passage of cyclonic depressions.
Incident waves are predominantly from the southwest or west. Deep-
water storm waves may exceed 10 m in height and poststorm decay swell
often reaches 12-14 s periods (Hogben and Lurnb, 1967) although longer
periods have been recorded (>20 s) affecting the gravel barrier at Chesil
Beach in Dorset (Golding, 1981). Table I shows the seasonal wave climate
Fig.1. Map of locations mentioned in the text. Fifty year H,,, for fully arisen sea are
also shown t o provide a picture of the extreme wave climate.
TABLE I
Seasonal wave conditions at Carnsore Point
Autumn 1.42 2.85 3.55 6.74 4.9 7.7 0.65

Winter 1.72 3.43 3.61 6.97 5.6 8.4 0.69
Spring 1.04 2.08 2.55 5.0 5.2 9.4 0.68
Summer 0.91 1.82 2.00 4.02 4.8 7.6 0.71
Unpublished _data from the Electricity Services Board, Dublin. Hs:mean significant wave
height (m), H,,,,,: mean maximum wave height (m), (95): 95th percentile value, T,:
zero-crossing wave period (s).
379
for southeastern Ireland. Spring tidal ranges on the Irish coasts vary from 1
t o 4 m. Graff (1981) has documented the return periods and heights of
annual sea-level maxima for Fishguard (on the east side of St. George’s
Channel, 100 km east of Carnsore - the nearest data source), which shows
that the 5-20 yr return period of total surge levels is in the range of 1.8-
2.0 m above MHWS. These values should be treated as crude estimates, as
coastal configuration is radically different for two sites. Storm surges on
more open coasts of the Celtic Sea rarely exceed 0.5 m (Pugh and Vassie,
1978; Flather, 1981).
EVOLUTIONARY FRAMEWORK
Compared t o the detailed work on fine clastic barriers (e.g. Kraft, 1971;
Moslow and Heron, 1981; Rampino and Sanders, 1981), the Holocene evolu-
tion of gravel barriers has not been extensively studied. However, a broad
framework may be suggested.
It would appear that most fringing gravel barriers have moved landward
under rising post-glacial sea levels. Transgressive sea levels lead to a rolling
over of the barrier (Carr and Blackley, 1973; Carter and Orford, 1980)
leaving little evidence of earlier shoreline positions.
The magnitude of sediment supply is crucial t o the development of gravel
beaches as there is limited sediment recycling once material is incorporated
in the barrier. Where sediment supply has failed, or is failing, as is the case
throughout the British Isles, barriers tend to develop as massive solitary
forms, rather than as multi-ridge complexes. The relative proportions of
gravel sizes t o finer sizes will determine the transportability of the gravel
mass. A small gravel component may become either trapped within the
larger gravel or boulder component, or buried under sand. Alternatively,
the coarse fraction may overpass the fine (Fahnestock and Houshild, 1962;
Everts, 1973); and under the asymmetrical wave stress move preferentially
onshore.
Through time, supply rate of gravel may change due t o the diminution of
sources and the maturing of the littoral drift system, the declining rate of
sea-level change, or the protuberance of headlands which act as natural
groins and disrupt the drift pattern. Examples of these changes have been
described around the British Isles; in southeast England, Edison (1983)
has argued that the longshore supply of gravel t o Dungeness declined asymp-
totically with a reduction in the rate of sea-level rise (after 5500 B.P.). Carr
and Blackley (1974) have described a similar situation for Chesil Beach in
Dorset. At Slapton Ley, Devon (Hails, 1975; Morey, 1983), and Carnsore
Point, southeast Ireland (Carter and Orford, 1980), the gradual emergence
of adjacent headlands has acted to cut off the longshore supply. In the
Irish example long-term sea level has been relatively static over the last
5000 yrs (Carter, 1983) and the relative maturity of the drift system, i.e. the
establishment of sediment cells (Lowry and Carter, 1982), has also contri-
buted to the diminution of lateral transport.
380
BARRIER FORM, STRUCTURE AND PROCESS
Wave climate modifications
Gravel-dominated barriers tend to fall within the reflective domain of the

morphological beach classifications of Short (1979) and Wright et al. (1979,
1982). This domain is distinguished by low values (<3) of both the surf-
scaling parameter, E (Huntley et al., 1977), and the beach stability index
(a< 1) of Dean (1973). Short (1979) and Wright et al. (1979) outline a
number of recognition criteria for reflective beaches, including presence
of a steep beach face slope, a break point step, a single swash berm and
high level cusps and absence of large-scale nearshore bedforms; points all
common to the southeastern Irish barriers. The transitional forms associated
with finer clastic beaches and barriers (Wright et al., 1982) do not occur on
gravel based barriers.
Reflective topography can engender sub-harmonic edge waves, usually of
zero mode (Guza and Davis, 1974). Under long-period swells these waves
are easily excited and increase in amplitude. Such long-period swells (>lo s)
are common in the North Atlantic (Hogben and Lumb, 1967) forming part
of the storm decay spectra. Manifestation of swell activity on the beach
appears in the accentuation of run-up, which may lead t o berm overtopping
and breaching (Short, 1979) or the formation of high-level erosional cusps,
with spacings of around half the edge wave length (Darbyshire, 1977; Wright,
1981). The development and persistence of edge-wave controlled topography
may assist in both directing subsequent overwash and controlling washover
spacing (Orford and Carter, 1984).
For a variety of reasons, gravel-dominated barriers remain relatively stable
in the face of wave attack. Of paramount importance is the inability of
particles within a gravel mass to become entrained, except under high-energy
events. Not only do the individual clasts resist transport, but they give rise
t o hydrodynamic rough surfaces capable of significantly affecting the in-
shore wave spectra (Wright et al., 1982).
Inevitably any beach face structures that do develop have a high sur-
vival potential into, or even through, normally destructive process phases.
Orford and Carter (1985) show that a fair-weather supratidal marine terrace
at Carnsore Point [a result of long-term (>1yr) development], not only
survived through a major storm, but also appeared t o control both the mode
of wave breaking and reformation, t o produce a spatially distinctive reflec-
tive/dissipative regime (Fig.2) across the upper beach. The terrace acts as a
step over which either waves are broken and reformed (similar to the
process acting on coral reefs; Suhayada and Roberts, 1977), or swash excur-
sions are amplified (Short, 1979). Ultimately, high-level swash excursions
during severe storms lead t o barrier crest sedimentation of marine gravels.
381
HR C l m kt0 >6rn
rn T 6-8sec dbCl5rn T 7-10rec
10
6-
’ --Ertr.rne MHWS
i 4
FINAL BREAKING
INITIAL BREAKING
I
Fig.2. Nearshore wave conditions a c r o s a ramp profile at Carnsore Point, Southeast
Ireland. Inshore storm waves, that may have already broken offshore, are partially broken
again by the outer “step” of the terrace. While a certain amount of wave energy is reflect-
ed at this point, low reformed waves also pass across the terrace constructing a swash
ramp (Orford and Carter, 1982a). Ultimately this may lead to dune overtopping, up to
6 m above the extreme level of the mean high water Spring tides.
Sedimentation control
Most coarse clastic barriers comprise varying mixtures of gravel and sand,
alongshore, crossbeach and at-depth. Where the gravel population dominates,
sand becomes a subsidiary, usually interstitial component. On the beach
face, sand may become packed into the gravel component, possibly via the
vibratory swash mechanism (kinetic sieving) of Middleton (1970). We favor
this process to produce the substantial upward coarsening units often en-
countered in gravel beach sections (especially on the upper beach face),
where the ratio of clast size to matrix size can be in excess of thirty times,
rather than the dispersive stress mechanism discussed by Sallenger (1979).
This latter process may assume more importance under swash/backwash
flows on the lower beach face. Continued introduction of fine material into
the coarse matrix will lead t o reductions in hydrostatic porosity and permea-
bility, so that water storage capacity and throughflow discharge (both from
swash inputs and sea-lagoon exchange) will fall. When fine particle infilling
reaches matrix capacity, excess sand will provide a source both for the for-
mation of swash-based bedforms, and when dry, eolian ones. Small barrier-
top dunes develop at this stage. Eventually the amount of sand, relative t o
gravel, will increase t o the point where it produces a low-angle beach face
sufficient for the transition from a stable reflective beach face to one more
liable to allow the intermediate reflective/dissipative stages (Wright et al.,
1982j. Should a small proportion of gravel remain, it may well become more
mobile, particularly in the upper flow regime, large clasts moving across
(overpassing) the finedominated bed. Carr (1971) notes that coarse particles
may “outrun” fine particles downdrift, as they are subject to constant
382
“rejection” (Moss, 1962). In circumstances where sediment supply is low,

this may lead t o coarsening-downdrift beach grading (Zenkovich, 1967).
Structural control
Orford and Carter (1982a) have shown how the barriers in southeast
Ireland build-up during storms. This is particularly important at the beach
crest, as vertical accumulations reduce the chances of swash breaching and
overwashing. Such modifications only arise, however, at a late stage in gravel
barrier genesis. Examination of barrier cross-sections and shallow borehole
records (Carter and Orford, 1982b) reveals a triparate structure. Figure 3
shows a lenticular lower barrier unit followed by washover gravels and over-
lain by overtop deposits. This division is related to gradually decreasing
volumes of overwash (Orford and Carter, 1982a) as the barrier agrades
vertically. The lower unit comprises a ridge-like structure of thin, repetitive
coarsening-upward lamina, with a persistent cross-barrier lenticularity. The
larger clasts are often imbricated seaward. Occasional massive boulders
(-10 4 ) occur in these laminations. The bottom unit is indicative of large-
volume swash bores traversing a lower barrier. As the height of the barrier
increased relative t o sea level, so these lower laminae were replaced by more
spatially discrete washover fans of cross-barrier wedge form. The focus of
sedimentation is at the ridge crest, rather than on the back of the barrier.
Finally the barrier is raised t o the point where washover is precluded and
only minor overtopping dominates.
The presence and distribution, or absence of certain morphological ele-
ments - barrier top dunes, high level cusps and/or supratidal terraces -
appear crucial in controlling the pattern of barrier crest and back barrier
sedimentation. On the Tacumshin section of the Irish barriers, the spacing
of contemporary washover fans appears related to the incidence of high
level beach morphology of a periodic nature (Orford and Carter, 1984).
Such structural determinism is especially important on gravel-dominated
barriers where inheritance from antecedent stages is the norm.
Barrier breaching
Gravel-dominated barriers often enclose brackish or freshwater lagoons.

Tidal passes or inlets are generally absent. As a consequence, sediment
transport on barriers is mainly onshore, with washover processes dominating.
Hayes (1979) considers that lack of tidal inlets is diagnostic of microtidal
conditions, yet this, as a rule can not obviously be extended to embrace
coarse barriers which appear on both meso- and macrotidal coasts. Absence
of tidal inlets is important, because without them the barriers are more
capable of moving steadily onshore.
Without tidal inlets, seaward discharge of back-barrier lagoons is effected
either by seepage through the barrier or by surface channel outlet. In the
case of cross-barrier seepage, discharge occur8 along the entire saturated
Fig.3. A typical transgressive barrier cross-section from Lady’s Island Lake southeast Ireland photographed in March 1980. The section
is 260 m long and 9 m hight a t the barrier crest. On the seaward (right-hand) side t h e landward-dippingwashover units have been truncated
b y beach erosion, and some of the material has been reworked into a small marine terrace formed under storm conditions. The washover
deposits comprise lenticular coarsening-upward imbricate units many of which may be traced across the entire section. Apart from slight
distal fining the units are not graded across the section. In places very large clasts have been transported within the overwash flow, pre-
sumably rolling across t h e mobile substrate. The upper units indicate a decreasing value of overwash, units pinching o u t regressively up-
barrier. In places small lobate fans have formed.
w
m
0)
384
barrier length that is in contact with the lagoon. N o distinctive or large-scale

morphological seepage forms develop, although the lower beach face may be
extensively rilled at low tide. Seepage discharge potential depends on both
textural and structural characteristics of the barrier and may be approxi-
mated by the Darcy formula, Q = AK(h/Z);where Q is discharge, A is the
longshore cross-sectional area of the barrier, K is a proportionality coeffi-
cient related t o the textural attributes of the barrier sediment and hydraulic
properties of the fluid, h is the hydrostatic head and 1 is the barrier width.
Interplay between semidiurnal tidal levels and seasonal lagoon levels deter-
mines the instantaneous hydraulic gradient and thus the ratio between land-
wardlseaward discharge.
Once the threshold seepage capacity of the barrier is exceeded due either
t o rising water level or t o falling permeability, a surface channel will form.
This may be of two types, spillover or incised. A spillover channel simply
overtops the barrier and expands out over the seaward face with minimal
incision. The seepage flow net remains undisturbed and the channel "closes"
as soon as the lagoon drawdown is sufficient. The shift t o a more major in-
cised channel would appear t o be a function of discharge; in small mixed
sediment embayments in Oregon, Clifton e t al. (1973) suggest that this
transition occurs around 0.01-1 cumecs. A value within the same order of
magnitude appears t o occur in southeast Ireland. In a survey of twenty
barrier dammed streams, with mean annual discharges ranging from 0.08 to
11.05 cumecs, Carter et al. (1984) were able t o show that the seepage t o
channel transition generally occured around 1-2 cumecs. Figure 4 suggests
a relationship between modal surface sediment size ( d ) and discharge as
d x Q2.5.
0 '"1 -200
01°C;
1,
100
Fig.4. Discharge versus grain size relationship for southeast Ireland barriers (between
Wicklow and Tacumshin). Permanent seepage outlets (solid circles), give way, with in-
creasing discharge and/or decreasing grain size t o ephemeral channels (solid squares) to
permanent channels (solid triangles).
385
Incised outlet (= dominant freshwater flow) channels may be deflected by

longshore drift and cause reworking of the barrier during migration. Closure
of incised channels tend to take place by infilling following a decline in seep-
age discharge associated with drawdown of the lagoon. Only if an outlet
becomes a tidal channel could it be maintained on a permanent basis.
Lagoon discharge by seepage removes the necessity for cross-barrier
channels and thus diminishes the role of tidalpass sediment recycling. Only a
very small proportion of the gravel-dominated barrier mass is stored in inlet
related morphology .
Seepage also plays an important part in the landward transfer of sediment
and in destabilising the barrier crest. When sea level is elevated relative to
lagoon level, significant volumes of water may pass through the barrier
emerging as surface flows on the landward side (Arkell, 1955). A t
Tacumshin, seepage flow has produced extensive back barrier channelling
and formed seepage fans along the barrier lagoon margin (Fig.5) Landward
discharges of 1 4 cumecs km-' h-' are feasible during storms (Carter et a].,
1984). Gradual excavation of seepage hollows acts to undermine the
barrier crest and may presage incipient breach points. The importance of
seepage has been recorded on other fringing barriers in the British Isles
(Arkell, 1955; Robinson, 1981; Edison, 1983).
Fig.5. Seepage and washover channels o n t h e barrier a t Tacumshin, southeast Ireland.

The seepage channel, t o t h e left, starts in an erosional hollow about 2 m below t h e beach
crest and expands i n t o a small fan. T h e washover channel, t o t h e right, is a larger feature
originating a t a breach in the barrier crest.
386
DISCUSSION
Gravel-dominated barriers form part of a barrier continuum based on

grain size, that has largely gone unnoticed and uninvestigated. Instead
efforts have concentrated on wave climate and tidal range variations to
explain barrier variation (Hayes, 1979). The barriers of the southwest
approaches to the British Isles have developed distinctive morphological
facies assemblages and planforms by virtue of the presence of coarse elements
in the sediment mass. The importance of seepage in effecting terrestrial
drainage has meant a lack of tidal passes, thus removing from the barriers
a mechanism by which material may be returned seaward. Although British
and Irish washover features are small in area when compared to many on the
US. coast (e.g. Andrews, 1970; Schwartz, 1975), they are left to dominate
the gross transport across the barriers. The inability t o achieve onshore/
offshore equilibrium means that barriers move steadily onshore. There is
no compensatory mechanism to balance the landward-directed effects of
overwash processes, while the various factors which lead to constant de-
stabilisation of the barrier crest (Carter and Orford, 1980j simply accen-
tuated any propensity to onshore transport. Lack of tidal inlet-associated
ebb deltas not only eliminates an important repository for barrier sediments
en route t o maintain the shoreface, but also rids the nearshore of significant
refraction perturbations.
Gravel-dominated barrier processes are strongly influenced by two factors.
One is the appearance of long-period swell waves in the nearshore climate
and the other is the inheritance of morphosedimentary structures into the bar-
rier through storm events. The reflective nature of gravel-dominated barrier
beaches leads to amplification of the sub-harmonic component of long-
period swell thus providing one mechanism for the extending swash run-up
up to, over and beyond the crest.
Gravel-beach structures have a high survival potential owing to their high
mass entrainment thresholds, aided and abetted by the fact that as strongly
reflective beaches, profile variability is low and only a small volume of
material moves. Thus fairweather structures (berms, terraces, swash bars)
sunrive well into storms, and act to modify the shoreline wave processes. The
presence of extensive berm terraces may, during storm surges, create condi-
tions whereby constructional spilling waves move sediment onto the barrier
crest. The comparative lack of exceptionally intense storms is important
inasmuch as there is rarely, if ever, a complete reorganisation of the barrier
morphosedimentary environments.
The persistent, if slow, landward migration of gravel-dominated barriers
probably continues even under stable sea levels, unlike the retreat of some
barrier islands discussed by Pilkey (1981). There are enough processes
capable of lowering the crest and causing overwashing of gravel-dominated
barriers without invoking the need for change in sea level.
387
CONCLUSIONS
The following conclusions may be noted:

(1)Gravel-dominated barriers constitute a distinctive morphosedimentary
environment.
(2) Gravel-dominated barriers fall within the reflective beach stage as
defined by Short (1979) and others. The amplification of wave run-up
during episodes of long period swell is one mechanism which causes over-
topping and over-washing of the barrier crest.
(3) Structural stability means that many features are inherited from ante-
cedent events. This exercises some control over back beach, barrier crest
and back barrier processes and sedimentation.
(4)Seepage through the barrier is important, both in draining back barrier
lagoons and streams, and during periods when sea levels are elevated above
lagoon levels, so that landward seepage takes place.
( 5 ) The comparative importance of seepage and surface outlet channels
leads to a lack of tidal passes, so that very little sediment is moved seaward
from the barrier. Morphology and structure may be preserved in the absence
of migrating inlets.
(6) Washover processes are the major methods by which the barrier mi-
grates. Migration is landward, and it is argued that migration may continue
even in the absence of sea level rise.
ACKNOWLEDGEMENTS
We would like to thank our respective universities and departments for

continuing field and technical support.
REFERENCES
Ali, M.T., 1976. The significance of a mid-Cretaceous cobble conglomerate, Beer Dis-
trict, South Devon. Geol. Mag., 13: 151-158.
Amin, M., 1982. On analysis and forecasting of surges o n the west coast of Britain. Geo-
phys. J.R. Astron. SOC.,68: 79-94.
Andrews, P.B., 1970. Facies and genesis of a hurricane-washover fan, St. Joseph's Island,
Central Texas Coast. Univ. Texas Bur. Econ. Geol., R.I. 6 7 , 1 6 7 pp.
Arkell, W.J., 1955. Effects of storms on Chesil Beach in November 1954. Proc. Dorset
Nat. Hist. Arch. SOC.,76: 141-145.
Armon, J.W., 1974. Late Quaternary shore lines near Lake Ellesmere Canterbury, New
Zealand. N.Z. J. Geol. Geophys., 17: 63-73.
Barnes, P.W., 1982. Marine ice-pushed boulder ridge, Beaufort Sea, Alaska. Arctic, 35:
312-31 6.
C a r , A.P., 1971, Experiments on the longshore transport and sorting of pebbles. J.
Carr, A.P. and Blackley, M.W.L., 1973. Investigations bearing on the age and development
of Chesil Beach, Dorset, and the associated area. Trans. Inst. Brit. Geogr., 58: 99-112.
Carr, A.P. and Blackley, M.W.L., 1974. Ideas on the origin and development of Chesil
Beach, Dorset. Proc. Dorset Nat. Hist. Arch. SOC.,95: 9-17.
Carter, R.W.G., 1983. Raised coastal landforms as products of modern process variations
388
and their relevance to eustatic sea level studies: examples from eastern Ireland. Boreas,
1 2 : 167-182.
Carter, R.W.G. and Orford, J.D., 1980. Gravel barrier genesis and management: a con-
trast. Proc. Coastal Zone 80, Am. SOC.Civ. Eng.: 1304-1320.
Carter, R.W.G. and Orford, J.D., 1981. Overwash processes along a gravel beach in
southeast Ireland. Earth Surf. Proc., 6: 413-426.
Carter, R.W.G., Johnson, T.W. and Orford, J.D., in press. Formation and significance of
stream outlets on the mixed sand and gravel barrier coasts of southeast Ireland. Z.
Geomorphol.
Clifton, H.E., Phillips, R.L. and Hunter, R.E., 1973. Depositional structures and processes
in the mouths of small coastal streams, southwestern Oregon. In: D.R. Coates (Editor),
Coastal Geomorphology. Publications in Geomorphology, State Univ. of New York,
Stoney Brook, N.Y., pp.115--140.
Darbyshire, J., 1977. An investigation of beach cusps in Hell’s Mouth Bay. In: M. Angel
(Editor), A Voyage of Discovery. Pergamon, Oxford, pp.405-427.
Dean, R.G., 1973. Heuristic models of sand transport in the surf zone. Proc. Conf. Eng.
Dynamics in Surf Zones, Sydney, N.S.W., pp.207-214.
Edison, J., 1983. Flandrian barrier beaches off the coast of Sussex and southeast Kent.
Q. Newslet., 39: 25-29.
Everts, C.H., 1973. Particle overpassing on flat granular boundaries. J. Water. Harbour
Div. Am. Soc. Civ. Eng., 99 (WW4): 425-438.
Fahnestock, R.K. and Houshild, W.L., 1962. Flume studies of the transport of pebbles
and cobbles on a sand bed. Geol. SOC.Am. Bull., 73: 1431-1436.
Flather, R.A., 1981. Practical surge prediction using numerical models. In: D.H. Peregrine
(Editor), Floods Due to High Winds and Tides. Academic Press, London, pp.21-43.
Glaesner, J.D., 1978. Global distribution of barrier islands in terms of tectonic setting.
J. Geol., 8 6 : 283-298.
Godfrey, P.J., Leatherman, S.P. and Zaremba, R., 1979. A geobotanical approach to the
classification of barrier beach systems. In: S.P. Leatherman (Editor), Barrier Islands.
Academic Press, New York, N.Y., pp.99-126.
Golding, B., 1981. A meteorological input to surge and wave prediction. In: D.H.
Peregrine (Editor), Floods Due to High Winds and Tides. Academic Press, London,
pp.9-20.
Guza, R.T. and Davis, R.E., 1974. Excitation of edge waves by waves incident on a beach.
J. Geophys. Res., 79: 1285-1291.
Graff, J., 1981. An investigation of the frequency distributions of annual sea level
maxima at Ports around Great Britain. Estuarine Coastal Shelf Sci., 1 2 : 289-349.
Hails, J.R., 1975. Some aspects of the Quaternary history of Start Bay, Devon. Field
Stud., 4: 207-222.
Hogben, N. and Lumb, F.E., 1967. Ocean Wave Statistics. H.M.S.O. London, 274 pp.
Huntley, D.A., Guza, R.T. and Bowen, A.J., 1977. A universal form for shoreline run-up
spectra? J. Geophys. Res., 82: 2577-2581.
Kaufman, W. and Pilkey, O.H., 1979. The Beaches are Moving. Anchor Press, New York,
N.Y., 326 pp.
transgression. Geol. SOC.Am. Bull., 82: 2131-2158.
Leatherman, S.P., 1979a. Barrier Islands. Academic Press, New York, N.Y., 325 pp.
Leatherman, S.P., 1979b. Barrier dunes: a reassessment. Sediment. Geol., 24: 1-16.
Leckie, D.A. and Walker, R.W., 1982. Storm and tide dominated shorelines in Cretaceous
MoosebarLower Gates interval - Outcrop equivalents of Deep Basin Gas Trap in
Western Canada. Bull. Am. Assoc. Pet. Geol., 66: 138-157.
Lowry, P. and Carter, R.W.G., 1982. Computer simulation and delimitation of littoral
drift cells on the south coast of Co. Wexford, Ireland. J. Earth Sci. R. Dublin SOC.,
4 : 121-132.
389
McCubbin, D.G., 1982. Barrier-island and strand-plain facies. In: P.A. Scholle and D.A.
Spearing (Editors), Sandstone Depositional Environments. Am. Assoc. Pet. Geol.
Tulsa, Okla., pp.247-279.
Middleton, G.V., 1970. Experimental studies related to problems of flysch sedimentation.
Geol. Assoc. Can., Spec. Publ., 7: 253-272.
Morey, C.R., 1983. Barrier stability. Q. Newslet., 40: 23-27.
Moslow, T.F. and Heron, D., 1981. Holocene deposition and history of a microtidal
cuspate foreland cape: Cape Lookout, North Carolina. Mar. Geol., 41: 251-270.
Moss, A.J., 1962. The physical nature of common sandy and pebbly deposits. Part I.
Am. J. Sci., 260: 337-373.
Orford, J.D. and Carter, R.W.G., 1982a. Crestal overtop and washover sedimentation on
a fringing sandy gravel barrier coast, Carnsore Point, Southeast Ireland. J. Sediment.
Petrol., 52: 265-278.
Orford, J.D. and Carter, R.W.G., 1982b. Geomorphological changes on the barrier coasts
of south Wexford. Ir. Geogr., 1 5 : 70-84.
Orford, J.D. and Carter, R.W.G., 1984. Models of longshore washover spacing on coarse
clastic barriers. Mar. Geol., 56: 207-226.
Orford, J.D. and Carter, R.W.G., 1985. Storm-generated dune armouring o n a sand-gravel
barrier system, southeastern Ireland. Sediment. Geol., 42 (in press).
Pilkey, O.H., 1981. Geologists, engineers, and a rising sea level. Northeast. Geol., 3:
150-158.
Pugh, D.T. and Vassie, J.M., 1978. Extreme sea levels from tide and surge probability.
Proc. 16th Conf. Coastal Engineering, Am. Soc. Civ. Eng., pp.911-930.
Rampino, M.R. and Sanders, J.E., 1981. Evolution of the barrier islands of southern Long
Island, New York. Sedimentology, 28: 37-48.
Randall, R.E., 1977. Shingle foreshores. In: R.S.K. Barnes (Editor), The Coast. Wiley,
London, pp. 4 9 - 6 1 .
Robinson, I.S., 1981. Salinity structure and tidal flushing of the Fleet. In: M. Ladle
(Editor), The Fleet and Chesil Beach. Dorset County Council, Dorchester, pp.33-38.
Sallenger, A.H. 1979. Inverse grading and hydraulic equivalence in grain-flow deposits.
J. Sediment. Petrol., 49: 553-562.
Schwartz, R.K., 1975. Nature and genesis of some storm washover deposits. U.S. Army
Corps of Eng., Tech. Memo 61, 70 pp.
Short, A.D., 1979. Three dimensional beach stage model. J. Geol., 87: 553-571.
Suhayada, J.N. and Roberts, H.H., 1977. Wave action and sediment transport on fringing
reefs. Proc. 3rd Int. Coral Reef Symp., 1: 65-70.
Wang, Y. and Piper, D.J.W., 1982. Dynamic geomorphology of the Drumlin coast of
southeast Cape Breton Island. Marit. Sediment. Atlant. Geol., 18: 1-28.
Wright, L.D., 1982. Field observations of long period surf zone oscillations in relation to
contrasting beach morphologies. Aust. J. Mar. Freshwater Res., 33: 181-201.
Wright, L.D., Chappell, J., Thom, B.G., Bradshaw, M.P. and Cowell, P., 1979. Morpho-
dynamics of reflective and dissipative beach and inshore systems: southeastern
and surf zones: A brief synthesis. Coastal Studies Unit, Univ. of Sydney, Rep. 82/5,
6 4 PP.
Wright, M.E. and Walker, R.G., 1981. Cardium Formation (Upper Cretaceous) and
Seebee, Alberta - storm transported sandstones and conglomerates in shallow marine
depositional environments below fairweather wave base. Can. J. Earth Sci., 18:
795-809.
Zenkovich, V.P., 1967. Processes of Coastal Development. Oliver and Boyd, Edinburgh,
738 pp.
Marine Geology, 60 ( 1 9 8 4 ) 391-412 391
SHOREFACE TRANSLATION AND THE HOLOCENE STRATIGRAPHIC

