Edited by D. Emery and K. J. Myers. ae a 4 — a bed a eSEDITED BY Dominic Emery and Keith Myers BP Exploration, Stockley Park Uxbridge, London Sequence Stratigraphy WITH CONTRIBUTIONS FROM George Bertram Cedric Griffiths Nick Milton Tony Reynolds Marcus Richards Simon Sturrock b Blackwell ‘Science© 196 by lgckorll Science Lid Editorial Offices: Osney Mead, Oxford OX2 OBL. 25 John Stree, London WCIN 2B. 23 Aina Place, Edinburgh EHS 64 350 Main Stret, Malden 'MA 02148 5018, USA 54 University Stree, Cason Victoria 3053, Australia 10, rue Casimir Delavigne 75006 Paris, France Other Editorial Offices Blackwell Wissenschafs-Verlag GmbH Kurforsendamm 57 10707 Berlin, Germany Blackwell Science KK MG Kodenmacho Building 7-10 Kedenmacho Nihombashi ‘Chou-ku, Tokyo 108, Japan “The right ofthe Author to be Identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988, Al rights eeserved. No part of this publcatian may be reproduced, stored ina retrieval system, or transmitted in any form or by any means, eletronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK ‘Copyright, Designs and Patents Act 1988, without the prior permission ofthe copyright owner. Firse published 1996 Reprinted 1997, 1988, 1999 Set by Setrte Typesetters, Hong Kong Peinged and bound in Great Britain atthe Alden Press Lid, ‘Oxford and Northampton “The Blackwell Science Lago is a trade mark of Blackwell Science Ltd, registered atthe United Kingdom Trade Marks Registry ‘Marston Book Services Led PO Box 209 Abingdon Oxon O14 4YN (Orders: Tel: 01235 465500 axe 01235 465555) USA Blackwell Science, In. Commerce Place {350 Main Street Malden, MA 02148 5018 (Orders: Tel 800 759 6102 G17 388 $250, Fax: 617 388 8255) Canada ‘Login Brothers hook Company 324 Main Street, Malden, MA 02148 S018 (Orders: Tel: 204 837-2987) Australia Blackwell Science Pry Ltd 54 University Street Calton, Vieworia 3053 (Orders: Tel: 3 9347 0300 Fax: 3 9347 5001) A catalogue record for this tle is available from the British Library ISBN 0.632.037067 Library of Congress Cataloguing-ia-Publieation Data Sequence stratigraphy/edited by Dominic E and Keith Myers; wich contributions from George Bertram. [eal Pp em. Includes bibliographical references and indes: ISBN 0-632.037067 1. Geology, Statigraphic, 1 Emery, Do T, Myers, Keith IL. Berean, George T. QESSL.SAS8” 1936 S517 — de20Contents List of Contributors, iv Preface, Historical Perspective 1D. EMERY: Historical Perspective, 3 Concepts and Principles 2. KJ. MYERS AND N.J. MILTON: Concepts and Principles of Sequence Stratigraphy, 11 Sequence Stratigraphic Tools 3 G.T. BERTRAM AND N,J. MILTON: Seismic Stratigraphy, 45 4 NJ. MILTON AND D. EMERY: Outcrop and Well Data, 61 5 NJ. MILTON: Chronostratigraphic Charts, 80 6 S.J. STURROCK: Biostratigraphy, 89 Applications to Deposi ional Systems 7 M.T. RICHARDS: Fluvial Systems, 111 8 A.D. REYNOLDS: Paralic Successions, 134 9 M.T. RICHARDS: Deep-marine Clastic Systems, 178 10 D. EMERY: Carbonate Systems, 211 11 KJ. MYERS: Organic-rich Facies and Hydrocarbon Source Rocks, 238 12. C.M. GRIFFITHS: Computer Modelling of Basin Fill, 258 References, 270 Index, 291List of Contributors GEORGE BERTRAM Stratigraphic Research Intemational, Brachead Avenue, Milngavie, Glasgow, UK DOMINIC EMERY BP Exploration, Uxbridge One, 1 Harefield Road, Uxbridge, London, UK CEDRIC GRIFFITHS NCPGG, Thebarton Campus, University of Adelaide, Adelaide, Australia NICK MILTON Bp Norge, Forus, Stavanger, Norway KEITH MYERS BP Exploration, Uxbridge One, 1 Harefeld Road, Usbridge, London, UK TONY REYNOLDS BP Exploration, Sunbury-on- ‘Thames, London, UK MARCUS RICHARDS BP Exploration, Anchorage, Alaska, USA SIMON STURROCK 56 Glowcesier Court, Kew Road, Kew, London, UKPreface In 1989, the chief geologist of BP Exploration and his, senior colleagues recognized the need to expand the company’s resource of sequence stratigraphers, and created, the Stratigraphic Studies Group. This group initially in- cluded a few experts, but was composed chiefly of a mixture of willing geophysicists, sedimentologists and bio- stratigeaphers, who were to train assequence stratigraphers, but more importantly, were to bring the expertise from their own disciplines to bear on sequence stratigraphy. This, merging of geological disciplines with sequence stratigraphic principles first saw the light of day as BP's “Introduction to Sequence Stratigraphy’ course. This course has been given internally since 1991, and has been presented in whole oF in part to over a dozen national oil companies and at several international geological and geophysical con- ferences. The ‘Introduction to Sequence. Stratigraphy” course manuals formed the basis for this book, because, as, wwe naively thought, it would not be too much trouble to, recast the manuals in the form of a textbook, which could, form the basis of university and professional courses. In- evitably, it was more difficult than we had imagined, Sequence stratigraphy is a rapidly evolving subject, new terminology was being added as we wrote, and the jargon we sought to demystify continues to grow. This book must thus be sven as a sequence stratigraphic synthesis for the mid-1990s, and not the final word on the subject. It is above all a practical guide and contains tools and tech- niques that the authors have found useful in their daily work. We have arranged the book to cover four main themes; a brief history of sequence stratigraphy, concepts and prin- ciples, sequence stratigraphic tools and finally the appli- cation of sequence stratigraphy 10 different depositional systems. For the last theme, we have tried to emphasize the importance of secing sequence stratigraphy in its sedimen- tological context, and it is recommended that the reader should have some familiarity with secimentological processes before tackling the last five chapters. Otherwise, the book covers ll the basics of sequence stratigraphy, and is intended to be a broad text suitable for undergraduate ‘geologists of all years, MSc and PhD sedimentologists and, stratigeaphers, and for oil company geoscientists who wish to broaden their knowledge of the stratigraphic methods available for solving problems with which they are routinely faced. We hope you find it interesting and useful. Acknowledgements Bob Jones contributed several sections on biostratigraphy and Neil Parkinson contribured to Chapter 2. We are srateful to the following reviewers who have helped improve the book at various stages of its development: David Roberts, Henry Posamentier, Maurice Tucker, Dan Bosence, Mike Bowman, Andy Horbury and Andy Fleet. We also acknowledge the support of BP Exploration, particularly David Roberts, Bob Rosenthal, John Wills and Peter Melville and we are grateful to BP's partners for permission to publish information on many of the areas and fields mentioned in the text. The following companies are also acknowledged for their assistance in providing seismic data; Lynx Information Systems, Trans-Asia Oil and Mineral Development Corp., Balabac Oil Exploration and Drilling Co. Inc., Crestone Energy Corp., Coplex. (Palawan) Ltd., Oriental Petroleum and Minerals Corp., The Philodrill Corp., Seafront Resources Corp., Unioil and Gas Development Co. Inc, Vulean Industrial and Mining Corp. George Bertram, Dominic Emery, Cedric Griffiths, Nick Milton, Keith Myers, Tony Reynolds, ‘Marcus Richards and Simon Sturrock BP Exploration London, Sunbury-on-Thames, Glasgow, Stavanger and AnchorageCHAPTER ONE Historical Perspective 1.1 What is sequence stratigraphy? 1.2 The evolution of sequence stratigraphy 1.1 What is sequence stratigraphy? Sequence stratigraphy is a subdiscipline of stratigraphy, the latter being defined broadly as “the historical geology of stratified rocks’. There have been many definitions of Sequence stratigraphy over the years, but perhaps the simplest, and that preferred by the authors, is ‘the sub- division of sedimentary basin fills into genetic packages hounded by unconformities and their correlative conform- ites’. Sequence stratigraphy is used to provide a chrono- stratigraphic framework for the correlation and mapping, of sedimentary facies and for stratigraphic prediction. Several geological disciplines contribute to the sequence stratigraphic approach, including seismic stratigraphy, bio- stratigraphy, chronostratigraphy and sedimentology. These are discussed in more detail in forthcoming chapters. Note that lithostratigraphy is not considered to contribute use- fully to sequence stratigraphy. Lithostratigraphy is the correlation of similar lithologies, which are commonly diachronous and have no time-significance (Fig. 1.1). Litho- stratigraphic correlation is useful provided the sequence stratigraphic boundaries enveloping the interval of interest are constrained. Fig. 1.1 The difference between sequence stratigraphy, which has a geological time significance, and lithostratigeaphy, which correlates rocks of similar type. A lithostratigraphic correlation would correlate conglomerate units 1 and 2, sandstone units 3, 4 and 5 and ‘mudstone units 6, 7 and 8. A sequence stratigraphic correlation would correlate time lines A—A', BB! and cc Tine 1.2 The evolution of sequence stratigraphy Sequence stratigraphy is often regarded as a relatively new science, evolving in the 1970s from seismic stratigraphy. In fact sequence stratigraphy has its roots in the centuries-old controversies over the origin of cyclic sedimentation and eustatic versus tectonic controls on sea-level. Much of this early debate has been summarized recently in a set of historical geological papers edited by Dott in 1992 (1992a), entitled “Eustasy: the Ups and Downs of a Major Geological Concept’, and the interested reader is referred to this, volume for more detail. Other historically important collec- tions of sequence stratigraphic papers include American Association of Petroleum Geologists (AAPG) Memoir 26, published in 1977, and Society of Economic Paleontologists and Mineralogists (SEPM) Special Publication 42, pub- lished in 1988. Sacred theories ‘The Deluge and the story of Noah is the most well-known, of the earliest references to sea-level change. To the early investigators of sea-level change, the veracity of the Delugewas not in question, but its origin was the subject of considerable debate by scientists and clergy alike. Pechaps the most popular of several theories were Burnet’s Sacred Theory of the Earth, published in 1681, and the Telliamed of de Maillet, published in 1748 (and recently revisited by Carozzi in 1992), De Maillet proposed that following the formation of the Earth by the aceretion of the ashes of burning suns over the cortex of an extinguished sun, a water envelope which developed around the planet gradu- ally diminished in volume through time, and in so doing created the topography we see today. In effect, de Maillet interpreted sea-level changes on Earth as a ‘single falling limb of a cosmic eustatic cycle’ (Carozzi, 1992). This ‘concept of a one-way sea-level fall was known as Neptunian theory. The erosion of primitive mountains by marine processes and the development of a series of offlapping ‘Suporcal bedes Prime and cssiferoue mourning oposisof stage A ‘Seatel, stage A _—Merine_srosion sediment packages as implied by de Maillet and other Neptunists is illustrated schematically in Fig. 1.2. ‘The eighteenth century The eighteenth century also saw the beginning of detailed stratigraphic analysis of rock units, and the recognition of unconformities as primary bounding surfaces. In 1788, Hutton first appreciated the significance of unconformities separating cycles of ‘uplift, erosion and deposition’, and unconformities were used by stratigraphers such as Sedgwick and Murchison in the following century to estab- lish physical boundaries for geological periods (Sedgwick and Murchison, 1839). As the great stratigraphers con- tinued with their practical approach, William Buckland (1823) proposed the concept of Diluvialism which was to Spongy erst of fetnguishod sun Siparical © Secondary Primitive Secondary Seposisof | mountains | mountains mountains sage 8 A ce ios ‘Seaove, stage B Marine 205100 ering erasion Secondary ‘secondary Fig. 1.2 Schomatc illustrations of the development of sedimentary units by ‘marine erosion and deposition during continuously falling sea-level, according to de Maillet (1748). After Caro (1992)‘eclipse Neptunian theory. In diluvial theory, the geological products immediately preceding the flood were referred to as antediluvial, and those following the flood were referred to as post-diluvial or alluvial. However, the attraction of dilavial theory also soon waned as further geological ‘evidence served to counter the simplistic notion of a single dramatic flooding event. The nineteenth century In the middle of the nineteenth century, the eustatic versus tectonic controlson sea-level change debate began in earnest with the glacial theories of Lyell and Agassiz. Lyell and others (including Celsius and Linnaeus) observed raised beaches along the coastline of Scandinavia and noted evidence of falling sea-levels from centuries-old marks on shoreline outcrops. Lyell concluded that the land was being, slowly and differentially elevated (Lyell, 1835), a fact confirmed by Bravais in 1840 who had observed tilted beaches along fjords of the Scandinavian Arctic coast. At about the same time, Agassiz (1840) was developing his theories of glaciation, and MacLaren, on reviewing ‘Agassiz’s glacial theory in 1842, saw the potential of melting ice-caps as a major control on global sea-level. Unfortunately, neither Agassiz nor MacLaren received acceptance for their ideas for at least two more decades, ‘until Croll (1864), in a forerunner of Milankovitch theory (1920), published the concept of orbitally forced glaciations. ‘fara sufaco ‘a Seen ® Fig. 1.3 Reproduction of figures feom ‘Chamberlin (1898) on diastrophism, tunconformities and geological time ivisions, From Dott (1992b) The early twentieth century By the late ninereenth century, glacial theory was thus able to explain eustatic sea-level change and isostatic uplift. However, it was to be several decades before glacial eustasy was resurrected as a control on sedimentary rhythmicity; other explanations of global eustasy took precedence, notably the work of Eduard Suess. Suess first coined the term eustasy in 1906, when he attributed the patterns of onlap and offlap of sedimentary units to global sea-level changes. Suess favoured a mechanism whereby sea-level was lowered by subsidence of the sea-floor, and raised by the displacement of seawater by oceanic sedimentation. He refused to believe the evidence for differential land uplift from Scandinavia, concluding that the Baltic was ‘gradually ‘emptying’ (Sucss, 1888). However, the majority of geo!- ogists in the early owentieth centuty still held the Lyellian view that the major control on sea-level at any point along, the coast was the movement of the land. Despite the general lack of support for Suess? ideas, a number of American geologists began to develop concepts of global controls on unconformity development. Foremost amongst these was Chamberlin, who in 1898 and 1909 published his theory on the ‘diastrophic control of stratigraphy by ‘world-wide sea level changes’. Three diagrams from his, first paper show this to be a precursor of modern sequence cepts (Fig 1.3). leas were developed by several American, stratigraphic cor ‘Chamberlin’s often ‘sualevel mares aanoadsiag peotiorsiyy | 1 4ardeqsgcologists in the following decades, particularly in Palaco- zoic systems of the Mid-west. Most notable amongst these were Ulrich and Schuchert, the latter using early palaco: ‘geographic concepts and facies theory to re-create past ‘environments bounded by global unconformities. However, the single most important publication from the ‘eustatic’ school was that of Grabau, a contemporary of Ulrich and Schuchert, whose ‘pulsation theory" postulated rhythmic transgressions and regressions caused by changing heat flow from inside the Earth. The resulting ‘pulse beat of the Earth’, published in Grabau's ‘The Rhythm of the Ages’ in 1940, had a periodicity of about 30 million years and used the development of global unconformities, which could be used to divide the stratigraphic record. Prior to Grabau’s work, European geologists, notably Stille (1924) had begun to develop ideas about global unconformities, caused by global rectonism with resulting eustatic effects, akin to modern low-order eustatie cycles. On a smaller scale, sedimentary chythms were being, ‘observed on a scale of metres in coal-beating Carboniferous (Pennsylvanian) strata in Mlinois and Kansas. In. 1935, following further studies on Pleistocene glacio-custatic changes, Wanless and Shepard proposed a control on the development of these Pennsylvanian ‘cyclothems’ by the accumulation and melting of Gondwana glaciers. This, study and others like it resurrected the glacial-custatie control on sedimentation developed by Croll many decades. carer. The case for periodicity at a variety of scales in the stratigraphic record was thus becoming compelling. How- ever, as with so many scientific bandwagons, it eventually ran out of steam, In a keynote address to the geological community in 1949, Gilluly argued for orogenesis asa continuous, rather than episodie process, and as a result the concept of rhythmicity of low order (tens of millions of years) gradually lost credibility. The Carboniferous, cyclothems were then reinterpreted as autocyclic products, resulting from delta lobe switching and the internal re- ‘organization of sedimentary systems. This latter point also ‘emphasizes the ascendancy of process sedimentology in the carly 1960s, Dott (1992a) amusingly points out that at that time many stratigraphers preferred to call themselves, sedimentologists! The late twentieth century Ironically, Sloss, Krumbein and Dapples (1949) first out lined the concept of stratigraphic sequences at the samg meeting that Gilluly proposed his ideas on orogenic con- tinaum. Sloss, Krumbein and