Acoustic backscattering at low grazing angles from the ocean

bottom. Part I. Bottom backscattering strength

H. Boehme,N. P. Chotiros,L. D. Rolleigh,S. P. Pitt,A. L. Garcia,T. G. Goldsberry,
and
R.A.

Lamb

Applied
Research
Laboratorie
TheUniversit
of Texasat Austin,P.O. Box8029,Austin,Texas78713-8029

(Received21 June1984;acceptedfor publication2 October1984}

Acousticbackscattering
measurements
ona sandbottomweremadeat grazinganglesin therange
of about2-10in waterdepthof approximately
15.5m nearSanDiego,California{reported
byT.
G. Goldsberry,
S.P. Pitt, andR. A. Lamb,104thMeetingof theAcousticalSocietyof America,
Orlando,FL, 8--12November1982).Data from thesemeasurements
havebeenanalyzedto
determinethe meanvalueand standarddeviationof the bottombackscattering
strengthper
squaremeterasa functionofgrazingangle,insonified
area,transmitsignaltype,andfrequency.
A
curvedray pathproportional
modelandmeasured
soundspeedprofileswereusedto determine
grazingangleversus
time.ThemeanvaluefollowedLainbert'slawfortherangeof grazingangles
measured
andfor all frequencies
used.No significant
differences
in meanvaluewereobserved
whenthe insonifiedareaand transmitsignaltype werevaried.The observedfrequency

dependenee
ofthebottom
backscattering
strength
persquare
meterfallsintherange
fromf.oto
f' for thisrelativelyfiat,sandybottom.
PACS numbers:43.30.Bp,43.30.Gv,92.10.Vz,91.50.Ey

chukandZhitkovskii
4 concluded
thatbackscattering
was

INTRODUCTION

Acousticbottom backscattering
measurements
were
madein May 1982aboutI mi offshore
fromMissionBeach,
California.The measurements
weremadeusingtransducers
mountedon a tripodassembly
about4 m tall that restedon

independent
of frequency,
or at mostonlyslightlydependent
on frequency,
for all bottomtypes.
The acoustic measurements made in shallow water near

SanDiegowerespecifically
plannedto providebottombackscattering
data
for
low
grazing
anglesovera relativelywide
the bottom. The horizontal and vertical orientation of the
range
of
frequencies.
The
transmitted
pulsewaveformswere
transducers
werecontrolledandmonitoredby testpersound
either
cw
(pulse
lengths
of
0.25-25
ms)
or linearFM (1-25 ms
on a nearbyoceanographic
tower.The acousticmeasurepulses
with
1-4
kHz
bandwidths}
and
weregenerallytransmentsweremadeovera rangeof grazinganglesof 2-10and
mitted
on
alternate
pings
until
approximately
75 pingsof
a rangeof frequencies
of 30-95kHz. Detailsof themeasureeach
pulse
type
had
been
transmitted.
mentsystemaswell asthe bottombackscattering
measurePhysicaloceanographic
measurements
were made by
mentsthatweremadehavebeenpresented
previously.

Naval
Ocean
Research
and
Development
Activity
A preliminaryobjectiveof the bottombackscattering
(NORDA}
during
the
same
period
of
time
that
acoustic
meameasurements
wasto providedata at low grazingangles
surements
were
made.
A
report
on
the
sediment
geoacousti
from which backscattering
strengthvaluesand reverberaproperties
hasbeenwrittenanddistributed?
tionstatistics
couldbeextracted.Severalcomprehensive
reviewshave beenpublishedwhich includebottombackI. DATA ANALYSIS DESCRIPTION
scatteringstrengthversusgrazinganglein the 20-100 kI-Iz

frequency
range.
2-4Themostnotable
ofthese,
byMeKinney
andAnderson
2 andbyShultz,
3 arealmost
20years
old;only

The acoustic measurement data were recorded on ana-

a few measurements
at grazinganglesbelow10are includ-

logmagnetic
taperecords,
converted
to digitaldatarecords
by useof generalpurlx)seanalog-to-digital
(A/D) eqhip-

ed.The morerecentreviewby BunehukandZhitkovskii

4

ment,and processed
by useof analysissoftwarewritten for a

includesinformationat low grazingangles.Thesereviews

generallyagreethat bottombackscattering
canbe broadly
categorizedaccordingto bottomcomposition,suchasmud
or silt, sand,and rock or gravel.However, relativelylarge

generalpurposecomputer(CDC CYBER 171).The digital

variations within each general bottom type are common,

with little or no correlation of bottom backscattering

strengthwith meanparticlesizewithin eachof the categories.For grazinganglesaboveabout2,backscattering

was

reported
toincrease
withgrazing
angleaccording
to'sin
t' 0,
where0 isthegrazingangleandk isa numberbetween1and
2. For sandsediments,
backscattering
wasfoundby McKin-

neyandAnderson
to increase
withfrequency
whileBun962

J. Acoust.Soc. Am. 77 (3), March 1985

data recordformat w_as

suchthat eachtransmittedpulseand
subsequentreverberationperiod was identified by pulse

type, frequency,time, and numberof sequentialsamples

composing
thedigitaldatarecord.An envelopewasgenerated for eachpingandwasthensmoothedby time averaging
witha movingtimewindowequalin lengthto thepulseduration.