RECORD: EXAMPLES FROM NOVA SCOTIA, THE MISSISSIPPI DELTA
A N D EASTERN AUSTRALIA
RON BOYD and SHEA PENLAND

Geology Department, Dalhousie University, Halifax, N.S. B3H 3J5 (Canada)
Louisiana Geological Survey, BOX G , University Station, Baton Rouge, LA 70803
(U.S. A .)
(Received April 11, 1983; revised and accepted October 7 , 1983)
ABSTRACT
Boyd, R. and Penland, S., 1984. Shoreface translation and the Holocene stratigraphic
record: Examples from Nova Scotia, the Mississippi Delta and eastern Australia. In:
Wave-Dominated Coastal Environments. Mar. Geol., 60: 391-412.
Classic descriptive models of barrier sedimentation have been developed with data
from t h e Atlantic and Gulf coasts of t h e United States. These models are dominated by
low t o moderate rates of relative sea level (RSL) rise and wave energy. Barriers respond by
landward recycling of sediment through t h e mechanism of shoreface retreat. Sedirnenta-
tion processes o n t h e central coast of New South Wales (N.S.W.), Australia, consist of
rapid RSL rise in early Holocene times followed by a stillstand since 6500 B.P. Wave
energy is relatively high year-round and sand sources for barrier formation are only
found o n the inner shelf. Barrier sedimentation o n the central coast of N.S.W. exhibits a
thick, composite sequence composed of a basal marine transgressive sand overlain by
regressive beach and d u n e facies.
T h e Louisiana coast surrounding t h e Mississippi delta is underlain by compacting
deltaic muds which generate very rapid rates of RSL rise. The Louisiana coast expe-
riences low wave energy punctuated b y high-energy tropical and extra-tropical storm
events. Barrier sediments accumulate from t h e erosion of deltaic headlands and undergo
a transformation from subaerial barrier island systems to subaqueous shoals located on
th e inner shelf. Drumlins experience coastal erosion on the Eastern Shore of Nova Scotia
and provide a sediment source for compartmented estuary m outh barriers. An ongoing,
moderate rise of RSL results from t h e passage of a glacial forebulge. Wave energy is inter-
mediate between Louisiana and N.S.W. and displays a seasonal pattern dominated by
frequent winter storms. Coastal barrier sedimentation is episodic, consisting of a period
of beach ridge progradation followed by barrier destruction and re-establishment further
landward.
The three contrasting sedimentary sequences found in examples from Louisiana,
N.S. W. and Nova Scotia indicate that presently available sedimentation models from
locations such as the middle Atlantic or Texas coasts of the United States may only
represent well-documented regional case studies. A true generalised coastal sedimentation
model is required which can identify t h e parameters controlling vertical and horizontal
translation of t h e depositional surface and provide relationships between these parameters
which quantitatively predict the genesis, distribution and geometry of coastal
sedimentary facies.

392
INTRODUCTION
A complex set of parameters control shoreface translation and the genera-

tion of coastal facies during sea-level regressions and transgressions. Recent
reviews indicate that existing coastal facies models are descriptive and repre-
sent generalisations based on a few well-documented case studies (Reading,
1978; Reinson, 1979). The next step in understanding coastal sedimentation
is to identify the dominant factors which control the vertical and horizontal
translation of the depositional surface. Once identified, relationships be-
tween these parameters can be established which are capable of quantitatively
predicting the geometry and distribution of coastal stratigraphic sequences
in modern and ancient environments.
Initial contributions by Sloss (1962) and Allen (1964) identified the
concepts of facies generation during sea-level regression and transgression.
Sloss (1962) expressed the shape of a body of sedimentary rocks as a func-
tion of the quantity (Q) of material supplied to the depositional site, the
rate of subsidence ( R )at the site, the rate of sediment dispersal ( D ) from the
site and the nature of the materials supplied ( M ) , or
shape = f(Q, R, D,M)
Allen (1964) derived geometrical expressions for the thickness of transgres-
sive and regressive layers under conditions of variable sediment supply (Q)
and variation in relative sea level depending on the basinward subsidence
from a clastic wedge hinge line. Transgressive units were found to be com-
monly thinner than regressive units and their thickness decreased with in-
creasing rates of transgression. Swift et al. (1972) grouped Sloss’s variables
in quasi-quantitative fashion
(~),-R=K
and replaced D with the concept of energy input ( E ) ,indicating that, for the
coastline position to remain constant ( K ) ,the ratio of sediment input to the
energy available for its dispersal must be balanced by an equivalent change
in relative sea level. Curray (1964) was able to plot a conceptual graph of
erosion and deposition versus change in relative sea level and hence classify
eight types of transgression and regression. Examples of variation in Q and
R were given but no attempt to quantify the variables was attempted. From
these earlier investigations, the two most significant parameters to emerge in
controlling coastal sedimentation were the rate of sediment input and
the relative sea level. Physical processes responsible for sediment dispersal
had not received widespread attention. Swift (197 5) restated the prob-
lem in the form of the sediment continuity equation
393
where the time rate of change of the coastal sediment-water interface

a h p t is proportional to the time rate of change of sediment concentration
in a unit volume a c / a t plus the net difference of sediment advected into and
out of the unit volume. Here C represents sediment concentration, V a
velocity vector, p sediment density, and d water depth. Swift (1975) also
identified the processes responsible for the sediment flux as: (1)wave driven
longshore currents; ( 2 ) a coast-parallel wind driven flow; and (3) wind and
wave setup driving shore normally oriented upwelling and downwelling
events. A more recent coastal sedimentation model (Belknap and Kraft,
1981) examined the preservation potential of transgressive coastal litho-
somes and extended the controlling factors to include pre-existing
topography, erosion resistance, and tidal range. Belknap and Kraft (1981j
concluded that the depth of shoreface erosion was related t o the rate of sea-
level rise with faster rates of sea-level rise capable of greater preservation-
the opposite conclusion t o that reached by Allen (1964).
Many studies have identified the characteristics of the shore-parallel and
shore-normal processes defined in Swift (1975). Wave-derived longshore
transport is widely accepted as controlling shore-parallel sedimentation in
the surf zone in response t o oblique wave approach (see for example Komar,
1976). Further seaward shore-parallel or shore-oblique transport derived
from wind forcing is known to occur across the shoreface (e.g. Boyd, 1980a;
Niederoda and Swift, 1981) and the continental shelf (e.g. Sternberg and
McManus, 1972) but the sedimentary responses t o this wind forcing are less
well documented. Bowen (1980) has recently summarised the processes con-
trolling shore-normal sediment dispersal on the shoreface in a revision of
Cornaglia’s (1898, in Munch-Peterson, 1938) original null point hypothesis.
Shore-normal transport here is seen as a balance between the force of gravity
and an oscillatory wave force combined with a perturbation component.
T o summarise, Sloss’s (1962) original variables (Q, R , D ,M ) seem concep-
tually appropriate for the formation of coastal stratigraphic models. Each
of the sediment supply, RSL and sediment dispersal variables is seen to
consist, in turn, of a complex subset of further variables (see for example
the list in Kraft, 1978, table 1).
Three regional examples from the central coast of New South Wales,
Louisiana and Nova Scotia are presented here to provide a wide spectrum
of contrasting case studies. For each example the variables controlling sedi-
mentation are identified. These variables range from: (1) a rising RSL
followed by stability under a moderate- to high-energy wave climate in
N.S.W.; t o (2) very rapid continuing rates of RSL rise, and low wave energy
punctuated by storm events on the low-gradient Louisiana coast; to (3)
drumlin point sources of sediment supplied to topographically controlled
estuaries by rising RSL and moderate levels of wave energy in Nova Scotia.
The resulting coastal stratigraphic record is also presented for each example
t o illustrate how contrasting process interactions can produce a spectrum of
barrier sedimentary sequences.
394
THE CENTRAL COAST OF NEW SOUTH WALES, AUSTRALIA
Relative sea level
RSL in southeast Australia rose from -130 m between 17,000 and 24,000
B.P. (Phipps, 1970) t o around -20 m at 9000 B.P. (Fig.1). RSL continued
to rise rapidly until reaching and maintaining a stillstand within a meter of
its present position by 6000 B.P. (Thom and Chappell, 1975). Using an
average distance of 1 4 km between the 130 and 20 m isobaths on this shelf
(Boyd, 1980b), the average rate of RSL rise between 17,000 and 9000 B.P.
was 1.40 m per century and the average rate of landward shoreline displace-
ment was 1.75 m yr-'. Between 9000 and 6000 B.P. the average rate of RSL
rise was 0.67 m per century and the minimum average rate of shoreline dis-
placement was around 0.35 m yr-'. Due to subsequent shifts in the shore-
face equilibrium profile and corresponding uncertainty regarding the original
location of the -20 m shoreline, rates of shoreline displacement between
9000 and 6000 B.P. are likely to have been faster by a factor of 2-3.
MSL 0
c
In
&
c
10
0)
r
u
I
I-
$ 15 SOUTHEAST AUSTRALIA
n
HOLOCENE SEA LEVEL RISE
20
25
10 8 6 4 '2 0
YEARS BP ( x i 0 3 1
Fig.1. Holocene sea level relative t o south east Australia. RSL rise was rapid prior to
establishing a stillstand within 1 m of its present position by 6300 B.P. (redrawn from
Thom and Chappell, 1 9 7 5 ) .
395
Sediment input
Since at least 9000 B.P., there appear to have been no external sediment
sources for most central N.S.W. coastal compartments except for onshore
transport from the inner continental shelf located directly seaward. This
situation arises from the inhibition of longshore sediment transport between
deeply embayed coastal compartments by shore-normal submarine bedrock
ridges and the trapping of fluvial sand input at the upstream estuary margins
of drowned river valleys. Numerous studies along the coastline north and
south from Sydney have established the presence of continued bedrock
ridges extending to the 60 m isobath from seismic data (MWSDB, 1976),
bathymetric profiles and Scuba surveys (Reffell, 1978; Hann, 1979; Boyd,
1980b). Water depths of over 20 m are commonly encountered on outcrop-
ping bedrock less than 200 m seaward of all headlands separating major
coastal compartments in the Sydney region. Adjacent coastal compartments
exhibit wide variations in sediment lithology as demonstrated by mean
grainsize statistics and calcium-carbonate content. (Hann, 1979; Boyd,
1980b). Only two central N.S.W. coastal rivers have succeeded in establish-
ing marine Holocene deltas. Of these the Hunter River does not supply sand
to the present coastline (Roy, 1975) while a minor sand source from the
Shoalhaven River (Wright, 1970; Boyd, 1980b) is contained within the Shoal-
haven Bight compartment. Headlands along the central N.S.W. coast are
mainly CarboniferousTriassic massive sandstones and volcanics (Packham,
1969) which contribute an insignificant amount of sediment to the coastal
sand budget.
Coastal processes
A high-energy wave climate and shore normal wind-forced bottom flow

are the dominant processes influencing sediment transport on the central
N.S.W. coast. The highest probability occurrence of wave height-period
combinations is for wave periods of 7-9 s and deep-water wave heights of
1-1.5 m from the southeast direction. Wave height exceeds 1 m around
80% of the time and exceeds 4 m for 1%of the time (Lawson and
Abernethy, 1975). However, the modal wave, for which the product of wave
power X frequency (see Fig. 2) is at a maximum is for waves of 2-2.5 m and
periods of 8-10 s (Wright et al., 1980) whereas the average wavepower is
2.7 X lo4 W m-'. Because of the steep, narrow continental shelf only 4%
of the deep-water incident wave power is dissipated before reaching the surf
zone of the central N.S.W. coast in contrast to 29 and 58% for comparable
east coast swell environments at Sergipe, Brazil and Cape Hatteras, North
Carolina (Wright, 1976). Boyd (1980a) found that for wind velocities in excess
of 25 km h-*, wind orientation t o the coastline controlled the near-bottom
flow field and the resulting sediment transport. Onshore winds dominantly
produced offshore bottom flows and offshore winds produced onshore flows
at the bed. The sandy barrier shoreface profile in this region (No. 3, Fig.3)
396
NEW SOUTH WALES WAVE CLIMATE
MODAL WAVE
HEIGHT 2 - 2 5 m
PERIOD 8-9sec
f AVERAGE WAVEPOWER 276 303 355

v 26,819 W/m
I-
2 26
E
I 4{
164 725
638 564 463 336

198 2 50
76 429 863 636 348 136 146

108 1228 892 607 831 181 98
W
t3 2-
31 617 1644 1286 762 445 242 196 70
13 48 228 1492 1616 1195 645 269 73

a
W
W 7 66 563 1008 1142 737 345 154 19
I-
22 130 241 225 137 51 23
o/ I I I I I I I I I I I I
0 I 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6
WAVE PERIOD (Seconds 1
Fig.2. A wave-climate matrix in which individual data values represent t h e product of

wavepower and probability of occurrence expressed in units of W m-' for t h e full range
of wave height and period combinations o n the central N.S.W. coast.
is concave up and decreases exponentially between depths of around 3 and

30 m, over a distance of less than 2 km. Gradients average 1:40-1:70 and a
corresponding decrease in mean sediment grainsize from 1.4 to 2.0 @ parallels
the increase in depth from the upper to lower shoreface.
Depositional his tory
Sedimentation along the central N.S.W. shelf is characterised by the

development of sandy bay barriers composed of beach-ridge plains and dune
complexes (Thom, 1965; Thom et al., 1978; Roy et al., 1980). Construction
of these barriers began around 6000-6500 B.P. and in most cases was over
75% complete by 3000 B.P. (Thom et al., 1981). Barriers are typically
20-30 m thick and occupy coastal compartments ranging from 1to 30 km
long. Barrier widths range from 1 to 3.3 km. Shoreline progradation slowed
during formation, from initial values of around 1.2-1.45 m yr-' to later
values of 0.5-0.25 m yr-' (Thom et al., 1981).
The pattern of barrier genesis and evolution on this coast appears to have
begun with establishment of the present RSL around 6000-6500 B.P.
During the following 3000 yrs sediment derived from the adjacent shelf and
397
MSL 0
\ @ Chandeleur Islands. Louisiana

5
10
-
I
n
L
d
a
z
v
15
\ \ o 2 Martinique Beach, Nova Scotia
I
t
W
20
n
25
30
35 I ! I I I 1 I !
0 1000 2000 3000 4000

DISTANCE OFFSHORE (Meters)
Fig. 3. Typical shoreface bathymetric profiles from. 1 : Chandeleur Islands, Louisiana; 2 :
Martinique Beach, Nova Scotia - note the boulder retreat shoal at the seaward termina-
tion of the profile; and 3 : Newcastle Bight, N.S.W. The increase in profile gradient from
top to bottom is paralleled by an increase in wave power.
shoreface was transported landward and accumulated for a finite period as

prograding bay barriers in estuary mouths. Since 3000 B.P. most barriers
have remained stable or undergone net erosion. The implication of this
evolutionary sequence is that landward rates of shoreline translation
exceeded the capabilities of landward sediment transfer in the period from
9000 t o 6000 B.P. Following the establishment of stable RSL a t 6000 B.P.,
wind- and wave-dominated sediment dispersal transformed a shallow dis-
equilibrium shoreface profile into a relatively steep, narrow profile which
achieved equilibrium with the high-energy wave climate. Excess sediment
accumulated as progradational shoreline deposits. Equilibrium was
apparently established by 3000 B.P. and major shoreline progradation on the
central N.S.W. coast ceased. A conceptual model (Fig.4) describing this barrier
evolution sequence for N.S.W. has been developed by Roy and Thom (1981).
MISSISSIPPI DELTA COASTAL BARRIER SYSTEMS, LOUISIANA U.S.A.
Relative sea level
RSL in the Gulf of Mexico is inferred to have risen from depths of -130 m,
15,000 B.P. (Curray, 1960) t o around -9 m by 8000 B.P. A subsequent rate
398
&'\2',3\ PSL
b 2 h 6 8 (0 I2 I;, I6 I8 io 22
Kilometers
BEDROCK OR PLEISTOCENE SUBSTRATE
ESTUARINE - BACKBARRIER DEPOSITS __---

--
TRANSGRESSIVE SAND ..................... . . . . . . . . . . . . . . . . . . .
...........
... :.-
.,.;,.:,..#
INNER SHELF SAND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !;<!j>..:;.::j
;.......:....I
\ \
STAGES OF HOLOCENE REDEPOSITION. .............. ix2p
DOMINANT DIRECTION OF SAND TRANSPORT .....
Fig.4. A conceptual model for barrier formation o n the N.S.W. coast (from Roy and
Thom, 1981). Two alternative mechanisms are identified t o produce stages of shoreline
progradation by landward sediment transfer from the shoreface.
of RSL rise around 20 cm per century then occurred (Fig.5) until eustatic
sea level reached its present position around 3650 B.P. (Coleman and Smith,
1964). Since this time RSL along the Mississippi Delta coastline of Louisiana
has continued t o rise in response to subsidence of the land surface. Subsi-
dence is primarily caused by the compactional subsidence of deltaic deposits
and varies as a function of sediment thickness and age. Present rates range
399
25 I I I 1 I I
10 8 6 4 2 0
YEARS BP ( ~ 1 0 3 )
Fig.5. Holocene sea level relative to the Louisiana coast. Eustatic components controlled
the rise in RSL until 3600 B.P. since when compactional subsidence has predominated
(based on Coleman and Smith, 1964, and Frazier, 1974).
from 7.5 cm per century for old, shallow-water delta complexes (Coleman
and Smith, 1964), to 60 cm per century for intermediate age deltas, to
600 m per century for the presently active deep-water Balize Delta (Kolb
and Van Lopik, 1958). Rates of shoreline retreat for abandoned delta
complexes exhibit a range from less than 1 to over 15 m yr-' with extreme
rates of over 50 m yr-' recorded in 1979 (Penland and Boyd, 1981), reflect-
ing the extremely low-gradient characteristic of deltaic coastlines.
Sediment input
Sediment has been supplied to the southeast Louisiana coast by the

Mississippi River in a series of spatially and temporally varying deltaic depo-
centres (Frazier, 1967). The only Holocene sand-sized sediment suitable for
coastal barrier formation is supplied from digitate or arcuate distributary
mouth bars (Fisk, 1955). Following upstream diversion, delta complexes
become abandoned and ongoing subsidence induces sea-level transgression.
During transgression the only source of sand for barrier formation is from
reworking (see Fig. 6) of distributary mouth bars and pre-existing shoreline
sand bodies by shoreface retreat (Swift, 1975). As subsidence and transgres-
sion continue, shoreface retreat becomes progressively less effective until
400
EROSIONAL HEADLAND STAGE

Shoreface
0 MSL
10
15 m
Overwash
INNER SHELF SHOAL STAGE Shoreface
0 MSL
10
15 rn
0 KM 5 10 15 20 25 30 35
REGRESSIVE FACIES TRANSGRESSIVE FACIES
Distributary Sand Barrier
Delta Front/Prodelta Marsh
Sediment Dispersal Pathway -+ Lagoon
Fig.6. Sediment input to coastal barrier systems of the Mississippi Delta is controlled by
the processes of subsidence and erosional shoreface retreat. The original sand source lies
in distributary and pre-existing barrier deposits. During transgression this source is re-
worked to form new barrier deposits by erosional shoreface retreat until ongoing com-
pactional subsidence removes the sand source below the base of the shoreface resulting
in eventual barrier submergence.
401
the base of the shoreface no longer penetrates to the underlying deltaic

sand source. Barriers are then decoupled from their only sediment input,
and subaerial barrier volume continuously decreases as sand is lost to sedi-
ment sinks located in spits, tidal inlets, washovers and the inner shelf
(Penland and Boyd, 1982).
Coastal processes
The Louisiana Gulf Coast is a storm-dominated environment and energy

levels resulting from wind and wave processes are low except for the winter
passage of cold front systems and the summer occurrence of hurricanes and
tropical storms. Modal wave conditions in deep water offshore from the
Mississippi Delta are characterised by wave heights of 1m and wave periods
of 5-6 s (Fig.7). On average, modal wave conditions occur for 4% of the
time (Bretschneider and Gaul, 1956) and average deep-water wave power is
only 1.8 X lo3 W rn-l Wright et al. (1974) indicated that over 99% of the
deep-water wave power offshore from the Mississippi Delta is dissipated
before reaching the shoreline. Locally generated high-energy storm wave
conditions are usually accompanied by strong onshore winds. Murray (1970,
LOUISIANA WAVE CLIMATE

7 ,
MODAL 'WAVE
HEIGHT 0.9-1.2m
6-
L
u)
PERIOD 5- 6 sec.
0)
f
0 AVER AGE WAVE POW E R
I 5-
v
1,843 W/m
t-
I
; 4-
59
w I50
5a 3-
95
120 27
w 97 79
t- 10 83 59 8
9 2- 50 192 5 5
Q
w 208 96 3 5
w 58 231 19 4 3 5 2
cl I-
133 13 II 4 6 3
01 1 1 I I I I I I I I I I 1 I I
0 I 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6
WAVE PERIOD (Seconds)
Fig.7. A wave climate matrix for Mississippi Delta barrier coasts (same method as Fig.2).
Low probabilities of occurrence result from the variable directions of locally generated
waves during hurricanes and winter fronts. Values shown represent waves propagating
onshore only.
402
1972) recorded mainly offshore and longshore-directed bottom currents

both during the passage of a moderately energetic front in 20 m depth east
of the delta and during Hurricane Camille. Shelf sediment surveys conducted
by Krawiec (1966) identified the deposition of a large sand sheet seaward
from Grand Isle between the 5 and 1 2 m isobaths supplied by offshore trans-
port during Hurricane Betsy. In contrast, wave-refraction analyses for the
Louisiana coast (Penland and Boyd, 1982) found significant sediment
transport in non-stonn conditions was restricted to the upper shoreface
landward of the 5 m isobath. The base of the shoreface lies at a depth of
5-8 m seaward from most Louisiana barrier islands and exhibits a flattened
gradient of 1:3000 (Fig.3). Mean sediment size decreases from 2.6 $I at the
shoreline to 3.9 @ at the base of the shoreface before passing into prodelta
or shelf muds. Tides in the region are mixed and predominantly diurnal
with a microtidal range of 30 cm. Tidal inlets develop as the delta lobe
subsides, exchanging the gradually increasing tidal prism volumes between
subsiding intra-deltaic lagoons and the Gulf of Mexico. Storm surges
frequently accompany tropical cyclones and winter fronts, which generate
storm surge elevations of 0.5 to over 8 m with concurrent intense ovenvash
events (Boyd and Penland, 1981).Dominant wave approach is from the south-
east but the local wind wave variability combined with a complex coastal orien-
tation results in uninterrupted longshore transport within most barrier systems.
Depositional history
Abandoned delta complexes undergoing transgression on the Louisiana

coast are characterised by sandy barrier system genesis and subsequent
evolution (Fig.8) through a well-defined three-stage cycle (Penland et al.,
1981). Distributary sand bodies reworked by shoreface retreat provide
sediment from an erosional headland source for longshore transport into
accumulating flanking spits and barrier islands (Stage 1).Subsidence of the
delta plain behind the barriers forms an intra-deltaic lagoon separating the
retreating transgressive barrier island arc (Stage 2) from the mainland. With
continued transgression and loss of the deltaic sand source, barrier migration
cannot keep pace with subsidence and lagoon shoreline retreat. The resulting
subaqueous inner shelf shoal forms the final phase, Stage 3 of the evolu-
tionary model. The presence of Mississippi Delta complexes of varying Holo-
cene age has produced barrier systems in all stages of evolution on the
present Louisiana coast. Typical time scales for the transition from one
evolutionary stage t o another lie between 600 and 1600 yrs.
EASTERN SHORE OF NOVA SCOTIA, CANADA
Relative sea level
Holocene RSL in Atlantic Canada, has been dominated by advance and

ablation of continental ice masses. Large volumes of water are withdrawn
SUBAQUEOUS
DELTA-FRONT
WMWOEOUS REWORKED M*Ry( WAVf ---c SEDIMENT DISTRIBUTARY LEVEE

T I E _r DISPERSAL PATHS
Fig.8. The transgressive depositional history of an abandoned Mississippi River Delta can be explained in three distinct evolutionary IP
stages generated by abandonment, subsidence and marine reworking. An abandoned delta complex is successively transformed from an
erosional headland with flanking barriers (Stage 1) to a transgressive barrier island arc (Stage 2 ) and finally to an inner shelf shoal
(Stage 3).
404
from ocean basins as ice thickness increases on land, causing a fall in RSL.
Areas under thick continental ice accumulation also experience an isostatic
sinking due t o the extra crustal loading. Areas such as Nova Scotia at the ice
margin experience less RSL fall due t o the development of a glacial fore-
bulge (Quinlan and Beaumont, 1981). Following ice recession the glacial
forebulge migrates inward through the marginal ice zone and initially causes
a fall in RSL followed by a rise after passage of the forebulge crest. A glacial
advance reached its maximum in Nova Scotia between 32,000 and 12,000
yrs B.P. During subsequent forebulge migration RSL on the Eastern Shore
fell t o a minimum of at least -27 m around 8000-12,000 yrs B.P. and since
that time has risen continuously (Fig. 9) at an average rate of 35 cm per cen-
tury (Scott and Medioli, 1982). Rates of shoreline retreat associated with this
transgression have averaged around 1 m yr-I but, as detailed below, tempo-
rary reversals and wide fluctuations have been common.
Sediment inp u t
Sea-level transgression on the Eastern Shore of Nova Scotia has confined

post-glacial, fluvial sediment supply to freshwater lakes and the upper
MSL 0
EASTERN SHORE NOVA SCOTIA