Dapples defined sequences as “assemblages of strata and formations’ bounded by promi- nent interregional unconformities. Despite the negative reaction 10 these ideas, Sloss (1963) published his major sequences correlateable across the North American Craton, the Indian Tribal names of which still appear as ‘super sequences’ on the Haq ef al. (1987) chart. Sloss’s ideas were developed further by his graduate students at North- western University, one of whom was Peter Vail. Also published at this time was Harry Wheeler's classic 1958 paper on time-stratigraphy which contains many of the concepts in use today, as well as an early attempt t0 introduce sequence stratigraphic terminology. Seismic stratigraphy ‘The next major breakthrough in sequence steatigraphy was in the 196s and 1970s, when the development of digitally recorded and processed multichannel seismic data made large scale two-dimensional images through basins avail- able, Vail etal. (1977a} in AAPG Memoir 26 is perhaps the most referenced work on sequence stratigraphy to date. It summarizes work carried out by Vail and his co-workers, first in the Carter Oil Company and subsequently at the Exxon Production Research Corporation, through the 1960s and early 1970s (Vail and Wilbur, 1966; Mitchum et al., 1976). This period of time marks a break where industry took the lead from academia in the development of sequence stratigraphy. Farther papers on seismic sequence stratigraphy followed, and the ideas were gradu- ally extended to incorporate both borehole and outcrop (Wail etal, 1984). In this work, eustatic sea-level was data emphasized as the controlling mechanism for sequence development. In 1985 AAPG Memoir 39 appeared, in which Hubbard et al. proposed a tectonic mechanism for the subdivision of basin fill nto ‘megasequences’, driven by changes in tectonic process. The tectonic versus eustatic debate was beginning afresh, although for many at this time seismic stratigraphy was synonymous with eusta sealevel change, possibly because of its appeal as a global predictive tool for hydrocarbon exploration. In 1987, the Hag etal. global sea-level cycle chart was published. This is possibly the single most contentious of all the ‘Exxon school” publications, chiefly because the supporting evi dence for the curves has not been released. It remains tunclear whether local corrections for tectonic uplift or subsidence have been applied, and the dating of uncon formities to the accuracy implied by the chart has be challenged (Miall, 1991). ‘The sequence stratigraphy bandwagon rolls Special Publication 42 of SEPM, Sea Level Changes — an Integrated Approach, was published in 1988 and inte: duced new concepts such as accommodation space and parasequences, and many of the concepts and principles described in Chapter 2 of this book. Special Publication 42 was important hecause it opened up the subject t0 a broader geological community beyond industrial seismic interpreters. In the late 1980s and in this decade, many sequence stratigraphic publications have appeared, some of which uncritically apply the cools and techniques, and some of which are strongly critical. Many question thevalidity of the interbasinal correlations upon which the Haq et al. (1987} curve is based, and others have questioned the validity of certain aspects ofthe sequence stratigraphic models presented in SEPM 42, such as Miall (1991) and Schlager (1992). Galloway (1989) presented an alternative model for the development of depositional units or ‘genetic stratigraphic units’ bounded by major flooding surfaces, unconformities, Pitman (1978) has suggested that the origin of sequences and onlap patterns can be explained by variations in subsidence at continental ‘margins, whereas Cloetingh (1988) and Kooi and Cloetingh (1991) proposed that relative sea-level changes and the formation of sequences of millions of years duration can be explained by intraplate stresses rather than eustatic sea- level changes. ‘The most recent developments in sequence stratigraphy have been in the area of high-resolution subseismic-scale sequence stratigraphy and computer modelling of sedimen- tary fll, Van Wagoner et al. (1990) led the way with the publication of a colourful text on high-resolution sequence stratigraphy from outcrops, logs and core. This stimulated excellent work in superbly exposed marine and marginal ‘marine settings, such as the Jurassic of the Yorkshire Coast and the Cretaceous of the Western Interior Seaway, USA {sce also Posamentier and Weimer, 1993, for review). High-resolution sequence stratigraphy also has been com- bined with work on metre-scale rhythmic successions, par- ticularly bedded platform carbonates and mixed siliciclastic carbonate units (Hardie e¢ al, 1986; Goldhammer et al., 1991). Milankovitch theory of orbital forcing has been revived by sequence stratigraphers in order to explain rather the origin of high-frequency subsequence-scale cycles, Computer modelling packages have been developed to analyse and replicate the sedimentary fill of basins, at scales, swide models include those developed by Royal Dutch/Shell, and pub- lished by Aigner e¢ al, (1990), and the SEDPAK program developed at the University of South Carolina. Smaller scale cyelicity has been modelled by software such as Mr Sediment (Goldhammer ef al., 1989) and by Bosence and Waltham (1990). from a few metres to entire basins. Bs The future The future direction of sequence stratigraphy is difficult t0 predict, given the turbulent history of the sea-level change debate. At leastin the short-term, carbonate systems require further case studies to demonstrate the importance (or otherwise) of controls other than sea-level change. Posamentier'and Weimer (1993) have also emphasized the need for further work on the applicability of the concepts, to non-marine and deep-marine settings, and further vali dation (or otherwise) of the sea-level cycle chart from outcrop and subsurface data. Schlager (1992) and others, also argue for a more sedimentological approach to sequence stratigraphy, accounting for the autocyclicity of sedimentary processes within the sequence stratigraphic framework. At the very least we can expect considerable debate and further critiques of the subject. This level of activity and debate is all a far cry from the early 1960s, when stratigraphy was unfashionable, before Peter Vail and others rescued the subject from its decline.CHAPTER TWO Concepts and Principles of Sequence Stratigraphy 21 Introduction 2.1.1 Basin forming processes 2.1.2 Basin-margin concepts 2.4 Sequences and systems tracts 2.4.1 Sequences and sequence acts boundaries ‘order) sequences and systems 2A.11. Genetic stratigraphic sequences 2.4.2 Systems tract definition 2.2 Relative sea-level, tectonics and custasy 2.43 Lowstand systems tract 2.5 High-resolution sequence steatigraphy 22:1 Definition of sea-level 2.414 Transgresive systems tact and parasequences 22.2 Accommodation 2.4.5 Highstand systems tact 25.1 Introduction 2.2.3 Accommodation through time 2.4.6 Type 2 sequence boundary and the 2.5.2 Parasequences and theit 22.4 Orders of eylicity and global slellémargin systems tract continental equivalents corceation 2.4.7 Lowstand systems tracts on a Parasequence sets amp margin Parasequence thickness trends 223 Sediment supply 2.4.8 Conteols on systems ect Sequence boundaries 23.1 Principles of clastic sediment boundaries Maximum flooding surfuces supply 2.4.9 Other possible systems tracts Ravinement surfaces 23.2 Filling of accommodation 23.3 Basin architec 2.41 Introduction The stratigraphic signatures and stratal patterns in the sedimentary rock record are a result of the interaction of tectonics, eustasy and climate. Tectonics and eustasy con- trol the amount of space available for sediment to accum late (accommodation), and tectonics, eustasy and climate interact to control sediment supply and how much of the accommodation is filled. Autocyclic sedimentary processes control the derailed facies architecture as accommodation is filled. The purpose of this chapter is to introduce the principles that govern the creation, filling and destruction dation. It then shows how these principles are ide the rock record into sequences and ‘systems tracts’, which describe the distribution of rocks in space and time. The chapter uses siliciclastic systems to introduce the concepts and principles of sequence stratigraphy. Carbonate systems differ from clastic systems in their ability to produce sediment ‘in situ’, and they respond in a different manner 1 accommodation changes. Carbonates are therefore dis- cussed separately in Chapter 10. 2.1.1 Basin forming processes Tectonism represents the primary control on the creation and destruction of accommodation. Without tectonic sul sidence there is no sedimentary basin. It also influences the ‘within a reatve sea-level cycle 2.4.10 Composite (second and third Problems and pitfalls of high resolution sequence stratigraphy rate of sediment supply to basins. Tectonic subsidence results from two principle mechanisms, either extension oF flexural loading of the lithosphere. Figure 2.1 illustrates, theoretical tectonic subsidence rates in extensional, foreland, and strike-slip basins. These curves in effect govern how much sediment can accumulate in the basin, modified by the effects of sediment loading, compaction and eustasy. Extensional basins form in a variety of plate tectonic settings, but are most common on constructive plate margins. In extensional basins, tectonic subsidence rates vary systematically through time, with an initial period of very rapid subsidence caused by isostatic adjustment to lithosphere stretching, followed by a gradual (60-100 million years) and decreasing thermal subsidence phase as the asthenosphere cools. This systematic change in tectonic subsidence rate has a strong influence on the geometry of the basin-fill, such that it may be possible to divide the stratigraphy into pre post- and syn-rift phases (these phases have been termed megasequences; Hubbard, 1988). In the simple syn-rift megasequence model the sediments are deposited in the active fault-controlled depocenires of the evolving rift and can show roll-over and growth into the active faults. Differential subsidence across the exten- sional faults may exert a strong control on facies distri- butions. In the post-rift megasequence, any remaining rift- related topography is gradually buried beneath sediments that fill the subsiding basin and onlap the basin margin, creating the typical ‘steers head” geometry (McKenzie, uujosiy aauapisqns uODaL PZ BL suyseg dys-oys pue pur] ‘auapsq7s pue mo4 eSH () ‘vapsans poe moy aH (@) (x ys 1w0d opus) couapleqns ue 1H (0) rs Med aint) a es a 1 2 fe oy mo l. cos ° el os os 104 he serow aue. 9) pou su001 (9) uowanoas ecto. on (1) ayions wroqun—{@u61) @zUB,ON1978). Thesyn-rift and post-rift megasequences in a marine rife will contain sequences in which development is con- trolled by higher frequency changes in relative se Foreland basins develop in response to loading of the lithosphere below thrust belts. The lithosphere bends in response to loading as the thrust sheets are emplaced, and creates a depression that is accentuated towards the load. ‘The sedimentary fil to this foreland basin has a character- istic wedge shape, thickening towards the thrust front and forming a foreland basin megasequence. The width of the basin is proportional to the rigidity of the underlying lithosphere, and the depth is proportional to the size of the load. Foreland basins formed adjacent to growing mountain belts are characterized by large, and initially rapidly inc ing, rates of sediment supply. Cessation of thrusti continued erosion of the mountain belt leads to an eventual decrease in load, and many foreland basins become uplifted. Strike-slip basins do not have a characteristic subsidence pattern, although in general, rates of subsidence (and uplift) are extremely rapid Tectonic subsidence curves provide a fundamental con- trol on sediment accommodation, upon which higher fre- quency controls, such as custasy, faule movement and iapirism, are superimposed. Figure 2.2 shows calculated tectonic subsidence curves for two real basins. In the Llanos Basin, Colombia, sediment supply has exceeded tecton subsidence, The basin has remained full to base level, with «excess sediment bypassed northwards to the sea. The sub- sidence curve shows slow subsidence ehrough the late Cretaceous and carly Tertiary, linked to thermal subsidence ina back-arc basin setting. Two distinct increases in subsidence rate occur in the mid—late Eocene and mid: Miocene, corresponding to two phases of mountain building inthe Andes. Inthe South Viking Graben example (Fig. 2.2), typical of a number of rifts, sedimentation has not always kept pace With true tectonic subsidence. This led to periods in the Cretaceous where water depths inereased and sediment starvation occurred. In the Tertiary, uplift of the Scottish mainland and adjacent North Sea Basin resulted in increased sediment input to the basin (Milton ef al.. 1990}, which locally filled to base level. The remainder of the basin, subsequently filled with sediment, resulting in the present- Separation of the syn-rift and post-rife in this basin is difficult, because the transition occurred during a period of sediment starvation (Milton, 1993). During periods of rapid basin subsidence, sequence boundaries generated by higher frequency eustatic sea-level falls will be obscured. In times of slow tectonic subsidence ‘or basin uplift, sequence boundaries will be enhanced. day shallow se 2.1.2 Basin-margin concepts Many of the concepts and principles of sequence stratigra- phy are based on the observation from seismic data that prograding, basin-margin systems often have a consistent LLANOS BASIN COLOMBIA Tk Tea [eo lo. [ws _[pr = sy i thermal subsidence phase F 1009 : E 509 SOUTH VIKING GRABEN 7 IK Tt cea Semapiiee, J and Se a >. em ate. Tuas bar+} ‘amet oe 150° 700 80, a Sou ing Graben eubseres cave Fig. 2.2 Calculated tectonic subsidence histories for two sedimentary basins (reproduced by permission of BP Exploration Led) depositional geometry (Fig, 2.3). Topset is a term used to describe the proximal portion of the basin-margin profile characterized by low gradients (<0.19). Topsets effectively appear flat on seismic data and generally contain alluvial, deltaic and shallow-marine depositional systems. The shoreline can be located at any point within the topset. It can coincide with the offlap break or may occur hundreds of kilometres landward. The proximal termin- ation of the topset is usually termed the point of coastal onlap, referring to the up-dip limit of coastal-plain or paralic facies. Clinoform is used to describe the more steeply dipping portion of the basin-margin profile (com- monly > 1°) developed basinward of the topset. Clinoforms, generally contain deeper water depositional systems charac- teristic of the slope. The slope of the clinoform generally 13‘Coastal ola (erbay ine) ——-Shoreine — ottlag break Key: 0 Suocessive postors ofthe ofp break can be resolved on seismic data. Bottomset is a term sometimes used to describe the portion of the basin-margin, profile at the base of the clinoform characterized by low gradients and containing deep-water depositional systems. The main break in slope in the depositional profile ‘occurs between topset and clinoform and is called the offlap break (Vail et al., 1991). The offlap break previously has been termed the shelf edge (Vail and Todd 1981; Vail et al, 1984), leading to a confusion with the shelf break, i.e. the edge of the modern continental shelf, which is usually a relict feature rather than depositional feature. The term depositional shoreline break (Van Wagoner et al., 1988) also has been used, but this implies that the main break in slope in a depositional profile coincides with the shoreline. The term offlap break is preferred here as it does, not imply coincidence of the main break in slope with the shoreline. The ropset—clinoform profile results from the interplay between sediment supply and wave, storm and tidal energy in the basin, Sediment enters the proximal end of the profile through river systems and is distributed across the topset area by wave- and/or current-related processes. These may include fluvial currents, tidal currents, storm, currents, et. However, these topset transport processes are effective only at relatively shallow depths of up co a few tens of metres, and to move sediment into deeper water a slope must develop in order to allow sediment transpor- tation by gravity processes. The clinoforms build to the angle needed to transport sediment at the required rate. Slope angle is strongly influenced by sediment calibre. Coarse-grained sediment, with a higher angle of rest, will build up steeper slopes than fine-grained sediment (Kenter, 1990). Also, carbonate systems generally can build steeper depositional slopes (up to 35°) than fine-grained clastic systems (0.5—3") owing to their greater shear strength. Steeper slopes in clastic systems generally are either made fof coarser grade material or are zones of erosion and sedimentary bypass. The importance of the offlap break on the depositional systems is most apparent during relative sea-level fall (see 2.2.1), When relative sea-level fall exposes the offlap break, rivers commonly incise in order to re-equilibrate to lowered base level, with the result that the river becomes entrenched “4 Fig. 2.3 Typical profile of a prograding basin-margin wait, comprising topsets and clinoforms separated by a break in slope; the offlap break. Bottomsets may also be present Botomate at its mouth (discussed in 2.4.3). The response of the depositional systems to this fallin relative sea-level depends con the nature of the basin margin (Fig, 2.4) Shelf-break margins are those with well developed depo- sitional clinoforms. Fluvial entrenchment during sea-level fall may result in focusing of the sediment load to discrete locations on the clinoform