A numberof sequential
pingcycles(usually30 to 50)
usingthe samepulsewaveformwereassembled
to form an
ensemble.Statisticaltestswere then performedin order to
assure
thattheassembled
envelope
recordsconstituted
a val-

0001-4966/85/030962-13500.80

1985 AcousticalSocietyof America

962

id ensemble.
An existingray tracingcomputerprogramwas
usedto relatetime afterinitiationof thepulsetransmission

to pathlength,horizontal
range,andgrazingangle.A horizontallystratifiedwater columnbasedupon a measured
soundspeed
profilewasassumed
withintheraytracingcomputations.
A meanbackscattering
strength
anda standard

HEIGHT
ABOVE
BO'l-rOM

deviation from the mcan were calculated for the ensembleof

envelopedatarecords.

Ais- Insonified
area

Themeanbackscattering
strength
BS,in dB/m2, was
calculatedaccordingto the followingequation:

8- (8]n
+8n_1)/2

perpendicularto

- eH0

BS---RL -- SL-- 2TL-- 101og[A,Do(O

Dv(O
2)], (1)
where

RL = equivalent
returnedsignallevelonthehydrophoneMRA in dB re:1pPa.
SL = acousticprojectoron-axissourcelevelin dB
re:l pPa at I m,
2TL = two-waypropagation
loss,in dB,
As = effective
insonified
area,in m2,
2
Dm{O1)= verticaldirectivity
functionof theprojector
intensityasa functionof 0 !. theangularseparation bctwcenthe maximumresponseaxis
(MRA) andthe launchangleof the outward
ray path,and
Dv(tz}= verticaldirectivityfunctionof the hydrophone
asa function
of0z. theangular
separa-

HEIGHT
ABOVE
BOTTOM

Ol
d

Ax

As
BOTTOM

rd ,

A Imonified
area

Ax = Projection
of As

rd- rn-rn-1

perpendicularto

tion between the receive MRA and the termi-

nationangleof thereturnray path.

FI. 1.Spreading
ofraysat pathlengthi, launchangle0. (n}Idealcase,(b)
Sincethe returnedsignallevelRL associated
with each practicalcas
pathlength,range,or grazinganglevariesfrompingto ping,
the power-averagedvalue was used for backscattering
strengthcomputations.
Thus,the ensemble
averageof the
The one-wayspreadinglossLs in dB wasexpressed
as
squareof theenvelope
wascomputed
andhalved,andthen
theidealspherical
spreading
losswith thecorrection
factor
the squareroot takento obtainRL sincenarrow-bandsigCs asfollows:
nals were assumed.
In the assumedhorizontally stratifiedmedium with
nonuniformsoundspeedversusdepth, the spreadingloss
portionLs of the propation lossTL was computedby
tracingtwo raysat slightlydifferentverticallaunchangles
and determiningtheft separationat the points of bottom
contact.In this manner,a correctionfactor Cs was deter-

minedwhicharisesdueto divergence
or convergence
of the
soundrays.The correctionfactor Cs was the ratio of the
actualareainsonified
perpendicular
to thetwoadjacent
rays
to theidealinsonified
areaassuming
spherical
spreading
at
the samepath length! and within the samelaunchangle
interval
Then, Cs wasiven by

Cs= r rdsin0,/120. cos0,,

where

= averagehorizontalrangeto the pointsof bottom

contactof the two adjacentrays,
rd = difference
in horizontalrangeof thetwo points,

0, = average
grazing
angleofthetworayswiththebottom,

! = pathlengthalongan idealstraightline ray path,

0d = differencein adjacentray launchangles,and
0t = averagelaunchangleof two rayswith respectto the
horizontal.

963

J. Acoust.Soc. Am,, Vol. 77, No. 3, March 1985

L s = -- 20 logl -- 10 logCs

(3)

Largenumericalvaluesof Cs areindicativeof shadowzones

whilesmallvaluesare indicativeof focusing.
The absorptionlosspart of the propagationlosswas
calculatedby combiningthe ray-tracedpath lengthand the
absorptioncoefficient
(in dB/m), wherethe absorptioncoef-

ficientwasdetermined
fromtheShulkinandMarsh6 equation. The total one-waypropagation
lossincludingabsorption losswasexpressed
as
TL = -- 20 log1-- 10logCs -- al.

(4)

Calculation of the effective insonified area also made

useof the ray-tracingcomputerprogram.Figure 1 shows

sketches
of (a)thegeometryfor ideal(isovelocity)
conditions
and(b)thegeometryusedwhenthemediumisassumed
to be
horizontallystratifiedwith depth-dependent
soundspeed.
Whencurvedraypathsandlowgrazinganglesareappropriate, the expression
usedfor the-insonified
areais asfollows:

A, =

o,] ,

whereisthehorizontalrange,. istheeffective
horizontal
beamwidthof the projectorandhydrophonearray,Cbis the
soundspeedjust abovethebottom,' isthe pulselength,and

Os isthegrazingangle.
Boehmeet al.: Acousticbackscatter'rig.Part I

963

AMBIENT

NOISE

SAMPLES

u,

TO

T1

T2

T3

RANGE OR TIME

FIG. 2. Three-dimensional
plot of an ensemble
of sequential
enveloperecordswhen propagationconditionswere rdatively stable.Transmitted
pulsewas0.2$-mscwandeventsontherange/timeaxisare
T O: transmit,
T. = acoustictargetat 70-m range,T2: acoustictargetat ] 10-m range,
and T 3 = acousfictargetat 210-mrange.