HOLOCENE SEA LEVEL RISE
5
-
tn
s
c
10
Q)
5
v
I
I-
% 15
0
20
25
YEARS BP ( x i 0 3 1
Fig.9. Holocene sea level relative t o t h e Eastern Shore, Nova Scotia. A regression prior
t o 7000 B.P. was followed by a n ongoing transgression resulting in rates of RSL rise of
around 3 5 cm per century. Unpublished and published d a t a courtesy D.B. Scott (Scott
and Medioli, 1982).
405
reaches of glacial estuaries. Longshore sediment transport is also ineffective

on this coast because of the highly irregular shoreline orientation and the
recessed nature of most estuary mouth barriers. Sediment input to the coastal
zone is derived almost exclusively from marine erosion of unconsolidated
glacial shoreline deposits.
Throughout the Eastern Shore of Nova Scotia resistant bedrock lithologies
of the Cambro-Ordovician Meguma Group are covered by 3-4 m thick
accumulations of glacial till (Stea and Fowler, 1979) except where drumlin
fields occur. Drumlins consist of glacial till and glaciofluvial sediments com-
posed of a wide range of grain size. Drumlin fields often contain several
hundred individuals with average sediment volumes in the order of 7.5 X
106m3.Major barrier systems on the Eastern Shore are restricted to locations
where the transgressing shoreline intersects a drumlin field. Cliff retreat
rates along exposed drumlins average 1-5 m yr-', supplying sediment
volume at the rates of between 5 X 10' and 7.5 X 104m3 yr-'. Sediment
supply from individual drumlins displays a skewed Gaussian distribution
with typical time scales of several hundred years for the total contribution
of any one drumlin.
Coastal processes
The Eastern Shore of Nova Scotia is located in an east coast swell environ-
ment (Davies, 1980) which experiences frequent winter storm conditions.
Data collected during the Canadian Marine Environmental Data Service's
Wave climate Study for locations on the Nova Scotian Atlantic coast show
that the most frequently occurring wave conditions are for 0.6-0.9 m
height and 8-9s period. The product of frequency and wave power is
greatest for modal wave conditions of 1.2-1.5 m height and 9-10 s waves
whereas the average year-round wave power is 1.68 X lo4 W m-l (Fig. 10).
Winter storms are frequently accompanied by strong onshore winds but
no detailed data on the near-bottom flow field have been collected.
Coastal sand accumulation on the Eastern Shore generally forms highly
embayed estuary mouth barriers with little sediment exchanged alongshore
between individual compartments. The location of barriers across the mouths
of large estuaries and the mesotidal (around 2 m) range results in the
development of tidal inlets and extensive flood tide delta systems which
often contain more sediment than the barrier itself.
The shoreface seaward of Eastern Shore barriers is frequently terminated
by till or bedrock, creating a ponded sediment accumulation (see Fig.3).
The base of the shoreface commonly is located at the 15-20 m isobath less
than 2000 m seaward of the shoreline, creating a gradient of 1:100 (Fig.3).
Sediment size ranges from fine to medium-grained sand (1-2.5 4 ) at the
beach t o fine-grained sand and some silt (mean size 2.8-3.5 4 ) at -20 m on
the shelf.
406
NOVA SCOTIA WAVE CLIMATE

7 ,
MODAL WAVE 155 169
HEIGHT 1.2-1.5m I28 I52
-,n 6 PERIOD 9-10 sec I I4 136 257
L
Q)
c 131
AVERAGE WAVE POWER 81 90 I07
5-
u 16,790 W/m I50
I-
I 55 108
CI
83 185 102 223 121 131
ii 4-
35 70 156 137 95 161
3:
W 25 103 58 130 144 203 257 145
4
(L
3-
36
88
42
67
123
213
160
213
93
236
129
52
280
168
119
121
50
w 19 68 131 106 451 131 184 175 263 31

5
a
2-
8
37
62
51
56
70
79
124
217
417
352
181
209
259
327
230
332
56
15 29
W 23 70 75 152 378 615 247 190 88 17
w 5 78 135 83 139 428 581 258 128 78
I-
2 25 57 78 88 185 335 272 107 77 76 48
3 3 3 8 13 31 68 81 30 26 19 30
01 1 I L I 1 I I I I I I I 1 I I
0 I 2 3 4 5 6 7 8 9 10 I1 12 13 1415 16
WAVE PERIOD (Seconds)
Fig.lO. Wave-climate matrix for the Atlantic coast of Nova Scotia. Data values represent
the product of wavepower and its probability of occurrence expressed in units of
W rn-'. Nova Scotian wavepower values are intermediate between those of N.S.W. and the
Mississippi Delta.
Depositional history
Nova Scotia coastal sedimentation has been dominated by repeated glacia-

tions (Fig.11, Stage l).The resulting coastal environment consists of thin
lodgement - ablation till and thicker drumlin till overlying a resistant
Meguma Group bedrock erosion surface of variable relief.
Sea-level transgression following passage of the glacial forebulge crest
resulted in exposure and reworking of till, formation of estuaries in base-
ment depressions (Fig.11, Stage 2) and concentration of sand and gravel-
sized glacial sediment in estuary mouth barriers (Fig.11, Stage 3). The
subsequent depositional history of the Eastern Shore has been a cyclic repe-
tition of barrier retreat, destruction and re-establishment further landward
(Fig.11, Stages 4-6).
This Eastern Shore example shows the input of a limited, spatially variable
sediment supply t o an ongoing regional transgression. Sediment temporarily
accumulates in compartment barriers while supply exceeds RSL rise. Fol-
lowing depletion of the sediment source, barrier systems are destroyed and
migrate landward. Seismic surveys of the inner shelf reveal only a thin
(1-2 m) sand veneer overlying till at the location of former barriers. This
2. SEA L E V E L RISE AND ESTUARY FORMATION 3. BARRIER GENESIS AND PROGRADATION 4. BARRIER RETREAT
~
SUBMERGED MORAINES
CONTINENTAL
GLACIER
====
I
1. CONTINENTAL GLACIER AND ICE SHELF 6. BARRIER RE- ESTABLISHMENT 5. BARRIER DESTRUCTION
MARSH OCEAN AND ESTUARY BARRIER MIGRATION PATH

BOULDER RETREAT SHOAL SUBTIDAL ESTUARINE SEDIMENT RON BOYD
DRUMLIN
DAL GEOLOGY
GLACIAL TILL SUBAERIAL BARRIER SUBAQUEOUS BARRIER
(SHOREFACE,TIDAL DELTAS)
Fig.11. Coastal sedimentation in Nova Scotia can be considered as a 6-stage transgressive sequence of barrier genesis, destruction and re- $
establishment dominated by localised glacial sediment sources, rising RSL and tidal inlet processes. Initial barrier genesis (Stages 1 and 2)
was related to advance and ablation of continental ice sheets. Subsequent evolution consists of a cyclic repetition of Stages 3--6.
408
indicates conservation of barrier sediment during transgression up the es-

tuary. Landward sediment transfer appears to be accomplished primarily by
transport through tidal inlets into flood tide deltas, and transport across the
barrier by ovenvash and aeolian processes. While under the influence of wave
action on the upper shoreface, the submerged barrier moves landward as a
series of subtidal and intertidal shoals until encountering a new set of head-
land anchor points which provide the stability for barrier re-establishment.
SUMMARY AND CONCLUSIONS
In qualitative terms, the problem of coastal sedimentation during trans-

gression and regression is determined by the parameters which control the
shape of the shoreface profile and the factors which result in profile trans-
lation. Vertical profile translation ( u ) is controlled by RSL. The vertical
profile translation may be taken as the vertical translation of an origin fixed
at the intersection of the mean sea level and the regional land surface. The
vertical position of the origin is determined by RSL since marine processes
are incapable of elevating sediments far above high tide level.
The horizontal profile translation may result from net vertical changes
within the profile and both net horizontal changes of and within the profile.
Profile shape and its horizontal translation ( u ) is controlled by regional
gradient (0 ) and the equilibrium relationship between sediment flux (Q)
across the profile, a wave climate parameter such as H(wT)-' and the inner
shelf oceanography OLs (primarily a result of wind and tidal forcing). This
may also be stated as (see Fig.12):
u = f (RSL)
u = f [Q,H (aT)-',0,,,8 1
Wright et al. (1982) have achieved considerable success using the dimen-
sionless parameter H ( w T)-' t o characterise surf-zone morphodynamics.
Fig. 12. Schematic representation of shoreface profile translation. The vertical compo-
nent of translation ( u ) from an origin at the shoreline (0)is seen t o be a function of rela-
tive sea level (RSL). The horizontal component of translation ( u ) is a function of sedi-
ment flux ( Q ) a parameter relating wave effects ( H b , local wave height; T, wave period)
and sediment response ( w ,sediment settling velocity), inner shelf oceanography (OIs)
and regional coastal gradient 0 .
409
Here H is local wave height, T wave period and w is sediment settling

velocity. It is felt that a parameter like H( wT)-' may also be suitable for shore-
face profile characterisation since H and T are dominant parameters deter-
mining wave mass transport effects (e.g. Longuet-Higgins, 1953) while w
and T strongly influence the location of initial sediment motion and hence
the base of the shoreface (see Komar and Miller, 1973; Hallermeier, 1981).
Calculations of the H(wT)-' parameter result in low values of 3-6 for
central N.S.W., intermediate values of 6-9 for Nova Scotia, and high values
of 20-25 for Louisiana. These values of H(wT)-' also correspond to the
sequence of decreasing shoreface gradient and decreasing depth to the base
of the shoreface observed (Fig.3) from N.S.W. to Louisiana. Since the maxi-
mum potential barrier thickness and the critical depth of shoreface erosion
both depend on the shoreface gradient and the depth at its base, the
H ( w T)-' parameter shows a potential for predicting barrier facies geometry.
Using the three contrasting regional examples presented here a sequence
of sedimentary processes and their resulting barrier stratigraphy is also
apparent. In Nova Scotia a continuing rise in RSL results in barrier
progradation followed by destruction. During the destruction phase estua-
rine headlands act as effective barriers to longshore transport and sediment is
moved landward by aeolian, overwash and tidal inlet processes. Sediment
remaining on the shoreface is still within the range of wave-induced transport
and most appears to move landward, leaving behind a thin shelf sand veneer.
Transgressive sedimentation on the Eastern Shore of Nova Scotia thus illus-
trates a form of Swift's (1975) shoreface retreat concept.
On the central coast of N.S.W. rapid RSL rise occurred prior to the
development of a stillstand after 6500 B.P. Sediment supplied from the
adjacent disequilibrium shelf formed regressive barrier deposits. Barrier
progradation slowed or ceased as the sediment supply diminished and the
resulting barrier sequence is one of a thin basal transgressive sand sheet over-
lain by regressive beach ridge and' dune complexes. Sediment dispersal on
the central N.S.W. coast seems to follow a pattern of rapid RSL rise during
which barrier sediments are retained on the inner shelf following the model
of Belknap and Kraft (1981). However, during stillstand, abundant wave
energy is available for landward transport of sediment from the inner shelf to
prograding barriers. This overall pattern of barrier sedimentation on the
central N.S.W. coast is therefore a form of punctuated, stepwise shoreface
retreat.
In Louisiana, extremely rapid rates of RSL rise generate barriers during
reworking of underlying deltaic sand sources. As rapid transgression con-
tinues over a very low regional gradient, the sandy barrier systems are trans-
formed into subaqueous shoals. Low wave energy in the northern Gulf of
Mexico is incapable of sediment transfer at a rate comparable t o that of
landward shoreline translation. The resulting barrier stratigraphy is a dis-
persed shelf sand lens overlying lagoonal muds (see, for example, Frazier,
1967). Coastal sedimentation processes in Louisiana therefore incorporate
reworked barrier sandbodies into the shelf stratigraphic record.
410
The presentation here of three contrasting case studies of coastal sedimen-

tation highlights some problems arising from the construction of coastal
facies models. Sufficient variability exists within the three studies to suggest
they are not fully described by presently available concepts of barrier genesis
and evolution. Future coastal sedimentation models require a more quanti-
tative approach t o firmly establish the relationships between dominant sedi-
mentary processes within a region and the resulting coastal stratigraphy.
ACKNOWLEDGEMENTS
Funding for the present study was provided through Canadian Natural
Sciences and Engineering Research Council Grant A8425, Canadian Depart-
ment of Energy, Mines and Resources Research Agreement 251 and from the
Coastal and Fluvial Geology Program of the Louisiana Geological Survey.
Useful data and discussions for the Nova Scotia case study were provided by
D.B. Scott and A.J. Bowen.
REFERENCES
Allen, P., 1964. Sedimentological models. J. Sediment. Petrol., 34: 289-293.

Belknap, D.F. and Kraft, J.C., 1981. Preservation potential of transgressive coastal litho-
somes on the U.S. Atlantic shelf. Mar. Geol., 42: 4 2 s 4 4 2 .
Bowen, A.J., 1980. Simple models of nearshore sedimentation; beach profiles and long-
80-10.
Boyd, R., 1980a. Sediment dispersal on the N.S.W. continental shelf. Proc. 17th Int.
Conf. Coastal Eng., Am. SOC. Civ. Eng., pp.1364--1381.
Boyd, R., 1980b. Sediment dispersal on the central New South Wales continental shelf.
Geology Department, University of Sydney, unpubl. Ph.D. thesis, 319 pp.
Boyd, R. and Penland, S., 1981. Washover of deltaic barriers on the Louisiana coast.
Trans. Gulf Coast Assoc. Geol. Socs., 31: 243-248.
Bretscheider, C.L. and Gaul, R.D., 1956. Wave statistics for the Gulf of Mexico off
Burrwood, Louisiana. U.S. Army Corps of Engineers, Tech. Mem. 87.
Coleman, J.M. and Smith, W.G., 1964. Late recent rise of sea level. Geol. SOC.Am. Bull.,
75: 833-840.
Curray, J.R., 1960. Sediments and history of the Holocene transgression, continental
shelf, Northern Gulf of Mexico. In: F.P. Shepard (Editor), Recent Sediments, North-
west Gulf of Mexico. Am. Assoc. Pet. Geol., Tulsa, Okla., pp. 221-266.
Curray, J.R., 1964. Transgressions and regressions. In: R.L. Miller (Editor), Papers in
Marine Geology (Shepard Commemorative Volume). McMillan, New York, N.Y.,
pp. 17 5-2 03.
Davies, J.L., 1980. Geographical Variation in Coastal Development (2nd ed.). Hafner,
Fischer, A.G., 1961. Stratigraphic record of transgressing seas in the light of sedimenta-
tion on the Atlantic coast of New Jersey. Bull. Am. Assoc. Pet. Geol., 45: 1656-1666.
Fisk, H.N., 1955. Sand facies of Recent Mississippi delta deposits. Fourth World
Petroleum Congress, Rome, Section 1-C: 377-398.
Frazier, D.E., 1967. Recent deltaic deposits of the Mississippi River: their development
and chronology. Trans. Gulf Coast Assoc. Geol. Socs., 27: 287-315.
Hallermeier, R.J., 1981. Seaward limit of significant sand transport by waves: An annual
zonation for seasonal profiles. Coastal Engineering Research Centre, US. Army Corps
of Engineers, Technical Aid 81-2.
41 1
Hann, J., 1979. Offshore bathymetry and sediment patterns off Sydney beaches. School
of Earth Science, Macquarie University, unpubl. B.S. (Hons.) thesis.
Hoyt, J.H., 1967. Barrier island formation. Geol. SOC.Am. Bull., 7 8 : 1125--1136.
Kolb, C.R. and Van Lopik, J.R., 1958. Geology of the Mississippi River Deltaic Plain,
Southeast Louisiana. Tech. Rep. 3-483. U.S. Army Corps of Engineers Waterways
Experiment Station, Vicksburg, La., 120 pp.
N.J., 429 pp.
Komar, P.D. and Miller M.C., 1973. T h e threshold of sediment movement under oscil-
latory water waves. J. Sediment. Petrol., 4 3 : 1101-1110.
Kraft, J.C., 1978. Coastal stratigraphic sequences. In: R.A. Davis (Editor), Coastal Sedi-
mentary Environments. Springer, New York, N.Y., 4 2 0 pp.
Krawiec, W., 1966. Recent sediment of the Louisiana inner continental shelf. Rice
University, unpubl. Ph.D. thesis, 50 pp.
Lawson, N.V. and Abernethy, C.L., 1975. Long term wave statistics off Botany Bay.
Second Australian Conf. on Coastal and Ocean Engineering, The Institute of Engineers,
7512: 167-176.
Longuet-Higgins, M.S., 1953. Mass transport in water waves. Philos. Trans. R. SOC.
London, Ser. A, 2 4 5 : 535-581.
Munch-Peterson, 1938. Littoral drift formula. U.S. Army Corps Eng., Beach Erosion
Board Bull. 4 ( 4 ) ( 1 9 5 0 ) : 1-36.
Murray, S.P., 1970. Bottom currents near the coast during hurricane Camille. J. Geophys.
Res., 7 5 ( 2 4 ) : 4579-4582.
Murray, S.P., 1972. Observations on wind, tidal and density driven currents in the
vicinity of the Mississippi River d e l t a In: D.J.P. Swift, D.B. Duane and O.H. Pilkey
(Editors), Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson and
Ross, Stroudsburg, Pa.
M.W.S.D.B. (Metropolitan Water Sewerage and Drainage Board), 1976. Report o n sub-
marine outfall studies. M.S.W.D.B., Sydney, N.S.W., 2 8 8 pp.
Niedoroda, A.W. and Swift, D.J.P., 1981. Maintenance of the shoreface by wave orbital
currents and mean flow: observations from the Long Island coast. Geophys. Res.
Lett., 8 ( 4 ) : 337-340.
Packham, G.H., 1969. The geology of New South Wales. J . Geol. SOC. Aust., 1 6 ( 1 ) :
6 5 4 pp.
Penland, S. and Boyd, R., 1981. Shoreline changes on t h e Louisiana barrier coast. IEEE,
Oceans '81: 209-219.
Penland, S.P. and Boyd, R., 1982. Assessment of geological and human factors respon-
sible for Louisiana coastal barrier erosion. In: D.F. Boesch (Editor), Proceedings of
the Conference on Coastal Erosion and Wetland Modification in Louisiana: Causes,
Consequences and Options. US. Fish and Wildlife Services, Biological Services Pro-
gram, Washington, D.C.,FWS/OBS - 8 2 / 5 9 , pp. 14-39.
Penland, S., Boyd, R., Nummedal, D. and Roberts, H.H., 1981. Deltaic barrier develop
ment o n the Louisiana coast. Trans. Gulf Coast Assoc. Geol. Socs., 31 (Suppl.):
47 1-476.
Phipps, C.V.G., 1970. Dating of eustatic events from cores taken in the Gulf of Carpen-
taria, and samples from the New South Wales continental shelf. Aust. J. Sci., 32:
329-330.
Quinlan, G. and Beaumont, C., 1981. A comparison of observed and theoretical post-
glacial sea levels in Atlantic Canada. Can. J. Earth Sci., 1 8 : 1146-1163.
Rampino, M.R. and Sanders, J.E., 1980. Holocene transgression in south central Long
Island. J. Sediment. Petrol., 5 0 : 1063-1079.
Rampino, M.R. and Sanders, J.E., 1981. Evolution of the barrier islands of southern Long
Island, New York. Sedimentology, 2 8 : 37-48.
Reading, H.G., 1978. Sedimentary Environments and Facies. Elsevier, New York, N.Y.,
557 pp.
412
Reffell, G., 1978. Descriptive analysis of the subaqueous extensions of subaerial rock
platforms. Geography Department, University of Sydney, unpubl. B.A. (Hons.) thesis.
Reinson, G.E., 1979. Facies models 6 : barrier island systems. In: R.G. Walker (Editor),
Facies Models. Geosci. Can., Reprint Ser. 1,Toronto, Ont., pp.57-75.
Roy, P.S., 1975. Does the Hunter River supply sand to the New South Wales coast
today? J. Proc. R. SOC.N.S.W., 110: 17-25.
Roy, P.S. and Thom, B.G., 1981. Late Quaternary marine deposition in New South Wales
and southern Queensland - an evolutionary model. J. Geol. SOC.Aust., 28: 471-489.
Roy, P.S., Thom, B.G. and Wright, L.D., 1980. Holocene sequences on an embayed high
energy coast: a n evolutionary model. Sediment. Geol., 26: 1-19.
Scott, D.B. and Medioli, F.S., 1982. Micropaleontological documentation for early Holo-
cene fall of relative sea level o n the Atlantic coast of Nova Scotia. Geology, 10:
2 7 8-2 8 1.
Sloss, L.L., 1962. Stratigraphic models in exploration. J. Sediment. Petrol., 32: 415-422.
Stea, R.R. and Fowler, J.H., 1979. Minor and trace element variations in Wisconsinan
tills, Eastern Shore, Nova Scotia. Nova Scotia Department of Mines and Energy,
Pap. 79-4, 3 0 pp.
Sternberg, R.W. and McManus, D.A., 1972. Implications of sediment dispersal from long-
term, bottom current measurements o n the continental shelf off Washington. In:
D.J.P. Swift, D.B. Duane and O.H. Pilkey (Editors), Shelf Sediment Transport: Process
and Pattern. Dowden, Hutchinson and Ross, Stroudsburg, Pa., pp.181--194.
Swift, D.J.P., 1975. Barrier island genesis: evidence from the central Atlantic shelf,
eastern U.S.A. Sediment. Geol., 1 4 : 1-43.
Swift, D.J.P. and Moslow, T.F., 1982. Holocene transgression in South Central Long
Island, New York - Discussion. J. Sediment. Petrol., 52(3): 1014-1019.
Swift, D.J.P., Kofoed, J.W., Saulsbury, P.J. and Sears, P., 1972. Holocene evolution of
the shelf surface, certral and southern Atlantic shelf of North America. In: D.J.P.
Swift, D.B. Duane and O.H. Pilkey (Editors), Shelf Sediment Transport: Process and
Pattern. Dowden, Hutchinson and Ross, Stroudsburg, Pa., pp.499-574.
Thom, B.G., 1965. Late Quaternary coastal morphology of the Port Stephens - Myall
Lakes area, N.S.W. J. Proc. R. SOC.N.S.W., 98: 23-36.
Thom, B.G. and Chappell, J., 1975. Holocene sea levels relative t o Australia. Search, 6 :
90-93.
Thom, B.G., Polach, H. and Bowman, G., 1978. Holocene age structure of coastal sand
barriers in New South Wales, Australia. Report of Department of Geography, Faculty
of Military Studies, University of N.S.W., Duntroon, N.S.W., 8 6 pp.
Thom, B.G., Bowman, G.M. and Roy, P.S., 1981. Late Quaternary evolution of coastal
sand barriers, Port Stephens - Myall Lakes area, central New South Wales,
Australia. Quat. Res., 1 5 : 345-364.
Wright, L.D., 1970. T h e influence of sediment availability on patterns of beach ridge
development in the vicinity of the Shoalhaven River delta, N.S.W. Aust. Geogr., 11:
336-348.
Wright, L.D., 1976. Nearshore wave power dissipation and the coastal energy regime of
the Sydney-Jervis Bay region. New South Wales: a comparison. Aust. J. Mar. Fresh-
water Res., 27: 633-640.
Wright, L.D., Coleman, J.M. and Erickson, M.W., 1974. Analysis of major river systems
and their deltas: morphologic and process comparisons. Louisiana State University,
Coastal Studies Institute, Tech. Rep. 156.
Wright, L.D., Coffey, F.C. and Cowell, P.J., 1980. Nearshore Oceanography and Morpho-
dynamics of the Broken Bay - Palm Beach Region, N.S.W.: Implications for Offshore
Dredging. Coastal Studies Unit, Department of Geography University of Sydney,
Tech. Rep. 80/1,210 pp.
and surf zones: a brief synthesis. Department of Geography, University of Sydney,
Coastal Studies Unit, Tech. Rep. 82/5.
Marine Geology, 6 0 (1984) 413-434 413
HOLOCENE SEDIMENTATION OF A WAVE-DOMINATED BARRIER-

ISLAND SHORELINE: CAPE LOOKOUT, NORTH CAROLINA
S. DUNCAN HERON, Jr., THOMAS F. MOSLOW', WILLIAM M. BERELSON',

JOHN R. HERBERT3, GEORGE A. STEELE 1114 and KENNETH R. SUSMAN5
Department of Geology, Duke University, Durham, N C 27708 ( U . S . A . )
(Received February 7,1983; revised and accepted July 14, 1983)
ABSTRACT
Heron Jr., S.D., Moslow, T.F., Berelson, W.M., Herbert, J.R., Steele 111, G.A. and Susman,
K.R., 1984. Holocene sedimentation of a wave-dominated barrier-island shoreline:
Cape Lookout, North Carolina. In: B. Greenwood and R.A. Davis, Jr. (Editors),
Geol., 60: 413-434.
The sedimentary record of 130 km of microtidal (0.9 m tidal range) high wave energy
(1.5 m average wave height) barrier island shoreline of the Cape Lookout cuspate fore-
land has been evaluated through examination of 3136 m of subsurface samples from
closely spaced drill holes. Holocene sedimentation and coastal evolution has been a func-
tion of five major depositional processes: (1)eustatic sea-level rise and barrier-shoreline
transgression ; ( 2 ) lateral tidal inlet migration and reworking of barrier island deposits;
(3) shoreface sedimentation and local barrier progradation; (4) storm washover deposi-
tion with infilling of shallow lagoons; and (5) flood-tidal delta sedimentation in back-
barrier environments.
Twenty-five radiocarbon dates of subsurface peat and shell material from the Cape
Lookout area are the basis for a late Holocene sea-level curve. From 9000 to 4000 B.P.
eustatic sea level rose rapidly, resulting in landward migration of both barrier limbs of
the cuspate foreland. A decline in the rate of sea-level rise since 4000 B.P. resulted in
relative shoreline stabilization and deposition of contrasting coastal sedimentary se-
quences. The higher energy, storm-dominated northeast barrier limb (Core and Portsmouth
Banks) has migrated landward producing a transgressive sequence of coarse-grained,
horizontally bedded washover sands overlying burrowed to laminated back-barrier and
lagoonal silty sands. Locally, ephemeral tidal inlets have reworked the transgressive
barrier sequence depositing fining-upward spit platform and channel-fill sequences of
cross-bedded, pebble gravel to fine sand and shell. Shoreface sedimentation along a
portion of the lower energy, northwest barrier limb (Bogue Banks) has resulted in shore-
line progradation and deposition of a coarsening-up sequence of burrowed to cross-
bedded and laminated, fine-grained shoreface and foreshore sands. In contrast, the
adjacent barrier island (Shackleford Banks) consists almost totally of inlet-fill sediments
Present addresses:
'Louisiana Geological Survey, Louisiana State University, Baton Rouge, LA 70893, U.S.A.
'University of Southern California, Los Angeles, CA 90007, U.S.A.
3Penzoil Company, Houston, T X 77001, U.S.A.
4Marathon Oil Company, Casper, WY 82601, U.S.A.
%an Francisco State University, San Francisco, CA 94132, U.S.A.