slope. Failure of the sediment ‘mass has the capacity for forming large turbidity currents and submarine fan deposits. Shelf-break margins are typical of passive continental margins at times of slow rise of relative sea-level, when the delta systems can easily prograde to the shelf edge Ramp margins are characteriied by relatively shallow water depths, where storms and current processes can ‘operate over much of the area of deposition. Depositional angles are generally less than 1° and seismic clinoforms (if, resolved) are shingled with a dip of around half a degree. ‘The offlap break on-a ramp margin is likely to be at the shoreline, where fluvial gradients pass into slightly steeper shelf or delta-front gradients. The response of the depo- sitional systems in a ramp setting to relative sea-level change is therefore different from the shelf-break margin. In particular, deep-water tusbidite deposition during low- stand may be absent, or of only minor significance. Depo- sitional systems will, instead, be translated basinward without significant slope bypass or basinal deposition. Any turbidites found on a siliciclastic ramp margin are likely to be delta-front turbidites, rather than detached submarine fans (Van Wagoner et al., 1990). Many modern delta systems can be considered to form ramp margins, as generally they are shelf deltas prograding, 6on to the drowned ropsets of a previous shelf-break margin (Fig. 2.4), Frazier (1974) has shown that deposition on the Continental shelf of the Gulf of Mexico is confined to the Mississippi delta, which is prograding into about 100m of water. The rest of the shelf is effectively an area of non- deposition. The Mississippi delta presently forms a ramp. margin, although very litle extra progradation is needed for the delta to reach the shelf edge, and for the margin to become a shelf-break margin. Rift margins characterize basins undergoing actiye crustal extension, Extensional faults have a strong, influence on both palacogeography and sediment influx rates. The spatial(a) Shel Break Margin Seaevel I 1008 Large watordops. Slope faces are wel developed esa Pe Spe lave are not elrican amp marin form in intacatoric basin, oF 5 chown on tho drowned topsets of previous shel-reak margins (oRKMagn ee ‘Basement opagrapy i high-eli, ‘and basement slopes may exco0d the Clinoform ange, forcing bypass Sediment supply rato may tow (@ Foreland Basin Suteldenceis greatest near tho mountain trot (poximal inthe system). Sedent supply ate canbe righ (6) Grow Faut Margin Fig. 24 Basin-margin types: (a) shelf-break margin; (b) ramp margin; (c) rift margin; (d) foreland basin, () growth-fault margin distribution of sediment accommodation within the rift is controlled largely by tectonics. Subsidence rates generally will increase from the margins to the centre of the rift, although each individual fault block will have its own pattern of accommodation, The foot-wall crest will see the least subsidence and may experience uplift and erosio whereas the hanging wall will experience progressively greater subsidence rates towards the controlling fault. The depositional systems that develop will depend on whether the rift is marine or continental. Transfer zones in the rift margin may control sediment entry points. Rift margins may be characterized by high topographic relief and rela- tive sediment starvation, because sediment is bypassed towards the rift centre. Basin-margin systems may build ‘uit into deep water with long clinoform slopes and rela- tively minor topsets (Fig. 24). There is little potential for trapping coarse material in the topsets, and much may be bypassed to the basin. Foreland-basin margins vary depending, on whether sedi ment is being fed axially along the foreland basin or directly into the foreland basin from the thrust belt. In the latter case, the rate of tectonic subsidence increases towards, the foreland thrust belt, ie. the sediment source area. In other words, sediment accommodation may be relatively high in proximal areas compared with the basin centre. ‘This has a marked affect on stratal geometries and may result in the aggradation of thick topset deposits, with litle ‘opportunity for seismic-scale clinoforms to develop (Posa- rmentier and Allen, 1993), Growth-fault margins are characterized by gravity driven syn-sedimentary extensional faults, The rate of subsidence is considerably greater on the hanging-wall side of the growth fault, resulting in an expanded sedimentary suc- cession. The effect of the growth fault in the depositional systems developed will depend on whether the fault had 18 topographic expression on the sea-bed. At times when the hanging wall was a topographic low relative to the foot wall, facies differentiation oceurs across the fault, with thick, deeper water clastic systems on the down- thrown side. Growth-fault margins are discussed further in section 9.3.3. 2.2. Relative sea-level, tectonics and custasy 2.2.1 Definitions of sea-level In order to understand the controls on sequence develop- ‘ment, it is first necessary to define what is meant by eustasy, relative 1 and water depth (Fig. 2.5, from Jervey, 1988). Global eustasy Eustasy is measured between the sea-surface and a fixed datum, usually the centre of the Barth, Eustasy can vary by changing ocean-basin volume (e.g. by varying ocean-ridge volume) or by varying ocean-water volume (c.g. by glacio- 1sFixod datum 0. contr oh the Earth ‘eustasy), The interpretation of eustatic changes from the rock record is accomplex and controversial topic, which, will be discussed in detail in section 2.2.4. For the moment it is important only to emphasize that it ean rise or fall, thus varying the base level for erosion on a global scale, where base level is defined as the level above which depo: tion is temporary and erosion occurs (see 2.2.2). Relative sea-level Relative sea-level is measured between the sea-surface and a local moving datum, such as basement oa surface within the sediment pile (Posamentier et al., 1988). Tectonic sub- sidence or upliftof a basement datum, sediment compaction involving subsidence of a datum within the sediment pile, and vertical eustatie movements of the sea-surface all con- tribute to relative sea-level change. Relative sea-level ‘rises’ due to subsidence, compaction and/or eustatic sea-level rise, and ‘falls’ due to tectonic uplift and/or custatic sea~ level fall. Relative sea-level should not be confused with water depth, which is measured between the sea-surface and the sea-bed in any given geographic location at a point in time. The term equilibriunt point is sometimes used 0 distinguish the point on a depositional profile where the rate of relative sea-level change is zero. The equilibrium, point will separate, at any given time, the zone at the basin margin where relative sea-level is falling from the zone where relative sea-level is rising. 2.2.2 Accommodation Eustasy and subsidence rate together control the amount of space available for sediment accumulation — this is con- ventionally termed accommodation, Accommodation is defined as the space available for sediment to accumulate at any point in time (Jervey, 1988). Accommodation is con- trolled by base level because in order for sediments t0 Accurusted sediment 2 Definitions of sea-level (after Jervey, 1988), ‘Eustatic sea-level’, base level for fluvial systems entering the accumulate, there must be space available below base level The base-level datum varies according to depositional setting (Fig. 2.6). In alluvial environments base level is controlled by the graded stream profile, which is graded to sea-level or lake-level at its distal end (Mackin, 1948; and Chapter 7). In deltaic and shoreline systems base level is effectively equivalent to sea-level. In shallow marine environments base level is ultimately also sea-level, although, fairweather wave base can form a temporary base level in the form of a “graded shelf profile’. Sediment supply fills the accommodation created and controls water depth: Aaccommodation = Acustasy + Asubsidence Sediment supply fills available accommodation, If the rate of sediment supply exceeds the rate of creation of accom= modation at a given point, water depths will decrease: Awater depth = Acustasy + Asubsidence + Acompaction — sediment deposited A series of cartoons in Fig. 2.7 illustrates the relationship between accommodation, relative sea-level and water depth in shoreline~shelf depositional systems, In these examples relative sea-level change and new sediment accommodation added are the same because base level is taken at the sea- surface. In the discussion that follows we examine the relation- ship between relative sea-level and accommodation in shoreline—shelf depositional systems. Fluvial systems and the controls on the graded stream profile are discussed in Chapter 7, paralic systems are discussed in Chaprer 8, submarine fans in Chapter 9 and carbonates in Chapter 10. 2.2.3 Accommodation through time In order to understand how accommodation varies through time it is useful to consider how different rates of tectonicwee (raced set pote Discharge — increases Secimertioad Decreases - Slope — Dectoases — : 2.6 Definitions of base level in fluvial, shoreline and shelf subsidence and, in this case, the same sinusoidal custatic sea-level curve combine 0 give different rates of addition and destruction of accommodation (rise and fall of relative sea-level) (Fig. 2.8; from Jervey, 1988). InFig. 2.8 subsidence is represented as a straight line, the gradient of which indicates the rate of subsidence at each Point. The different gradients can be thought of as rep- tesenting positions in a basin with increasing subsidence tates or changes in subsidence rate through time. Eustasy is, represented by the same smooth curve in each case. The change in relative sea-level through time is found simply by the addition of the two curves. Relative sea-level is, in this case, equivalent to accommodation because the curves begin at zero water depth. Where slow subsidence occurs, maximum accom- modation is developed near the custatie maximum, When eustasy falls to its original position, accommodation falls to a value representing that created only by subsidence. With increased rates of subsidence, the time of maximum accommodation is progressively later. Points in the basin where subsidence rates are very high experience no decrease in accommodation even though eustatic fall may be occur- ring. Note that the same curves could be produced theor- «tially by adding variable rates of tectonie subsidence and uplift ro a flat eustatic curve. 2.24 Orders of eyclicty and global correlation A depositional sequence represents a complete cycle of deposition bounded above and below by erosional uncon- formities, The sequence has a maximum duration, which is measured between the correlative conformities to the bounding unconformities. Thus, the duration of the sequence will be determined by the event controlling the creation and destruction of accommodation, ic. tectonic subsidence and/or eustasy. Tectonic cycles of subsidence and uplift and eustatic cycles of rising and falling sea-level, can operate over different time periods, and itis useful to, classify sequences in terms of their order of duration, commonly termed first, second, third, fourth order, etc (Fig. 2.9), A basin-il ean then be divided into a hierarchy ‘of sequences, each representing the product of a particular order of tectonic or eustatic cycle. In Fig. 2.9, from Duval ef al. (1992), four orders of stratigraphic cycle are depicted. The continental encroach- ment cycle is defined by the very largest scale (> 50 million years) cycles of sedimentary onlap and offlap of the super- continents. There are only two such cycles in the Phanero oie, according to the Hag et al. (1987) sea Firstorder continental encroachment cycles are considered to be controlled by tectono-eustasy, ie. changes in ocean basin volume related to plate tectonic cycles (Pitman, 1978).@ Rolaive seaevel (= Accommodation water doth) "aidaive Seniovel + Polatve sea-level ses om 11 0 (2 asa result of subaaance + Seciment supply > rate of oatve Soave! ee (accommedation increase) ‘+ Sediment accommedition increases rom 1 t 2 but water depth decreases resulting n facies bet regression + Relaive sea-ovel ries rom f1 to (2s. osu of subsidence + Seciment supply < rate of oativeseatevel ise (accommedation ncroaso) *+ Water and deptn sediment accommodation inerease from ¢1 to 12,1. transgression cttaces bets occurs Second-order (3~50 million years) cycles are the building, blocks of the first-order sequences and represent particular stages in the evolution of a basin. They may be caused by changes in the rate of tectonic subsidence in the basin or rate of uplift in the sediment source terrane. Thitd-order (0.5~3 million years) sequence ey the foundation of sequence stratigraphy because they are often of a scale well-esolved by seismic data. They are identified by the recognition of individual cycles of accom: modation creation and destruction. These cycles are con- sidered by Vail et al. (1991) to be controlled by glacio- eustasy, although other tectonic mechanisms are possible (Cloctingh, 1988) Composite sequence isa term sometimes used to describe second- or third-order sequences made up of higher order sequences (Mitchum and Van Wagoner 1991; and see 2.4.10). Fourth-order (0.1=0.5 million years) ‘parasequence’ cycles represent individual shallowing upward facies cycles 18 Sea surtaoe Fig. 2.7 This series of diagrams illustrates how eustatie se fall and subsidence/uplift destroy accommodation, The rate at which sediment fills the space created controls water depth and whether progradation ox retrogradation of facies belts is observed, In (a) relative level rises and accommodation increases from time 1 to time 2 owing to subsidence, but the rate of sediment accumulation at this point is gi than the rate of relative sea-level rise and so water depth decreases from ‘ime 1 to time 2. In the depositional record, the interaction of these parameters would result in an overall regressive character to the vertical facies succession. (b) Relative sea-level rises and accommodation increases from time 1 to time 2 owing to subsidence, but the rate of sediment accumulation is less than the rate of relative sea-level rise and so water depths increase from time 1 co time 2. In the depositional record this may be apparent as facies succession. ‘The same patterns would occur for a relative sea-level rise due to a eustatic rise of the same rate. {c} (opposite) Sea sutoce nsgressive vertical bounded by surfaces of abrupt deepening, These may be related in part to autocyclic processes within the sedimen- tary system, ‘The theory of eustatic control on deposition is a unifying stratigraphic concept that has attracted geologists for many ‘generations (see Chapter I and papers in Dott, 1992a,b). If itwere true that a global eustatic signature was overprinted oon all stratigraphic suecessions, then it would be possible to date a stratigraphic section from the pattern of sequences and systems tracts, and to predict stratigraphy in unsampled areas from a knowledge of the global standard. A proposed global sea-level chart was first published by Vail et al (1977a), and updated by Hag et al. (1987), based on ‘measurements from basins around the globe. This chart is taken to support the theory that third-order relative sea level variations are mostly custatic in origin. Seepties would argue that the chart is based on thar theory rather than proof of it. More discussion and controversy have been caused by© Sea surace (base iver EUSTATIC FALL “+ Sediment has filed (accommodation) to sea surace (Lo. base lve) in #4 + Folate ceasevel ale rom #1 fo 2 a8 reeut of eustaic tat “+ Base ovals by arelatve soatovat + Segment eroded 0 new base lvel Fig, 2.7(6)Ilastrates time when accommodation space is destroyed. This can occur by two mechanisms — tectonic uplift (and more locally salt anu diapirism) and/or custatic sea- lee fall. The effect of relative sea-level falls manifest at the basin margin (where accommodation space is limited) by erosion, and atthe basi centre where accommodation space is not imited} by ar increase in sediment supply. The increase in sediment : supply to the more basinal areas is due 10 both the erosion of previously deposited sediment and bypass of areas + where accommodation space i now filled the concept of a global eustatie signal than any other aspect of sequence stratigraphy. It is beyond the scope of this, work t0 cover this discussion in detail and the following is, a brief summary only. There are a number of aspects of the Hag et al. (1987) chart that have excited comment: 1 The data to support the Haq chart have never been published fully, in particular the evidence for the global correlatability of the sequence boundaries. Miall (1986, 1992) has been a consistent critic of the Haq et al. curve, stating: ‘The basic premise of the Exxon eyele chart, that there exists a globally correlatable suite of third order euistatic cycles, remains unproven... There are some specific cases where global synchroneity is suggested by derailed stratigraphic documentation (e.g. 4th- and, Sth-order glacioeustatic cycles in the Neogene and, possibly the late Palaeozoic; Ist- and 2nd-order cycles, upurt ‘Sediment eroded senteve, ‘Sediment has filed accommodation to sea surface (base lve) in (1 + elaive sea-level folsom 1110 #245 result o pit ase ovl(eustatcseadeve) doesnot change + Sediment yplited above bas eves ou of equilbum and is eroded related to changing rates of global sea floor spreading), but for the greater part of the Phanerozoic column no such proof is available. (Miall, 1991) Miall also points out that it is arguable whether global biostratigraphic control is accurate enough to correlate third-order relative sca-levels unambiguously. Thus, for the moment, the global synchroneity of eustatic cycles has t0, remain to a degree an article of scientific faith rather than, scientific fact. 2 The mechanism for generating third-order-scale eusiatic, sea-level changes is problematic in certain periods of geo- logical time. Increases in the global ice volume dur glaciations provide a mechanism for eustatie falls in sea- level in the late Cenozoic and late Palaeozoic, but no such mechanism exists for the (presumed) ice-free Cretaceous and Jurassic, Cloetingh et al. (1985) have proposed intra- plate stress as a tectonic mechanism for generating third- order plate-wide relative sea-level cycles. Finally, itis not 19SLOW SUBSIDENCE ° Yourgest 0 Time Tine Fig. 2.8. Accommodation space through time (from Jervey, 1988) accepted universally that the eustatic signal will be visible in all basins, and several stratigraphers believe that it may be obscured by a tectonic signal (eg. Hubbard, 1988). However, work is continuing to date basin-margin tunconformities more accurately and to correlate these with the Neogene oxygen isotope record and hence directly to ice volume changes (eg. Miller et al., 1991, 1993). Also, various projects are underway to accurately date and corre late sequence boundaries on a regional scale in Europe, €.8. De Graciansky et al. (1993) 2.3 Sediment supply ‘The rate of sediment supply controls both how much and where accommodation is filed. The balance between sedi ment supply and relative sea-level rise controls whether facies belts prograde basinward or retrograde landward, and the calibre of sediment supplied to the basin has a strong influence on sedimentary facies. The first part ofthis section considers the principles controlling siliciclastic sedi- ‘ment supply to the basin margin and how sediment supply may vary through time. The second part considers how accommodation is filled in locations with high, moderate and low rates of sediment supply. The principles of carbon- ate sediment production and supply are discussed in Chapter 10. 