The effective
horizontalbeamwidthr of theprojector
andhydrophone
arrayisapproxitnated
in termsof the -- 3
dB horizontalbeamwidths
of the hydrophone
arrayh and

thelrojector
, according
tothefollowing
expression:

t1.065('2+-2)

-n/z, in radians,

(6)

where, and, arealsoexpressed

in radiansandthenumericalfactorarisesfrom an assumption
that the mainlobes
of the actualhydrophoneand projectorhorizontaldirectivity functionsare approximatedby Gaussianfunctions.
The expression
givenin Eq. (5) for the insonifiedarea
assumes
that grazingangleis a constantovertheinstantan-

TO

T1

T2

T3

RANGE OR TIME --

FIG. 3. Three-dimensional
plotof an ensemble
of sequential
envelope
recordswhenpropngation
conditions
wereunstable.
Transmitted
pulsewas
0.25-ms
cwandevents
ontherange/time
axisarethesameasthose
inFig.2.

grazingangle,effective
horizontal
beamwidth,
transmitsignaltype,frequency,
andbottomtype.Sincebackscattering
strengthwas calculatedby averagingover an ensembleof
enveloperecords,an attemptwasmadeto selectdatarepresentativeof the characteristic
beinginvestigated
overa time
intervalduringwhichpropagationconditionsremainedrelativelystable.In additionto statisticaltestsusedto indicate
validensembles,
three-dimensional
plotsweregenerated
al-

lowingvisualindications
ofthestabilityof themedium.Figure 2 is an exampleof sucha plot,showing
the envelope
recordsof sequential
pingsof a particularpulsetypewhen
eousinsonified
area;this is an acceptable
approximation propagation
conditions
wererelativelystable.Figure3 is a
onlyif thesoundspeedprofiledoesnotresultin focusing
or
similarplotshowing
unstable
propagation
conditions
asevishadowzoneswithinthe acousticmeasurement
regionof dencedby thevariationin amplitudes
for adjacentpingcyinterest.Inaccuracies
in grazingangleestimation
arepropagatedto esthnatesof the inson/fledarea.

The acousticprojectoron-ax/ssourcelevel{$L ) versus

transm/telectricalcurrentwasdeterminedby acousticcalibrationat the LakeTravh TestStation{LTTS)calibration
facilitypriorto makingtheacoustic
backscattering
measurements.The transmitcurrentwasmonitoredduringthemeasurementsand recordedon data log sheets.
The projectorandhydrophone
verticaldirectiv/tyfunctionswerealsomeasured
andrecordedon datalogsheetsat
LTTS. The verticaldirecfiv/tyfunctionsweredigitizedand
stored in a digital calibration data file which could be accessedas neededduring the analysisof measurementdata.
II. DATA ANALYSIS RESULTS

The acousticdata wereanalyzedto detetainethe behaviorof bottombackscattering

strengthasa functionof
964

J.Acoust.
Soc.,m.,Vol.77,No.3, March1985

cles.The time betweenpingsin bothfigureswas 1 s.

A. Bottombackscatteringstrengthversusgrazing
angle

Initialestimates
ofbackscattering
strength
werecarried
outusingtheoretical
estimates
of thetransducer
beampatternsandassuming
a constant
soundspeedversusdepthprofile.The resultsobtainedwerereliableonlyfor rangesless
thanabout25 m, whichcorresponded
to grazinganglesof
10or more.For lowergrazingangles
andcorrespondingly
longerranges,the soundspeedprofilewasfoundto havea
significant
influence
on the results.Examplesareshownin
Figs. 4 and 5, in whichthe estimatedbackscattering
strengths
asa function
ofgrazing
anglearepresented.
Figure
4 represents
estimates
based
ona constant
soundspedversusdepthassumption,
while Fig. 5 represents
estimates
based
ona measured
sound
speed
profileinconjunction
with
Boehme
otal.:AcousMc
backscattering.
PartI

964

-10 -

T - CALIBRATED

TARGET

FIG. 4. An exampleof backscattering

strengthestimates
asa functionof grazinganglewherea constant
soundsled
versusdepthwasassumed.
Frequency

-60
0

10

GRAZING ANGLE - deg

a raytracingcomputer
program.Theresultsshownin Fig.4
indicatethatgrazinganglesof 10-1
werebeingmeasured,
whilethoseof Fig. 5 indicatethat, in reality,thedownward
refraction
caused
bythesoundspeed
profilelimitedthegrazinganglesto valuesaboveabout2.5.
Raytracingwasalsousedto determine
therange,propagationdelay,andgrazingangleof a fewkeyraypaths,such
asthe raysat the -- 3 clBpointsof the verticalbeampatterns,andthe raysfrom the centerof the mainlobeand the
sidelobes.
The informationwasprovidedin the form of a
printoutwhichaccompanies
the graphicalresult.Someof
thisinformationis illustratedin Fig. 5..The informationis

particularlyusefulfor checkinganomalous
featuresin the
bottombackscattering
strengthestimates.
Other propagation anomalies,includingfocusingand shadowzones,are
alsoprintedout aswarningmessages
astheyaredetected.
In orderto determinethesensitivity
of grazingangleto
input soundspeedprofile,a particularblockof data was
processed
usingsoundspeedprofilesmeasuredon two
success/re
days(:5-6May)asinputs.Thereverberation
data
usedwererecordedapproximately
2 1/2 h afterthe sound
speedprofilewas measuredon 5 May. A comparisonof
backscattering
strengthversusgrazingangleisshownin Fig.
6 for inputsoundspeedprofilesmeasured
on 5-6 May. Since