414
deposited by lateral tidal inlet migration. Holocene sediments in the shallow lagoons
behind the barriers are 5-8 m thick fining-up sequences of interbedded burrowed,
rooted and laminated flood-tidal delta, salt marsh, and washover sands, silts and clays.
While barrier island sequences are generally 1 0 m in thickness, inlet-fill sequences may
be as much as 25 m thick and comprise an average of 35% of the Holocene sedimentary
deposits. Tidal inlet-fill, back-barrier (including flood-tidal delta) and shoreface deposits
are the most highly preservable facies in the wave-dominated barrier-shoreline setting. In
the Cape Lookout cuspate foreland, these three facies account for over 80% of the sedi-
mentary deposits preserved beneath the barriers. Foreshore, spit platform and overwash
facies account for the remaining 20%.
INTRODUCTION
The Outer Banks of North Carolina are a classic wave-dominated micro-

tidal barrier island shoreline (Fig.1). For the past two decades numerous
investigations have examined the morphology, processes, vegetative patterns
and historical development of this shoreline setting (Fisher, 1962, 1967;
Pierce, 1969; Knowles et al., 1973; Godfrey and Godfrey, 1976; Mixon and
Pilkey, 1976; Cleary and Hosier, 1979). Prior to 1975 there had been only
one subsurface study of Holocene sediments on the North Carolina coast-
line (Pierce and Colquhoun, 1970). Since this time, however, there have
been several coordinated investigations of the Holocene stratigraphy along
the 130 km of barrier shoreline that forms the Cape Lookout cuspate fore-
land (Susman, 1975; Moslow, 1977; Herbert, 1978; Berelson, 1979; Steele,
1980). This paper is a synthesis of these studies and provides insight into the
Fig.1. Morphology of the southeastern United States Atlantic coastline showing major
cuspate forelands and cape systems. The Cape Lookout cuspate foreland is enclosed
within the rectangular box and shown in detail in Fig.3.
415
record of Holocene sedimentation for a clastic wave-dominated barrier-

island shoreline.
Earliest stratigraphic studies of barrier islands were conducted on the
Texas Gulf Coast (Fisk, 1959; Bernard et al., 1962), and resulted in the
first depositional models of barrier island facies. Even today Galveston
Island is considered by some to be the “type” example of barrier island
stratigraphy (McCubbin, 1982). Since the development of these early models,
however, subsurface studies of U S . East Coast barriers have shown that
there is a wide disparity of stratigraphic motifs and facies relationships
within barrier island complexes (Hoyt and Henry, 1967; Kraft, 1971; Kraft
et al., 1978; Susman and Heron, 1979; Leatherman, 1979; Moslow and
Colquhoun, 1981). From these studies it has been recognized that the Gulf
Coast barriers are one of a wide spectrum of stratigraphic types. As with
other clastic depositional systems the morphology, depositional environ-
ments and internal stratigraphy of barrier islands vary dramatically as a func-
tion of waves, tides and sediment input.
This paper is a synthesis of closely spaced drill hole data gathered to
examine the Holocene stratigraphy and sedimentation of a chain of barrier
islands in a low tidal-range, high wave-energy (wave-dominated) setting. The
major objectives of this study are twofold: (1)document the wide variation
in preserved vertical sedimentary sequences within the wave-dominated
shoreline of the Cape Lookout cuspate foreland; and (2) provide a process-
response depositional model for wave-dominated barrier islands in sand-rich
shoreline environments.
Several coordinated investigations of barrier island stratigraphy in the
study area have provided an extensive data base that includes detailed exam-
ination of approximately 3000 m of Holocene and Pleistocene sediments
from 230 auger, wash-bore and vibracore holes. Sedimentary structures,
lithology, grain size and fossil assemblages were analyzed to determine en-
vironments of deposition. Analysis of historic maps and charts, and a late
Holocene sea-level curve constructed from 25 carbon-14 dated samples
(Fig.2) provided the basis for interpreting the depositional history and
shoreline evolution of the study area.
Detailed explanations of drill hole methodology and sediment analyses
are provided in Susman and Heron (1979) and Moslow and Heron (1979,
1981).
GEOLOGIC SETTING
The barrier islands of the Cape Lookout area form one of the “cuspate
forelands” that are dominant features of coastal North and South Carolina
(Fig.1). These more prominent cape systems are associated with large sub-
aqueous shoals that protrude into the Atlantic Ocean at roughly 90” to the
shoreline. Cuspate forelands such as Cape Lookout are best developed in
clastic shoreline settings where wave energy prevails over tidal or fluvial
processes. Classic examples of cuspate-foreland systems on the east coast of
416
Year Before Present (BP) x lo3

9 8 7 6 5 4 3 2 1 0
Fig.2. Holocene sea-level curve for the Cape Lookout cuspate foreland as determined
from radiocarbon-dated samples of peat, wood and shell material cored beneath Bogue
Banks, Shackleford Banks, Back Sound, Core Banks and Portsmouth Island. Peat samples
are given preferential consideration as more accurate indicators of past sea-level positions.
Note the decreased rate in sea-level rise since approximately 4000 B.P.
the United States (Capes Hatteras, Lookout and Fear) are all found in the
wave-dominated Outer Banks of North Carolina (Fig.1). The origin and even-
spacing of these “Carolina Capes” (120-140 km) have been related to a
number of phenomena including eddy patterns of ocean currents (Dolan
and Ferm, 1968; Shepard and Wanless, 1971) and erosional remnants of
Pleistocene deltas (Hoyt and Henry, 1971). Recent studies on the North
Carolina continental shelf, however, suggest that the shoreline orientation,
origin and distribution of the Carolina capes and cuspate forelands are con-
trolled by the structure and pattern of pre-Holocene erosion-resistant strata
(Blackwelder et al., 1982). It is therefore quite conceivable that antecedant
topography and subtle structural features are an important control on
present shoreline configuration and sedimentation even along a passive,
trailing-edge continental margin. Pleistocene beach ridges and shoreline
orientation suggest that the Cape Lookout area was the site of at least one
pre-Holocene cuspate foreland (Fisher, 1967; Blackwelder and Cronin,
1981; Mixon and Pilkey, 1976; and Fig.3).
COASTAL PROCESSES AND MORPHOLOGY
Waves, tides and s t o r m s
The Cape Lookout cuspate foreland consists of two barrier island limbs
whose contrasting morphologies reflect local variations in wave energy,
tidal range and storm response. Mean tidal range in the Cape Lookout area
varies from 0.3 to 1.1 m (US.Department of Commerce, 1982), and there-
fore is classified as microtidal (<2 m tidal range; Davies, 1964). These are
417
77” 76”
Holocene Barrier Islands

0 Pleistocene Coastal Plain
Fig.3. Map showing location and names of the Holocene barrier islands that form the
“limbs” of the Cape Lookout cuspate foreland. Pleistocene shoreline trends are shown
by the dashed and solid line patterns on the Coastal Plain. These Pleistocene shoreline
features form a roughly cuspate orientation approximating the present day shoreline
trend, suggesting that antecedant topography is a major control on Holocene shore-
line configuration. Holocene barriers are black; Pleistocene deposits are stippled.
some of the lowest tidal ranges on the United States east coast (Nummedal
et al., 1977). A significant aspect of tidal processes in the Cape Lookout
cuspate foreland is that the 0.47 m mean tidal range for the northeast-
trending barrier limb (Core and Portsmouth Banks) is roughly half the
0.89 m range for the southwest-trending limb (Boque and Shackleford
Banks, Fig.3). The difference in tidal range is at least partly responsible for
many of the variations in morphology and sedimentation patterns observed
for the two barrier limbs. While the tidal range is relatively low, tidal
processes should not be discounted as unimportant in the Cape Lookout
area, Tidal currents have been measured in excess of 115 cm s-’ (Sarle,
1977), and are responsible for transport and deposition of fine- and coarse-
grained sediment especially in the tidal inlets, tidal deltas and lagoons. The
size of the tidal prism is an important factor in controlling the magnitude
and duration of tidal currents, and in determining flood or ebb dominance
of sediment transport in tidal inlets and backbarrier environments.
Although tides and tidal prism are important, the most significant coastal
processes affecting Holocene sedimentation in the area are waves and storms.
Mean annual wave height is 1.7 m with wave heights exceeding 2.0m
approximately 30% of the year (Fig.4; Nummedal et al., 1977). These values
are among the highest for the U.S. east coast and are primarily responsible
for the wave-dominated barrier shoreline morphology (Fig.5).
418
Fig.4. Deep-water wave-energy flux diagram and wave height-frequency histogram for
portions of t h e North Carolina continental shelf and shoreline. Energy flux is in units of
10'' ergs m l s-'. Only the onshore wave-energy flux components are shown. Wave
height histogram indicates t h at wave heights in excess of 2 m occur 30% of the time.
Average wave height ( h ) is 1. 7 m f o r t h e study area (modified from Nummedal e t al.,
1977).
WA S H O VER, M
FLOOD- TIDA
DELTA
LAGOON
FORESHORE
I SEA CHI
EBB-TIDAL
SALT.
MARSH
...-
' "W
TIDAL INLET
[CHANNELI
1
-
W A V E DOMINATED
KM
5.0
I
[BARRIER S H O R E L I N E ]
Fig. 5. Diagram depicting barrier-island and tidal-inlet morphology and depositional

environments in a wave-dominated (high average wave height, low mean tidal range)
setting. This model is primarily based o n observations of the study area.
419
An evaluation of the nearshore wave climate for North Carolina indicates

that significantly more wave energy is focused on the Core and Portsmouth
Banks barrier limb (Figs.3 and 4). These islands are directly affected by the
northeast, east and southeast onshore components of wave-energy flux
which account for approximately 75% (5.8 X lo1' ergs m-l s-') of the total
onshore wave energy for North Carolina (Fig.4; Nummedal et al., 1977).
Bogue and Shackleford Banks barrier-limb is relatively sheltered with respect
to wave energy because of its shoreline orientation. These islands are in-
fluenced only by the southerly onshore component which accounts for 25%
(1.9 X 10" ergs m-l s-') of total onshore wave energy flux for the study
area.
Extratropical storms and hurricanes have an important impact on sedi-
mentation patterns in the Cape Lookout area. Storm-related processes re-
distribute coarse-grained sediment on the ocean margin of tidal inlets, tidal
deltas and lagoons (Sarle, 1977). Large volumes of sand are transported
across low-lying barriers during storms and deposited in the form of wash-
over fans (Godfrey and Godfrey, 1976). The North Carolina coast has a
history of 150 recorded hurricanes since 1585. An average of 1.64 hurricanes
per year affect the Cape Lookout area (Crutcher and Quayle, 1974) with
recorded storm surge up to 2.3 m above mean sea level ( U S . Corps of
Engineers, 1976).
Barrier morphology
The barrier islands show a morphology typical of a wave-dominated,

microtidal environment (Fig. 5; Hayes, 1979). The barriers are relatively
long, linear and narrow and backed by wide shallow, open, lagoons. Tidal
inlets are rare and ephemeral but migrate laterally at very high rates (Moslow
and Heron, 1978). Tidal inlets on the higher-energy northeast barrier limb
(Core and Portsmouth Banks) are associated with large flood-tidal deltas and
small ebb deltas (Fig.6). The most prominent aspects of barrier morphology
of Core and Portsmouth Banks are the extensive washover fans, fringing salt
marsh and wide open lagoons.
In morphological contrast to the northeast limb the lower-energy south-
west limb of Shackleford and Bogue Banks, while still falling within the
wave-dominated shoreline morphology model, possess some unique
features. Parallel and recurved beach ridges are common to both Shackleford
and Bogue Banks and washover fans are much less common (Fig.7). The
tidal inlets of the southwest lower-energy limb are associated with more
prominent ebb-tidal deltas. Also, adjacent back-barrier lagoons are partially
filled with salt marsh and tidal flat sediments.
CONTROLS ON HOLOCENE SEDIMENTATION
While the hydrographic regime (waves, tides and storms) is an important

control on sedimentation patterns in the study area, other natural controls
120
Fig.6. Low-altitude oblique air p h o t o of Core Banks and New Drum Inlet. Core Sound is
to t h e right (west), Atlantic Ocean is to t h e left. Note t h e series of coalescing sand lobes
t h a t form t h e large flood-tidal delta o n t h e sound side of t h e inlet throat. On t h e ocean
side of t h e inlet, shoaling waves are breaking over the ebb-tidal delta. Note that the
barriers are narrow, relatively flat, and featureless.
are equally important. As with any depositional system, sediment supply

is important. The Cape Lookout area is a sand-rich environment with most
of the sediment supplied from the Coastal Plain and Piedmont rivers or
from the shoreface and inner shelf (Pierce, 1969; Mixon and Pilkey, 1976).
Holocene sedimentation in the area records several depositional processes.
Each produced a specific shoreline response (Fig.81, the combination of
which yielded the present observed sedimentary sequences. Perhaps the
most important process has been the Holocene rise of sea level (Fig.2), a
process that has resulted in the shoreline transgression most evident in the
CorePortsmouth barrier limb. Tidal-inlet migration has reworked large
portions of Core, Portsmouth and Shackleford Banks. Shoreface sedimenta-
tion has resulted in a seaward progradation of Bogue Banks over the past
4000 yrs. Storms are important in washover sedimentation, inlet formation,
and lagoonal infilling. Flood-dominated inlet sedimentation is responsible
for partial infilling of shallow water lagoons. Flood-tidal deltas weld onto
updrift portions of adjacent barriers resulting in abnormally wide, thick
accumulations of flood-delta sediments (Fig.8).
121
Fig.7. Low-altitude air p h o t o of Bogue Banks, Bogue Inlet an d Bear Island (foreground).
View is to t h e east. N o t e t h e series of shore-parallel beach ridges an d greater w i d t h of
t h e barrier a t t h e western e n d of Bogue Island. This morphology contrasts with t h at
typically found along t h e higher energy barrier limli as sh o w n in Fig.6.
HOLOCENE SEDIMENTATION
(Cape Lookout Cuspate Foreland)
I DEPOSITIONAL EVENT 1 SHORELINE RESPONSE ,

~~
Sea Level Rise Transgression
-
-+
Inlet Migration qtc

/ Reworking
Shoreface
T- Progradation
Sedimentation
li- - Washover U Inlet
-
Storms
Formation
Flood-Delta , "i>zi '_
i Back-Barrier
Attachment ' i- Accretion
Fig.8. Major depositional events a n d resulting shorellne responses effecting Holocene

sedimentation in t h e s t u d y area
422
SEDIMENTARY SEQUENCES AND FACIES RELATIONSHIPS
The two foreland limbs have different facies that can be related to the
processes of transgression, regression and inlet filling. Although many facies
are common to both limbs, the general vertical and horizontal facies-
associations of each limb are quite different.
Higher-energy transgressive limb
Core Banks and Portsmouth Island form the transgressive limb (Fig.3).
Both barriers have a common transgressive stratigraphic sequence although
the southwest and northeast ends reflect other processes. Four prominent
facies are associated with the transgressive limb, each of which represents a
number of sedimentary environments (Table I).
TABLE I
Facies characteristics of t h e high-energy transgressive limb and t h e lower-energy regressive

limb of t h e Cape Lookout foreland
Depositional Lithology Shells a n d organics Sedimentary Largescale features

environm r n t s truc t imcs
_______ ___.
Overwash and Clean, modcratelv Vv'liole and abraded Horizontal a n d planar Caps inlet and
foreshore sorted, fine to shells in layers: laminations barrier sequences
medium sand variable assemblage
(1 ow diversity )
Shoreface Wellsortcd. fine Abundance of Cross-bedded (upper Coarsening-upward

t o medium sand s a n c l s i ~ e dshell half) a n d burrowed sequence; increase
a n d silt material G e i n m a (lower half) sequence i n mud content
gemmu. Arcopecten towards base
sp., Oliuella s p .
Bac kb arrier Wellsorted, fine Organic rich: Burrowed: thin Capped by salt
(lagoon, t o medium silty Sparfirla sp. and parallel clay marsh: increasing
tidal flat. sand and sandy other plant material; laminations m u d and organic
salt marsh) clay Ensis s p . . Crassostrea content upwards
s p . , C r c p i d u l a sp.
(mollusks)
Flood-tidal Moderately sorted. Coarse shell frags. Gently dipping cross- Interbedded with
delta medium t o coarse c o m m o n ; Ee hino~ laminae; burrowed backbarrier facies;
siltv sand d e r m frags. coininon cyclic fining-
upward sequences
Inlet Clean, wellsorted. Mollusks rare; low Planar a n d horizontal Caps fining-upward
margin fine t o medium diversity laminations inlet sequence
sand
Inlet Moderately sorted, Mixed mollusk Cross-bedded (trough Thickest unit of

channel medium t o coarse assemblage of shelf a n d planar?) inlet sequence;
sand and shell; a n d backbarrier sps.. fines upward
pebbly sand shells common a n d
comni on abraded
Inlet Poorly sorted, Large. worn a n d Rip-up elasts: Basal scour lag
floor coarse t o pebbly abraded shell frags. graded bedding
sand and shcll common
Overwash and foreshore. The beach and berm, depositional environments
of the foreshore have been combined with overwash facies in cross-section
Figs.9 and 11,and in the composite vertical section, Fig.10. The facies charac-
teristics are described in Table I. Sedimentary structures and other diagnostic
criteria associated with each facies in vertical sequences are shown in Table 11.
Although not shown on the cross-sections, dune sands may locally overlie
the overwash unit. Isolated wind-shadow or barchan dunes are scattered
along the northeast limb of the foreland. Hosier (1973) attributes this to
winds that parallel the island so that sand moves up and down rather than
across the barriers. The highest dunes are 7 m and occur as isolated mounds
just south of Ocracoke Inlet.
Backbarrier. The two dominant backbarrier depositional environments are

the lagoon and flood-tidal delta, but others can be identified including tidal
flat, salt marsh and distal overwash (Table I). The backbarrier facies under-
lies the overwash-foreshore deposits and commonly comprises the lower
half of the transgressive barrier sequence.
Shoreface. At the Cape Lookout apex of Core Banks, the ovenvash-