20 [MODERATE SUBSIDENCE RAPID SUBSIDENCE Youngest 0. Youngest 2.3.1 Principles of clastic sediment supply River transport is. the principle means of transporting ‘material from the continental interior to the depositional basin. The volume and grade of sediment delivered to the basin margin is a complex function of hinterland physi geography, tectonics and climate, Studies of modern rivers show huge variations in the rate of sediment supplied to the continental margins (Fig. 2.10). Around 70% of the total load is supplied from only 10% of the land area, and just three rivers, the Ganges, the Brahmaputra and the Huang He (Yellow) supply 20% of the total lavial load (Summerfield, 1991). ‘The amount of sediment supplied to the basin margin is a function of both the fluvial drainage basin area and the mechanical denudation (erosion) rate. Tectonism at both local and regional scale affects fluvial drainage basin shape, size and relief and also the geology of the provenance area, and will control the calibre of sediment eroded. The rate of fluvial denudation is a complex function of relief within the drainage area and climate. Climate influences not only the erosive power of the river by controlling discharge but also the erodibility of soils in the drainage basin, and the presence or absence of stabilizing vegetation. Present-day mechanical denudation rates vary from less than 1 mm per 1000 years in the St Lawrence river drainage basin 10 640 mm per 1000 years in the Brahmaputra drainage basin.A tributary of the Huang He river in China is an extreme example, with a denudation rate of 19800mm per 1000 years, because it drains over 3000 km? of loess-covered ter- rain in a semi-arid region of sparse vegetation (Summerfield, 1991). It clearly would be unwise to think of sediment supply to the basin margin as being either spatially or temporally constant. Local sediment supply will depend on the prox- imity to a fluvial entry point to the basin margin, There also may be a linkage between glacio-eustatically controlled relative sea-level cycles and climate in the fluvial drainage basin (Blum, 1990). This will mean that sediment supply may vary through a sea-level cycle ina fashion characteristic of the drainage basin. 2.3.2 Filling of accommodation The amount of sediment supplied to locations in the basin ‘sa function of both the general rate of sediment supply to the basin and the proximity to sediment entry points to the basin. Figure 2.11 (from Jervey, 1988) considers the relationship of facies, relative sea-level and rates of sediment accumulation at three fixed points in the basin with identical relative sea-level curves but with differing rates of sediment, supply. These could represent points along a continental Pi emorreax MODERATE SEDIMENT INFLUX 1 Sesiment 1 bypassed margin at varying distances from a point source, Each model begins at time 0 with zero water depth, i.e. the shoreline is exactly located at the model location. For the purpose of illustration, Jervey distinguishes marine ‘mud- prone’ from coastal plain ‘sand-prone’ facies in Fig, 2.11. Sediment grain-size is clearly a function both of sediment supply “type” as well as sediment supply ‘rate’. The sediment supply rate is held constant through the relative sea-level cycle in this simple model. At the location with low rates of sediment influx, accom ‘modation always exceeds sediment accumulation, the coast= line migrates landward and transgression ensues, with considerable water depths being developed. Mud-prone marine facies may be expected to accumulate at some distance from a coastline located marginward of the figured depositional site. The rate of accumulation in this case reflects rate of supply of sediment to this point in the basin. With a moderate rate of sediment influx, the sea-floor can aggrade to sea-level (base level). The rate of increase of accommodation initially exceeds the ability of sediment supply to maintain the sediment surface at sea-level and a transgression ensues. During the transgression warcr depth increases at this location and marine shales are deposited, AAs the rate of relative sea-level rise diminishes, regression of the shoreline commences. Regression of the shoreline [RAPID SEDIMENT INFLUK ° Youngest Time Fig. 2.11 Water depth and facies relationships with varying sediment supply rates (from Jervey, 1988) 23‘continues until marine facies have aggraded to sea-level and the coastline again reaches the location shown. Thereafter, sediment supply rate exceeds the rate of creation of accom- ‘modation, the sediment surface is maintained at sea-level and coastal-plain facies accumulate. Excess sediment which cannot be accommodated in the coastal plain is transported, basinwards. When accommodation decreases (celative sea- level falls) there is the potential for crosion of sediment deposited previously ‘Where sediment influx is rapid, sediment supply rate always exceeds the rate of creation of accommodation and coastal/delta-plain sediments will accumulate. Regression of the shoreline will be continuous through the sea-level cycle. The rate of accumulation at this point in the basin is, limited by the rate of accommodation increase. Erosion is likely when accommodation is removed during relative sea- level fall, 23.3 Basin architecture In order to understand the behaviour of a topset/clinoform margin through time itis neeessary to consider the balance between the rate of sediment supply and the rate of creation, of topset accommodation volume (sometimes termed ‘shel- fal accommodation volume’). The rate of change of accommodation volume is a function of the magnitude of the sea-level rise multiplied by the topset area (Milton and, Bertram, in press). If, during the same interval, the basin margin is supplied with a greater volume of sediment, then, topset accommodation volume will be completely filled, and sediment will deposit on the clinoforms allowing the offlap break to prograde basinwards (Fig. 2.12). Progradational geometries therefore occur when sedi- ment supply exceeds the rate of creation of topset accom modation volume and facies belts migrate basinward. ‘On seismic data progeadation is expressed as clinoforms that show the basinward migration of the offlap break Regression is a term that will be used here to refer specifi- cally to basinward movement of the shoreline. Ageradational geometries occur when sediment supply and rate of creation of topset accommodation volume are roughly balanced. Facies belts stack vertically and the offlap break does not migrate landward or basinward. Retrogradational geometries occur when sediment supply is less than the rate of creation of topset accommodation volume, Facies belts migrate landward and the former depositional offlap break becomes a relict feature. Trans- gression is used here to refer specifically to the landward movement of the shoreline. These phases of progradation, aggradation and retro- gradation are not continuous but are made up of smaller (subseismic) scale progeadational units called parasequences (sce 2.5). Parasequences stack together in. parasequence sets to make up the depositional geometries observable on seismic data. The next section will show that the principles of cyclic 24 changes in accommodation through time can be used to divide the sedimentary record into packages deposited during characteristic phases of the sea-level cycle. 2.4 Sequences and systems tracts | 24.1. Sequences and sequence boundaries The term “sequence”, as applied in sequence stratigraphy, was defined originally by Mitchum et al. (1977) as: A stratigraphic unit composed of a relatively conformable succession of genetically related strata bounded at its top and base by unconformities or their correlative conformities. This generalized definition docs not specify the scale or duration of the sequence, nor does it imply any particular ‘mechanism for causing the unconformities. The term ‘un- conformity’ in this definition was an initial cause of con- fusion, because the precise usage of the term can vary. Mitchum et al. (1977a) initially included marine hiatuses and condensed intervals in the term ‘unconformity’, but as models of eyclic deposition driven by relative sea-level variations developed, it became clear that basin-margin subaerial unconformities needed to be distinguished from basin-centre marine hiatuses. For the purpose of defining sequences, the term “unconférmity’ is now restricted to @ much narrower definition, namely ‘a surface separating younger from older strata along which there is evidence of subaerial erosion and truncation (and in some areas corre- lative submarine erosion) and subaerial exposure and along which a significant hiatus is indicated” (Van Wagoner etal, 1988). ‘Thus sequences are units bounded by significant subacrial ‘erosion surfaces. Units bounded by marine condensed sur- faces, surfaces of transgression, or marine onlap surfaces, are not sequences by this definition. Iris interesting to note that the Exxon workers ‘scriously considered using the term “synthem” instead of “sequence” (Mitchum ef al. 1977a). In retrospect this might have avoided a lot of confusion with the sedimentological use of the term “sequence’, and with other eycle-defined ‘sequences’ (such as the genetic depositional sequences of Galloway, 1989), However ‘synthem stratigraphy’ does not have the same ring to itt At frst sight the definition quoted above seems simple enough. However, in practice, itis not so simple to apply. Ieis difficult to demonstrate non-marine exposure from a well log or from a seismic data set and tracing an erosion surface offshore into its ‘correlative conformity’ is often problematic. The term ‘significant’ in the clarification of ‘Van Wagoner et al.(1988), quoted above, is not particularly helpful because it gives no indication of what magnitude of discontinuity is significant. Composite sequences (discussed in 2.4.10) are allowed to contain unconformities, provided these unconformities are of a higher ‘order’ than the ones which bound the sequence, and are therefore not significant.Fig. 2.12 Depositional architecture as 4 function of accommodation volume and sediment supply (after Galloway, 1989) ‘ice nla soatove ‘nrease in opset accommodation volume AV Topseis ——_Clncorme ‘The increment o ops aecommodaticn volume Via caused by aise Invlaive searfvel aR ls equal othe product of AR and he topset area = ‘TRANSGRESSIVE oH + seonene iruux —+ Low NONE =~ SUBSDENCE + FAST ‘Sano «= SEALEVEL—~ RISE 3 2 PROGRADATIONAL sojdiounsg pur sadaouen | z 42sdeq9)‘The stratigrapher must define ‘signficane” on the scale of the particular study. Nevertheless the restricted principle is straightforward, and a sequence represents one cycle of deposition bounded by non-marine erosion, deposited during one ‘significant™ (on the scale of the study) cycle of fall and rise of base level. In the majority of basins, base level is controlled by sea- level, and a sequence is the product of a cycle of fall and tise of relative sea-level. An idealized sequence resulting from one cycle of base level change is shown in Fig. 2.13, after figures in Van Wagoner etal. (1988). This is a type 1 sequence, where the fall in relative sea-level is sufficiently, large that the first topsets within the sequence onlap the clinoforms of the previous sequence, implying a fall in relative sea-level at the position of the offlap break. Type 2 sequences are described below in section 2.4.6. According to Van Wagoner et al. (1988), a type 1 sequence boundary is characterized by subaerial exposure and concurrent subaerial erosion associated with stream, rejuvenation, a basinward shift in facies, a downward shift in coastal onlap, and onlap of overlying strata. Coastal onlap is a term used to describe the onlap point on topset strata at the basin margin (see Chapter 3). Asa result of the basinward shift in facies, non-marine or marginal marine rocks, such as braided-stream or estuarine sandstones, may. directly overlie shallow-marine rocks, such as lower shore face sandstones or shelf mudstones, across a. sequence boundary with no intervening rocks deposited in inter mediate depositional environments, This facies super- position is termed a facies dislocation. A type 1 sequence boundary is interpreted by Van Wagoner e¢ al. (1988) to form when the rate of eustatic fall exceeds the rate of basin subsidence at the offlap break, producing a fall in relative sea-level at that position. 24.2 Systems tract definition The idealized type 1 sequence shown on Fig. 2.13 is rep- resentative of a shelf-break margin, It can be seen to be comprised of a number of distinct depositional packages. It Highland systoms tact Tranagressive systems tact was observed in the early days of seismic stratigraphy that deposition in a basin was not uniform and continuous but ‘occurred in a series of discrete ‘packets” bounded by seismic reflection terminations (see Chapter 3). Workers in Exxon found that these packages generally were arranged in a predictable fashion in the majority of sequences they observed on seismic data, These packages are known as systems tracts. ‘This term systems tract was first defined by Brown and Fisher (1977) as a linkage of contemporaneous depositional systems, where a depositional system is a three-dimensional assemblage of lithofacies, genetically linked by active (modern) or inferred (ancient) processes and environments (after Fisher and MeGowen, 1967). A systems tract is therefore a three-dimensional unit of deposition, and the boundaries of a systems tract are depositional boundaries of onlap, downlap, ete. The seismic expression of a systems tract is a unit of conformable reflections bounded by surfaces of reflection termination (Cseismic-stratigraphic units’ of Brown and Fisher, 1977; “seismic sequences’ of Mitchum et al., 1977a; referred to as “seismic packages’ in Chapter 5). Systems tracts are recognized and defined by the nature of their boundaries and by their internal geometry. Within any one relative sea-level cycle, three main systems tracts that characterize different parts of the relative sea-level cycle are frequently developed (Fig. 2.13). It is easy to become fost in the complexities of systems tract terminology, and it is always worth remembering the purpose of stratigraphic division into systems tracts. The systems tract represents the fundamental mapping unit for stratigraphic prediction, because it contains a set of depo- sitional systems with consistent palacogeography and depositional polarity, and for which a single palacogeo- graphic map can be drawn. 