-I0-

T = CALIBRATED

TARGET

-40 FIG. 5. An exampleof backscattering

strengthestimates
asa functionof grnzing angle,with ray tracinganda measuredsoundspeedprofile.Frequency

-50UPPER
-3 dB LIMIT
I

HITSSURFACE

MAINLOSE

-3 dBLIMIT

IUPPER
SlOELOBE
[
LOWER
I

BOOM-U.FACE
0

tO

GRAZING ANGLE - deg

965

J. Acoust.Sec.Am.,VoL77, No.3, March1985

goehrneeta/.: Acousticbackscattering.
PartI

965

-10 -

LEGEND:
FREQUENCY:

T{b)

T(b)

!
==-3o-

FIG.. 6. An exampleof backscattering

strengthvsus
asle estimated
b
upon the sam data set and two
differentsoundspeedprofile Curve(a)

AZIMUTH: 26.60
VERTICAL: -5.0

30 kHz

ARRAY:
STAVE 5, LF
PULSE: 0.40 ms cw

-r

DATE:

5 MAY

TIME:

13:04:57

1982

I--

ii

='-4o-

ii

ii

input soundspeedprofilemeasured
proximately
2 h beforetheacoustic
men-

mate8madewith a soundspeedprofile
measuredappxmntelyoneday later

n..

--.
I--

I0

GRAZING ANGLE - deg

thesamereverberation
datawereusedfor bothplotsin Fig.
6, thetwoinputsoundspeedprofilesusedcausedifferences
of lessthan0.5 in grazingandesforcorresponding
bottom
features(andcalibratedacoustic
targets,asindicated).
The bottombackscattering
measurements
wereintendedto provideinformationat grazinganglesbelowabout1$o.
The lowgrazinganglelimit, corresponding
to longerranges,
was observedto dependupon the propagationconditions
existingat the timethe particularbackscattering
measuremeatsweremade.In particular,the soundspeedprofilewas
suchthat downwardrefractionprevented
meaningful
measurements
belowabout2, sinceenergybackscattered
from
the bottombecamecont:amipated
by energybackscattered
fromtheseasurfaceat thelongerranges.
The surface reverberation contribution

resulted from a

directstirfacepathaswellasa bottombounceto surfacepath

with thesametwo-waytraveltimeasthedirectbottompath
for bottom backscattering
at longerranges.The sketchin
Fig. 7 is helpfulin describing
the relationship
betweenbottom andsurfacebackscattering.
At shorterrangesthebottom-surface
and directsurfacepaths(solidlines)with the
sametwo-waytraveltimeasthedirectbottompathareasso-

ciatedwith thesidelobe
regionof theprojectorandreceiver
verticalbeampatterns.The surfacereverberation
levelsare
not sufficientto seriouslyaffectthe bottomreverberation
level. However,at longerranges,the bottomsurfaceand
directsurfacepaths(dashedlines)areassociated
with beam
patternregionsthat are migratingtowardthe beamMRA.
The directbottompath, conversely,is associated
with beam

pattern regionsthat are migratingaway from the beam

MRA as the range increases.At some range, depending
uponexistingpropagation
conditions
andseasurfaceconditions,the surfacereverberationlevelcan becomecomparablewith or greaterthanthe bottomreverberation
Icyel.This
conditionimposes
a lowergrazinganglelimit on the analysesof bottombackscattering
strength.
To confirmthat the background
noiselevelassociated
with low grazingangleswasindeedcontaminated
by surface
reverberation,the backgroundlevelswere first compared
with measured ambient noise levels. The observed back-

groundlevelsweresignificantly
aboveambientnoiselevels
andwereobserved
to dependupontransmittedsignalpulse
length.The observed
background
levelat thelongerranges
wasalsofoundto be correlatedwith wind speed.On occa-

SURFACE

FIG. 7. Cont,mit/on at longerrangesof bottom backscatterinmeaattrement

by surface
bac.atterins.

BOTTOM

966

d. Acoust.Soc.Am.,VoL77, No. 3, March1985

Boehrneeta/.: Acousticbackscatter'ng.
PartI

966

-10

FREQUENCY:

45 kHz

PULSE TYPE:

0 5 mscw

T - RETURNS FROM FLUID-FILLED

SPHERES

FIG. 8. An exampleofob.Tvedbottom
backsctterin$strengthwMch
nificantcontnmin=tion
by

ingangle.