foreshore sequence is underlain by a coarsening-up sequence of well-sorted,
fine-grained shoreface sediments (Fig.9 and Table I). The shoreface facies
is relatively rare within the higher-energy transgressive barrier limb.
CAPE DRUM
LOOKOUT INLET
CORE BANKS
\ I
Om
10m
tom
3Om
0 1 2 3KM U Overwash
6a
aBackbarrier UShoreface
0Pre-Holocene
Fig.9. Shore-parallel cross section of Core Banks from Cape Lookout (south) t o New
Drum Inlet (north). Note shoreface sediments beneath Cape Lookout and five isolated
inlet-fill sand bodies. Greatest volume of sediment is a transgressive sequence of fore-
shore and overwash overlying backbarrier facies, a sequence typical for most of the
higher-energy barrier limb (modified from Moslow and Heron, 1979).
424
~~~
Mud I Sand
I
C S V F F M C
OVERWASH
(Foreshore)
TIDAL FLAT c3
FLOOD DELTA
(Backbarrier)
5 MARSH (Peat)
LAGOON
PLEISTOCENE
Fig.10. Composite vertical sequence of Holocene and uppermost Pleistocene sediments

f o r Core Banks and Portsmouth Island. An explanation key t o sedimentary features
shown in all sequences is given in Table 11.
Om Om
1Om 1Om
20m 2Om
-
30m 30m
UForeshore U DTldal Inlet
Overwash 0 1 2 3
I Backbarrier OPre-Holocene
Fig.11. Shore-parallel cross-section of Portsmouth Island from New Drum Inlet (south)
t o Ocracoke Inlet ( n o r t h ) . Note the extremely thick Holocene inlet-fill sequence at
northern end of cross-section. The thick hackbarrier sequence adjacent t o Ocracoke
Inlet is primarily a sequence of flood-tidal delta and lagoonal sediments (modified from
Herbert, 1 9 7 8 ) .
425
TABLE I1
Key to lithologies and physical and biogenic sedimentary structures shown in vertical
sedimentary sequences (Figs.10, 1 4 , 1 5 , and 16)
PHYSICAL SEDIMENTARY
LITHOLOGIES
STRUCTURES
Sand Horizontal Laminations

,-,
OJ0, oo,
~ Pebbly Sand and Gravel zSz2 Planar Lami no ti ons
2
2s
--
", Plant Material Trough Cross-bedding
W
-=;
Shell Material Mud Lenses
Rip-up Clasts Parallel Laminations
BlOGENlC SEDIMENTARY
STRUCTURES
4 4 Rooting
& Sand-filled Burrows
@ Mud-filled Burrows
Tidal inlet. Ten historical relict inlets and one modern but artificial inlet
have deposited expansive inlet-fill sand bodies within the transgressive limb
(Figs.9 and 11).Fifteen t o 20% of Core Banks and nearly 40% of Portsmouth
Island is underlain by inlet-fill.
The stratigraphic sequence of the five relic inlets on Core Banks has been
documented by Moslow and Heron (1978) and the facies details of the three
sub-environments are shown in Table I. Though not very common, the inlet-
fill deposit can consist of a series of stacked fining-up sequences, as observed
beneath Portsmouth Island (Herbert, 1978).
The transgressive barrier sequence. Transgressive barriers do not form an

easily recognizable thick vertical sequence of sediments. On the northeast
limb, transgressive barriers are underlain by about 10 m of lagoon, marsh,
tidal flat, flood delta and ovenvash-foreshore sands with some silts and muds
(Fig.10). Lagoon, marsh, tidal flat and flood delta facies are not really
characteristic of the barrier island per se although they may be associated
with a barrier shoreline. The transgressive barrier proper consists of 2-3 m
of overwash-foreshore sands as linear bodies. Associated facies occur as non-
linear or arcuate-shaped sand bodies (that is, flood-tidal deltas), overlying
widespread lagoonal silts and muds. Thus, even though the transgressive
barrier has a typical vertical sequence, recognition of ancient barriers would
be difficult based on observation of this sequence alone. The presence of
fining-upward inlet sequences associated with the other barrier-related facies
326
would be the clue t o identifying ancient wave-dominated transgressive

barriers.
Lower-energy depositional limb
Bogue Banks and Shackleford Banks (Fig.3) are prograding and inlet-
modified barriers of the lower-energy southwest limb. Both barriers are
morphologically different from the higher-energy northeast limb. Whereas
Core Banks and the southwest part of Portsmouth Island form a long,
narrow, low barrier system (Fig.61, Bogue and Shackleford Banks are gen-
erally wide with prominent beach ridges and often extensive dune fields
(Fig.7). The topography reflects the complex progradation and inlet migra-
tion evolution of these barriers. Three prominent facies are associated with
the lower-energy limb of the Cape Lookout foreland.
Shoreface. The shoreface facies is combined with the foreshore facies in

stratigraphic cross-section of Bogue Banks (Fig.12). Shoreface and foreshore
sediments are a coarsening-upward sequence of fine- t o coarse-grained sand
with an increasing percentage of silt and clay towards the base. These facies
thicken in a seaward direction beneath the barrier and erosionally overlie
backbarrier deposits. The sedimentary characteristics of shoreface and fore-
shore facies are shown in Table I
\
Bogue Beaufort
Inlet Inlet
0 Dune/Overwash 1 Backbarrier
0 Shoreface Pre-Holocene
0 Inlet
0 lokm
Fig.12. Shore-parallel cross section of Bogue Banks from Bogue Inlet (west) t o the tip
of Shackleford Banks adjacent t o Beaufort Inlet (east). In sharp contrast t o the trans-
gressive barrier limb, t h e majority of t h e Holocene subsurface of Bogue Banks is a thick
sequence of shoreface deposits erosionally overlying a thin sequence of basal backbarrier
deposits (modified from Steele, 1980).
\
*A-
. -,.
Beaufort /
Inlet
\
O m O m
10 10
20 20
30 30
I €3 Dune-Overwash 0 Backbarrier 1
1
I
Q Shoreface
Spit-Foreshore
Inlet-fill
0 Pm!*HOlOCX?nel
I -
km
- -
7
Fig.13. Shore-parallel cross-section of Shackleford Banks from Beaufort Inlet (west) to

Barden Inlet (east). Most of the Holocene subsurface is a thick, laterally extensive se-
quence of tidal inlet-fill. The dashed lines forming an arcuate pattern o n the sound side
of the island enclose relict flood-tidal deltas (from Susman and Heron, 1979).
Tidal inlet. As much as 25 m of inlet fill underlies most of Shackleford

Banks (Fig.13 and Table I). This inlet fill does not represent deposition from
a single inlet that migrated westerly toward the present Beaufort Inlet as
implied by Susman and Heron (1979). Instead, there are several stacked
inlet sequences deposited by a minimum of two migrating inlets. Inlet se-
quences are crossbedded, poorly sorted, fine- t o very coarse-grained sand and
shell that are capped by dune and foreshore sands (Fig.14).
Flood-tidal delta. Vibracores from the relict flood-tidal delta behind

Shackleford Banks in Back Sound show a three-part fining-upward sequence
very similar t o the inlet fill sequence except that the uppermost unit is
usually an organic (that is salt marsh) mud (Table I). Flood-tidal delta sands
thin in a landward direction and are interbedded with lagoon and tidal flat
muds.
Vertical sequences. The general vertical sequence of prograding Bogue

Banks is shown in Fig.15. The basal tidal flat/lagoon facies was deposited
during an early Holocene transgressive phase. Regression started about
4000-5000 B.P. The general coarsening-up nature of this sequence is typical
of prograding barriers (Bernard et al., 1970; Moslow and Colquhoun, 1981).
The vertical sequence of inlet-fill on Shackleford Banks is shown in Fig.14.
The fining up trend is characteristic of inlet-fill. Flood-tidal delta sequences
also fine up, and are capped by muddy, marsh deposits. A distinct difference
in sand body geometry is observed between the somewhat isolated bodies
428
Mudl Sand
C S V F F M C V C
1 4 1 1 1 1 1
Dune
Spit
~
Platform
~~
Act3ve
Inlet
Channel
Inlet Floor
Pleistocene
Fig.14. Composite sedimentary sequence of Holocene inlet-fill t h a t is representative of
an inlet-modified barrier island, such as Shackleford Banks (modified from Susman and
Heron, 1979).
of inlet-fill in the higher-energy limb (Figs.9 and 11)and the extensive lateral
body of inlet-fill in Shackleford (Fig.13).
DISCUSSION
Vertical sequence variability
Perhaps the most striking aspect of the subsurface stratigraphy of the

Cape Lookout cuspate foreland is the extreme variability in facies relation-
ships and vertical sedimentary sequences (Fig.16). The Holocene sedimentary
deposits of the higher-energy transgressive barrier limb (Core and
Portsmouth Banks) are a mixture of: (1)coarsening-upward, transgressive
barrier-island sequences; ( 2 ) isolated, fining-upward inlet-fill sequences; and
(3) interbedded, fining- and coarsening-upward sequences of flood-tidal
delta, inlet-fill and lagoonal sediments (northern Portsmouth Island; Fig.16).
In contrast, the lower-energy depositional limb consists of: (1) laterally
extensive, fining-upward sequences of stacked inlet-fill and flood-delta
sediments (entirety of Shackleford Banks) ; and (2) a coarsening-upward
foreshore-shoreface sequence, also laterally extensive (majority of Bogue
Banks). The Cape Lookout apex and adjacent subaqueous shoals are typified
also by coarsening-upward sequences of primarily upper and lower shoreface
sediments (Fig.16). Therefore, while the surficial geomorphology and ob-
served depositional environnients contrast somewhat for most of the Cape
Lookout coastal area, this contrast does not reflect the variability of sedi-
mentary sequences and facies relationships that occur in the Holocene
stratigraphy.
429
Mud,Sand
c S'VFF M
4mm
bm
FORESHORE
UPPER
8m SHOREFACE
[Om
LOWER
SHOREFACE
Ltm (Off shore)
I A m I--(+ 7 , 0 8 O TIDAL FLAT
BP -LAGOON
PLEISTOCENE
Fig.15. Composite vertical sequence of Holocene sediments from Bogue Banks. The
coarsening-upward shoreface and foreshore sequence was deposited by seaward pro-
gradation of t h e barrier island. Sedimentary structures shown here are inferred from
beach trenches, d u n e scarps and core data of shoreface sequences beneath Cape Lookout.
Basal Holocene tidal flat and lagoon sediments were deposited during an earlier trans-
gressive phase of island evolution (modified from Steele, 1980).
While not serving as an all-encompassing depositional model, Fig.16 does

outline the variability and types of vertical sedimentary sequences that one
can expect to find in a wave-dominated barrier island shoreline. This aspect
of subsurface complexity is important to the sedimentologist examining
other wave-dominated barriers or their ancient equivalents.
Facies preservation potential
The facies correlations and cross-sections developed from closely spaced

drill holes in the study area have been used to calculate the relative percen-
tages of Holocene sedimentary facies presently found in the subsurface
(Fig.17). Previous subsurface studies in wave- and tide-dominated shorelines
elsewhere have identified the more preservable barrier-island related sedi-
mentary facies (Hoyt and Henry, 1967; Kumar and Sanders, 1974; Hayes,
1980; Belknap and Kraft, 1981). In general, sediments deposited in a sub-
aqueous environment, primarily those deposited below wave base or
protected from wave erosion, should have a much better potential for preser-
vation in the coastal environment. This generalization is applicable t o the
Cape Lookout barrier islands, where tidal-inlet, backbarrier and shoreface
facies account for approximately 80% of the Holocene sediments presently
430
0
TI
Fig.16. Diagram showing t h e variability in vertical sedimentary sequences found within the
Holocene barrier islands of t h e Cape Lookout cuspate foreland. Vertical sequences are
keyed by shaded patterns t o those portions of the shoreline where they are most prevalent.
Transgressive barrier ( T B ) ,tidal inlet ( T I )and flood-tidal delta ( I - F T D )sequences comprise
the higher-energy barrier limb. Prograding barrier-shoreface ( P B S ) and barrier-inlet ( B - I )
sequences are found within the barrier islands of the lower-energy limb.
FORESHORE %= I"-
U SPIT
20% ~
SHOREFACE c
BACKBARRIER 25% 1111
(U FLOOD DELTA)-,5%
- .-
INLET - _~
_.
-.-c
. ~ - .
Fig.17. Diagram showing the relative amounts of t h e sedimentary facies presently found in
the Holocene barrier and backbarrier subsurface. The solid black bars indicate t h e percen-
tages existing a t t h i s m o m e n t i n time as determined from cross-section analysis. T h e stippled
bars in the diagram depict t h e possible volumetric increase of a sedimentary facieswith time.
found in the subsurface (Fig.17). Tidal inlet deposits account for the greatest
percentage (3570) of Holocene sediments. This should not be surprising
because inlet channels in a wave-dominated shoreline migrate laterally over
great distances reworking the previously deposited barrier sediments, some-
times scouring through the entire Holocene section, depositing thick inlet-
fill sequences. It is likely therefore, that as the inlets continue to migrate
431
along the shoreline, an even higher percentage of the Holocene sediment

package will be represented by the tidal inlet facies.
The backbarrier facies (flood-tidal delta, lagoon and tidal flat deposits)
account for 25% of the Holocene subsurface. Rapid sedimentation and pro-
tection from wave erosion in backbarrier environments allows for enhanced
accumulation, rapid burial and therefore greater preservation potential. High
rates of sedimentation along the seaward margin of Bogue Banks and the
Cape Lookout apex have preserved thick shoreface sequences. The shoreface
facies accounts for approximately 20% of the Holocene subsurface (Fig.17).
Those sediments deposited in intertidal or subaerial environments (fore-
shore, spit platform, ovenvash and dune) have a low preservation potential,
and combined, account for approximately 20% of the Holocene subsurface.
CONCLUSION
The wave-dominated Cape Lookout cuspate foreland consists of two

barrier-island limbs and a massive cape-shoal complex that display a variety
of vertical sedimentary sequences and facies relationships within the Holo-
cene subsurface. The diverse nature of the Holocene sedimentary record
has been a function of five major depositional processes and shoreline re-
sponses: (1)sea-level rise and barrier-island transgression; (2) tidal inlet migra-
tion and barrier reworking; ( 3 ) shoreface sedimentation and barrier prograda-
tion; (4)storm washover and lagoonal infilling; and (5) flood-tidal delta
attachment and accretion at updrift portions of barrier islands. Sediment
(that is, sand) supply, hydrographic regime, structural setting and antecedent
topography are important controls of Holocene sedimentation and facies
preservation.
Inlet-fill and transgressive barrier island sequences form the bulk of the
higher-energy, storm-dominated limb of the cuspate foreland (Core and
Portsmouth Banks). While inlet-fill is a prevalent facies, a significant portion
of the lower-energy depositional barrier island limb (Shackleford and Bogue
Banks) is a thick regressive sequence of prograding shoreface sediments.
Likewise, the massive Cape Lookout shoal complex is mostly a coarsening-
upward shoreface sequence. The variety and types of sedimentary sequences
found within the Holocene subsurface of the Cape Lookout cuspate foreland
should be similar to those found in other modern wave-dominated shorelines
or their ancient equivalents.
The tidal inlet, backbarrier (including flood-delta) and shoreface facies
account for approximately 80% of the Holocene subsurface deposits. These
facies have the highest preservation potential among those observed in the
wave-dominated barrier shoreline.
ACKNOWLEDGEMENTS
The majority of the research incorporated into this manuscript was

funded by grants from the National Park Service to the Duke University
432
Department of Geology. Other partial funding was provided by the Cities

Service Oil Company and the American Association of Petroleum Geologists.
Logistical support was provided by the Duke University Marine Laboratory
and the University of North Carolina Marine Science Consortium. Albert C.
Hine is especially thanked for his support of the research conducted on
Bogue Banks. Identification of mollusk assemblages and their related environ-
ments of deposition were identified by Blake Blackwelder and William C.
Miller, 111. Figures were prepared by the Cities Service Research Graphics
Department and the Geology Department of Louisiana State University.
Bruce Wilkinson and Albert Hine are thanked for their helpful review of the
manuscript.
This manuscript is a synthesis of several M.S. thesis research projects con-
ducted from 1974 t o 1980 at the Duke University Department of Geology.
Contributions of the authors are as follows: S.D. Heron, Jr. coordinated all
research and helped prepare the manuscript; T.F. Moslow examined the sub-
surface stratigraphy of southern Core Banks and Cape Lookout, and helped
prepare the manuscript; W.M. Berelson performed a vibracore study of Back
Sound sediments; northern Core Banks and Portsmouth Island was studied
by J.R. Herbert; K.R. Susman conducted a study of Shackleford Banks;
Bogue Banks was examined by G.A. Steele 111.
REFERENCES
Belknap, D.F. and Kraft, J.C., 1981. Preservation potential of transgressive coastal litho-
somes on the U.S. Atlantic shelf. Mar. Geol., 42: 429-442.
Berelson, W.M., 1979. Barrier island evolution and its effect on lagoonal sedimentation:
Shackleford Banks, Back Sound, and Harkers Island : Cape Lookout National Seashore.
Thesis, Duke University, Durham, N.C., 132 pp. (unpublished).
Berg, R.G. and Davies, D.K., 1968. Origin of Lower Cretaceous Muddy Sandstone at Bell
Creek Field, Montana. Bull. Am. Assoc. Pet. Geol., 52: 1888-1898.
Bernard, H.A., LeBlanc, R.J. and Major, C.F., 1962. Recent and Pleistocene geology of
southeast Texas. In: E.H. Rainwater and R.P. Zingula (Editors], Geology of the Gulf
Coast and Central Texas and Guidebook Excursions. Houston Geol. Soc., Field Trip
Guidebook, pp. 175-224.
Bernard, H.A., Major, C.E., Parrott, B.S. and LeBlanc, R.J., 1970. Recent sediments of
southeast Texas. In: A Field Guide to the Brazos Alluvial and Deltaic Plains and the
Galveston Barrier Island Complex. Univ. of Texas, Bur. Econ. Geol. Guidebook 11,
pp. 1-16.
Blackwelder, B.W. and Cronin, T.M., 1981. Atlantic Coastal Plain geomorphology illus-
trated by computer-generated block diagrams. U S . Geol. Surv., Misc. Field Stud. Map
MF-1242.
Blackwelder, B.W., MacIntyre, I.G. and Pilkey, O.H., 1982. Geology of continental shelf,
Onslow Bay, North Carolina, as revealed by submarine outcrops. Bull. Am. Assoc. Pet.
Geol., 66: 44-56.
Cleary, W.J. and Hosier, P.E., 1979. Geomorphology, washover history, and inlet zona-
tion: Cape Lookout, N.C., to Bird Island, N.C. In: S.P. Leatherman (Editor), Barrier
Islands. Academic Press, New York, N.Y., pp.237-271.
Crutcher, H.D. and Quayle, R.G., 1974. Mariners worldwide climate guide to tropical
storms at sea. Naval Weather Service Command, US. Dept. Commerce, Washington,
D.C.
Davies, J.L., 1964. A morphogenic approach to world shorelines. Z . Geomorphol., 8:
27-42.
433
Dolan, R. and Ferm, J.C., 1968. Crescentic landforms along the Atlantic coast of the
United States. Science, 159: 627-629.
Fisher, J.J., 1962. Geomorphic expression of former inlets along the Outer Banks of N.C.
Thesis, University of North Carolina, Chapel Hill, N.C., 120 pp. (unpublished).
Fisher, J.J., 1967. Development pattern of relict beach ridges, Outer Banks barrier chain,
N.C. Diss., University of North Carolina, Chapel Hill, N.C., 325 pp. (unpublished).
Fisk, G.N., 1959. Padre Island and LagunaMadre flats, coastal south Texas. Proc. Second
Coastal Studies Institute Conf., Louisiana State University, Baton Rouge, La.,
pp. 103-1 5 2.
Godfrey, P.J. and Godfrey, M.M., 1976. Barrier island ecology of Cape Lookout National
Seashore and vicinity, North Carolina. National Park Service Scientific Monograph
Series, no. 9, 1 6 0 pp.
Goff, T.W., 1977. Sedimentation in a modern tidal inlet: Bogue Inlet, North Carolina.
Thesis, Univ. of North Carolina, Chapel Hill, N.C., (unpublished).
Hayes, M.O., 1980. General morphology and sediment patterns in tidal inlets. Sediment
Geol., 26: 139-156.
Hayes, M.O. and Kana, T.W., 1976. Terrigenous clastic depositional environments. Tech.
Rep. 11-CRD, Dept. of Geology, Univ. of South Carolina, Columbia, S.C.
Herbert, J.R., 1978. Post-Miocene stratigraphy and evolution of northern Core Banks,
North Carolina. Thesis, Duke University, Durham, N.C., 166 pp. (unpublished).
Hosier, P.E., 1973. The effects of oceanic overwash on the vegetation of Core and
Shackleford Banks, North Carolina. Diss., Duke University, Durham, N.C., 230 pp.
(unpublished).
Howard, J.D. and Reineck, H.E., 1972. Georgia coastal region, Sapelo Island, U.S.A.:
sedimentology and biology. V. 111. Conclusions. Senckenbergiana Marit., 4:81-123.
Hoyt, J.H. and Henry, V.J., 1967. Influence of inlet migration on barrier island sedimen-
tation. Geol. SOC.Am. Bull., 78: 77-86.
Hoyt, J.H. and Henry, V.J., 1971. Origin of capes and shoals along the southeastern coast
of the United States. Geol. SOC.Am. Bull., 82 : 59-66.
Knowles, C.E., Langfelder, J. and McDonald, R., 1973. A preliminary study of storm
induced beach erosion for North Carolina. Rep. 73-75, Center for Marine Coastal
Studies, North Carolina State University, Raleigh, N.C.
transgression. Geol. SOC.Am. Bull., 82: 2131-2158.
Kraft, J.C., Allen, E.A. and Maurmeyer, E.M., 1978. The geologic and paleogeomorpho-
logical evolution of a spit system and its associated coastal environments: Cape
Henlopen Spit, Delaware. J. Sediment. Petrol., 48(1): 211-226.
Kumar, N. and Sanders, J.E., 1974. Inlet sequences: A vertical succession of sedimentary
structures and textures created by the lateral migration of tidal inlets. Sedimentology,
21: 291-323.
Leatherman, S.P., 1979. Barrier Islands from the Gulf of St. Lawrence to the Gulf of
Mexico. Academic Press, New York, N.Y., 325 pp.
McCubbin, D.G., 1982. Barrier island and strand-plain facies. In: P.A. Scholle and
D. Spearing (Editors), Sandstone Depositional Environments. Am. Assoc. Pet. Geol.,
Tulsa, Okla., pp.247-279.
Mixon, R.B. and Pilkey, O.H., 1976. Reconnaissance geology of the submerged and
emerged Coastal Plain province, Cape Lookout area, North Carolina. U S . Geol. Sum.,
Prof. Pap., 8 5 9 , 4 1 pp.
Moslow, T.F., 1977. Quaternary evolution of Core Banks, North Carolina from Cape
Lookout to New Drum Inlet. Thesis, Duke University, Durham, N.C., 132 pp.
(unpublished).
Moslow, T.F. and Colquhoun, D.J., 1981. Influence of sea level change on barrier island
evolution. Oceanis, 7 : 439-454.
434
Moslow, T.F. and Heron Jr., S.D. 1978. Relict inlets: preservation and occurrence in the
Holocene stratigraphy of southern Core Banks, North Carolina. J. Sediment. Petrol.,
48: 1275-1286.
Moslow, T.F. and Heron Jr., S.D. 1979. Quaternary evolution of Core Banks, North
Carolina: Cape Lookout t o New Drum Inlet. In: S.P. Leatherman (Editor), Barrier
Islands. Academic Press, New York, N.Y., pp.211-236.
Moslow, T.F. and Heron Jr., S.D. 1981. Holocene depositional history of a microtidal
cuspate foreland cape: Cape Lookout, North Carolina. Mar. Geol., 41: 251-270.
Moslow, T.F. and Tye, R.S., 1985. Recognition and characterization of Holocene tidal
inlet sequences. In: G.F. Oertel and S.P. Leatherman (Editors), Barrier Islands. Mar.
Geol., 6 3 (in press).
Nummedal, D., Oertel, G.F., Hubbard, D.K. and Hine, A.C., 1977. Tidal inlet varia-
bility - Cape Hatteras to Cape Canaveral. Proc. Conf. on Coastal Sediments 77,
A.S.C.E., Charleston, S.C., pp.543-562.
Pierce, J.W., 1969. Sediment budget along a barrier island chain. Sediment. Geol., 3:
5-16.
Pierce, J.W. and Colquhoun, D.J., 1970. Holocene evolution of a portion of the North
Carolina coast. Geol. SOC.Am. Bull., 81 : 3697-3714.
Sarle, L.L., 1977. Processes and resulting morphology of sand deposits within Beaufort
Inlet, Carteret County, North Carolina. Thesis, Duke University, Durham, N.C.,
149 pp. (unpublished).
Schwartz, R.K., 1975. Nature and genesis of some storm washover deposits. Tech. Memo.
No. 61. U.S. Army Corps of Engineers, 60 pp.
Shepard, F.P. and Wanless, H.R., 1971. Our Changing Coastlines. McGraw-Hill, New
York, N.Y., 579 pp.
Steele, G.A., 1980. Stratigraphy and depositional history of Bogue Banks, North Carolina.
Thesis, Duke University, Durham, N.C., 201 pp. (unpublished).
Susman, K.R., 1975. Post-Miocene subsurface stratigraphy of Shackleford Banks, Carteret
County, North Carolina. Thesis, Duke University, Durham, N.C., 85 pp. (unpublished).
Susman, K.R. and Heron Jr., S.D., 1979. Evolution of a barrier island, Shackleford Banks,
Carteret County, North Carolina. Geol. SOC.Am. Bull., 90: 205-215.
United States Corps of Engineers, 1976. Morehead City harbor, North Carolina, final
environmental statement. U.S. Army Engineer District, Wilmington, N.C.
U.S. Department of Commerce, 1982, Tide Tables, East Coast of North and South
America, Admin., Natl. Ocean Survey, Rockville, Md., 222 pp.
Marine Geology, 60 (1984) 435-454 435
RECONSTRUCTION O F SALEO-WAVE CONDITIONS DURING THE

LATE PLEISTOCENE FROM MARINE TERRACE DEPOSlTS, MONTEREY
BAY, CALIFORNIA
WILLIAM R. DUPRE
Department of Geosciences, University of Houston, Houston, TX 77004 (U.S.A.)
(Received February 24, 1983; revised and accepted July 29, 1983)
ABSTRACT
Dupre, W.R., 1984. Reconstruction of paleo-wave conditions during the Late Pleistocene
from marine terrace deposits, Monterey Bay, California. In: B. Greenwood and R.A.
Davis, Jr. (Editors), Hydrodynamics and Sedimentation in Wave-Dominated Coastal
Environments. Mar. Geol., 60: 435-454.
The Santa Cruz coastal terrace fringes much of the northern Monterey Bay region,
California. It consists mainly of a regressive sequence of high-energy, barred nearshore
marine sediments deposited during the last (Sangamonian) highstand of sea level. This
sequence can be sub-divided into several depth-dependent facies on the basis of paleo-
current data and vertical sequence of sedimentary structures. These include a lower
shoreface facies deposited in 10-16 m water depth, an upper shoreface facies (including
both a storm-dominated assemblage and a surf zone assemblage) deposited in 0-10 m
water depth, and a foreshore facies deposited in the swash zone, up to 3.6 m above high
tide.
The magnitudes of the storm events responsible for depositing these sediments were
estimated by calculating paleo-wave heights using a variety of criteria (e.g., critical thresh-
old equations, breaker depths, berm heights). In addition, the climate and paleogeography
during the deposition of these sediments were essentially the same as today, allowing the
use of present-day wave statistics t o estimate the frequency of these storm events. The
largest storms formed offshore-flowing currents (e.g., rip, wind-forced, and possibly storm-
surge ebb currents) that resulted in the deposition of approximately 30% of the sediments
seaward of the surf zone; however, the magnitude and frequency of these events
are unknown. The remaining 70% of the sediment beyond the surf zone was de-
posited in response to smaller storm waves which were, on the average, at least 1.6 m high;
such waves presently occur no more than 15% of the time. Sediments deposited during
“fainveather” conditions (i.e., the remaining 85% of the time) have a low preservation
potential, and are generally not preserved in this facies. In contrast, surf zone sediments
were deposited by a variety of processes associated with waves whose maximum offshore
heights were probably G2.2 m; such waves presently occur up to 92% of the time. Sedi-
ments within the swash zone were deposited by waves up t o 3 m high, the largest of
which presently occur approximately 2% of the time.
Most of the sediments were deposited by storms of intermediate magnitude and
frequency; different facies, however, appear to preferentially record events of different
recurrence intervals. In particular, surf zone sediments were deposited under relatively
small storm and post-storm conditions, whereas sediments deposited farther offshore
record increasingly larger, less frequent storm events. Relatively rare events (e.g., the 100
or 1000 yr events) do not appear to have significantly affected sedimentation in these
nearshore environments.

436
INTRODUCTION
The recognition of characteristic vertical sequences of sedimentary struc-

tures associated with prograding, wave-dominated shorelines has been an
active area of research in the past few years (e.g., Clifton, 1969; Campbell,
1971; Ryer, 1977; Vos and Hobday, 1977; Roep et al., 1979; Hunter et al.,
1979; Howard and Reineck, 1981; Clifton, 1981). These and other studies
of modern and ancient marine sediments have led to an increasing awareness
of the significance of storm deposits preserved in the stratigraphic record. In
fact, it appears evident that storm deposits are preferentially preserved in
many depositional settings (e.g., Wolman and Miller, 1960; Kumar and
Sanders, 1974; Kreisa, 1981). The magnitude and frequency of the storm
events responsible for these storm deposits are, however, less certain.
The purpose of this paper is two-fold. The first is to describe the vertical
sequence of sedimentary structures in Late Pleistocene marine terrace de-
posits in the Monterey Bay region, California (Fig.l), emphasizing the