243. Lowstand systems tract ‘The basal (stratigraphically oldest) systems tract in a type 1 depositional sequence is called the lowstand systems tract. Maximum taeding Fig. 2.13 Stratal geometries ina type 1 sequence on a shelf-break margin, Five separate sedimentary packages are shown, traditionally assigned to three systems tracts; lowstand, transgressive and highstandThe lowstand systems tract is deposited during an interval of relative sea-level fall at the offlap break, and subsequent slow relative sea-level rise. Falling relative sea-level at the offlap break of a shell break margin will have an extreme effect on the river systems. Prior to the fall in relative sea-level, the rivers will have more-or-less maintained a graded river profile with an pDDd>D | >>> L ele oe lee ae sca EE ll ae Variable sensy Vaviole areargale Varabie erstyvigge Fig. 3.6 Examples of various types of seismic trace display. a, amplitude Gain is set at the plotting stage of the seismic line and it is a very important parameter for emphasizing amplitude variations. It affects the ‘darkness’ of the section and so ow values of gain may be used to help bring out amplitude anomalies from the background signal, Trace equalization The pronounced loss of amplitude with depth on seismic data can be overcome by application of a technique called equalization, the aim of which is to try to produce a more balanced and interpretable section. The process adjusts the amplitude of the section to make the average amplitude constant over a predetermined interval or window on the trace. This window may be the whole section, as in single gate equalization, or multiple, as in the procedure termed, automatic gain control (AGC). Fast AGC is a special application of equalization in which the gate is set at a particularly short interval, perhaps ‘one to ten times the sample rate (4—40 ms). It is applied to ‘equalize all amplitudes as a way of enhancing reflector continuity and terminations. It should be used with some caution because fast AGC distorts the wavelet and empha- sizes noise. Nevertheless the technique does have value, especially when applied to bring out detail adjacent to high amplitudes. Complex attributes Conventional seismic information is displayed as frequency and amplitude plots. However, these data, which have been recorded at the geophone or hydrophone, can be manipulated mathematically to produce an additional series, of attributes. Although no new information is produced in this process, the display and interpretation of these complex. seismic trace attributes can sometimes give new insights 50. into the geology that would not have been obtained from conventional data, Theit potential use in seismic strati- graphic interpretation is shown in Table 3.1, and described in more detail by Tanner and Shervif (1977). Table 3.1 Seismic stratigraphic application of seismic attributes (after Sonneland et al., 1988). ‘Aerbute ‘pplication ed thickness Lthological contrasts Fuld content Instantaneous frequency Instantaneous phase Bedding continuity Cosine of instantaneous phase Bedding continuity etifying sequence boundaries Reflection strength Litholegieal contrast Bedding continuity Bed spacing Visual enhancement techniques ‘This term covers some simple, cheap but often very effective redisplay options, which can be readily applied even when tapes are not available. Colour. There is no doubt that the use of colour improves the interpretability of sections. Historically this was an expensive option for paper, but with the advent of inexpen- sive plotters this is no longer true. Moreover, workstations are eminently suited to colour display and allow easy experimentation to find the best scales to suit the users preferences, Colour is probably best used as a background intensity display on which wiggle traces are superimposed,The increase in interpretability associated with colour is kar in the presence of noise. It is better not to use too ‘many colours and always to be aware that a significant proportion of the male population is colour blind. Squash plots. Seismic sections were displayed originally for sirutural interpretation and the aim was to have as near a Ame scale, with no vertical distortion, as the time display would allow. Although this may be an ideal scale for trap Afinition itis often less than optimum for seismic stratigea- Riwhere sometimes the interpreter is interested in ex. tremely subtle geometric relationships. Sedimentary dips ae generally low. Submarine fans may exhibit dips of 1-3’ on the levees and depositional lobes. Clastic shallow- imine prograding complexes rarcly have dips greater than ‘few degrees on the slope and the coastal plain is very lose to horizontal. In order to see marine onlap, downlap and coastal onlap relationships in these environments, pecially when there is little syn-sedimentary structuring, itis necessary to enhance the apparent dips. Indeed this is what the interpreter has been doing optically by squinting abiquely long the section. The redisplay can be done very simply and cheaply by foreshortening the horizontal scale, wiile keeping the vertical scale fixed, commonly 5 or ems". The advantage of keeping the vertical scale fixed isthat ron-squeezed lines still tic. However, with modern dda the close trace spacing means that the reduction in horizontal scale must be accompanied by dropping traces Ortrice somming. An alternative approach to resolve this problem isto reduce the horizontal scale until the minimum, horizontal trace spacing is reached. The horizontal scale «an be kept fixed and further vertical exaggeration is achieved by increasing the vertical scale. In the past the ierpreter could request sections to be displayed at com pressed horizontal scales (eg, 1: 100.000 or 1 : 200000) or thedita could be optically compressed in a squeeze camera inwhich the film and the section pass the lens at different speeds. Nowadays, workstations permit easy scale changes and all interpreters should be encouraged to experiment to find the vertical exaggeration that best brings out the statigraphic relationships in their data set. 3.2 Scismic reflection termination patterns 3.2.1 Marking up a seismic section The firs step in the stratigraphic interpretation of aseismic line isto determine the vertical and horizontal scale of the section. An appreciation of both is essential to constrain the geological models that will be constructed later. It is ao worthwhile to find out from the header or the seismic data itself ifthe section has been migrated, and whether it ismarine or land data. Both are rich in multiples, even after optimum processing, but it is generally easier to identify mmukiples on marine data, where the acoustic contrast at thesea-floor generates both simple water-bottom multiples and peg-leg mutiples. Ifa reflection’s origin as a multiple is suspected then it can be marked in some distinguishing colour, usually light blue by convention, In Fig. 3.7, from the Tertiary succession of the Outer Moray Firth, North, Sea, a reflection is seen at the right-hand (eastern end) side of the data, intersecting the edge of the section at 0.3s. This. reflection is probably a water-bottom multiple, caused by the sound waves bouncing twice between the sea-surface and sea-bed, and being recorded at a two-way-time (TWT) twice that of the sea-bed. This reflection does not have any geological significance, and can be ignored in the rest of the interpretation. ‘The next step is to divide the seismic data into the discrete natural stratigraphic packages that make up the section, To do this, first identify and mark reflection termin- ations. It will be obvious that all reflections do not go on for ever; most stop, often against another reflection. Mark these terminations with an arrowhead (by convention use a red pencil). This is not always a straightforward procedures sometimes two reflections can appear to merge, and it may. beunclear which ofthe reflections terminates against which. In noisy, chaotic, multiple-prone or low-amplitude data it may be unclear whether reflections terminate, fade out, and/or reappear somewhere else. Generally it is a good idea to start by ignoring zones of broken or chaotic reflec- tions and to concentrate on the better data areas. The chaotic zones and poor data zones can be interpreted later with the aid of the model derived from the good data. On, Fig. 3.7, the terminations ofthe continuous, high-amplitude flat reflections above 0.7 are hard to see. These reflections, look quite continuous, except where they are truncated by channels. However, on closer inspection, often helped by ooking along the seismic line at a low angle, subtle termin- ations can be recognized within this interval. In contrast, the zone between 0.7 and 1 contains numerous dipping reflections, which consistently terminate up-dip to the left and down-dip to the right. ‘Where reflections terminate in a consistent manner they define a line on the section (and a surface in three dimen- sions). This is known as a seismic surface. The next step is to use the red arrowheads to identify and highlight the seismic surfaces on the line, The number of surfaces will vary depending on the stratigraphic complexity, and there generally will be several major seismic surfaces of consistent reflection termination, and other more minor ones. These surfaces are by convention coloured yellow in the initial stages of a seismic stratigraphic exercise prior to them being assigned an identification colour based on their type oor age. In Fig. 3.7 these surfaces of consistent reflection termination have been marked. The most prominent are ‘one or two surfaces of high relief in the zone 0.2-0.5s. However, a good seismic surface is also seen at around 0.5's. Others are present in the interval of flat-lying reflec- tions, and a complex set of surfaces are present within the interval of dipping reflections from 0.7 t0 1s. A strong reflection at around 0.75 is also classified as a seismic slWesT 1400 1000 1200 1200 1400 ~~ Seismic surface —<— Reflection termination surface, owing to the onlap of overlying reflections around, shotpoint (SP) 9000. ‘Once all the seismic surfaces have been picked it is necessary to perform a similar exercise on the other lines in the seismic data set, and to tie the interpretations (ie. ensure that the interpretation is consistent where lines ross) to generate a three-dimensional grid of surfaces. Some of the seismic surfaces so defined will be important regional surfaces, others will be more local and may be less significant. By picking the surfaces of reflection termination, the interpreter has divided the stratigraphy on the seismic data $2 ML spuoseg “LMA spuoses, Fig. 3.7 Seismic daca from the Outer Moray Firth, central North Sea, showing the seismic stratigeaphy of the post-Palaeocene section; (top) uninterpreted, (bottom) interpreted. The surface at around 0.75 is dated as close to the top Eocene, the surface at 0.5 s may mark the top of the Miocene, and the high-relict surfaces in the shallower section are interpreted as glacial lowstand surfaces in the Plio-Pleistocene set into a number of depositional packages containing relatively conformable reflections of a similar or gently changing character and geometry, which are bounded by susfaces that mark the reorganization of reflection geomety 3.2.2 Categorizing reflection terminations Reflection terminations are characterized on a two: dimensional seismic section by the geometric relationship between the reflection and the seismic surface against which it terminates. Mitchum e¢ al. (1977a) introduced the terms “lapout, truncation, baselap, toplap, onlap and downlap' toerellecion termination styles (Fig. 3.8). These terms, described below. Most are based purely on geometry, i others involve some interpretation of whether the fon termination is an original depositional limit. The are applied to describe the reflection’s: current spot isthe lateral termination of a reflector (generally ding plane) at its depositional limit, whereas truncation pls that the reflector originally extended further but either been eroded (erosional truncation) or truncated fault plane, a slump surface, a contact with mobile salt le,oran igneous intrusion (Mitchum et al., 1977a,b). lap isthe lapout of reflections against an underlying smi surface (which marks the base of the seismic pack- Basclap can consist of dowurlap, where the dip of the less than the dip of the overlying strata, or onlap, the dip of the surface is greater (Fig, 3.8). lap commonly is seen at the base of prograding forms, and usually represents the progradation of a snsmargin slope system into deep water (either the sea lake). Downlap therefore represents a change from (or lacustrine) slope deposition to marine (or lacus: jeondensation or non-deposition. The surface of down- Tepresents a marine condensed unit. It is extremely Ito generate downlap in a subaerial environment, however, that it may be easy to confuse true de- tional downlap and original onlap rotated by later m. he rection terminations interpreted as downlap may yy eases be apparent terminations, where the strata distally below seismic resolution ap is recognized on seismic data by the termination wrangle reflections against a steeper seismic surface. o types of onlap are recognized; marine and coastal. ine onlap is onlap of marine strata, representing a nee from marine deposition to marine non-deposition ondensation, and results from the partial infill of space maine sediments. Patterns of marine onlap cannot be alto determine changes in relative sea-level, because the ‘of marine onlap is not related directly to relative I. “Trneation 38 Types of reflection ‘Marine onlap reflects a submarine facies change from significant rates of deposition to much lower energy pelagic drape. Wells drilled beyond the marine onlap limit of a rock unit will encounter a time-equivalent condensed unit co hiatus (time gap). The seismic surface of marine onlap represents a marine hiatus or condensed interval Coastal onlap is onlap of non-marine, paralic, or marginal marine strata and represents a change from a zone of, deposition to basin-margin (subaerial or shelf) erosion and non-deposition. Coastal onlap generally is inferred from seismic data as the landward onlap of topset reflections, {section 2.4), where these are assumed or are demonstrated, to represent littoral, paralic or non-marine deposits. These topset deposits are assumed to have accumulated close to sea-level, and patterns of coastal onlap with respect to the onlapped surface indicate changes in relative sea-level. A. landward progression in coastal onlap results from rising relative sea-level, whereas a downward or basinward shift in coastal onlap results from a fall in relative sea-level, (described in more detail in Chapter 2) Coastal onlap does not necessarily occur at the coastline, and progressive landward coastal onlap may accompany. either regression or transgression depending on sediment supply. Wells drilled landward of the coastal onlap limit of arock unit do not encounter a time-equivalent succession, but may encounter an unconformity, a palacosol, or a karst horizon. ‘Toplap is the termination of inclined reflections (clino- forms) against an overlying lower angle surface, where this, is believed to represent the proximal depositional limit. In marginal marine strata, it represents a change from slope deposition to non-marine or shallow marine bypass or erosion, and the toplap surface is an unconformity. An apparent (oplap surface can occur, where the clinoforms pass upwards into topsets that are too thin to resolve seismically. In a deep marine setting, an apparent toplap surface is much more likely to be a marine erosion surface, such as seen in contourites. In this case the surface is localized and rarely flat over a large area. Erosional truncation is the termination of strata against an overlying erosional surface. Toplap may develop into UPPER BOUNDARY (eran) LOWER BOUNDARY 53erosional truncation, but cruncation is more extreme than toplap, and implies either the development of erosional relief or the development of an angular unconformity. The erosion surface may be marine, such as at the base of a canyon, channel or major scour surface, or a non-marine erosion surface developed at a sequence boundary. Apparent truncation is the termination of relatively low- angle seismic reflections beneath a dipping scism‘e surface, where that surface represents marine condensation. The terminations represent a distal depositional limit (or thinning, below scismie resolution), generally within topset strata, but sometimes also within submarine fans. Many reflec~ tion terminations in marine strata fall under the heading, ‘apparent’, because it is likely that thin condensed units, extend beyond the limit of seismic resolution (Fig. 3.9). Fault truncation represents the termination of reflections, against a syn- or post-depositional fault, slump, glide or intrusion plane. Termination against a relict fault scarp is, onlap. Frequently it appears that reflections overlying a se surface terminate against it, whereas those below appear to be conformable (or vice versa). This conformity is often apparent only, and is due to a very low angle between the surface and the apparently conformable reflections or to condensation along the surface. Several types of reflection termination can be seen on Fig. 3.7. There is obvious truncation beneath a high-relief channelled surface between 0.3 and 0.5 s. Short reflections within the channels onlap the channel margins. The dipping reflections in the Eocene package between 0.7 and 1s. downlap to the right and either toplap or are truncated to the left. The overlying seismic surface is relatively flat between shotpoints 950 and 1100, and is a toplap surface. A unit of reflections below 0.8 s and east of shotpoint 1200 can be seen to onlap to the west and downlap to the east. mic 3.2.3 Seismic facies and attribute analysis ‘Once the seismic data has been divided into its component depositional packages, using the procedure described above, further geological interpretation’ of the stratigraphy may ‘Shale-out be attempted. This usually is done through seismic facies mapping, deseribed by Sangree and Widmier (1977) as the interpretation of depositional facies from seismic reflection data, It involves the delineation and interpretation of reflee tion geometry, continuity, amplitude, frequency, and interval velocity, as well as the external form and three-dimensional associations of groups of reflections. Each of these seismic information of stratigraphic reflection parameters contains significance. ‘One of the easiest attributes 10 map and define is the ‘geometry of the reflections. Prograding basin-margin units are commonly seen on seismic data to consist of topsets and clinoforms (Fig. 3.10; and as described in Chapter 2, section 2.1.2). Examples are shown on Fig. 3.7, with the late Eovene package east of SP 1200 and below 0.7s displaying well-developed topsets and clinoforms. The dipping, reflections to the west of this package are clino- forms, with minor or absent topsers. A well-developed topset~clinoform package can be interpreted as represent: ing a systems tract of paralic to shelfal units (the topsets) and slope sediments (the clinoforms). The break in slope between the two is the offlap break (Fig. 2.3) ‘Other sedimentary units, such as submarine fan lobes, may sometimes have a similar shape, with relatively flat reflections becoming steeper in a basinward direction. The key to