LIMITING
GRAZING
ANGLE

IO

GRAZING ANGLE - deg

sionswhentheseasurfacewasve rough,the trendof the monfor observed

backscattering
at low grazinganglesfor
estimated
backscattering
strengthdecreased
with decreasing bottomconditionswherethe roughness
scaleis the same
anglefromabout10 downto 3' or 4; thistrendwouldthen orderas the acousticwavelength.
For comparison,
a solid
a trendof 10log(sin0 }isalsoshownin Fig.
abruptly
change
andincrease
withdecreasing
grazing
angle. linerepresenting
An example of observed bottom backscattering 8.
A measureof the variation of bottom backscattering
strengthversusgrazingangleat 45 kHz isshown'inFig. 8 for
whichcontamination
by surfacereverberation
becomes
sig- strength
versus
grazing
.angle
wascalculated
foreachensemnificantat about3 grazingangle.Thus,the usefulrangeof
ble of envelope
records.
The measure
usedwasthe coeffideviation
fromthemeandibottombackscattering
strengthinformationlies between cientof variation(thestandard
videdby themeanvalue}at eachgrazingangle.An example
about3 and9.5. Withinthisregionof grazinganglesthe
isshowninFig.8wherethequantity10logV
trend(dashed
line)ofthebottom
backscattering
strenhfol- ofthismeasure
grazinganglehasbeenplotted,whereVii thecoeffilowsthat
of 10log(sin
2 0), where0isthegrazing
angle.
This versus

Lambert's
rulebehavior,
discussed
byUrick,?isnotuncom-

cient of variation.

-10-

T -- FLUID-FILLED

SPHERES

FIG. 9. Comparison.
of backscattering trengthestimates
at 30 kHz
ing a combinednimuthal
widthof(a) 21.1 (solidline)and(b)
2.8 (chhedline).In bothcases
a cw
pulseof 0.4 mswasused.

INSONIFIED
AREAS
SURFACE
I
ATTHIS
PONT
CONTAMINATED

1.03m2 -- SUMBEAM
7.73m2 -- STAVE

-70
0

1o

GRAZING ANGLE--d

967

J. Acoust.Soc.Am.,Vol.77, No.3, March1985

Boehmoota/.: Acousticbackscattering.
PartI

967

-10T -

FLUID-FILLED

SPHERES

FIG. 10. Comparisonof backscatteringstrengthestimatesat 45

kl-[z using a combinednimuthal
beamwidthof (a) 14.2 {solidline)
and (b) 2.0 (dashedne). In both
casesa O.5-mscw pulsewasused.
SURFACE
CONTAMINATED

INSONIFIEDAREAS

AT THIS POINT

0.92m2 - SUMBEAM
6.5 m2 -- STAVE

-700

GRAZING ANGLE -cleg

an accompanying
paper.)In all casesthe variationof the
curvesfromthegeneraltrendwithgrazinganglewasnoticeablylessforthelargerbeamwidths
associated
withthestaves

B. Beamwidthdependence of bottom backscattering

strength

Both sum
and 'individualreceivin
s array stave
outputs
wererecorded
duringtheexperiments.
Figures9-13
showcomparisons
of estimatedbackscattering
strengthfor
thebeafnwidths
of thesumbeamsandstavesof thereceiving

than for the beamwidths associatedwith the sum beams. The

generaltrend with grazingangle for both staveand sum

beamresultsat all frequencies
wasobserved
to followLambert'srule, as shownin Fig. $.

arraysfor frequencies
of 30-95 kHz. The corresponding
insonifiedareasareindicatedat a commonrangepoint of 70 m
on all the figures.For the examplesshown,and for other
pulsetypesanalyzed,therewasno observeddependenceof
meanbottombackscattering
strengthonbeamwidth.(Minor
differences
notedin the reverberationstatisticsare givenin

C. Pulse type dependence

The averagebottom backscatteringstrength versus

grazingangleexhibitedno dependence
uponpulsetype or
pulselength.Examplesof resultsObtainedusingthe sum

T - FLUID-FILLED

SPHERES

-I0-

FIG. 11. Comparisonof backscatteringstrengthestimatesat 60

kHz usinga combinedndmuthal
beamwidthof (a 17.7 (olid Line}
and (b) 2.3 (dashedline). In both
casesa 0.5-mscw pulsewasused.
; j''

,,

1; -'

, 1,

INSONIFIED AREAS
AT THIS POINT

SURFACE

-50 -

,.qli

CONTAMINATED

1.05m2 - SUMBEAM
8.1 m2 - STAVE

-60

10

GRAZING ANGLE - deg

968

J. Acoust.Soc,Am.,Vol.77, No.3, March1985

Boehmeet al.: Acousticbackscattering.

PartI

968

-10-

T -

FLUID-FILLED

SPHERES

FIG. 12. Comparison of backscatteringstrengthestimatesat 80

kHz using a combinedazimuthal
beamwidthof (a) 13.2 (soLidline)

(b) 1.7 (c:Lhcdt_).In both

casesa O.-mscw pulsewasused.
INSONIFIED AREAS
AT THIS POINT

SURFACE
CONTAM
NATED

1.56m2 - SUMBEAM
12.1m2 - STAVE

-70
0

!0

GRAZING ANGLE -- deg

B, = 101oBp+ 10log(sin
0),

beamoutputsareshownin Figs.14-18forfrequencies
of 3095 kHz. In each ease,it can be seenthat the bottom back-

(7)

where0 = grazingangleand 10 log/t = backscattering

strengthin dB at normalincidence
if Lainbert'srule were

scattering
strengthassociated
with eachpulsetypetendedto
varyrandomlyaboutthe samemeanvalue.The resultsfor
the FM slidepulsetypes,with a time-bandwidth
product
greaterthanunity,weresmoother;
all the data havebeen

valid at nomml incidence.