characteristics and relative abundance of storm deposits. The second is to
calculate the magnitude and frequency of the storm conditions under which
these sediments were deposited.
The magnitudes of paleo-storm events are difficult to estimate, with
investigators typically relying on either ripple characteristics (e.g., Tanner,
1971; Komar, 1974; Allen, 1979; Miller and Komar, 1980), or critical
1
L TE PLEISTOCENE
FLOODPLAtN
0 5
km
MONTEREY BAY
KM '
/
- -,', fi MEASURED SECTIONS
1
Fig.1. Location of study area. M indicates Manresa measured section (see Fig.2). Hach-
ured lines indicate fluvial and marine scarps.
437
threshold criteria (e.g., Komar and Miller, 1974; Clifton, 1976; Dingler,
1979). The use of ripple morphology is a useful tool, particularly in low to
moderate wave-energy environments. They are of less use in high-energy
environments where large-magnitude storm events rarely produce ripples
except at the very top of the deposits, only to be usually removed by the
next storm event (Clifton, 1976; Howard and Reineck, 1981).Thus the use
of velocity-grain size diagrams (e.g., Komar and Miller, 1974; Clifton,
1976, fig.4) to determine critical threshold boundaries often provides the
best criteria for estimating the magnitude of ancient storm conditions. Other
criteria such as breaker depth and berm height may also be used, and are
discussed elsewhere in this paper.
Estimates of the frequency of paleo-storm events are even more tenuous,
relying either on the assumption that the ancient wave regime was similar
to that of the present (e.g., Bourgeois, 1980), or that the recurrence interval
is equal to the number of storm deposits divided by the time interval of the
unit as a whole. The latter method produces typical storm frequencies on the
order of 100’s of years for Holocene storm sands (e.g., Hayes, 1967; Morton,
1981; Nelson, 1982), and on the order of 100@-40,000 yrs for ancient
storm sandstones (e.g., Brenchley et al., 1979; Hamblin and Walker, 1979;
Kreisa, 1982).
Each method of estimating the recurrence interval of waves has potential
problems. The use of modern wave statistics to estimate ancient wave con-
ditions becomes increasingly unreliable with increasingly older deposits
where wind patterns, storm tracks, and shelf and nearshore bathymetry may
differ significantly from the present. The second method requires estimating
the time interval represented by a part of the stratigraphic record, an
estimate that is often unreliable. In addition, it should be noted that the
recurrence interval of a storm event capable of forming a storm deposit is
not necessarily the same as the number of storm deposits per unit time, as
such an estimate cannot take into account storm deposits which were sub-
sequently reworked or removed by erosion. For that reason such estimates
tend t o overestimate the actual recurrence interval. Lastly, the results of
this method are very much an “average” storm frequency, without the
ability to discern different recurrence intervals for different storm deposits.
This may be particularly significant in nearshore sediments where the preser-
vation potential of different frequency events may vary in different parts
of the nearshore zone (Hunter et al., 1979).
Calculations of the magnitude and frequency of storm events are probably
most accurate only where modern wave statistics can be extrapolated t o the
ancient wave conditions with some degree of confidence. This appears to be
a valid assumption for the Late Pleistocene terrace deposits in the Monterey
Bay region of California (Fig.1). The youngest of these sediments were de-
posited during the last (Sangamonian) interglacial period, during which time
the climate was approximately the same as the present, albeit slightly
warmer (Flint, 1971, p. 439). In addition, the configuration of the shoreline
was essentially the same as it is today, thus wave refraction patterns are
438
assumed to be the same as well. I t seems reasonable, therefore, to assume

that the nearshore wave spectrum during the deposition of these marine
sediments was essentially the same as the present. Wave statistics for the
Monterey Bay region were summarized by Arnal (1973) using Synoptic
Meteorological Observations (SSMO) data for the period 1939-1971, and
were used here as representative of the present-day wave climate. It should
be noted that these data were obtained for a relatively small area (1" X 1")
near Monterey Bay, and differ from the SSMO data summarized by Harris
(1972) for a 10" X 10" Marsden square for the area. The wave climate of the
central California region is much higher than that in the relatively more
sheltered Monterey Bay area, thus the use of the more generalized wave
statistics would have resulted in significant errors with the generation of
much larger recurrence intervals than those calculated in this study.
SANTA CRUZ COASTAL TERRACE DEPOSITS
Much of the central California coastal zone is rimmed by a series of up-

lifted marine terraces. The youngest of these is the Santa Cruz coastal
terrace formed during the rise and fall of sea level associated with the last
(Sangamonian) interglacial period (Duprh, 1975a; Duprh et al., 1980). These
deposits are relatively well exposed in sea cliffs along the northern part of
Monterey Bay, providing an opportunity to compare the preserved pro-
gradational sequence with that predicted on the basis of studies of modern
nearshore marine facies (e.g., Clifton, 1969; Davidson-Arnott and
Greenwood, 1976; Hunter et al., 1979; Howard and Reineck, 1981). Several
sections were measured, however only one is described here as all showed
basically the same vertical sequence of sedimentary facies illustrated in
Fig.2. Associated non-marine facies are not described in this paper, but are
discussed elsewhere (Dupr6, 1975a; Dupr6 et al., 1980).
Lower shoreface deposits
The base of the terrace deposits is marked by a roughly planar erosional

surface (Fig.3A) formed by marine planation mainly during a late stage of
rising sea level (see Bradley and Griggs, 1976, for a more detailed discussion
of the origin of the wave-cut platform). Some irregular scour depressions
with up to a meter of relative relief are locally present, as are erosional
ripples (cf. Reineck and Singh, 1975, p. 41) with wave length ( L ) = 30 cm
and wave height ( H ) = 5 cm. Transgressive marine deposits, where present,
are thin (typically less than 30 cm thick), and consist of landward-dipping
(Fig.3B) ripple cross-stratified gravels (L = 1.2 m, H = 10 cm) and relict
concretions of eolian deposits that appear to represent palimpsest sediments
of an ancestral Monterey Bay.
The transgressive lag deposit grades upward into the lower shoreface
deposits (Fig3C). These deposits are approximately 1.5 m thick, and con-
sists of highly bioturbated, parallel-laminated, fine t o very fine grained sand
439
PALEO- SEDIMENTARY
GRAIN SIZE
CURRENT STRUCTURES
d I n d l v l d u a l p a l o o o u r r e n t mea.ur.m.nt
generallz.d paleocurrmnt meaauremmnt
FORESHORE DEPOSITS
TOE-OF-BEACH
W
0
SURF ZONE DEPOSITS a
LL
W
a
0
I
& LOWER SHOREFACE

-w a v e-c u t p Iat f orm-
EOLIAN DEPOSITS
VERTICAL SCALE IN METRES

4 3 2 1 0-1-2-3
@
Fig.2. Vertical sequence of sedimentary structures and grain-size variations within the
Santa Cruz coastal terrace, as exposed at the Manresa measured section. Individual and
generalized paleocurrent data are from the Manresa section, however the rose diagrams
include paleocurrent data from nearby measured sections as well (see Fig.1 for locations).
Grain size determined by sieve analysis.
440
Fig.3. Outcrop photographs of the Santa Cruz coastal terrace sediments, northern
Monterey Bay, California (arrow indicates direction of land when sediments were being
deposited). A . Outcrop photograph of the Manresa measured section (see Fig.2). Dashed
line delineates the wave-cut platform, separating eolian ( E ) deposits from overlying
marine ( M ) deposits. B. Wave-cut platform separating eolian deposits ( E ) from overlying
marine deposits ( M ) . Note the onshore-dipping, cross-stratified gravels a t the base of the
lower shoreface deposits. C. Intensely bioturbated lower shoreface deposits of sand and
scattered pebbles, overlain by graveliferous storm deposits. D. Landward-dipping, cross-
stratified sand and gravel near t h e base of the storm-dominated sediments. Note t h e
truncated vertical A s t e r o s o m a ( ? ) burrows. E . Parallel-laminated sand deposited in the
upper flow regime, grading upwards into onshore-dipping, cross-laminated sand, overlain
b y storm deposited gravels. Note t h e A s t e r o s o m a ( ? ) burrows. F. Macraroiiichus segregotis
burrows within the lower shoreface deposits.
441
with scattered pebbles. High-angle cross-stratification is not present, how-

ever some of the sets of parallel-laminated sand have landward dips of 5-15'.
Scattered pebbles (long axis up to 2 cm) within the fine sand (Fig.3C),
combined with the complete lack of ripple-drift stratification, suggests
deposition under relatively high velocity, upper flow regime conditions
(see also Hunter and Clifton, 1982, p.136). This unit is also characterized
by extensive bioturbation (locally up to loo%), although body fossils are
very rare. Recognizable trace fossils include polecypod burrows (up to
30 cm deep), thin, vertical burrows (Fig.3D, E ) similar to Asterosoma (cf.,
Howard, 1972), and extensive, lined, worm-like feeding traces of
Macruronichus segregatis (Fig.3F) described by Clifton and Thompson (1978).
Upper shoreface deposits
This unit is approximately 5 m thick, and consists of complexly inter-

bedded sand and pebble gravel which can be further subdivided into a storm-
dominated assemblage and a surf zone assemblage (Fig.2).
The storm-dominated deposits are approximately 2.5 m thick, and consist
of parallel-laminated and cross-stratified, fine- to medium-grained sand and
interbedded pebble gravel. The base of the gravel beds is marked by a sharp
erosional surface (e.g., Figs.3D and 4A) which is planar where exposed
perpendicular to the paleo-shoreline, but which is locally scoured with up to
a meter of relative relief where exposed parallel t o the paleo-shoreline. The
gravel is typically 0.2-2 cm in diameter (long axis), moderately well sorted,
and occurs in beds which commonly range from 10 to 30 cm thick. The
gravel beds may be structureless, cross-stratified, or graded; most have a
sandy matrix which appears to have formed by infiltration from overlying
sandy units. Paleo-current data indicate most of the gravel beds were d e
posited by offshore-moving currents (some of which were channelized) during
periods of maximum storm activity. The tops of some of the gravels have
onshore-dipping ripple cross-laminations indicating reworking by onshore
waves during periods of smaller storms or post-storm recovery.
The gravels are overlain by fine- t o medium-grained sand which occurs in
beds 0.1-1.2 m thick. Most of the sand is parallel-laminated, although
medium-scale sets of cross-stratified coarse-grained sand approximately
30 cm thick (Fig.4A) formed by landward-migrating megaripples are also
present. The sets show an overall fining-upward sequence due to avalanching
down the slipface forsets. These forsets dip up to 30' in the landward direc-
tion, have an angular base, and show a distinct brinkpoint (Fig.4A), similar
t o that produced by migrating megaripples described by Clifton et al. (1969).
These units occur as single sets of cross-stratified sand (i.e. no cosets), sug-
gesting a large spacing between adjacent megaripples and/or relatively slow
rates of deposition.
Most of the sand shows evidence of onshore directions of sediment trans-
port, however some offshore movement of sand occurred as well. In some
cases the gravel grades upwards into the parallel-laminated sands, suggesting
442
Fig.4. Outcrop photographs of the Santa Cruz coastal terrace sediments, northern
Monterey Bay, California (arrow indicates direction of land when sediments were being
deposited). A . Cross-laminated sand deposited by a n onshore-migrating, lunate megaripple
( L M R ) with well developed brink point. Note the overlying graveliferous storm deposits.
B . Parallel-laminated t o hummocky cross-stratified sand overlain by a thick, graveliferous
storm deposit. C. Complexly cross-stratified sand and gravel within t h e surf zone deposits
(viewed parallel t o the paleo-shoreline). D . Cosets of cross-stratified pebbly sand
deposited by longshore-migrating megaripples (viewed parallel t o t h e paleo-shoreline). E .
Rip channel complex ( R C C ) within t h e inner part of the surf zone assemblage, overlain
by foreshore deposits. F . Poorly stratified pebble gravels deposited a t the “toe-of-beach”.
deposition during the waning current of a single storm event (cf., Kumar and
Sanders, 1974). In other cases, the contact is abrupt, and appears to record
separate storm events of different magnitudes. Hummocky cross-stratifica-
tion (Fig.4B) is present in only one area, where it grades both laterally and
vertically into parallel laminations. The absence of “fairweather” deposits
is indicated by the lack of any bioturbation or ripple cross-stratification.
443
The storm-dominated sediments are overlain by a 3.5 m thick surf zone

assemblage of sediments which can be further subdivided into an outer and
inner zone on the basis of dominantly longshore and offshore sediment
transport, respectively. The outer surf zone assemblage is approximately
1.5 m thick, and mainly consists of cross-stratified coarse sand and pebble
gravel (Fig.4C). In some places the gravel occurs as isolated linear ripple
form sets ( L = 0.6-1 m, H = 5-7 cm); elsewhere they are present at the base
of 40-60 cm thick cosets of cross-stratified sand (Fig.4D) deposited by the
migration of two- and three-dimensional megaripples ( H = 30 cm, I, = 3 m).
The cross-stratified sediments all show a dominantly longshore direction of
sediment transport, indicating formation in a longshore trough landward
of the breaker zone (cf., Davidson-Arnott and Greenwood, 1974, 1976;
Hunter et al., 1979). Both two- and three-dimensional megaripples of similar
dimensions ( H = 25 cm, L = 2.5 m) can be seen in longshore troughs along
the northern parts of Monterey Bay today. In contrast t o cross-stratified
sands deposited farther offshore, the strata deposited within the surf zone
typically form cosets of cross-stratified sand, and appear to have been
deposited by rapidly migrating megaripples with relatively small wavelengths
(approximately 3 m). Where one megaripple climbed over another, the base
of the upper set of cross laminations is typically concave t o tangential,
rather than angular. In addition, the forsets have more shallow dips (typically
20‘), and show no evidence of a distinct brinkpoint.
These sediments are overlain by u p t o 2 m of inner surf zone deposits
which are characterized by coarse pebble gravels and interbedded, fine t o
coarse sand. The gravel beds range in thickness from a few centimeters t o
0.75 m; they are generally thicker and more abundant at the top of the sec-
tion. The gravels may be structureless, horizontally stratified, or form sets
of cross-stratification up t o 75 cm thick (Fig.4E). Where present, paleo-
current data is predominantly offshore, suggesting the gravels were deposited
by offshore-flowing currents within the inner rough zone of Clifton et al.
(1969). The coarsest gravels (long axis up t o 10 cm) are found at the top of
this section, which is interpreted to represent “toe-of-beach” deposits
(Fig.4F) from which estimates of paleo-water depth can be made. Some
thin interbedded zones of planar laminated sands (up t o 30 cm thick) are
also present within this zone, locally dipping up t o 25” landward. These
probably record the partial preservation of small bars migrating through the
surf zone.
Foreshore deposits
This unit consists of up t o 5 m of inclined parallel-laminated, well-sorted,

medium- t o coarse-grained sand and pebble gravel. The coarsest sediment
occurs at the base of this unit, where layers of imbricated pebbles are locally
common. Bioturbation is also restricted t o the lowermost 0.5-0.75 m of
this unit, where it may approach 100% in sandy sediments. The seaward-
dipping laminated sands show abundant low-angle truncations (Fig.5A),
Fig.5. Swash zone deposits within the Santa Cruz coastal terrace sediments, northern Monterey Bay, California. A . Parallel-laminated
swash zone (foreshore) deposits. Note the gentle seaward dips. B . Discontinuous laminae deposited by migrating antidunes within the
swash zone.
typical of swash laminations (Thompson, 1937); inverse grading within these
laminations is also relatively common (Clifton, 1969). Most of the lamina-
tions are characterized by extreme lateral continuity, however, some form
discontinuous lenses (typically 1-2 cm thick and 15-30 cm long) in which
some internal laminations dip at a low angle landward (Fig.5B). This is
characteristic of laminations deposited by antidunes on the beach (Hayes
and Kana, 1976). Thin laminations of heavy minerals are locally present,
typically merging landward t o form placer deposits up to 0.6 m thick. These
deposits are typical of those formed within the swash zone of a high-energy,
microtidal shoreline.
Vertical sequence of sedimentary structures
The Santa Cruz coastal terrace deposits are the result of sedimentation
during a glacio-eustatic rise and fall of sea level. The transgressive marine
deposits are almost completely lacking, however, because of their selec-
tive removal during the erosional transgression which accompanied the
rise in sea level (Fisher, 1961; Dupre, 1975a; Ryer, 1977). Essentially all
of the preserved sediments are those deposited during the interglacial high-
stand and subsequent lowering of sea level. The resultant regressive sequence
of sedimentary structures (Fig. 6 ) is essentially that predicted for a moderate-
t o high-energy, barred nearshore marine environment (Davidson-Arnott and
Greenwood, 1974, 1976; Hunter et al., 1979).
There are, however, some exceptions worthy of note. The gradual
coarsening-upwards trend of mean grain size is almost completely obscured
by the repeated introduction of coarse-grained sands and gravels during
storm events (Fig.2). The graveliferous sediments were introduced into the
shoreface by winter floods along the Pleistocene Pajaro River which was
approximately 5 km to the southeast of the Manresa measured section
(Fig.1). The gravels were transported laterally by longshore currents within
the surf zone and then beyond the surf zone during storm events. In addi-
tion, the general lack of ripple and lunate megaripple cross-stratification
which characterizes the modern zone of wave buildup (outer rough of
MAX BERM HT = 3 5 M
SWASH ZONE
ER H T = 2 7 M
TIDAL RANGE - 1 1M (Hi 2 2M) DEPOSITS
-toe-oI-beaCh-
inner SURF ZONE
outer DEPOSITS
DEPTH
___-
STORM-DOMINATED
DEPOSITS
Fig.6. Facies associations and environmental reconstruction of the Santa Cruz coastal
terrace deposits (hb = depth of bar, h, = depth of trough) Water depths were measured
from mean low tide ( M L T ) , elevation of the berm was measured from mean high tide
(MHT).
446
Clifton et al., 1971), is evidence of the low preservation potential of these fair-
weather features. Storm deposits, characterized by sharp erosional bases
(with or without a basal gravel), overlain by parallel-laminated sand, have
been preferentially preserved stratigraphically below the surf zone deposits,
as previously noted by several authors (e.g., Clifton, 1976; Howard and
Reineck, 1981j. The lack of any significant hummocky cross stratification
in these wave-dominated storm deposits is also noteworthy, however it prob-
ably simply reflects the relatively shallow depth (mainly <10 m) in which
they were deposited. The highly bioturbated, lower shoreface deposits
represent a zone where the rate of biogenic reworking is greater than the
rate of storm deposition. It appears to correspond to the “transitional zone”
which presently occurs in water depths of 9-18 m off the coast of southern
California (Howard and Reineck, 1981).
RECONSTRUCTION O F PALEO-WAVE CONDITIONS
The main variables controlling nearshore marine sedimentation along wave-

dominated coastlines are: (1) rate of sediment supply; (2) grain size; (3)
water depth; and (4)wave height and period (e.g. Komar, 1974; Clifton,
1976). Grain size and associated sedimentary structures are observable on
the outcrop; the other variables must either be assumed or calculated, as
discussed be1ow.
Calculation of paleo-water depths
The reconstruction of paleo-depth conditions is relatively easy if we can

assume no change in the relative position of sea level during deposition of
the sediments, and no compaction following deposition. Under such condi-
tions, the water depth is approximated by the vertical distance a deposit
occurs below sediment deposited at sea level (cf., Klein, 1974). Thus storm-
generated sediments located 3 m below the top of the “toe-of-beach”
deposits would have been deposited 3 m below low tide. Such a thickness/
sea-level curve is shown on Fig.7 (curve A).
The sediments of the Santa Cruz terrace are relatively young sands and
gravels with no visible signs of significant compaction. In addition, the
amount of tectonic uplift was negligible over the relatively short interval of
time during which the sediments were deposited. The assumption that sea
level did not change during the deposition of the sediment seems less valid,
however. Where exposed, the marine sediments are approximately 1.5 km
seaward of the maximum transgressive shoreline (Duprk, 1975b). If we
assume no change in sea level occurred during the deposition of these sedi-
ments, then the maximum water depth 1.5 km offshore would equal the
maximum thickness of the sediments deposited below sea level (i.e., 6 m).
Studies by Bradley and Griggs (1976) show that the gradient of the wave-
cut platform of the Santa Cruz coastal terrace is essentially the same as that
of the modem day shelf (0.01-0.015), where the maximum water depth
441
0
SEDIMENTATION
h SEA LEVEL RATE
: l
L
c
A STABLE HIGHSTAND CONSTANT
a 2 B CONTINUOUSLY DROPPING C O N S T A N l
E
c
~
Y
CONTINUOUSLY OROPPING INCREASINO RATE WITH
v)
3 DECREASING DEPTH
v)
B
gY 4
Non-bioturbated
-5
0
r
+ 6 WAVE-CUT PLATFORM
5 10 15
PALEO-DEPTH ( m e t r e s )
Fig.7. Selected curves relating stratigraphic thickness below toe-of-beach deposits with
paleo-water depths. Curve C was used for the calculation of water depths in this paper.
Dashed line indicates contact between highly bioturbated (lower shoreface) and non-
bioturbated (upper shoreface) deposits.
1.5 km offshore today is approximately 16 m. This suggests that much of

the deposition must have occurred during the late interglacial lowering of
sea level, with a drop of up to 1 0 m having occurred during the deposition
of the sediments (Fig.8). This is consistent with the fact that there has been
almost no seaward progradation of the shoreline during the last 5000 yrs
of the present highstand of sea level, presumably because of the relatively
high wave regime of the Pacific Coast and low sediment input (Duprh,
1975a). Why progradation should have preferentially occurred during the
early stages of sea level lowering is unclear; however, it could be the result
of the increased rate of sediment influx due to fluvial entrenchment which
accompanied the sea level lowering (Dupr6, 1975a), increased wave atten-
uation due to the decreased gradient of the newly emergent shelf (Bradley
and Griggs, 1976), or some element of both.
A . SANGAMONIAN (?) H I G H S T A N D OF S E A L E V E L
Shoreline
Angle max water depth-16 m
Wavecut Platform
MEASURED SECTION
7 J - p
1.5 Km.
6. E A R L Y WISCONSINAN DROP IN S E A L E V E L 8 m. ( m i d
Fig.8. Configuration of the shelf and water depths during deposition of the Santa Cruz
coastal terrace. Overlying eolian deposits are not shown for the purpose of clarity.
448
If the sediments at the base of the section were deposited in approximately

16 m water depth, and if the rates of sedimentation and sea-level drop were
constant, then a linear thickness/water-depth curve can be made (Fig.7,
curve B) t o estimate maximum water depths. It seems more likely, however,
that the rate of sedimentation was not constant, but rather was greatest
near the shoreline, as implied by the non-linear curve (C) in Fig.7. This type
of curve is also preferred on the basis of the predicted depth of the highly
bioturbated, lower shoreface deposits. The transition from non-bioturbated
t o highly bioturbated sediment in sou them California presently occurs at
approximately 9 m water depth (Howard and Reineck, 1981). This transi-
tion is predicted to have occurred in water depths of 9-10 m using the non-
linear curve (curve C), as opposed to 1 4 m water depths using the maximum
water depth curve (curve B). For these reasons, I have chosen the non-
linear curve (curve C) to determine paleo-water depths.
Calculation of paleo-wave heigh ts
Several methods exist for the determination of ancient wave conditions

from the stratigraphic record. Perhaps the most accurate of these are the
equations which relate wave height ( H ) to maximum orbital velocity
(Urnax).Two such equations are provided below:
H = T sinh (27r h / L ) Uma/m (1)
H = T sinh (2 n h l L ) (U,,, -A Urn,,/2)/n (Clifton, 1976) (2)
The first equation is based on Airy wave theory, whereas the second is a
modified Stokes (second-order) equation which assumes negligible mass
transport. Preliminary calculations using probable wave periods, wave
heights, and water depths indicated that the waves responsible for the storm
layers in the terrace deposits were relatively shallow-water waves, thus
strictly speaking the Airy wave theory is inapplicable (see US.Army Coastal
Engineering Research Center, 1975, fig. 2-7). In fact, the use of eq. 1 pro-
duces relatively small differences (see Table I), typically less than 5%,none-
theless, the Stokes equation was used in this study.
In order t o calculate wave heights using either equation, it is necessary to
know or t o approximate the water depth ( h ) , wave period ( T ) , and maxi-
mum orbital velocity ( Urn,,). Paleo-water depths were estimated using
curve C (Fig.7), as discussed previously. The wave period was assumed to be
10 s, inasmuch as approximately 85% of the waves in Monterey Bay today
have a period of 8-12 s (based on SSMO data summarized by Arnal, 1973).
Only 8.5% of the waves have periods less than 8 s. It should be noted that
the use of 8 or 12 s waves (rather than 10 s waves) produces relatively
small differences for wave heights in shallow water (e.g., 3%deviation in 5 m
water depths).
In the absence of data on ripple morphology, the minimum critical
threshold value of Urn,, for the movement of sand can be determined as a
function of wave period, grain size, and associated bedform (Komar, 1974;
449
TABLE I
Estimate of minimum storm wave heights (assuming T = 10 s), using eqs. 1 (Airy) and 2
(Stokes), for 10 selected storm deposits (A-J). Urn,, = maximum orbital velocity;
AU,,, = maximum orbital velocity asymmetry; h = water depth; H = wave height;
Ho = offshore wave height (ignoring effects of refraction and frictional attenuation)
A 0.7 0.05 11 1.7 1.7 1.7

B 1.25 0.05 8 2.5 2.4 2.4
C 1 .o 0.2 7 1.7 1.6 1.6
D 1.0 0.2 6.8 1.7 1.6 1.6
E 1.1 0.05 6.5 1.9 1.9 1.6
F 1.0 0.05 5 1.5 1.5 1.3
G 1.0 0.05 4 1.3 1.2 1.0
H 2.0 0.4 3.5 2.5 2.2 1.9
I 2 .o 0.4 3.4 2.5 2.2 1.9
J 0.9 0.05 3.2 1.1 1.0 0.8
Average H, = 1.6 m.
Clifton, 1976; Dingler, 1979). For the purpose of this study, mean grain
size was determined by sieving; bedforms (e.g., lunate megaripples, upper
flow regime plane beds) were inferred on the basis of preserved sedimentary
structures. Given these data, critical values of U,,, (and AU,,, in the case
of the Stokes equation) were determined using graphs provided by Clifton
(1976).
The wave height as calculated by this procedure is a minimum wave height
at the site where the sediment was deposited. In fact, this minimum height
is typically greater than the minimum offshore wave height (H,) because of
the shoaling effect. This can cause .some confusion unless nearshore wave
heights are converted to their offshore wave equivalents for the purpose of
comparisons. Hb represents the offshore wave height ignoring the effects of
wave refraction and frictional attenuation. It can be calculated from the
ratio H/HL provided in tables in the Corps of Engineers Shoreline Protection
Manual (US.Army Coastal Engineering Research Center, 1975, vol. 3).
The assumption of no significant changes in wave height due to refraction
appears to be valid, as most of the paleo-waves were from the southwest, and
southwesterly waves today have a refraction coefficient (K,) of approxi-
mately 1 (i.e., no refraction; Johnson, 1953). The effect of decreasing wave
heights due to frictional attenuation was considered. negligible, in part be-
cause of the relatively steep offshore gradients (0.01-0.015); however, this