recognizing a true topset—clinoform package is t0 recognize a clear offlap break, and for the topset reflections to be concordant and parallel. Both these criteria are met by the late Eocene package in Fig. 3.7. Occasionally, lino form reflections can be seen to decrease in slope downward, and pass basinward into flatter lying reflections known as bottomsets. In other cases the more distal low-angle reflec: tions are not in depositional continuity with the clinoforms, but form separate depositional packages onlapping the clinoform front. Ramasayer (1979) presented a rigorous methodology for two-dimensional seismic facies mapping, known as the “A,B,C technique’. Three characteristics of each seismic package are recorded, given code letters (Table 3.2) and mapped. These are the nature of the reflection terminations against the upper boundary, the nature of the reflection pean St | GEOLOGY (distal shale-out of parasequenses) ‘Apparent uncation ‘SEISMIC RESPONSE (termination of reflectons below a cpping suriace) Fig. 3.9. Apparent truncation; the termination of reflections against an overlying surface, which looks like ‘erosional truncation, but represents the original depositional limit of the strata‘Simos Oblique tangential Sigma: bla Obique pale shingled Fig. 3.10 Types of clinoform profile falter Mitchum et al., 19774] Table 32 Scismic facies classification (from Ramasayer, 1973) Hummecsy So penetrations through the mapped interval. Without well control, a seismic facies map generally remains open to several geological interpretations, ad gst 2-81C Using modern geophysical workstation technology a range A aon of parameters can be quantified and mapped for any seismic ee, peta package. The amplitude of the seismic reflections at the top a pom or base of the package can be mapped (Enachescu, 1993), Goren G enaote ‘Workstations can ‘snap a pick tothe seismic data, ensuring Wy, wavy that the horizon interpreted always lies on a seismic trough a” es at mrusdal or peak, and can map the amplitude of that horizon eu, OnSotguepogrdatonat through the data set. An amplitude map can be read De crien Sorsigmoid progedatonat directly as a geological facies map where the amplitude of RE, reflection free Sh shingled Jeminations against the lower boundary, and the internal ‘configuration of the reflections. Thus the late Eocene pack agein Fig. 3.7 mentioned above would be characterized in {tsproximal portion as: C-On/P; and in its distal portion 5: CDwn/Ob. These codes can be marked on a map, and distributions athe various seismic facies can be constructed using the Ante seismic grid, With calibration from well data, it is tien possible to make a reasonable geological facies map from seismic facies. The late Eocene seismic package de- serbed above has not been drilled, but the seismic strati- agaphic and seismic facies analysis presented here predicts that it contains a basin-margin—slope assemblage. How- tver, the nature of the facies within the topsets is still tancertain, and could be either alluvial plain, coastal plain, panic or shelf facies. An example of 2 seismic facies map isshown as Fig. 3.11, reproduced from Mitchum and Vail 97). There is no unequivocal link between seismic facies and depositional systems, with the probable exception of the link berween clinoforms and slope systems. Sangree and Widmier (1977) lise seismic facies with their geological interpretation, but without well control this link is tenuous. Continuous flat-lying reflections may, for example, reflect eep-marine shales, coastal-plain topsets, alluvial plain, or hhastrne facies. However, a seismic facies map may be ised to construct one or more tentative geological models, ‘which should then be tested against and calibrated by well the reflection is related to the geology. For example, under- lying sands may result in a low-amplitude reflection, whereas underlying shales result in achigh amplitude. The amplitude map would thus show the sand—shale distribution. How- ever, ina seismic stratigraphic interpretation, where the top of a seismic package is not a reflection but a surface of reflection terminations, it may not be possible to ‘snap’ the pick to any one reflection, and amplitude maps of seismic surfaces may not be possible. The average amplitude of the whole package is often a useful attribute. This generally is measured as a root-mean- square of the amplitude within the package (RMS ampli- tude), or as a mean square amplitude (‘average energy’) This attribute can be quantified, mapped and contoured using a workstation, and used to differentiate zones of Reflection termination Fig, 3.13 A lowstand systems tract on seismic data. On this part of this line, only the lowstand prograding wedge is seen. The underlying sequence boundary is recognized by a downward shift in coastal onlap. Late Eocene, Outer Moray Firth, central North Sea lapping on to a top fan surface (Fig, 3.14). This lowstand, systems tract therefore may be divided into two parts, a Jowstand fan and a younger lowstand wedge. This systems tract is dated as late Palacocene, and forms part of the T45 unit of Jones and Milton (1994). The linoforms would represent the Domoch Formation, whereas the basinal shales would be the Sele Formation. Larger Sele Formation fans form attractive reservoir targets in the North Sea, This lowstand represents the time of maximum, uplift of the Scottish Mainland and the adjoining North Sea during the Palaeocene uplift episode (Jones and Milton, 1994). ~~ Seismic surface Reflection termination 3.3.2 Recognition of transgressive systems tracts ‘Transgressive systems tracts are bounded below by a tr gressive surface and above by a maximum flooding su and consist of retrograding topset parasequences. Tr ‘gressive systems tracts are often very thin, and may co of no more than one reflection. Figure 3.15 shows a tr gressive systems tract from the early Eocene of the cet North Sea. It is recognized as a transgressive systems t as its base marks the transition from an underlying int« of mainly clinoforms, to an interval of mostly or ent topsets. It also clearly shows internal retrogradati ‘geometries. Two high-amplitude reflections are seen wi the systems tract, the higher (younger) of which is placed landward with respect to the lower. These reflect come from two retrograding coal intervals; one or bot which is equivalent to the Top Dornoch Formation co: Deegan and Scull (1977). This figure shows the diffic of using time-transgressive facies such as coals to subd stratigraphy. ‘The transgressive systems tract is overlain by a maxin flooding surface, recognized by the downlap of overt clinoforms. A further coal overlies these clinoforms ‘Top Beauly Formation coal), marking the next transgres No clear apparent truncation can be seen below the m ‘mum flooding surface, except perhaps at the extreme ti hand end of the systems tract (Fig. 3.15) 3.3.3 Recognition of highstand systems tracts Highstand systems tracts are bounded below by a maxin flooding surface and above by a sequence boundary exhibit progradational geometries. Figure 3.16 illustrat highstand systems tract from the early Eocene of the cer Wsicin aNeaa sat Seconds TWT Fig. 3.14 A lowstand systems tract on seismic data, Here the lowstand wedge can be scen as a clinoform unit, with a toplap significantly below the previous highstand offlap break. A lowstand fan unit is seen as a mounded unit with bi-directional downlap, Earliest Eocene, Outer Moray Firth, central North SeaAML spuos9s, AML spuosag Sqn NEUILLIE} UOnSoyoy s0eyins o1uisiag uoneutui9, uonse4y aoeins o1sieg,North Sea. Ir consists of prograding topscts and clinoforms, representing progradation, and it overlies a maximutn flood- ing surface. Clinoforms within the systems tract downlap topsets of the underlying systems tract. Apparent truncation can be seen beneath this surface, and the underlying systems tract infils erosional relief on an older sequence boundary, ‘The systems tract also underlies a sequence boundary. Flat. lying topsers of the subsequent systems tract onlap the last highstand clinoform (representing a downward shift of about 100m). The nature of the two boundaries and the progradational architecture confirm this as a highstand systems tract. In detail this highstand systems tract could be interpreted to consist of two or perhaps three down- stepping progradational wedges (each one displaced basin- ward and below the previous wedge). These are probably higher order (fourth order) sequences,.and the systems ‘tact is a composite systems tract, as discussed in Chapter 2. 3.4 Pitfalls in interpretation There are many pitfalls and ambiguities inherent in seismic stratigraphic interpretation, The more important ones are outlined below. 1 Seismic data have a relatively coarse resolution, and stratal relationships in thin successions may be impossible to resolve, 60. Fig. 3.17 Not every seismic line encounters every systems tract. Seisme line A—A" here misses the incised valley and lowstand fan 2 Not every systems tract will be present on every line Any one line may, for example, completely miss the low stand fan system. This is illustrated in Fig. 3.17, where @ lowstand fan is shown developed at the mouth of a canyon incised into a highstand slope. A seismic section along line A-A’ would encounter neither the incised valley nor the lowstand fan, 3 A.common mistake is to assume that all seismic surfaces which have been identified by reflection terminations must be sequence boundaries sensu Van Wagoner et al. (1988) 4 The key to successful seismic stratigraphy is the appre ation of the significance of coastal onlap and being able ta recognize it on seismic data. However, itis easy to config marine and coastal onlap. Although coastal onlap is cot fined to topset reflections, it is not always easy to determi Which reflections are truly topsets. Topsets can be dential confidently only where they are parallel and lie landwa of an offlap break. $ Fluvial incision and marine canyons may be easily « fused. Fluvial incision is an indicator of a sequence bound ary; canyon cutting is not necessarily so. 6 In a clinoform succession with extensive bottom sets is easy to misidentify the downlap surface. Many of clinoforms will erminate against older bottomsets, and true downlap surface is at the bortomset terminationsIniroduction and historical perspective Rosoaton of well daca Sequence stratigraphy of outcrops and 41 Parasequences in outcrops and 432 Parasequence stacking patterns an ystems tracts 433 Key stratigraphic surfaces in AA Sequence stratigraphy of wieline logs 4.4.1 Log suites used in sequence steatigraphy 44.2 Log tens 4.43 The log response of elinoforms 44.4 The log response of parasequences AS Log responses from basil 4.46 Estimation of depositional ‘controls and sequence stratigraphy from log response 44.7 Key surfaces AAS Lentifcation of systems tracts from log response 449 Piehlls and ambiguities in sequence analysis of log data 44.10 Cheat for sequence sratigraphic inerpretation of @ welklog data base ‘uterops and cores 41 Introduction and historical perspective sequence stratigraphy of outcrop, core and wireline data was developed in the early 1980s with the testing seismic stratigraphic techniques against rock data. These niques have evolved considerably over the last decade, lyin the last few years, with the publication of two volumes. Firstly, Society of Economic Paleontologists Mineralogists (SEPM) Special Publication 42 (Wilgus al, 1988) contains a series of papers in which Exxon shes laid out a series of key sequence stratigraphic itions and conceptual models. ‘The definitions and Is together provide a framework linking basin-scale ie stratigraphic models and sedimentary facies models. the second significant publication, Van Wagoner et al. 1990) dealt specifically with high-resolution sequence graphy from outcrop, core and wireline log data. ih the publication of these volumes, outcrop, core and ine log sequence stratigraphy has reached a wider ice and much progeess has been made in the last few . However, at the same time a range of views, some- contradictory, has emerged. Miall (1991), for example, questioned many aspects of the models presented in SEPM Special Publication 42, and Walker (1990) has Aocumented many challenging stratigraphic geometries from welleonstrained subsurface data sets in western Canada that do not fit conventional models easily. This chapter is organized into an outcrop and drill-core section, material with which the reader is likely to have some familiarity, followed by a longer section on wireline Jogs and combinations of logs and drill-core. The outcrop and core section is deliberately brief as it is covered in eater depth in Chapters 7—10. 4.2. Resolution of well data Not all the techniques described in this chapter yield the same scale or resolution of stratigraphic information. Figure 3.2 shows a gamma-ray log and simplifed lithological log through the reservoir section of the North Sea Beatrice Cilfeld compared with a seismic waveform. The logs can resolve bedding detail that the seismic cannot, and hence permit a more detailed stratigraphic analysis. However, the gamma-ray log and other electric logging tools described in, more detail in section 4.4 do not have perfect resolution, and cannot ‘see’ beyond the vicinity of the well bore. Furthermore, converting log properties to lithology does not always provide a unique geological answer, and wher- ever possible logs should be calibrated by core data, Core data from wells is good material for sequence stratigraphic analysis, but is rarely acquired in significant quantity Usually the core is cut in porous units over hydrocarbon- bearing intervals, which usually means several tens of ‘metres of core at most. In addition, cores, like wireline logs, see only a one-dimensional view of the geology at the well bore. As well as log and core data, wells also yield biostratigeaphic information, discussed more fully in Chapter 6. 4.3. Sequence stratigraphy of outcrops and cores 4.3.1 Parasequences in outcrops and cores Parasequences are defined in section 2.5 as relatively con- formable successions of beds and bed sets bounded by marine flooding surfaces and their correlative surfaces. At present, parasequences can be recognized with confidence 61‘only in shallow marine and shoreline successions, and are much more difficult to identify in deeper shelf and basinal settings, and in non-marine environments. ‘The precise nature of the parasequence geology depends on the facies associations, and several end-members are described in section 2.5 and by Van Wagoner et a. (1990). The most common is the coarsening-up signature (Fig. 4.1), widely recognized in parasequences from marine set- tings. Here, the shale content decreases upwards, but sand content and bed thickness may increase upwards. The marine flooding surfaces can be recognized by abrupt deepening, for example marine shales lying erosively on sandstones with rootlets. In addition, a number of non- diagnostic features characterize the outcrop and core ex- pression of parasequence boundaries: 1 marine carbonate, phosphate or glauconite may be pre- sent, suggesting low siliciclastic sedimentation rates: 2 lags, which may record the transgression of the shoreface, are common, but are generally thin (< 10cm thick) and contain only sediment reworked from below; 3 where flooding surfaces pass into amalgamated marine sandstones they sometimes can be traced as zones of pref erential marine sedimentation; 4 if outcrop exposure is good enough, or cores are spaced sufficiently closely, parasequence boundaries may be rec: ognized as essentially flat, with only a few centimetres (but rarely up to a metre or two) of sediment eroded at the boundary. Dirtying-up parasequences may rarely also be recognized from tidal estuary facies, and parasequences that shallow- up to a non-marine facies may have a fine-grained unit of paralic shale at the top, and possibly a coal bed. Parasequences should have correlative units in contem: porancous alluvial systems. Cyclicity in alluvial/flavial systems generally results in stacked channel facies, which 62 individually may show fining upward trends in core and outcrop. The link berween these channel cycles and the pparasequences further down the systems tracts is discussed in more detail in Chapter 8 4.3.2 Parasequence stacking patterns and systems tracts ‘The stacking patterns (or ‘architecture’) of parasequenses are discussed in section 2.5.3. The three patterns observed are progradational, in which the facies at the top of each parasequence becomes progressively more proximal, aggra: dational, in which the facies atthe top of each parasequence is essentially the same, and retrogradational, where the facies become progressively more distal. These stacking patterns can be recognized both in outcrop and in core, and the position of the parasequences with respect to a majot stratigraphic surface can help to constrain the systems tract represented by the parasequence architecture. “The example illustrated in Fig. 4.2 is from the Viking Formation, Alberta, Canada, A core cut in well 6-11-48 21-W4 shows two complete parasequences, considered te hhave been deposited in a tidal sand-sheet environmen) (Chapter 8, and Reynolds, 1994a). Ac its base the con penetrates the top of a strongly bioturbated, fine-grainec sandstone with rare ripple-scale cross-lamination. An abrup, decrease in the sandstone content marks the first loodin, surface, which is also bioturbated and contains scatterec medium-grade sand grains. Upward, there is a steady in crease in sandstone content and change from horizontal t¢ vertical burrows. These changes, along with an increase it the preservation of unburrowed cross-beds, are interprete, to indicate shallowing and progradation within a para sequence, The top of this parasequence is marked by decrease in sandstone content and by scattered medi grade sand grains, followed again by burrowed system Fig. 4.1 Coarsening upward parasequence in outcrop, Panther ‘Tongue Formation Cretaceous, Midwest USA. Telegraph pole for scaleWELL 6-11-48-21W4 VIKING FORMATION, ALBERTA teal 42 Parasequences in core, Viking tion, Alberta, Canada. See text Aketails. FS, flooding sequence; i [msi 3] ns own Bs p [ess oS | ves ns smosesret | g 2 8] e ower shat oo TeaNsGRESSVE [seers vs > ves nner sl madesret |B 8 8] ex 8 & | ouersrat les | 8 Ji g e Bade shat 5 8 4 82 outer she ; fs Zé sequence boundary o lac to the first parasequence, but towards the top of second parasequence, dune-scale cross-bedding (senst cy, 1990) predominates. This cross-bedding is intepreted indicate daily current activity, the development of a proximal facies and hence a progradational pars- nce set. Above the second parasequence is a third nding surface, marked by a thick lag deposit, then a atic change in grain size and se Ye change in. sedimentary processes. suggest that this surface may be an interfluve sequence boundary (Chapter 8), which from regional mapping evidence correlates lat- crally with the development of an incised valley. The most, likely interpretation of the progradational parasequence set in this example is as a highstand systems tract, overlain by a sequence boundary. Retrogradational parasequence architectures are well- demonstrated in an outcrop example from the Jurassic Scarborough Formation in Yorkshire (Fig. 4.3, and Gowland, and Riding, 1991). Below the Scarborough Formation, the > owsuns ONIOOOTS WNAIKYA_— spy snaisis aNwsHoik ‘Auvannoa SONaNOSS 907 Avie OUYEHOS LOvULSWAISAS ‘NYISHOH Tanaty3s 3AuyIaH‘At the basin margin the maximum flooding surface of a condensed section is recognized by the sudden influx of low diversity, open marine plankton, sandwiched between shal lower marine benthic or terrestrial fossil assemblages. On ihe shelf the maximum flooding surface is recognized by the presence of more diverse open marine plankton and Possibly deeper water benthic fauna, whereas in a deep basin, sediment starvation may result in the developmene of highly fossilferous deposits. In the case of exereme clastic starvation, pelagic carbonates, composed of the Femains of calcarcous microfossils, may develop. Alter. natively, slow sedimentation may allow time for corrosive bortom-water to dissolve the calcareous fossils, 6.4.6 Highstand systems tracts Aggrading highstand systems tracts occur when the rate of sediment supply is equal to the amount of accommodation space being excated by rising relative sea-level. They are characterized by thick accumulations containing, stacked shelfal or terrestrial fossil assemblages with no overall shallowing-up observed. Progradational highstand system tracts occur when the rate of sediment supply exceeds the amount of accommo. dation space being ereated by rising relative sea-level. Ip essence they superimpose proximal fosil assemblages upon more distal ones (Fig, 613). In a vertical succession they ass upwards, in a complete example, from deep marine, Shallow marine, through marginal marine to non-marine, Possibly with an oscillatory character reflecting para, Sequences and minor flooding surfaces. During early highstand, shelf deltas or the coastal margin tmust first advance across the drowned shelf of the under. lying transgressive systems tract to the margin created by the previous lowstand wedge. Ar this time, shelf width i & maximum and wave energy is at a minimum. In the absence of powerful tidal currents, muddy sediments tend 0 predominate. The fossil assemblages of these broad ‘muddy shelves are dominated by distinctive, fine-sediment- dwelling, ‘turbid water’ benthic assemblages in shelf to slope environments. Tidally dominated shelves with coarse {ae deposits predominantly have an attached epifsanal Penthos and common derived planktonic elements owing to increased transportation across the shelf, Sheltal fossil assemblages are strongly influenced by the presence of shelf deltas and associated rapid sedimentation, increased turbidity of the water and reduced salinity. In this nutrient-rich environment the benthic fossil ascem, pages are often abundant, diverse and dominated by in. faunal species. Planktonics are generally scare, although certain groups, such as dinocysts and acritarchs, have adapted to these conditions, and nannofossils, more easly tzansported in from the open ocean owing to their very simall ‘size, have important biostratigraphic correlation potential. sufficient ime and sediment are available, the highstand 104 progradation may be able to advance to the margin cre by the previous lowstand wedge. These deltas now b shelfedge deltas and capable of supplying sediment, also terrestrial and shelfal organisms, directly to the basin, Topsethighstand deposits may comprise shell, pa and fluvial deposits and associated shallow and non- hari fossil assemblages, with the proportion of each depen upon the nature of the progradation, In extreme obi Progradation the highstand will comprise mostly slopea shelfal deposits, with significant sediment bypass and li evidence of deposition in paralic or fluvial environme Accommodation space created during the highstand by relative rise in sea-level will result in the accumal of deposits with shallow marine and terrestrial assemblages. The prograding highstand slope comprises both gra flow and hemipelagic deposits that are often a slumped and contorted. Consequently they may con mixed indigenous and derived fossil assemblages, Hi sedimentation rates can cause the dilation of indigen fossil assemblages. A prograding highstand slope nay inferred in a vertical section, but not exclusively definedson a gradual shallowing-up signature in the benthic fos through different, shallower biofacies as the hishsta slope advances into the basin, produces apparent extine tions in the fossil record that may cause diachronous correlations (Atmentrout, 1987), Within the deep basin, slow sedimentation basinwans of the highstand toesets produces condensed sections th ‘may contain abundant deep-marine fossil assemblage similar to those in the condensed sections of the tran Bressive systems tract and maximum flooding. surfae (Armentrout and Clement, 1991). More rapid sedinen tation in the deep basin suggests erosion of the slope through mass wasting via slumping, debris and turbidiy flows or possibly sediment bypass. These may introdue both slope and shelfal, or non-marine, fossil components the deeper basin, diluting the indigenous fossil assemblage and inhibiting a true palacoenviconmental determination, Turbidites are commonly barren of indigenous fossil (McNeil et al, 1990), and often contain only reworked fossils and faunas derived from further up the slope, 6.5 Conclusions The biostratigraphie character of sedimentary sequences controlled by the interplay of basin-specific environmental conditions, the evolution of organisms and the oy changes in depositional style related o base-level changes Consequently, there are few concrete laws on the relation, ship between biostratigraphy and sequence stratigraphy, However, there are general observations and trends thee dy apply and are summarized as follows:dividual fossil groups do not provide sufficient age palaeoenvironmental determinations in all depositional ments and throughout Phanerozoic time. The com- d use of different groups can complement each other improve biostratigraphic resolution. Individual fossil isolation may result in incorrect palaeo- pnmental and age determination, with severe impli for geological modelling. he late appearance or premature disappearance of a il from a stratigraphic section may be due to local ronmental constraints, and relate to the biofacies ofthe ilrather than true extinction (Fig. 614(a)). Correlations 4(a) The impact of biofacies, sedimentation dilution and fossil dissolution on planktonic fossil distribution and abundance First downhole occurence (ast appearance datum ‘ttn Plankonie mance abundant Condensed section Dilton of planktonic ‘marke in tapily Geposted fon sands ‘noroased sedimentation ‘ate of progadng HST utes panitonc sbuncance Disoluton of calcareous —— planktonic marerby aide botom water Last downhole occurence {Fret appoaranco datum = evohitonary inepton} aan sean Distal condensed secon ‘ancnsne ee wit numerous bernie mpcnca and planktonic index species See teece || cer toe Stable deop marine Botacios with tow raponte to progradaton st and revogradaton Hats i Derived fos assemblages 5 ‘Yom botacies Aard 8 se sandwiched n between ce oop marine biofacos C vst“ Dissolution ot ‘calcareous fase by ie low oxygen attr watar A4(b) Biostratigraphic resolution within different systems tracts based on biofacies are generally diachronous and reflect progradation or retrogradation, 3 Fossil resolution is hampered by rapid sedimentation and by the degree of diagenesis (Fig. 6.14(b)). The highest resolution possible with fossil events (the smallest resolvable stratigraphic unit) may not be sufficient in all circumstances. 4 The ability to both identify and date sequence bound- aries, transgressive surfaces or maximum flooding surfaces using biostratigraphy, depends upon the actual fossil resol- ution and the apparent resolution recorded by the sample programme. Sample spacing should be designed to solve the geological problem and, ideally, closely spaced samples 10sag Semple ection and tp Gane oan rue arenas | pe acl [Tae cs Tm] Tom [somo ant seo] SOToaieeees | Keomcme | Noam | i = b ab a i as nico . ede s ii ils etait Ison 7 ; i | i ik alee Ww : Lo Ala jet La li ah ak 1 a Leo nop in| e moe SE tieoe > Wi é | Fig. 6.15 The effects of sample type and spacing on biostratigraphic resolution and impact for sequence and systems tract recognition. HST, highstand systems tract; TST, transgressive systems tract; MFS, maximum flooding surface; SB, sequence boundary. Palacoenvironmental in 1, 0, oceanic planktonics are recommended in the vicinity of all potential boundaries. Figure 6.15 summarizes the limitations and advantages of different sample types. Note particularly the limited resol tution of cuttings compared with recovered cores. 5 Caution is advised when tying fossil and seismic events, as both may have errors associated with depth conversion, It is particularly important to recognize that fossil and seismic ties in condensed sections may diverge significantly if correlated into a thicker sedimentary package, such as when correlating from condensed distal sediments into a prograding highstand systems tract. 6 Biostratigraphy and isotope stratigraphy are particularly useful for calibrating and correlating sequence boundaries and maximum flooding surfaces where there is poor seismic cortelatability due to tectonic complexity (e.g. North Sea — pre-rift, Gulf of Mexico —growth faults) and poor oF 106 cators j terrestrial; a, d ef |, py shallow marine; b, m, deep marine; k, marginal marine; & sparse data (e.g. Papua New Guinea fold belt kearstified carbonates affecting seismic resolution, and tropical rainforest making seismic acquisition expensive 7 Biofacies trends can be used to define progradationa, aggradational and retrogradational trends and predict the timing of clastic accumulation on a shelf or bypass to @ basin. An overall shallowing-up trend in biofacies applies equally to both highstand and lowstand progradation whereas the transgressive systems tract is marked by an overall deepening-up trend in biofacies. 8A maximum flooding surface is characterized by common and diverse planktonic assemblages at their widest distribution. 9A sequence boundary is associated with erosion, « biostratigraphic hiatus and reworking. 10 The widespread distribution of planktonic markersmaximum flooding surface of a condensed section the preferred correlative surface for biochrono- hic purposes. tecognition of palacoenvironments within systems tracts using fossil assemblages can provide a general indi- cation of the potential type, distribution and sand content of reservoir facies. AydexSnensorg | 9 s2rdqyHAPTER SEVEN luvial Systems 1 Introduction 72 Hil processes and channel styles 72.1 Straight and anastomosingcivers 722 High-sinwosity channel systems 723 Low-sinuosiy channel systems 724 Classification of fluvial systems 73 The concept ofthe graded stream profile 7.4 Fluvial architcctare 7.4.1 Conteols on fuvial architecture 7.42 Sequence boundaries and lovrstand systems eracts in alluvial 7.43 Transgressive systems tects and flooding sueaces it alluvial srata 74 Highscand systems eracts in alluvial strats 7.5 Reconstructing fh 75.1 Eocene South Pyrenees, Spain 7.5.2 Triassic, Wshak Formation, Prudhoe Bay Field, North Slope, Alaska, USA 7 Introduction je value of sequence stratigraphy as a predictive tool in theanalysis of clastic shoreline and shallow marine systems is well documented in the literature and is reviewed in pter 8. Recent attemprs to apply sequence stratigraphic coaceps to flavial systems has met with less success because the role of relative sea-level fluctuations in fashioning the fluvial stratigraphic record is less clear (Posamentier and Vail, 1988; Posamenticr, 1993; Shanley, 1991; Shanley and McCabe, 1989, 1990, 1991, 1994; Westcott, 1993). Fluvial systems respond to a bewildering set of external flloclic) and internal (autocyclic) controls (Schumm, 1968, 1981; Schumm and Ethridge, 1991). The problem is tompounded by rapid lateral facies changes and a lack of inernal features in thick alluvial successions which allow them to be sub-divided into time stratigraphic units. As a result, the application of sequenee stratigraphy to fluvial syste is stil in its infancy, with concepts and models still thesubject of lively debate (Galloway, 1981; Miall, 1986, 1991; Boyd et al., 1989; Walker, 1990; Posamentier and James, 1993; Schumm, 1993; Westeott, 1993; Koss et a, 1994; Shanley and McCabe, 1994). The early sections of this chapter focus on depositional Jrocesses and fluvial channel patterns. The concept of the faded river profile is then introduced as the primary ‘ontrol on the development of accommodation in fluvial systems. Finally, the potential role of relative sea-level shange in influencing the fluvial record is reviewed and ilusrated with a number of examples. 7.2 Fluvial processes and channel styles Fluvial systems are one ofthe best studied ofall depositional environments. A detailed review of fluvial depositional environments is beyond the scope of this chapter and the reader is referred to a number of excellent summaries oon this subject (Cant, 1978b; Miall, 1978, 1992; Ethridge and Flores, 1981; Collinson and Lewin, 1983; Collinson, 1986; Ethridge et al., 1987). Four traditional styles of fluvial channel are commonly recognized; braided, meandering, anastomosing and straight (Leopold and Wolman, 1957; Leopold et al 1964; Rust, 1978). These categories are useful but must be considered only as tendencies within a spectrum of channel types or classes (see discussion in Miall, 1992). River channel pat- terns are controlled by discharge, sediment supply and gradient (Bridge, 1985). As a result, changes between ‘channel classes are gradational, with one or a number of different channel patterns containing similar morphological elements. 7.2.1 Straight and anastomosing rivers Straight and anastomosing river systems ate relatively rare in the recent and geological record. Straight rivers have well-defined single channel courses with stable banks flanked by levees. Anastomosing rivers form interconnected networks of low gradient, relatively deep and narrow channels of variable sinuosity, characterized by stable, vegetated banks composed of fine-grained siltor clay (Smith and Smith, 1980; Patman, 1983; Rust and Legun, 1983; Nanson et al 1986; Fig, 7.1). Lateral channel migration is limited by the development of fine grained bank material miMap viow of channel oomety a (channel sand ‘Soestg or pot Flow dtetion crevasse splay / Teed to verleally stacked channel sande wit ow wth to depth ratios t Fluvial aggradation and channel continemant hal Mulistory channel fis encased in mud and clay ‘Saquence dominated by Fig. 7.1 (a) Block diagram of an anastomosing fluvial system illustrating the facies associations, channel belt and floodplain subenvironments. Channel belts confined by fine-grained, vegetated overbank and floodplain systems. Preserved channel belt systems form an interconnected network of low width-to-depth ratio sand bodies encased in overbank fines. Channel type common in low-gradient drainage systems and where increasing accommodation space leads to vertical rather than lateral accretion of alluvial deposits (after Galloway, 1981; Miall, 1992). OF, overbank fines; SB, sandy bedforms; CH, channels and vegetation. Changes in channel course occur through avulsion (Smith, 1983); a process where successive major flooding events of the river result in progressive breaching and crevassing of the channel banks and alluvial ridge, leading to the formation of a new channel course along the lowest segment of the flood plain. Width to depth ratios of ‘modern anastomosing channels are noticeably lower than for meandering systems (Smith and Smith, 1980; Smith and Putman, 1980; Smith, 1983; Tornqvist, 1993). The overbank and floodplain areas separating channel courses consist of narrow natural levees, numerous crevasses, veg- tated islands and wetlands (Smith and Smith, 1980). ‘Anastomosing channels have been documented in humid, tropical, semi-arid and arid climate settings, and appear to be typical of low-gradient downstream drainage systems dominated by cohesive sediments such as coastal plains and delta tops (Smith and Smith, 1980; Rust, 1981; Rust and Legun, 1983; Smith, 1983; Caimeross e¢ al., 1988). This channel style may be developed preferentially in times 2 of rapidly rising base level, where increasing accommo- dation space leads to a fluvial succession dominated by vertical rather than lateral accretion deposits (Smith and Smith, 1980; Smith, 1986; Kirschbaum and McCabe, 1992; Torngvist, 1993). The depositional record of these systems is therefore dominated by isolated, shoestring-like sand bodies separated by levee—overbank, erevasse splay and flood-plain fines (Fig. 7-1a; Friend et al, 1978; Friend, 1983). 