A sin2 0 functionwasfittedto eachbackscattering

strength
versus
grazinganglecurveandthevalueof 101ogH
smoothed
byaveraging
overthepulse
length.
wasestimated.The quantity101ogp wasthenplottedasa
functionof frequency;
theresultsareshownin Fig. 19.
D. Frequency dependence
Within theparticularbottomregionfor whichFig. 19
Thebottom
backscattering
strength
asa function'ofapplies(finesand}andoverthefrequencyrangeconsidered
strengthwith fregrazing
anglewasfoundto fit Lambert's
rulefairlywellfor here,anincreasein bottombackscattering
wasobserved.
Duetothescatterin thedatapoints,a
all frequencies
andpulsetypesused.Therefore
theback- quency
frequencydependence
of 10logf n, where1<n< 1.$,canbe
scattering
strengthB, maybeexpressed
as
-10 -

T -

FLUID-FILLED

SPHERES

I
-r
I-

FIG. 13. Comparisonof back-

-40-

scatteringstrengthctimatcaat 95

kHz usinga combincdnmltha.!

beamwidthof (a) 11.1 (solidline)
and {b) ]. IdashedUne).n both

_ -5oINSONIFIED

casesa 0.S-mscw pulsewasused.

AREAS

AT THIS POINT

1.37m2 -- SUMBEAM
10.2m2 - STAVE

SURFACE
CONTAMINATED

-0
0

i '

10

GRAZING ANGLE -- deg

969

J. Acoust.Soc.Am.,Vol. 77, No. 3, March1985

Boehmeetal.:Acoucbackscattering.Partl

969

-10-

(a) ..........

0.25 ms cw

(b) .......

0.5 ms cw

(c) ....

I ms 4 kHz FM

(d)

5ms4kHzFM

F[O.

14. Compadson of back-

scattering
strengthat 0 kHz using
(a}0.25-mscw, (b)0.5-mscw, ()lms,4-kHz FM, and(d}5-ms,4-kl-Iz
FM pulsetypc

,,z,-40i=.

i,=,

-60 SURFACE
CONTAMINATION

'70o

;0

GRAZING ANGLE- deg

thanresults
reported
byCrispetaL9 at 30and48 kHz for

inferred.This frequencydependence
is consistent
with that
reportedby McKiuney and Anderson(Ref. 2) of approxi-

two sitesin thePugetSound.No particularfrequencydependencewasobserved

in theresultsfromthePugetSoundmca-

mately10 1ogfL6for fieldmeasurements

in sandbottom

surements
(at20 grazingangle9).
Bottombackscattering
measurements
werereported
2

regions.
Two pointsarealsoshownin Fig. 19at 100kHz that

wereestimated
fromdatabelow10 grazingangle
2 forsand
of aboutthe sameparticlesizeas reportedin Ref. 5. These
pointscompareverywell with dataplottedat 95 kHz from

for a water/sandboundarywherethesandhadbeencarefully smoothed.Valuesof 10 log/t for grazinganglesfrom

the current measurements.
about3--10
andfrequencies
of 57 and90 kHz werededuced
Datareported
byWongandChesterman
s formeasure- fromRef.2, andareshownin Fig. 19.A straightlinethrough
mentsat low grazinganglein a siltysandregionat 48 kHz
thesetwo pointsindicatesa frequencydependence
of 10log
are 15-20 dB higher than data plotted at 45 kHz from the
f Le,ingood
agreement
withtlecurrent
measurements.
The
level of the backscatteringfrom smoothsand is 10-12 dB
currentmeasurements.
The observed
valuesof 10 1ogpfor
The high30 and45 kHz shownin Fig. 19arealsosubstantially
lower belowthat observedin the currentmeasurements.
-10-

zuJ

(a) ....

0.5 ms cw

lb)--

lms2kHzFM

-40-

FIG.

15. Comparison of back-

scatteringstrength at 45 kI-tz using

{a) 0.5-ms cw and (b) 1-ms,2-kHz

FM pulsetypes.

FF-

-'70
0

'l
8

I0

GRAZING ANGLE -- deg

970

J. Acoust.Soc.Am.,VoL77, No.3, March1985

Boehmeetat: Acouslicbackscattering.
PartI

970

(a) ..........
(b! .......
(c)--

-10-

{3.25ms cw
0.5 rn cw
lms4kHzFM

FIO. 16. Comparisonof bckscattering

strengthat 60 kHz using
(a)0.25-ms
cw,(b)0.5-mscw,and(c]
l-ms,4-kHz P'Mpulsetypes.[Note:
Curves(a)and(c)olXainedwithverficultilt angleof -- 3 whilecurve
(hiobtained
withverticaltilt angleof
-- o.]
-50SURFACE
CONTAMINATION

-600

-I

!0

GRAZING ANGLE - deg

er values observed in the current measurements are unstartedat a rangeof about25 m, at a grazingangleof 10.
doubtedlydueto the interfacereliefwhichwaspurposely Unfortunately,dueto strongsurfaceactivityat the time of
missingin the smoothsandmeasurements.
the experiment,the bottombackscattering
data, beyonda

rangeof about70 m, wereseverelycontaminatedby surface

backscatter;
thereforeonly data corresponding
to range.