could represent a source of error (see Goldsmith, 1976).
Ten storm deposits were studied in order to calculate the height (and
frequency of occurrence) of waves responsible for their deposition (Table I).
These storm deposits are representative of the approximately 70% of the
sediment beneath the surf zone assemblage that was deposited in response
t o shoreward propagating wave trains. The remaining 30% of the sediment
450
deposited beyond the surf zone consists of pebble gravels and sands de-
posited by offshore-flowing storm currents (e.g., rip currents, wind-forced
currents, and perhaps storm-surge ebb currents),
The minimum offshore wave height (Yb) as calculated for the ten storm
deposits averaged 1.6 m (range 0.8-2.4 m); such waves presently occur in
Monterey Bay no more than 15% of the time (Arnal, 1973). Presumably
these waves represent relatively frequent, small storms and post-storm
recovery periods. The smaller waves which occurred during the remaining
85% of the time (i.e. “fairweather conditions”) were probably capable of
reworking the tops of the deposits, however their effects were not preserved
because of subsequent reworking by storm waves. The coarser-grained storm
sediments deposited by offshore-flowing storm currents presumably reflect
larger storm events, however there is presently no method by which the
height of the waves associated with these storm-generated currents can be
derived.
The methods described above allow an estimate of the minimum size and
recurrence interval of waves responsible for the deposition of most of the
sediments deposited beyond the surf zone. It is also possible to estimate the
wave conditions under which more shallow water sediments were deposited
based on: (1)thickness of the surf zone assemblage; and (2) thickness of the
swash zone assemblage.
The maximum thickness of the surf zone deposits is a function of the
maximum depth of scour within the surf zone (either in the longshore
trough or rip channels). In the case of the Santa Cruz coastal terrace deposits
(Fig.2), the maximum thickness is 2.8 m, which corresponds to a maximum
depth of scour in the longshore trough of 4.4 m below low tide (using curve
C, Fig.7). If the associated longshore bar (Fig.5) during storms was approxi-
mately 1 m higher than the adjacent trough (cf., Bascom, 1964; Davidson-
Arnott and Greenwood, 1976; Hunter et al., 1979; Sallenger et al., 1983),
then the maximum depth of the bar crest during storms was approximately
3.4 m. The calculated depths to the trough (4.4 m) and t o the bar (3.4 m)
are similar to the observed depths to the storm-generated trough (3.2 m
below mean sea level) and bar (4.4 m) measured by Sallenger et al. (1983)
in the southern Monterey Bay region. The ratio of trough depth (h,) to bar
depth (h,) is virtually identical (Pleistocene = 1.29; Modern = 1.31), and is
within the range of many modern and experimental bar systems (see
summary by Greenwood and Davidson-Arnott, 1975, p. 144).
It is possible to estimate the height of the breaking waves by using the
depth of the bar crest t o approximate the depth at which the waves are
breaking. The relationship between the height of breaking waves (H,)
and breaker depth (h,) for solitary waves is commonly approximated by the
following:
yb = Hb/hb = 0.78 (Weigel, 1964) (3)
Assuming hb = 3.4 m below low tide, we can use this value to estimate that
the sediment in the surf zone was deposited in response to breaking waves no
45 1
larger than 2.7 m high. This is equivalent to a maximum offshore wave

height (Hb) of 2.2 m, which presently occurs approximately 8%of the time.
The actual value of y b depends on wave shape, offshore wave steepness,
and beach slope, however, with experimental values ranging from 0.7 to 1.2
(Galvin, 1972; Komar, 1976). Such values serve to limit the likely range of
maximum wave conditions (i.e. H b = 2-3.4 m), with the 2.2 m waves being
used as a representative value.
The thickness of the swash zone deposits also provides an indication of
the height of waves responsible for their deposition. The maximum thick-
ness is a function of the paleo-tidal range and the maximum height of the
berm above high tide. The mean tidal range during the last interglacial is
assumed to have been the same as the present (1.1m), thus the maximum
berm height can be estimated as being equal to the total thickness of the
swash zone deposits (4.5 m) minus the paleo-tidal range (Fig.5). This sug-
gests a maximum berm height of approximately 3.5 m, which can be related
t o associated offshore wave heights by the following equation:
Max. berm ht. = 1.3 H,K, (Bascom, 1964) (4)
If the characteristic refraction coefficient (K,) = 1 (as discussed earlier),
then the swash zone sediments were deposited in response t o waves whose
offshore heights were no greater than approximately 3 m; the largest of these
waves presently occur approximately 2% of the time.
C 0NCL USION S
The Santa Cruz coastal terrace deposits consist of a regressive sequence of

marine sediments deposited in a high-energy, barred nearshore marine
environment during the end of the last interglacial highstand of sea level.
This sequence can be subdivided into several depth-dependent facies on the
basis of paleocurrent data and vertical sequence of sedimentary structures.
The lower shoreface deposits consist of highly bioturbated, parallel-
laminated sands deposited in water depths of approximately 10-16 m. These
are overlain by a sequence of interbedded sand and gravel deposited in r e
sponse to storm waves in water depths of approximately 5-10 m. These are,
in turn, overlain by a surf zone assemblage of sediments consisting of com-
plexly cross-stratified sands and gravels deposited in water depths of 0-5 m.
The top of the marine section consists of sediments deposited in the swash
zone, where storm berms were built up t o 3.5 m above high tide.
An analysis of the magnitude and frequency of storm conditions respon-
sible for the deposition of these deposits indicates that they preferentially
record storm events of intermediate frequency and magnitude (cf., Wolman
and Miller, 1960). In addition, it appears that different facies within the
high-energy, nearshore environment record events of different recurrence
intervals. Sediments within the surf zone were deposited under relatively
small storm and post-storm conditions, whereas sediments farther offshore
were deposited by relatively larger, less frequent storm events (cf., Hunter
et al., 1979).
452
ACKNOWLEDGEMENTS
This study was initiated as part of my dissertation at Stanford University,

under the supervision of E.I. Rich; J.C. Ingle, Jr. and W.R. Dickinson also
served as committee members. Additional work was done as part of a
regional study of the Quaternary sediments in the Monterey Bay region,
under the supervision of E.E. Brabb (U.S. Geological Survey). Many other
people aided me throughout the course of this study, however I would
like to especially thank H.E. Clifton (U.S. Geological Survey), who showed
me that many sediments record the processes by which they were formed,
and Brian Greenwood (Univ. of Toronto), for his many useful editorial com-
ments. I would also like to acknowledge partial funding from the University
of Houston Geology Foundation.
REFERENCES
wavelength, textural composition, and shape. J. Geol. SOC. London, 136:
6 7 3-6 82.
Arnal, R.E., 1973. Marine geology of Monterey Bay near Moss Landing. In: Sand Trans-
port Studies in Monterey Bay, California, Moss Landing Tech. Rep. 73-5, pp.III-1-
111-21.
Bascom, W., 1964. Waves and Beaches. Garden City, Doubleday, 267 pp.
Bourgeois, J., 1980. A transgressive shelf sequence exhibiting hummocky stratification:
The Cape Sebastian sandstone (Upper Cretaceous), southwestern Oregon. J. Sedi-
ment. Petrol., 50: 681-702.
Bradley, W.C. and Griggs, G.B., 1976. Form, genesis and deformation of Central
California wave-cut platforms. Geol. SOC.Am. Bull., 87 : 433-449.
Brenchley, P.J., Newall, G. and Stanistreet, I.G., 1979. A storm surge origin for sandstone
beds in an epicontinental platform sequence, Ordovician, Norway. Sediment. Geol.,
22: 185-217.
Campbell, C.V., 1971. Depositional model, Upper Cretaceous Gallup beach shoreline,
Ship Rock area, northwestern New Mexico. J. Sediment. Petrol, 41: 395-409.
Clifton, H.E., 1969. Beach lamination: nature and origin. Mar. Geol., 7: 553-559.
Clifton, H.E., 1976. Wave-formed sedimentary structures - a conceptual model: In:
Clifton, H.E., 1981. Progradational sequences in Miocene shoreline deposits, southeastern
Caliente range, California. J. Sediment. Petrol., 51 : 165-184.
Clifton, H.E. and Thompson, J.K., 1978. Macraronichnus segregafis: a feeding structure
of shallow marine polychaetes. J. Sediment. Petrol., 48: 1293-1302.
Clifton, H.E., Hunter, R.E. and Phillips, R.L., 1971. Depositional structures and proces-
ses in the non-barred high-energy nearshore. J. Sediment. Petrol., 41: 651-670.
coast, Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davis, Jr. and R.L.
453
Dupr6, W.R., 1975a. Quaternary history of the Watsonville lowlands, north-central

Monterey Bay region, California. Unpubl. Ph.D. Diss., Stanford University, Palo Alto,
Calif.
Dupr6, W.R., 1975b. Maps showing geology and liquefaction potential of the Quaternary
deposits in Santa Cruz County, California. U.S. Geol. Surv., Misc. Field Studies Map,
MF-643, 2 sheets, scale 1:62,500.
Dupre, W.R., Clifton, H.E. and Hunter, R.E., 1980. Modern sedimentary facies of the
open Pacific coast and Pleistocene analogs from Monterey Bay, California. In: M. E.
Fields et al. (Editors), Quaternary Depositional Environments of the Pacific Coast.
Pacific Coast Paleogeography Symposium 4, Pacific Section, SOC. Econ. Paleontol.
Mineral., Los Angeles, Calif., pp.105-120.
Fisher, A.G., 1961. Stratigraphic record of transgressing seas in light of sedimentation on
the Atlantic coast of New Jersey. Bull. Am. Assoc. Pet. Geol., 45: 1656-1666.
Flint, R.F., 1971. Glacial and Quaternary Geology. Wiley, New York, N.Y., 892 pp.
Galvin Jr., C.J., 1972. Waves breaking in shallow water. In: R.E. Meyer (Editor), Waves
on Beaches and Resulting Sediment Transport. Academic Press, New York, N.Y.,
pp.4 13-456.
Goldsmith, V., 1976. Continental shelf wave climate models: Critical links between shelf
hydraulics and shoreline processes. In: R.A. Davis, Jr. and R.L. Ethington (Editors),
Beach and Nearshore Sedimentation. SOC. Econ. Paleontol. Mineral., Spec. Publ.,
24: 24-47.
tary processes, Kouchibouguac Bay, New Brunswick, Canada. In: J. Hails and A. Cam
(Editors), Nearshore Sediment Dynamics and Sedimentation. Wiley, London,
pp. 1 23-1 50.
Hamblin, A.P. and Walker, R.G., 1979. Storm-dominated shallow marine deposits: the
F e r n i e K o o t e n a y (Jurassic) transition, southern Rocky Mountains. Can. J. Earth Sci.,
1 6 : 1673-1690.
Harris, D.L., 1972. Wave statistics for coastal regions. In: D.J.P. Swift, D.B. Duane and
O.H. Pilkey (Editors), Shelf Sediment Transport: Process and Pattern. Dowden,
Hutchinson and Ross, Stroudsburg, Pa., pp.99-125.
Hayes, M.O., 1967. Hurricanes as geologic agents: case studies of hurricane Carla, 1961,
and Cindy, 1963. Rep. Inv. 61, Univ. Tex., Austin, Bureau Econ. Geol., 54 pp.
Hayes, M.O. and Kana, T.W., 1976. Terrigenous Clastic Depositional Environments.
Tech. Rep. 11-CRD, Coastal Research Div., Univ. S. Carolina, Columbia, S.C., 301 pp.
Howard, J.D., 1972. Trace fossils as criteria for recognizing shorelines in the stratigraphic
record. In: J.K. Rigby and W.K. Hamblin (Editors), Recognition of Ancient Sedimen-
tary Environments. SOC.Econ. Paleontol. Mineral., Spec. Publ., 16: 215-225.
Howard, J.D. and Reineck, H.-E., 1981. Depositional facies of high-energy beach-to-
offshore sequence: Comparison with low energy sequence. Bull. Am. Assoc. Pet.
Geol., 65: 807-830.
Hunter, R.E. and Clifton, H.E., 1982. Cyclic deposits and hummocky cross-stratification
of probable storm origin in Upper Cretaceous rocks of the Cape Sebastian area, south-
western Oregon. J. Sediment. Petrol., 52: 127--143.
Ingle Jr., J.C., 1966. The Movement of Beach Sand. (Developments in Sedimentology, 5)
Elsevier, Amsterdam, 221 pp.
Johnson, J.W., 1953. Engineering aspects of diffraction and refraction. Trans. Am. SOC.
Civ. Eng., 118: 617-648.
Klein, G. deV., 1974. Estimating water depths from analysis of barrier island and deltaic
sedimentary sequences. Geology, 2: 408- 412.
tions and environments. J. Sediment. Petrol., 44: 169-180.
N.J., 429 pp.
454
Komar, P.D. and Miller, M.C., 1974. Sediment threshold under oscillatory waves. Am.
SOC.Civ. Eng., Proc. 14th Conf. on Coastal Engineering, pp.756-775.
Kreisa, R.D., 1981. Storm-generated sedimentary structures in subtidal marine facies with
examples from the middle and upper Ordovician of southwestern Virginia. J. Sediment.
Petrol., 51: 823-848.
Kumar, N. and Sanders, J.E., 1974. Characteristics of shoreface storm deposits: modern
and ancient examples. J. Sediment. Petrol., 46: 145-162.
Miller, M.C. and Komar, P.D., 1980. A field investigation of the relationship between
oscillatory ripple spacing and the near-bottom water orbital motions. J. Sediment.
Petrol., 50: 183-191.
Morton, R.A., 1981. Formation of storm deposits by wind-forced currents in the Gulf of
Mexico and the North Sea. In: S.-D. Nio, R.T.E. Shuttenhelm and Tj.C.E. van Weering
(Editors), Holocene Marine Sedimentation in the North Sea Basin. Int. Assoc. Sedi-
mentol., Spec. Publ., 5 : 385-396.
Nelson, C.H., 1982. Modern shallow-water graded sand layers from storm surges, Bering
shelf: a mimic of Bouma sequences and turbidite systems. J. Sediment. Petrol., 52:
537- 54 5.
Roep, Th.B., Beets, D.J., Dronkert, H. and Pagnier, H., 1979. A prograding coastal se-
quence of wave-built structures of Messinian age, Sorbas, Alameria, Spain. Sediment.
Geol., 22: 135-163.
Ryer, T.A., 1977. Patterns of Cretaceous shallow-marine sedimentation, Coalville and
Rockport areas, Utah. Geol. SOC.Am. Bull., 88: 177-188.
Sallenger Jr., A.H., Howard, P.C., Fletcher 111, C.H. and Howd, P.A., 1983. A system
for measuring bottom profile, waves and currents in the high-energy nearshore en-
vironment. Mar. Geol., 51: 63-76.
Tanner, W.F., 1971. Numerical estimates of ancient waves, water depths and fetch.
Sedimentology, 1 6 : 71-88.
Thompson, W.O., 1937. Original structures of beaches, bars, and dunes. Geol. SOC. Am.
Bull., 48: 723-752.
U.S. Army Coastal Engineering Research Center, 1975. Shore Protection Manual. U.S.
Govt. Printing Office, Washington, D.C., 3 vols.
Vos, R.G. and Hobday, D.K., 1977. Storm beach deposits in the late Palaeozoic Ecca
Group of South Africa. Sediment. Geol., 1 9 : 217-232.
Weigel, R.L., 1964. Oceanographical Engineering. Prentice Hall, Englewood Cliffs, N.J.,
532 pp.
Wolman, M.G. and Miller, J.P., 1960. Magnitude and frequency of forces in geomorphic
processes. J. Geol., 68: 54-74.
Marine Geology, 60 (1984) 455-473 455
RECONSTRUCTION OF ANCIENT SEA CONDITIONS WTH AN

EXAMPLE FROM THE SWISS MOLASSE
PHILIP A. ALLEN
Department o f Geology, University College, P.O. Box 78, Cardiff C F l I X L (United
Kingdom)
(Received March 15, 1983; revised and accepted August 29, 1983)
ABSTRACT
Allen, P.A., 1984. Reconstruction of ancient sea conditions with an example from the
Swiss Molasse. In: B. Greenwood and R.A. Davis, Jr. (Editors), Hydrodynamics and
Sedimentation in Wave-Dominated Coastal Environments. Mar. Geol.. 60: 455-473.
Ancient sea conditions can be estimated from the grain size, spacing and steepness
of preserved ripple-marks. The element of greatest uncertainty in such reconstructions is
the relationship between near-bed orbital diameter of water particles and the ripple
spacing. This relationship is simple for vortex ripples of high steepness but is problem-
atical for the low-steepness forms known as post-vortex, rolling-grain or anorbital ripples.
The existence field for wave ripples is between the threshold velocity for sediment
movement and the onset of sheet flow, most low-steepness forms occurring close to the
bed planation threshold. A range of maximum period of formative waves can be obtained
using combinations of orbital diameter and orbital velocity, assuming linear wave theory
to be a reasonable approximation.
Probable wave heights, wave lengths and water depths can be investigated using the
transformation of wave parameters in shallowing waters and the constraints on wave
dimensions provided by the wave-breaking condition. Given reasonable estimates of
wave height, crude estimates of wave power allow a comparison of ancient wave-
influenced sequences with modern counterparts.
Wave ripple-marks preserved in the Upper Marine Molasse of western Switzerland have
been investigated. Results, which are in agreement with regional geology, suggest deposi-
tion in a seaway of approximately 100 km width, where moderate period waves (2' =
3-6 s) were generated. The depositional facies belts were adjusted to the prevailingwaves,
tides and fluvial outflows.
INTRODUCTION
Since Harms (1969), Tanner (1971) and Komar (1974), with varying
degrees of rigour, proposed the use of preserved wave ripple-marks in re-
constructing ancient wave conditions, surprisingly few ancient sequences
have been analysed in this way. Further encouragements both from theo-
retical and empirical (Clifton, 1976; Allen, 1979; Dingler, 1979; Miller and
Komar, 1980a) and field studies (Newton, 1968; Cook and Gorsline, 1972;
Stone and Summers, 1972; Dingler and Inman, 1977; Miller and Komar,
1980b) have not yet resulted in a flourish of case-studies of ancient marine

456
or lacustrine sedimentary basins. Allen (1981a) analysed a Devonian lacus-

trine basin-fill and Homewood and Allen (1981) studied the marine sedi-
ments of the “Upper Marine Molasse” of Switzerland, but to the author’s
knowledge very few oceanic coastline deposits have been comprehensively
treated.
There are, of course, formidable assumptions and approximations which
must be made in quantifying ancient wave conditions (Allen, 1981b), but
with knowledge of the pitfalls, important insights into palaeoenvironments
and processes can be acquired. Without knowledge of the assumptions
and approximations, spurious results will be obtained. The purpose of this
paper is to describe the methods used by Homewood and Allen (1981) in
their analysis of the Upper Marine Molasse of western Switzerland and to
synthesize some of the more important elements of particular uncertainty
in the reconstruction of ancient sea conditions. This contribution therefore
acts as something of a cookbook which may encourage workers t o analyse or
re-analyse quantitatively their ancient wave-influenced sequences.
METHODS
Are they wave ripple-marks?
Boersma (1970) distinguished a number of features thought t o be charac-

teristic of wave-generated ripple cross-lamination, including irregular or
catenary-arcuate lower bounding surfaces, bundled upbuilding of cross-sets
within ripple cross-laminated lenses, chevron structures, cross-stratal off-
shoots and form-discordancy. Tanner (1967) and Reineck et al. (1971)
likewise summarized wave ripple-marks in terms of ripple steepness, sym-
metry and crestal arrangement in plan. Wave ripple-marks characteristically
have straight crestlines which commonly bifurcate, low ripple indices and
symmetrical to near-symmetrical profiles.
There have been relatively few studies of the natural variability in mor-
phology and structure of wave ripple-marks. A very wide spectrum of ripple
morphologies has been described from experimental studies (Bagnold, 1946;
Manohar, 1955; Dingler, 1974; Sleath, 1976), ranging from the very flat
varieties which have variously been termed “rolling grain ripples” (Bagnold,
1946; Sleath, 1976; Allen, 1979) and “post-vortex ripples” (Dingler, 1974;
Dingler and Inman, 1976) to the very steep varieties termed ‘vortex ripples’
(Bagnold, 1946; Sleath, 1976) or ‘orbital ripples’ (Clifton, 1976). The iden-
tification of these ripple types in natural environments has not always been
unequivocal.
The relationship between orbital diameter of water particles and ripple

spacing
Miller and Komar (1980a) summarized laboratory experiments seeking

t o determine the relationship between near-bed orbital diameter and ripple
457
spacing. Oscillating beds (Bagnold, 1946; Manohar, 1955; Kalkanis, 1964;

Sleath, 1975, 1976), oscillating water tunnels or U-tubes (Carstens et al.,
1969; Chan et al., 1972; Mogridge, 1973; Brebner and Reidel, 1973;
Lofquist, 1977, 1978) and wave flumes (Yalin and Russell, 1962; Horikawa
and Watanabe, 1967; Mogridge and Kamphuis, 1973; Dingler, 1974; Dingler
and Inman, 1977) have been used to study oscillatory flows. From these
data, for small orbital diameters, the relationship between near-bed orbital
diameter ( d o )and ripple spacing ( A ) is very simple (Fig.1):
X = 0.65 do (1)
(Miller and Komar, 1980, p.178), the flume and U-tube data providing the
best fit to this curve. Equation 1 has a weak Reynolds number dependence
demonstrated independently by Sleath (1976) and Japanese workers (e.g.
Hom-Ma et al., 1965), but can be neglected here for simplicity. As near-bed
orbital diameter increases eq.1 no longer holds, becoming invalid as a
function of grain size:
, , ,X = 0.0028 D'.68 (2)
Orbital Diameter based on H (m)

113
0.1
I I I I I I I I I I I 1 I I I l l I 1 I l l
I GRAIN SIZE
I
1 0
8
0.088-0.177
0.250-0.350 mm
0.500- 0.-
mm
's
0 .
500 -710pm
lnman 1957,Di - ~
Mlller and Komar 1 9 8 0 b y I

8 O A
. 0 0
-
L -
e
P
P
E 0.1 - - 0
MOST EXPERIMENTAL o
DATA
1 I
, . .;
,
1 I I I I l l ( I I I 1
I
I I I I I 1 I 1
1 0.1 1
Orbital Diameter based on H,,,(rn)

Fig.1. Relationship between near-bed orbital diameter of water particles ( d o )and ripple
spacing ( h ) with data sources indicated. Note the poor correlation between d o and h for
d o and h 2 0.2 m. H,,? is significant wave height, H,,, is root mean square of wave
height. Modified from Miller and Komar (1980b).
458
(Miller and Komar, 1980, p.180) where D is the median grain-diameter in

microns and A,, is in centimetres.
Allen (1979) showed that ripple steepness is an important factor influen-
cing wave reconstructions. He plotted the data of Inman (1957), Kennedy
and Falcon (1965), Manohar (1955), Inman and Bowen (1963), Hom-Ma
et al. (1965), Carstens et al. (1969), Lofquist (1978) and Sleath and Ellis
(1978) on a dimensionless ripple spacing ( h / D )versus dimensionless orbital
diameter (d,/D) diagram with ripple steepness as a further variable (Fig.2).
A range of dimensionless orbital diameters could be obtained for a given
ripple steepness. Equation 1 then merely represented one limit on Allen’s
fig.2 (1979, p.676) for vortex ripples, the other limit being determined for
post-vortex or rolling grain ripples (Sleath, 1975, 1976) as:
0.036 < A/do < 0.059 (3)
Vortex ripples
LOW - steepness ripples
101
lo’ lo3 lo4
-
Non dimensional orbital diameter ( d J D )
Fig.2. Occurrence of wave ripple-marks as a function of dimensionless ripple spacing
( h / D ) and dimensionless near-bed orbital diameter of water particles (d,/D). The lower
limit of ripple occurrence is given by h / d , = 0.65 or 1.0 and the upper limits for given
ripple steepnesses (vertical form indices) are given by the series of curves. Note that the
curves represent the upper limit of d o/Dand are not isopleths. After Allen (1979, p.676).
459
Allen (1981) suggested that the use of a wide range of ripple steepness
inevitably led to an unacceptably wide range of estimated orbital diameters,
making wave reconstructions hazardous.
The relationship between ripple spacing and grain and flow parameters
can be expressed as:
= JYdo, v , D ,Ps, P, g, w ) (4)
Since Uo = o d o / 2 , Sleath (1976) was able to express the relationship be-
tween orbital diameter and ripple spacing for “rolling grain ripples” as a func-
tion of four dimensionless variables:
do
- =F
2h W , R , P J P , (Ps - P ) g D / P w v l (5)
where R = U o / ( ov)”’ is a form of wave Reynolds number and p = (o/ 2 ~ ) ~ ’ * .

It can be seen that three of these dimensionless groups contain w and one
contains Uo.Only prior knowledge of U,, and an iterative solution of Sleath’s
equations (1976, p. 78) would allow a solution to be made.
Clearly, the best results will come from those ripple-marks for which an
unambiguous relationship between orbital diameter and ripple spacing exists.
Such vortex ripple-marks are defined by the incidence of flow separation
over a steep crest. Bagnold (1946), J.F.A. Sleath (pers. commun., 1979)
and Allen (1979) suggested limiting ripple steepnesses (expressed as a vertical
form index, VFI; Bucher 1919) of 8.0, 8.3 and 7.5, respectively, and Dingler
and Inman (1977) stated an average value of 6.7 for vortex ripples. It is
worth emphasising that ripple steepness is a function of several fluid, sedi-
ment and flow-related variables but that ripple steepness alone is the pre-
dominant control on the existence of vortex ripples. Extreme caution must
be exercised in analysing low-steepness ripple-marks for wave reconstructions
and it is recommended that attention is focused on vortex ripple-marks
possessing vertical form indices of less than 7.5 and certainly less than 10.
The critical velocities for wave ripple formation
The critical threshold for entrainment under waves is given by a modified

Shields parameter, and is determined by grain and fluid density ( p s , p ) , fluid
viscosity (p), grain diameter (D) and near-bed orbital diameter of water
particles (do). For grain sizes of less than 0.5 mm, Komar and Miller’s (1973)
relation is:
p V:/(p, -p ) g D = 0.21 (do/D)”’ (6)
which is based exclusively on Bagnold’s (1946) data and corresponds to
laminar boundary layers. For grain sizes greater than 0.5 mm, Komar and
Miller (1973) suggested the expression:
pU?/(ps -p ) g D = 0.46 71 ( d 0 / D ) 1 / 4 (7)
derived from the data of Rance and Warren (1969) which applies to
turbulent boundary layers.
460
Ripple steepness is intimately related t o processes in the wave boundary

layer. It is the near-bed curvature-related drift velocities (Sleath, 1975,
1976) which are directly responsible for wave ripple-mark formation
(Bagnold, 1946; Kaneko and Honji, 1979), whereas it is the purely oscil-
latory (simple harmonic) component which causes most grain movement.
Allen (1979, fig.1, p. 675) plotted the purely oscillatory component, U,,,,
against grain diameter for a range of steepness values (Fig.3). Together with
the analysis of Komar and Miller (1976), Allen's compilation shows that
wave ripple-marks occur at orbital velocities well above those at the thresh-
old condition, up to the point at which ripples are destroyed and sheet flow
commences (Fig.3).
Komar and Miller (1976) gave the critical relative stress for ripple disap-
pearance as a function of grain size alone, as:
ec = 0.413 D - ~ . ~ ~ ~ (8)
0 1 1 ' 1 1 1 1 1
NO SEDIMENT MOVEMENT
1 1 1 1 1 1 1 1 1 1 1 1 1 ' 1 1
i
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Sediment particle diameter D ( m m )

Fig.3. Occurrence of wave ripple-marks as a function of maximum near-bed orbital
velocity of water particles ( U - ) and grain size of sediment ( D s o). The curves for ripple
steepness (vertical form index) enclose the existence-field of ripple-marks with a given
steepness. After Allen (1979, p.675).
46 1
or in terms of grain Reynolds number as:

e = 4.40 (urn,,
~ i 4 - O . ~ ~ ~ (9)
both based on Manohar's (1955) data (Fig.4). The simpler method is to go
directly to Allen's (1979) fig.1 which is based on a far wider range of data.
Calculation of wave period
Using linear wave theory, which is adequate for these purposes (Le
Mkhautk e t al., 1969), it is possible to relate near-bottom orbital diameters
and velocities to wave period ( T )with the simple expression:
urn,, = ndoIT (10)
Equation 10 indicates that although the entire range of conditions from the
threshold velocity to the destruction velocity should be studied, it is the
threshold condition (minimum U o ) which gives the maximum wave period.
Maximum wave periods are of greater interest in reconstructing ancient sea
conditions. If one experiments with the limits of the range of do obtained
above, a range of T for both the threshold and destruction condition is
obtained.
The rationale behind calculating T is that the period of gravity waves is
determined by wind strength, duration and fetch. Placing reasonable limits
on wind speed and duration (for example, for typical fair-weather and
typical storm conditions), it is possible to use wave forecasting methods to
derive fetch limits for calculated wave periods (Sverdrup and Munk, 1946;
refined by Bretschneider, 1970, or Darbyshire and Draper, 1963). In fetch-
limited seas the analysis of Neumann (1953) is particularly useful (Fig.5).
Fig. 4. Occurrence of wave ripple-marks as a function of a Bagnold (1963)-Shields (1936)

relative stress ( 0 ) and grain Reynolds number ( R e g ) .The ripple existence-field was based
on the laboratory data of Manohar (1955)' Horikawa and Watanabe (1967) and Carstens