E. Azimuth dependence

between 25-70 m were used.

For the purposeof measuring

azimuthdependence,
a
set of data was taken at 30 kHz in which the sonar beam was

Thebackscatter
dataappeared
to followLambert'srule
exceptwheretherewasa transitionfromonetypeof sandto
another.Thedatawereblockedintoninegroups
oftenpings
eachwhich,at a scanrateof 1.7 betweenadjacentpings,
wouldcorrespond
to 17 sectors.The total sectorcovered
was153. Withineachblock,the averageLambertnormal
incidencebackscattering
strength10 logp wasestimated.

slowlyscanned
overa largesectorofthebottom.Thebottom
may be separated
into two regions--freesandand coarse
sandregionswitha discernible
boundary
between
them.s
The scandataincludedmeasurements
in bothregions.The
sonarbeamwastiltedverticallyat a depression
angleof 5.
At this depression
angie,the bottomreverberation
returns

-10(a} ....

0.5 ms cw

(b)

! ms 2 kHz FM

4
II

FIG. 17. Comparisonof backscatteringstrengthat 80 kHz using

(a) 0.5-mscw and (b) 1-ms,2-kHz
FM pulsetypes.

I Ill f

-50-

ISURFACE
CONTAMINATION

-diDO
0

10

GRAZING ANGLE - deg

971

J. Acoust.Soc.Am.,VOL77, No. 3, March1985

Boehrneeta/.: Acousticbackscattering.
Part I

971

-10 {a! ....

0.5 ms cw

{b}

lms4kHzFM

FTG. 18. Co

of back-

scatterlag
streagthat $ kI
(a) O.$-mscw and (b) l-ms,
FM puhetypes.

SURFACE
CONTAMINATION

'60
o

I
I

I
.2

I
3

]
4

I
5

[
6

I0

GRAZING ANGLE -- deg

Wheretherewasa cleartransitionfromonetypeof bottom

to another,theblockwassubdivided
at thetransitionrange.
The resultsareshownin Fig. 20. The measured
valuesof 10
log/z rangedfrom -- 26 to -- 34 dB;bothextremevalues
weremeasured
overthecoarsesandregion.At theboundary
betweenthe two typesof sand,surprisingly,
the fine sand
showeda highervalueof 101ogp.
Somedegreeof azimuthalangledependence
is indicatedby theanalysis
resultswhichmaybeassociated
with the

Ill. CONCLUSIONS

tom profile measurementsmade by NORDA indicate an

It wasobserved
that the influenceof the soundspeed
profilewassubstantial
withregardtoboththenormalization
of backscattering
strengthto unitareaandtheassociation
of
backscattering
strengthper unit area to grazingangle.In
bothcasesthe ray bendingat relativelymodestrangeswas
sufficientto producenoticeabledifferencesin levelsand
grazinganglesfromthoseresulting
fromisovelocity
conditionassumptions.
A moresubtleinfluenceuponpropagation
losscalculations
is possible
asa resultof energyfocusingor
divergence
accompanying
soundrefraction.

rms roughness
of about I cm in the fine sandregionand

Sound
speed
profiles
were
taken
about
three
time
aday

orientation and structure of sand waves on the bottom. Bot-

about2.5-3cmin thecoarse
sandregion.
IDiversreported duringthetestperiodandin onelocationonly.This resulted
that sandwavesin the finesandregionappearedto be ran-

in a rathersparsesamplingandundoubtedly
contributedto
inaccuracies
in the data analysisresults;however,practical
considerations
invariably prevent sufficientsamplingto

domly orientedand only a few centimetersin lateral extent.

The wave heightsas well as lateral extentof wavesin the
coarsesandregionweregreater.

characterize the acoustic environment.

-10-

FIG. 19. Estimated value of the bot-

tom bacsag

craoto'i

1o

log/ versusfrequencyfor the free

md bottomreginanearSanDiego

azimuthtowardthe 15-cm{6 in.)diamsphere.

Thedataincludegraz-

O -- SUM BEAM

-40-

X -- STAVE

ing angle from about 2.5 to about

- MUDDY
FINE
SAND

10'.(See
text for detailsof compari-

-- CLAYEY
MEDIUM
SAND , FROMRef.2
--SMOOTHED
MEDIUMSAND.,)
-5O

4O
FREQUENCY

972

J. Acoust.Soc.Am.,VoL77, No.3, March1985

son of authors' reaults with those of

5O

60

80

9O

'

tO0

-- kHz

Boehmeoral.: Acoustic
backscattering.
PartI

972

BOUNDARY

FINE SAND

COARSE SAND

- 28 dB
-29

FIG. 20. Measuredvaluesof Bo from a

dB

slow azimuthal

scan of the bottom.

Groupsof ten pings,whichspansectors

of 17, w'e blackedandaveraged.

The
measurements were made at 30 kHz
- 29 dB

with a systembeamwidthof 2.8. The

scanned area covo-ed both fine and
-26dB

coarsesand regions,with mean grn/n

sizes
9X 10-sinand5X 10-4m,respec
tivcly.