et al. (1969) and the field data of Inman (1957). The bed planation threshold was con-
structed from the data of Manohar (1955) and supported by the theoretical criterion of
Bagnold (1956, 1966). After Komar and Miller (1975, p.701).
462
10
-
(v
E
7
b
r 6
L
k
v)
s5
Q
L.
Q
s4
2
ru
a
3
1
100 km
I0 m/sec
- L k m -
I I l l I I I
20 1098 7 6 5 I
WAVE PERIOD I T ) sec
Fig.5.The energy density spectra for sea conditions where fetch is limiting, after
Neumann (1953).
Calculated fetches for ancient waves are an important asset in palaeogeo-

graphical studies. Homewood and Allen (1981) used fetch length calculated
from Miocene wave ripple-marks to confirm traditional views, based on the
distribution of marine facies, on the size of the peri-Alpine sea in Switzerland.
Allen (1981) used fetch data to postulate the extent of a Devonian lacus-
trine basin where outcrops were incomplete.
Simulation of wavelength, wave height and water depth
The orbital diameter of water particles near the bed is the result of a wave
of period T, height H and wavelength L acting in water depth h . It follows
that it is impossible t o obtain a unique solution t o H , L and h ; it is only
possible to obtain combinations of parameters. This is why estimation of H,
L and h is well suited to computer simulation (see Komar, 1974, for instance).
463
Because deep-water waves are unaffected by water depth, it is possible to

construct deep-water wavelength based purely on wave period, as follows:
L, =gT2/2n (11)
which is a simplification of the general case:
L = (gT2/2n)tanh
(3
As the wave travels into shallower water its form changes so that orbital
diameter of water particles near the bottom is given by:
do = H/sinh(2nh/L) (13)
It is necessary therefore to study the transformations which take place in
a wave of known period T as it moves into shallower water. This can be
done in various ways, but a relatively painless method is to calculate hlL,
for the primary field of interest for water depth and use the graphs for Airy
wave transformations provided by Wiegel (1964) (Fig.6). Alternatively Eckart
(1952) gave the approximation:
L = L, abs[tanh(2nh/L, ) ] (14)
The variation of W is of particular interest since it affects near-bed orbital
diameter. Assuming that in order to form the observed ripple-marks the
waves possessed finite near-bottom orbital diameters (i.e., the formative
waves were not deep-water waves), eq.13 can be used t o obtain wave height.
Wave breaking
There is another limit which is critical to the validity of the reconstructed

wave parameters, that of wave breaking. Miche (1944) gives the limiting
steepness for waves in water of finite depth as:
(H/L),, = 0.142 tanh(2nh/L) (15)
and in progressively shallower water (McCowan, 1894) waves break at:
(H/h)lim= 0.78 (16)
Equation 16 is sensitive to beach slope (Ippen and Kulin, 1955). For small
slopes (tan 0 < 0.07):
0.72 < (H/h)iim < 1.18 (17)
and for a reasonable beach slope of 0.003:
(H/h)lim = 0.88 (18)
which is in close agreement with eq.16. Only rarely do geologists have data
on ancient beach slope, so it is normal to simply implement eqs. 15 and 16.
464
I I I I I I I
l 0 C
-
-
-
-
-
-
0.01 - I I I I 1 1 1 1 . I I I l l 1 I I I I I I I
1
Fig.6. T h e shoaling transformations for Airy waves as functions of the ratio of water
depth to deep-water wavelength, h/L,. C,C , and C , are wave celerity (phase velocity),
group velocity and deep-water celerity. H and H , are wave height and deep-water wave
height. L and L, are surface wavelength and deep-water wavelength. n is a shoaling coef-
ficient relating wave celerity to group velocity. After Wiegel (1964).
Energy flux (wave p o w e r )
The motion of waves produces a transfer of energy over the sea surface
which is of great interest to physical oceanographers and coastal engineers. A
wave train possesses a total energy made up of two components. The poten-
tial energy component is caused by water particles being displaced from the
still water level, and the kinetic energy component accounts for the orbital
motions of water particles. The total energy is given by:
E = pgH2/8 t 19)
(Teleki, 1972, p.38; Madsen, 1976, p.72) where E is a surface energy den-
sity per unit width of wave crest. The rate at which this energy is propagated
in the direction of wave advance is the energy flux (or wave power), and is
directly related t o the velocity of the wave train (group velocity Cg) rather
than the phase velocity of individual waveforms. The energy flux per unit
length of wave crest is:
465
P = pgH2Cg/8 (20)
Group velocity can be obtained from Wiegel’s (1964) Airy wave transforma-
tions as the dimensionless ratio Cg/C, (Fig. 6) by use of the expression for
wave phase velocity (celerity):
C = L/T = (gT/2n) tanh(2nhlL) (21)
where C = C, = gT/2n in deep water.
Energy flux for reconstructed wave conditions should decrease in shal-
lowing waters, the ratio of deep-water energy flux to shallow-water energy
flux providing an index of power attenuation. Homewood and Allen (1981)
used estimates of energy flux to provide modern analogues to the Miocene
Sea of Switzerland by comparison with the energy flux (wave power) data of
Coleman and Wright (1975) and Wright and Coleman (1973).
T H E UPPER MARINE MOLASSE O F SWITZERLAND
The wedge of Tertiary clastic sediments north of the Swiss Alps is tradi-
tionally subdivided into four units (Matter et al., 1980; and Fig.7A). The
lowermost, of Oligocene age, is termed the Lower Marine Molasse and
represents offshore mudstones with storm sandstones and culminates in a
wave-dominated shoreline sequence. The overlying unit, the Lower Fresh-
water Molasse comprises predominantly fluviatile clastic sediments with
some playa and lacustrine sediments. The third unit is the Upper Marine
Molasse of Miocene (Burdigalian) age, consisting of wave- and tide-dominated
shallow marine sandstones and the conglomerates of fringing fan-deltas. The
uppermost unit, the Upper Freshwater Molasse is composed of alluvial fan
and fluviatile elastics and lacustrine deposits.
The Upper Marine Molasse was deposited in a peri-Alpine depression north
of the Alps (Fig.7B) which extended eastward to the Austro-Vienna basin
and southwestward into France. Although wave-formed structures occur
throughout the Upper Marine Molasse, the present study concerns the area
in the vicinity of Fribourg where road cuttings and river gorges provide
spectacular sections through the marine sand bodies.
In the Fribourg area Homewood (1978, 1981) described four facies belts
in the Upper Marine Molasse (Fig.7C). The proximal facies belt, restricted
to the south (Hoffman, 1960), is composed of thick fan-delta deposits which
represent the major feeder systems of sediments from the Alpine hills to the
marine seaway. A coastal facies belt contains abundant tidal sandwaves
(Allen and Homewood, 1984) interbedded with intertidal sandflat deposits
and distributary and tidal channel sandstones. The nearshore facies belt is
constructed of thick, elongate subtidal shoals with shoal crevasse deltas and
intershoal swales. The offshore facies belt contains sandy and pebbly
coquinas deposited as giant-sized flow-transverse tidal bedforms.
Wave ripple-marks are very common in the coastal, nearshore and off-
shore facies belts. Homewood and Allen (1981, pp. 2540-2543) summarized
466
PROXIMAL FLUVIAL NEARSHORE SUBTIDAL

C DlSTRl BUTARY SHOALS
WITH TIDAL INFLUENCE
COQUINA BANKS
TIDAL CHANNELS
Fig.7. A . The four lithostratigraphic subdivisions of the Swiss Molasse after Matter e t al.
(1980). B. Palaeogeographic map of the Upper Marine Molasse of Switzerland with loca-
tion of major fan-deltas, after Hofmann (1960), Rigassi, in Matter et al. (1980) and
Lemcke (1981). F marks city of Fribourg. C. Diagrammatic reconstruction of distribution
of facies belts in the Fribourg area during Burdigalian (Miocene) times (not to scale).
GI = Gibloux fan; G U = Guggisberg fan; TB = transverse bars in tidally influenced dis-
tributaries; TS = tidal sandwaves in coastal belt; S = elongate subtidal shoals in nearshore
belt; C = coquina banks in offshore belt. After Homewood and Allen (1981).
the major findings of a study of ancient wave and tide conditions from these
facies belts.
Reconstructed sea conditions
The wave ripple-marks measured in the field possess the symmetries and
steepnesses shown in Fig.8. Care was taken to omit ripple-marks which were
of questionable origin, in particular those resulting from probable combined
flows of waves superimposed on tidal currents. Such ripple-marks were
somewhat more asymmetric and were commonly associated with the flanks
of tidal sandwaves. A large number of ripple-marks have steepnesses (large
VFI) that make estimation of ancient wave conditions hazardous because of
the wide range of possible orbital diameters. Fifty-three out of 150 ripple-
467
0.5- 0 .. ... . . . . 0
*.
0
......... .. ... ..
.
.
0 .
. : ..:.
0.
...to- ..
. . : . . .
0 .
0 . 0 .
0.4 -
. :. 0
.
.
.
0.3
0.2 -
-
. .
0.1 -
Fig.8. Plot of an index of symmetry against ripple steepness (vertical form index) for
wave ripple-marks from the Upper Marine Molasse near Fribourg. After Homewood and
Allen (1981).
marks are clearly of the vortex type, where eq. 1 can be used with confi-
dence, and only three of these 53 ripple-marks had a spacing greater than
, ,A in eq.2 (Fig.9). Bearing in mind the original scatter of data from eq.1
and difficulties of accurate field measurements, it is justifiable to simply use
the threshold condition for vortex ripples in estimating wave period from
orbital diameter and orbital velocity (Fig.10).
For each locality a table was constructed giving lengths and heights of
formative waves over a range of water depths. Unreasonable combinations
of H, L and h were then eliminated according t o wave breaking criteria. In
this way, an impression of the maximum water depths at which the wave
ripple-marks formed was obtained. An example for one locality is given in
Table I.
Water depths under formative waves were in most cases less than 25 m,
but in extreme cases very high waves near the breaking limit may theoreti-
cally have been responsible for the wave ripple-marks in deeper waters,
perhaps up to 60 m. Such large water depths are unlikely from the facies
associations. Furthermore, the association of wave ripple-marks with tidal
sandwaves showing shallow-stage run-off patterns and miniripples (as at
Illens, map co-ord. 574.50/176.50, Swiss topographic maps No. 252) and
468
I
% = 0.00280'68
. I.
I
I
0-20 . I *
I
I
..
0
0
... 1
I
I
0
..
0 .
. I
1 0 0
O.l5{ . . .'. . I
1 0
0
0
0 t
.. I
.
r 0 0 0 3
./ ../
0 0
0
0 2
0 2
./.
0 0 0
0.5- 1@ 0 0 2 -
/ 0
0
/
Q VORTEX R I P P L E S , ~M ~
A X~
0 I
Fig.9. Plot of median grain size against ripple spacing showing the ripple spacing at which
the linear relationship between orbital diameter and ripple spacing (eq.1) breaks down
(dashed line). Double black circles, vortex ripples, VFI < 7.5; double open circles, vortex
ripples with h > Am=; small dots, lowsteepness ripples, VFI > 7.6. Number of superim-
posed data points also indicated for vortex ripples.
swash bars or flood ramps (as at Fribourg, 578.90/184.70 Map 242) suggests
that water depths were shallow, and most wave ripple-marks may have been
produced under modest fair-weather waves.
The variability of reconstructed wave conditions between localities is
not great, but it may be explained by the relative exposure or sheltering of
sub-environments from wave attack and to the viscissitudes of depth during
the tidal cycle. In the case of the offshore facies comprising coquina banks,
the reconstructed wave periods are generally about 3 s and water depths
must have been less than 10 m for formative waves. Since these ripple-maxks
originate from the facies most distal from the shoreline in the south, it is
inconceivable that more proximal facies were deposited in water depths
469
'COASTAL' ' NEARSHORE ' 'OFFSHORE '
1 2 3 4
T (secs)
5 6 7
L
1 2 3 L
4
T (secs)
5 6 7 7 1 2 3
T (recs)
L
4
TOTAL hloo I
1 2 3 4 5 6 7 8 9
T (secs) T (secs)
Fig.10. Histograms of calculated maximum wave periods of formative waves for the
coastal, nearshore and offshore facies belts and histogram and cumulative frequency
curve for total. Assumes the threshold condition and eq.1, utilizing vortex ripple-mark
data only.
TABLE I
Example of work-sheet for each locality showing wave transformations in shallowing

water depths. Asterisk at h = 20 m indicates unstable wave conditions. Terms defined
in text and caption to Fig.6; h, L and H are in metres
Water (h/L,) (L/L,) (HIH,) Shoaling (C,/C,) Surface Surface wave

depth coefficient wavelength height
( h) (n) ( L) (H)
0.5 0.019 0.33 1.30 0.97 0.33 8.58 0.06

1.0 0.038 0.48 1.10 0.93 0.42 12.48 0.08
5.0 0.192 0.85 0.91 0.68 0.59 22.10 0.31
10.0 0.385 0.99 0.99 0.53 0.53 25.74 0.91
20.0 0.769 1.00 1.00 0.50 0.50 26.00 9.8*
Locality: Cates vers le Lac, Estavayer - Yverdon Road. Mean of maximum wave periods:
4.08 s. Deep-water wave length: 26.0 m.
47 0
substantially greater than this. The apparent landward increase in period of

formative waves (Fig.10) may be due to a steadily increasing fetch for the
northerly or westerly winds, as the Jura coast became more distant and the
Alpine shore was approached.
The calculated values of wave power (or energy flux) fall within a wide
range of approximately 108-106 erg s-’. I t must be stressed that calculated
wave power values are highly sensitive to estimated water depth since h
determines the height of formative waves. Nevertheless, bearing in mind
the probable errors, it is possible to compare the wave power at the coast
of the Burdigalian Sea in the Fribourg area to that of the Danube, Ebro,
Niger and Nile deltas (Table 11; Coleman and Wright, 1975). In contrast,
the wave-dominated deltas such as the SZo Francisco and Senegal are
moulded by considerably more powerful waves than those estimated for
the Burdigalian Sea. Delta morphology is also a function of river input
(discharge effectiveness index) and tidal range. Homewood and Allen’s
(1981) plots using wave power, discharge effectiveness index and tidal range
as the three controlling parameters suggested that the closest affinities of the
Burdigalian coastal systems in the Fribourg area lie with the present-day
deltas of the Niger and possibly Burdekin and Klang.
Comparing the wave power calculated for the offshore coquina bank
facies to that of the coastal swash bar or flood ramp facies, representing the
passage from the open Burdigalian seaway to the Alpine coast, there appears
t o have been a wave attenuation of between 50:l and 1 O O : l .
CONCLUSIONS
Further case-studies of wave-influenced sedimentary sequences are re-

quired to broaden the presently inadequate data base. Substantial progress
TABLE I1
Year-average wave powers for seven of the world’s major deltas and estimated values for
the Burdigalian Sea of western Switzerland (Fribourg area). Wright and Coleman’s ( 1 9 7 3 )
data have been converted to metric c.g.s. units. The Fribourg molasse values are derived
from formative waves in 1 0 m water depth and in depths of less than 2 m for shoreline
wave powers
Delta Year-average wave Year-average wave

power a t 10 m contour power a t shoreline
ergs-’ x 1 0 7 ergs’ x l o 7
Mississippi 190 0.041

Danube 49 0.037
Ebro 172 0.155
Niger 107 2.01
Nile 128 10.17
Sao Francisco 598 30.37
Senegal 285 114.72
Fribou rg M o lasse 10-50 0.1-1.0
471
in the quantification of ancient wave conditions can only be made possible

by advances in our knowledge of the relationship between near-bed orbital
diameter of water particles and ripple-mark spacing. In this respect, linkage
of characteristic ripple geometry or internal structure with flow parameters
is necessary.
The reconstructed sea conditions of the Miocene (Burdigalian) Sea in the
Fribourg area of western Switzerland are of moderate period waves (3-6 s).
Fetch lengths for such waves were of the order of 100 km, placing some
constraint on the minimum size of the seaway. The estimated wave power
of the Burdigalian Sea indicates an affinity with sea conditions off deltas
such as the Ebro and Danube, located in semi-restricted seas, but tidal range
was substantially larger than in these two examples.
ACKNOWLEDGEMENTS
I am grateful to John Allen, Peter Homewood and Darwin Spearing for

their useful comments on the manuscript. Data were collected with the
financial support of the Swiss National Science Foundation, Project 2.242-
0.79.
REFERENCES
Allen, G.P., Laurier, D. and Thouvenin, J., 1979. Etude SCdimentologique du Delta du
Mahakam. Notes e t MCmoires No.15, Compagnie Franqaise des PBtroles, Paris.
wavelength, textural composition and shape. J. Geol. SOC.London, 136: 673-682.
Allen, P.A., 1981a. Wave-generated structures in the Devonian lacustrine sediments of
SE Shetland, and ancient wave conditions. Sedimentology, 28: 369-379.
Allen, P.A., 1981b. Some guidelines in reconstructing ancient sea conditions from wave
ripple marks. Mar. Geol., 43: M59-M67.
Allen, P.A. and Homewood, P., 1983. Mechanics and evolution of a Miocene tidal sand-
wave. Sedimentology, 31: 63-82.
Bagnold, R.A., 1946. Motion of waves in shallow water. Interactions between waves and
sand bottoms. Proc. R. SOC.London, Ser. A, 187: 1--15.
Boersma, J.R., 1970. Distinguishing features of wave-ripple cross-stratification and
morphology. Ph.D. Thesis, University of Utrecht, Utrecht, 6 5 pp. (unpubl.).
Brebner, A. and Reidel, P.H., 1973. A new oscillating water tunnel. J. Hydraul. Res.,
1 1 : 107-121.
Bretschneider, C.L., 1966. Wave generation by wind, deep and shallow water. In: A.T.
Ippen (Editor), Estuary and Coastline Hydrodynamics. McGraw Hill, New York, N.Y.,
pp. 133-196.
Bucher, W.H., 1919. On ripples and related sedimentary surface forms and their palaeo-
geographic interpretations. Am. J. Sci., 47 : 149-210, 241-269.
Carstens, M.R., Nielson, F.M. and Altinbilek, H.D., 1969. Bedforms generated in the
laboratory under an oscillatory flow: analytical and experimental study. U.S. Army
Coastal Eng. Res. Centre, Tech. Memo. 28.
Chan, K.W., Baird, M.H.I. and Round, G.F., 1972. Behaviour of beds of dense particles
in a horizontally oscillating liquid. Proc. R. SOC.London, Ser. A, 330: 537-559.
Clifton, H.E., 1976. Wave-formed sedimentary structures -- a conceptual model. In:
R.A. Davies and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SOC.
Econ. Paleontol. Mineral., Spec. Publ., 24: 126-148.
47 2
Coleman, J.M. and Wright, L.D., 1975. Modern river deltas: variability of processes and
sand bodies. In: M.L. Broussard (Editor), Deltas, Models for Exploration. Houston
Geol. Soc., pp.99-149.
Cook, D.O. and Gorsline, D.S., 1972. Field observations of sand transport by shoaling
waves. Mar. Geol., 1 3 : 31-56.
Darbyshire, M. and Draper, L., 1 9 6 3 . Forecasting wind-generated sea waves. Engineering
(London), 195: 482-484.
Dingler, J.R., 1974. Wave-formed ripples in nearshore sands. Ph.D. Thesis, Univ. of
California, San Diego, Calif., 1 3 6 pp.
wave channel. J. Sediment. Petrol., 49: 287- 294.
Dingler, J.R. and Inman, D.L., 1977. Wave-formed ripples in nearshore sands. Proc.
1 5 t h Conf. Coastal Engineering, pp.2109-2126.
Eckart, C., 1952. The propagation of gravity waves from deep t o shallow water. U.S.
Nat. Bur. Stand. Gravity Waves, Circ., 5 2 1 : 165-173.
Harms, J.C., 1969. Hydraulic significance of some sand ripples. Geol. SOC. Am. Bull.,
80: 363-396.
Hoffman, F., 1960. Materialherkunft, Transport und Sedimentation im Schweizerischen
Molassebecken. Jahrb. St. Gallischen Naturwissensch. Ges., 76: 5-28.
Hom-Ma, N., Horikawa, K. a n d Hajima, R., 1965. A study o n suspended sediment due t o
wave action. Coastal Eng. Jpn., 8: 85-103.
Homewood, P., 1978. Exemples d e sequences de faci& dans la molasse fribourgeoise e t
leur interpretation. Soc. Fribourg Sci. Nat. Bull., 6 7 : 73-82.
Homewood, P., 1981. Facies et environnements d e depbt d e la Molasse de Fribourg.
Eclogae. Geol. Helv., 7 4 : 29-36.
Homewood, P. and Allen, P.A., 1981. Wave-, tide- and current-controlled sandbodies of
Miocene Molasse, western Switzerland. Bull. Am. Assoc. Pet. Geol., 6 5 : 2534- 2545.
Horikawa, K. and Watanabe, A., 1967. A study of sand movement due t o wave action.
Coastal Eng. Jpn., 10: 39-57.
Inman, D.L., 1957. Wave generated ripples in nearshore sands. U S . Technol. Memo.
Beach Erosion Board, 100, 4 2 pp.
Inman, D.L. and Bowen, A.J., 1963. Flume experiments o n sand transport by waves and
currents. Proc. 8 t h Conf. Coastal Engineering, 1: 137-150.
Ippen, A.T. and Kulin, G., 1955. T h e shoaling and breaking of the solitary wave. Proc.
5th Conf. Coastal Engineering, pp.27-49.
Kalkanis, G., 1964. Transportation of bed material d u e to wave action. U.S. Army Corps
Engrs, Coastal Eng. Res. Centre, Tech. Memo., 2, 38 pp.
Kaneko, A. and Honji, H., 1979. Initiation of ripple marks under oscillating water.
Sedimentology, 2 6 : 101-113.
Kennedy, J.F. and Falcon, M., 1965. Wave generated sediment ripples. M.I.T. Dept. Civ.
Eng. Hydro. Lab. Rept. 8 6 , 5 5 pp.
tions and environments. J. Sediment, Petrol., 44: 169-180.
N.J., 4 2 9 pp.
Komar, P.D. and Miller, M.C., 1973. The threshold of sediment movement under oscil-
latory water waves. J. Sediment. Petrol., 4 3 : 1101-1110.
Komar, P.D. and Miller, M.C., 1976. The initiation of oscillatory ripple marks and the
development of plane-bed a t high shear stresses under waves. J. Sediment. Petrol.,
4 5 : 697-703.
Le M6haut6, B., Divoky, D. and Lin, A , , 1969. Shallow water waves: a comparison of
theories and experiments. Am. SOC.Civ. Eng., Proc. 1 1 t h Conf. Coastal Engineering,
pp. 86-9 6.
Lofquist, K.E.B., 1977. A positive displacement oscillatory water tunnel. Coastal Eng.
Res. Centre, Misc. Rep., 77-1, 2 6 pp.
473
Lofquist, K.E.B., 1978. Sand ripple growth in an oscillatory-flow water tunnel. Tech.
Pap. U.S. Coastal Eng. Res. Centre, 78-5, 1 0 1 pp.
Madsen, O.S., 1974. Wave climate of the continental margin (its mathematical descrip-
tion). In: D.J. Stanley and D.J.P. Swift (Editors), The New Concepts of Continental
Margin Sedimentation - Sediment Transport and its Application to Environmental
Management. Short Course Lecture Notes, Am. Geol. Inst., pp.42-108.
Manohar, M., 1955. Mechanics of bottom sediment movement due to wave action. U.S.
Army Corps Engrs, Beach Erosion Bd., Tech. Memo., 75, 1 2 1 pp.
Matter, A. et al., 1980. Flysch and Molasse of central and western Switzerland. In:
R. Trumpy (Editor), Geology of Switzerland, a Guide Book. Exc. 126A, 26th Int.
Geol. Congress, Paris, Schweiz. Geol. Komm., pp. 261-293.
McCowan, J., 1894. On the highest wave of permanent type. Philos. Mag., 5: 351-357.
Miche, R., 1944. Undulatory movements of the sea in constant and decreasing depth.
Ann. Ponts Chauss, May-June, July-August, 25-78, 131-164, 270-292,
369-406.
Miller, M.C. and Komar, P.D., 1980a. Oscillation sand ripples generated by laboratory
apparatus. J. Sediment. Petrol., 50: 173-182.
Miller, M.C. and Komar, P.D., 1980b. A field investigation of the relationship between
oscillation ripple spacing and the near-bottom orbital motions. J. Sediment. Petrol.,
50: 183-191.
Mogridge, G.R., 1973. Bedforms generated by wave action. DME/NAE Q . Bull., 1973(2),
4 1 PP.
Mogridge, G.R. and Kamphuis, J.W., 1973. Experiments on bedform generation by wave
action. Am. SOC.Civ. Engr., Proc. 13th Conf. Coastal Engineering, pp.1123--1142.
Neumann, G., 1953. On ocean wave spectra and a new way of forecasting wind-generated
sea. U.S. Army Corps Eng., Beach Erosion Board, Tech. Mem., 43, 42 pp.
Newton, R.S., 1968. Internal structure of wave-formed ripple marks in the nearshore
zone. Sedimentology, 11: 275-292.
Rance, P.J. and Warren, N.F., 1969. The threshold of movement of coarse material in
oscillatory flow. Am. SOC. Civ. Eng., Proc. 11th Conf. Coastal Engineering,
pp.487-491.
Reineck, H.E., Singh, I.B. and Wunderlich, F., 1971. Einteilung der Rippeln und anderer
mariner Sandkorper. Senckenbergiana Marit., 3: 93-101.
Sleath, J.F.A., 1975. A contribution t o the study of vortex ripples. J. Hydraul. Res.,
1 3 : 315-328.
Sleath, J.F.A., 1976. On rolling grain ripples. J. Hydraul. Res., 1 4 : 69-80.
Sleath, J.F.A. and Ellis, A.C., 1978. Ripple geometry in oscillatory flow. Univ.
Cambridge Dept. Engr. Rept. A/Hydraulics/TR2, 15 pp.
Stone, R.O. and Summers, H.J., 1972. Study of subaqueous and subaerial sand ripples.
Univ. S. Calif. Dept. Geol. Sci., Final Rept., 72-1.
Sverdrup, H.U. and Munk, W.J., 1946. Empirical and theoretical relations between wind,
sea and swell. EOS, Trans. Am. Geophys. Union, 28: 823-927.
Tanner, W.F., 1967. Ripple mark indices and their uses. Sedimentology, 9 : 89-104.
Tanner, W.F., 1971. Numerical estimates of ancient waves, water depth and fetch. Sedi-
mentology, 16: 71-88.
Teleki, P.G., 1972. Wave boundary layers and their relation to sediment transport. In:
D.J.P. Swift, D.B. Duane and O.H. Pilkey (Editors), Shelf Sediment Transport: Process
and Pattern. Dowden, Hutchinson and Ross, Stroudsbourg, Pa., pp.21-60.
Wiegel, R.L., 1964. Oceanographical Engineering. Prentice Hall, Englewood Cliffs, N.J.,
532 pp.
Wright, L.D. and Coleman, J.M., 1973. Variations in morphology of major river deltas
as functions of ocean wave and river discharge regimes. Bull. Am. Assoc. Pet. Geol.,
57: 370-398.
Yalin, S. and Russell, R.D.H., 1962. Similarity in sediment transport due to waves. Proc.
8th Conf. Coastal Engineering, pp.151-171.

Hydrodynamics and Sedimentation in Wave-Dominated Coastal -- B_ Greenwood and R_A_ Davis (Eds_) -- Developments in Sedimentology 39, 1984 -- Elsevier -- 9780080869537 -- 75d36386eed2b37eb5e4bb0e8dc1b259 -- Anna’s Archive

Uploaded by

Copyright:

Available Formats

Hydrodynamics and Sedimentation in Wave-Dominated Coastal -- B_ Greenwood and R_A_ Davis (Eds_) -- Developments in Sedimentology 39, 1984 -- Elsevier -- 9780080869537 -- 75d36386eed2b37eb5e4bb0e8dc1b259 -- Anna’s Archive

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hydrodynamics and Sedimentation in Wave-Dominated Coastal -- B_ Greenwood and R_A_ Davis (Eds_) -- Developments in Sedimentology 39, 1984 -- Elsevier -- 9780080869537 -- 75d36386eed2b37eb5e4bb0e8dc1b259 -- Anna’s Archive

Uploaded by

Copyright:

Available Formats

DEVELOPMENTS IN SEDIMENTOLOGY 39

HYDRODYNAMICS AND SEDl MENTATION

B. GREENWOOD and R.A. DAVIS, Jr.

Reprinted from Marine Geology, Vol. 60 (1-4)

Distributors f o r the United States and Canada.

ELSEVIER SCIENCE PUBLISHING COMPANY INC.

Library or ('ongrr\s Cataloging in Publication Data

Hydrodynamics and sedimentation in wave-dominated

(Developments in sedimentoloa ; 39)

@ Elsevier Science Publishers B.V., 1984

Printed in The Netherlands

Preface ................................................... VII

Waves, long waves and nearshore morphology

This volume evolved from a symposium held at the Eleventh International

BRIAN GREENWOOD and RICHARD A. DAVIS, Jr. (Editors)

WAVES, LONG WAVES AND NEARSHORE MORPHOLOGY

A.J. BOWEN and D.A. HUNTLEY

0025-3227/84/$03.00 0 1984 Elsevier Science Publishers B.V.

0 100 200 300 400

WAVES AND LONG WAVES

There has recently been considerable interest in the direct measurement

SURF BEAT AND EDGE WAVES

Figure 3 schematically illustrates the principle types of waves that may

u2 = g h sin(2n + 1)0, n = 0, 1, 2 . . . (2)

LONG WAVES AND NEARSHORE MORPHOLOGY

As previously mentioned, surf beat, whether a leaky mode or an edge

nowhere as large as the beach slope, no “bar” will be formed, while on a

beach profile in a subtle way without necessardy providing obvious length

SPATIAL AND TEMPORAL VARIATIONS IN SPECTRA OF STORM

ROBIN G.D. DAVIDSON-ARNOTT and DAVID C. RANDALL

On many gently sloping sandy coastlines the nearshore area is charac-

0025-3227/84/$03.00 0 1984 Elsevier Science Publishers B.V

Fig.1. Location of study area.

INNER BAR VY 1.08

A meteorological depression moving northeastward above Georgian Bay

SPATIAL VARIATION IN WAVE CHARACTERISTICS

observed to be primarily in the spilling range. A t the peak of the storm

Wave form and number

Wave and current spectra

Spatial and temporal changes in the magnitude and distribution of energy

WENDAKE BEACH MAY 31,1982 22:OO HRS

Temporal and spatial changes in energy distribution

Wave staff Time

* N o significant primary peak.

It is evident that, under storm conditions, the portion of the nearshore

WAVES, CURRENTS, SEDIMENT FLUX AND MORPHOLOGICAL

BRIAN GREENWOOD and DOUGLAS J. SHERMAN*

A shore-normal array of seven, bi-directional electromagnetic flowmeters and nine

*Present address: Department of Geography, University of Southern California, Los

0025-3227/84/$03.00 0 1984 Elsevier Science Publishers B.V.

Since the first scientific description of wave-formed bars by Elie de

The study site is at Wendake Beach in Southern Georgian Bay, Ontario

Wym bol wood

Fig.1. Location of study area.

110 m from the shoreline, approximately 0.50 m high, relatively straight,

LINE 1 NORTH SWL

The field experiment was designed specifically to examine the two-

Wave and current sensors

piercing, continuous resistance wave staffs following the basic design of

Data acquisition and analysis