BOUNDARY

Thelackofanindependent
measure
ofpropagation
loss

An azimuthaldependence
wasobserved
in thebottom
backscattering
strength.The acousticmeasurement
equipeachfrequency.
Althoughfluid-filledspherical
acoustic
tar- mentwaslocatedneara transitionregionbetweenfineand
gets were calibrated under free-fieldconditionsand decoarsesand;dataweretakenin bothregionsasthesonarwas
ployedin thebottommeasurement
region,the deployment scannedin azimuth.The highestvariabilityin backscattergeometry,environment,
and systemparameterscombined ing strengthwasobserved
in the coarsesandregionwhere
to preventtheuseof thisinformation
,tohelpreduceuncer- diversreportedlargersandwavesthan in the finesandreare limitedat
tainties
in propagation
loss.Theacoustic
targets
werevery gion.Analysesresultson bottomroughness
usefulasreference
pointsin rangeandbearingduringdata this time; however,the bottombackscattering
resultsobacquisition
andagainduringdataanalyses
efforts.
servedare expectedto be attributableto sandwavesand,
The estimated
bottombackscattering
strengthversus particularly,to theirorientation.
grazingangleplotswereoftenobserved
to increase
with deis believed to have contributed to the scatter of the data at

creasing
grazing
anglebelowabout3ashasbeenreported.
2
Theobserved
background
levels
at thelowergrazing
angles
werefoundto dependuponpulselength,to beaboveambient
noiselevels,and to correlatewith wind speed.The ranges
involvedwhenbackgroundlevelswereobservedto increase
with decreasing
grazinganglewere consistent
with backscattering
fromtheair-watersurface.The spatialandtemporalcorrelations
of datafromlongrangesandlowgrazing
angleswere differentfrom similarcorrelationsat shorter
rangesandhighergrazingangles.The authorsfeelthat the
observed
behaviorat lowgrazinganglesisa resultof energy
backscattered
fromthewatersurfaceduringtheseacoustic

ACKNOWLEDGMENTS

The work wassupportedby NAVSEA, Code63R, and

NORDA, Code113.The authorsalsowishto acknowledge
thesupportof NOSC for theuseof theoceanographic
tower
during the acousticmeasurements.We wish also to acknowledge
theconsultation
provided
byDr. JohnM. Huckabay, Dr. C. Robert Culbertson,and Garland R. Barnard
duringall phases
of thiswork,and the effortsprovidedby
PaulaTaylorandDeloresHigdonin thepreparation
of the
manuscript.

measurements,and that it is not an anomalouscharacteristic

of bottombackscattering.
The bottombackscattering
strengthwasobserved
to be
independent
of beamwidth
andpulselengthsat all frequencies used in the acoustic measurements.

'T. O. Goldsberry,
S.P.Pitt,andR. A. Lamb,"Acoustic
Backscatter
from
the Ocean.Bottom,"preentedat The AcousticalSocietyof Ame.a
Meeting,Orlando,Florida,9-12 November
1982.[Copies
of thispaper
areavailable
fromARL:UT andmaybeobtained
byrequesting
document

The frequencydependence
of the bottombackscatter- number ARL-TP-82-46.]
andC.D. Anderson,
"Measurements
ofBackscattering
ingstrengthovertherangeof frequencies
usedwasobserved ZC.M. McKinney
of Soundfrom the Ocean Bottom,"J. Acoust.Soc. Am. 36, 158-163
to follow a 101ogf n, wheren wasbetweenI and 1.5. This
(1964).
observed
behavioris consistent
with resultsreportedin Ref.
T. L Schultz,
"Undersca
Revcrberation
(U),"BoltBeranek
andNewman,
2.

973

Rep. 4081255(December1965),confidential.

J. Acoust.
Soc.Am.,VoL77, No.3, March1985

Boehme
etal.: Acousc
backscattering.
PartI

973

4A.V. Bunchult
and.Y. Zhitkovskii,
"Sound
Scattering
bytheOcean

L I. Urik,PrinCll!es
of ndenoager
Sound(McCn'aw-Hill,
NewYork,

Bottomin Shallow-WatRegions
(ReviewL"
Soy.Phys.Acoust.26,363-

sH.-I WongandW. D. Chesterman,

"Bottom
Backscattering
NearGraz-

370

SM.D.Richardson,
D. K. Yom andR.I. Ray,"Environmmtal
Support
for High FrequencyAcousticMeasurements
at NOSC Oceanographic

ing Incidencein ShallowWater,*' I. Acoust.Scc. Am. 44, 1713-1718

0968

"FtrstAnnualReporton
Tower,
26April-7May1982;
Pat.I: Sediment
Geoacoustic
Properties," 9I. I. Cl'l Y. Ii,arahi,andD. R. Jackson,
NORDA Tech.Note219,NavalOceanResearch
andDevelopment
Activity,NSTL Station,Mii,ipl [tune1983}.

6H.
Acousc
ScbulUi,
Soc.Am.
and34,
H. 864-S65
W. Marsh,
{1962}.
"SoundAbsorption
in Sawater."

rCP BottomScatteringMeasurements,"
The TechnicalCooperation
Program,Subgroup
O {June1980}.

*M.D. Richsrdson,
NavalOceanResearch
andDevelopment
Activity
(pvatecommunication}.

974

J. Acoust Soc. Am.,Vol. 77, No. 3, March1985

Boehrneet .: Acoustic
backscattering.
PartI

974