7.2.2 High-sinuosity channel systems High-sinuosity channels develop in drainage areas domi- nated by low-gradient slopes, and where the river is com: monly dominated by a high suspended load to bedload ratio (Leopold and Wolman, 1957; Allen, 1965; Schumm, 1971, 1977, 1981). They are typically organized into channel and overbank segments (Fig. 7.1b). Channels lie within a broad meander belt dominated by a complex‘onanne map view plain sediments, or deposi bation of active and abandoned channels. Active depo- is confined largely to the channel belt, resulting in a developed, raised alluvial ridge which stands above I level of the flood plain. The distal margins of ial ridge form overbank areas that interfinger with jaent flood plain, The course of the channel may be ned by abandoned channel plugs or be relatively to migrate laterally, developing point bars and associ- lateral accretion deposits. sediment load of modern high-sinuosity channels is variable, ranging from systems dominated by a fine- suspended load (Jackson, 1976, 1978; Stewart, 3) 0 coarse sand, gravel, pebble (Bernard and Major, j MeGowen and Garner, 1970; Levey, 19755 Arche, 3} and gravel/pebble-rich systems (Gustavson, 1978; 1982; Forbes, 1983; Campbell and Hendry, 1987). A variey of facies and vertical facies successions may Point baraccrtion Sancoody arentecte Low tt gga aa set ho, tla ‘Sdn Sos wn mted dope “oti ine with choral ava. Fring war! jl nay be tune channel course ate fi! aoa od 0 toon snore bes Wi Wo vp arorng channel an rear es. ing omar ces pected ark snd bed aerton etn pes soon 71 (h) Block diagram of a high-sinuosity fuvial system illustrating the facies associations, channel belts and flood-plain ronments, The channel bet is confined within a raised alluvial ridge. The character of the channel-fill may be highly = Channels may migrate laterally to develop tabular to sheet-like sand bodies separated by fine-grained overbank and n may be confined by channel plugs, resulting in ribbon sand bodies. In these cases the width-to- ratios may therefore be variable, The stacking patterns of channel sand bodies will be affected significantly by rates of lood- subsidence (accommodation space). (Modified after Galloway 1981; Miall, 1992) develop in the ancient record and facies successions may grade into the deposits of low-sinuosity. systems (Puigdefabregas, 1973; Miall, 1983, 1987; Stewart, 1983). Channel facies form tabular to sheet-like sand bodies separated by fine-grained overbank and flood:plain deposits (Fig. 7.1b; Friend, 1983). Areally restricted ribbon sand bodies may be more prevalent where channel belt migration is impeded (Puigdefabregas and Van Vliet, 1978). Associ ated overbank deposits comprise wedged shaped accumu- lations that thin and fine uniformly from the channel belt into the adjacent lacustrine or flood plain (Tornqvist, 1993). 7.2.3 Low-sinuosity channel systems Low sinuosity or ‘braided’ channel systems occur where the coarser geain sizes, such as gravels and/or sands, form the dominant load of the system. In these cases, the lack of 113Muti-storey channet fs wi ow proportion of ‘vetbank tines. Channa fil dominstes by gfavelly tar fers Vertical sequences eek tinier Leck ot Say Sao se oe ; i ) ; Hee Sf meni —, AT Cg \ fo. 7 — Downstream sooraen OA) (Channel sand ‘eth Sandbody architecture SS Mut-storey chara is vith ow ropoion of ‘verb ines. Channel 1 dominated by sandy bedloes Fig. 7-1 (¢) Block diagram of a low sinuosity (upper) gravelly and (lower) sandy fluial system illustrating the facies association channel belts and flod-plain subenvironments. The lack of well-developed fine grained overbank sediments within thes systems leads to mobile channel courses and well-developed sheet sandstones (SB) and gravel bodies (GB} (modified after Galloway, 1981; Mill, 1992) significant cohesive bank material leads to extremely mobile 1969; Collinson, 1970; Smith, 1974; Cant and Walker channel courses (Fig. 7.1c) Individual channels continually 1976, 19785 Miall, 1977; Cant, 1978a,b). shift and bifurcate producing a rapidly evolving network of The high bedload content of the low-sinuosity rivers variable-scale braid channels with submerged in-channel combined with the highly mobile nature of channel systems migrating bedforms (Leopold and Wolman, 1957; Coleman, leads to a stratigraphic record dominated by lenticular 4ave-upward sand bodies characterized by variable scale lication, lateral accretion deposits and a lack of wel-margin facies (Moody-Stuart, 1966, Campbell, 16 Hazeldine, 1983; Bristow, 1987). 124 Classification of fluvial systems basic channel styles described above are often difficult fewognize in the geological record. The grain size of the al system has been suggested as a parameter to sub- ile flaial systems because it can be measured in both ancient and moder, at outerop and in the subsurface. this basis, river systems can be broadly subdivided into principal types comprising; high-bedload, bed- id, mixed-load and suspended-load dominated rivers yumm, 1977; Schumm and Brakenridge, 1987; Orton Reading, 1995), Each of the four system-types displays acteristic channel-fil geometries, facies assemblages id vertical successions (Fig. 7.2 and Table 7.1). 3 The concept of the graded stream profile all depositional systems, deposition, burial and erosion controlled by an equilibrium surface or base level ichdefines and influences the amount of accommodation ce (see Chapter 2), The equilibrium surface separating, jn from deposition in fluvial systems may reflect the ence of numerous base levels (Miall, 1987, 1992; umentier, 1988; Westcott, 1993). These include lake- fable 7. Summary characteristics of fluvial depositional syst after Mill (1978, 1985). level, the level of trunk-stream drainage, the position of riick points, the position of ground-water tables and relative sea-level. This is in contrast to shoreline and shallow marine systems, where base level generally equates to sea level. For this reason, the concept of a graded stream profile has been applied to fluvial systems to define an equilibrium surface that separates erosion from deposition and controls available sediment accommodation (Mackin, 1948; Sloss, 1962). Mackin (1948) defined a graded steeam or river as: ‘One in which, over a period of years, the slope is delicately adjusted co provide, with available discharge and with prevailing channel characteristics, just the velocity required for the transport of load supplied to the basin, The graded river is a system in equilibrium: its diagnostic characteristic is that any change in one of the controlling factors will be transmitted throughout the whole profile A graded profile can be considered as representing a balance between erosion and deposition; it will be graded such that the stream can transport its load along the profile without significant erosion or deposition. The slope at any point on the graded river profile will be a function of river discharge and sediment load (volume and capacity), Downstream reductions in slope relate to increasing discharge and down- valley decreases in grain size (Fig. 7.2). The overall shape of the slope will be modified through time to approximate a concave upward graded profile, flat at the river mouth and, steepening towards the source. ms according to grain size (from Orton and Reading, 1993). Facies 1 2 3 4 Gravel ‘Gravel and sand Fine sand Mudie small (< 10?) Intermediate (= 10°?) Intermediate (= 10°km) Large High Intermediate Intermediate tow Ari, artic Temperate Temperate Humid, topical Small Intermediate Intermediate Large Very steep (»Smkm-") Intermediate (>0.Smkm-") Intermediate (=0.05mkm-") Low High to very high Intermediate Intermediate Low low (<100m's") Intermediate (< 10°95") Intermediate (= 10¢m?s-") High | Very irequlae ulor—regulae Regular-iregular ‘Very regular Low (= 108 tons year") Intermediate (= 10" tonsyear: (= 10° tone High (= 10° tons ye Highivery high intermediate | Low Bedload Bed load Mixed toad Suspended load Braided/sbeont Braided Meandering/braided Staighvmeandering Moderate Lom-moderate tow High Incermediate High-intermediate High tow Incermediate High-itermediate High Low (fined Joditetual elements 68, 56 35, FLA, LA 38, Fa, OF OF, LA,S8 66, yal barbedtocms 56, sediment gravity flows; sandy bedforms; FM, foreset macroforms; LA, lateral action; OF, overbank fines 11sMupysitt FINE SAND COARSE SAND GRAVEL or TROLLHEM = SG came! ouney NY ‘ BEET sand | gravel bars a Flow pattern je to decharge capacity Increase instream power Increase in sediment toad rt ‘SUSPENDED LOAD MIxED LOAD. BED-LOAD VERY HIGH BED- LOAD eee EE. E> PROPORTION OF BED-LOAD KE 1 of a 1 Increasing size of drainage basin area, dscharge regulary, and sorting =o HOTT mi [TI conoees Chane i wor DEPTH nario Ae ‘om (nn (C0 ccc Cee ]) wo vow OCERATE (CHANNEL STABILITY or BANK STRENGTH Fig. 7.2. Relationship between grain size and channel pattern (after Orton and Reading, 1993). Note general tendency of jnercasing sinuosity, confinement of channel belts and discharge regularity with decreasing grain size and low-gradient slopes Position of fluvial examples approximate (based on Schumm, 1981; Ferguson, 1987; Miall, 1985) (Fluvial facies and architect clements, after Miall, 1985). Architectural elements: GB, gravel bar/bedforms; SB, sandy bedforms; FM, foreset macroforms; LA lateral accretion; SG, sediment gravity flows; and OF, overbank fines ‘Through time, rivers will attempt to establish a stable and internal attributes of the river (Schumm and Ethridss graded profile for a given discharge of water and sediment. 1991; Germanoski and Schumm, 1993; Schumm, 1993) Disturbances in this system, such as sea-level fluctuations, This can be achieved by altering a river’s characteristis climatic change and tectonics, cause the stream to re- such as (Table 7.2): establish a new state of equilibrium by changing heexternal___ channel width 16sediment calibre relocty boundary roughness depth sediment discharge slope lanform. ugh autocyclic mechanisms may modify the short- character of the fluvial system, only long-term allo- licevents may alter fundamentally one or a number of Yariables, leading to major changes in fluvial architec- Studies of modern river systems suggest that these jastments may take considerable periods of time. Indeed modern rivers exhibit graded profiles still adjusted to wernary glacial events (Wilcox, 1967; Church and immaker, 1989). Recognition of these large-scale system- changes, and their attendant stratigraphic boundaries, provide a sequence stratigraphic framework for the lation of ancient flavial deposits. 14 Fluvial architecture vial architecture can be defined as the three-dimensional etry and interrelationships of the deposits of channel, crevasse splay and flood-plain and other sub: ronments of the fluvial depositional system (Miall, 83, 1985). Fluvial architecture can be regarded as the lt of the interaction between river process and lone- UPSTREAM CONTROLS timate Er MiD-STREAM CONTROLS ‘Tectonic ting, faulieg, drainage basin age system 73 Key controls influencing the graded profile within the upstream, mi term allocyctic controls. The term can be applied at varying scales from the deposits ofa single river to the non-marine depositional record of a basin fill. Most recent studies suggest that all fluvial deposits can be subdivided into a number of key architectural elements, defined by overall shape, distinctive facies assemblages and intemal geometries (Miall, 1985, 1988). These elements form the basic building blocks of all fluvial systems, yer their overall geometry and stacking patterns vary between different fluvial channel styles. Variations in the proportions of particular architec- tural clements can be used to interpret different favial styles in the geological record. 7.4.1 Controls on fluvial architecture In modern systems, channel morphology evolves down: stream in response to changes in valley slope, sediment load, bank materials, climate and tectonic regime, These same controls may change abruptly or evolve over time to cause significant or subtle morphological changes through- out the whole, or part, of the fluvial system (Table 7.2) (Burnett and Schumm, 1983; Schumm, 1993). For this reason, fluvial channel patterns in the geological record are unlikely to remain constant through a stratigraphic unit To review how these changes take place, is it useful co divide the fluvial system into three geographical arcas (Fig 7.3). These comprise: (1) the upstream area forming the hinterland and source area for the river; (2) the mid-stream region forming much of the drainage basin over which DOWNSTAEAM CONTROLS 7 eustasy ase evel S suosidence cate stream and downstream portions of the fluvial 117Table 7.2 Channel adjustment to changes in hydrology and sediment load based upon empirical equations derived from Schumm (1968, 1977). The data illustrate how decreases and increases in annual discharge rates (Q) and percentage sediment load transported as bed-load (QS) influence channel depth (D), meander wavelength (I), sinuosity (P), slope (s) and width : depth ratio (F) Percentage: wisthreo- ‘Annual bed-oad Channel channel Meander depth Impact on fluvial discharge transport = width degth—— wavelength Sinuosityt Slope ratio system @ 2) @) o o ° “6 a) Increase Increase == Increase Increase Increase —=««Decreaie—increate/_-—inceate Broadening of channek, decease decrease ecreasng sinusty, incense in channel width ard ope Decrease Decrease Decrease Increase Decrease —ncrease—sncrease/ ‘Decrease Narrowing of channels, decrease decease increas in channel depth, increasing sinuosity, decreasing slope Incroace Decrease Increase Increase Increase ——trerease-—«‘Decreases Decrease Narrowing and deepering ef decrease decease ‘channels, increased sist, decreased slope Decrease increase Increases Decrease _Inerease/_—«Dacreate “Increases ncreaues Widening of channels, decrease decroxe increase in slope, decease In sinuosiy £ Sinuosty (/)= channel length (i/valleylenath (.)= valley slope (5,)/ehannel slope (5) sediment is transported; and (3) the downstream or lower reaches of the river in which sediment is deposited (Schumm, 1977). The controls influencing each of these regions of the system are discussed below. Upstream controls The graded profile within the upstream portion of the fluvial system is affected by tectonics, climate and bedrock. geology. Tectonics plays a significant role in influencing the type of channel system, the nature of the sediment load and, the calibre of sediment deposited (Cant, 1978b; Miall, 1981). For example, source areas undergoing rapid uplift may generate large volumes of coarse-grained clastic sedi- ments and steep gradients. These conditions may favour the development of high-gradient braided or low-sinuosity channel systems within the upstream portions of the drain- age basin, In contrast, lack of relief in the stable hinterland and river catchment areas may result in redueed regional ‘gradients and the transport of finer grained clastic sediments by river systems. In these cases, meandering or anasto- mosing river systems may dominate the upstream region. The impact of tectonics on the architecture of upstream fluvial systems i illustrated in studies of modern systems (Coleman, 1969; Alexander and Leeder, 1987), outcrop, (Heward, 1978; Gloppen and Steel, 1981; Lawrence and, Williams, 1987; Nichols, 19875 Jolley et al. 1990; Turner, 19925 Garcia-Gil, 1993) and in laboratory flume exper~ iments (Burnett and Schumm, 1983; Ouchi, 1985; Fig, 7.4). Active tectonics in the upstream areas of the fluvial 18. system may lead to stream rejuvenation, incision by nick: point retreat, river capture, and the erosion and cannibal ization of basin-margin clastic sediments. These events may alter the discharge and load characteristics of the fluvial system, leading to changes in shape of the graded profile, ‘Where these modifications are sustained, abrupt changesin channel pattern, style and architecture may be recorded in the stratigraphic record (Blakey and Gubitosa, 1984 Butler, 1984; Turner, 1992). The upstream portion of the fluvial system is affect additionally by climate and bedrock geology (Schumm, 1977; Westcott, 1993). Climate influences the vegetation types, the degree of rainfall and run-off and the long-term discharge of the river (Knighton, 1984). Similarly, vat ations in bedrock geology control the type of material available and the load transported. Mid-stream controls Intrabasinal tectonics exert the dominant control on the id-stream portion of the flavial system by affecting the location and types of drainage patterns on the alluvial plin (Miall, 1981; Alexander and Leeder, 1987; Kraus and ‘Middleton, 1987). Tectonic tilting associated with extension for thrust-related asymmetric subsidence superimpose a new slope on the pre-existing flood-plain topography (Alexander and Leeder, 1987; Wells and Dorr, 1987) ‘These changes influence the graded profile of the rivet, leading to stream deflection and the shift of the channel course in the direction of maximum subsidence. Tilting(2) Bedload Braided River Sipe dotormaton iver adjustment ‘Slope deformation iver sovatment up (i) Subsidence rosradation rate : fi Iw Wa i Hd eet 7th te KBR ean lacaracsation a gradation £ ar-braided or Terraces Atom Gue Tanto fate bralaee wah wre ars braced bars Bars meandering channel xed Load teandering River ‘Sopa cetermation Fiver adusiment Slope detomation Fiver adjustment win ‘chan sutscorce Channel er = caren] — Raieuato + Suspended Load Meandering River ‘Spe deformation Fiver agjustment Slope deformation fiver aciustment | Subsloence ‘Channa i { Eee Reticulate ' 1 eticulate — 1 Anestomosng t DIM Fig. 74 Response of contrasting fluvial systems ¢0 uplift Adjstmene of (a) bed-load-dominated braided, (b) atilinal uplift, (2) synclinal subsidence across it, The ‘emission of Geological Society of America) al flume studies by Ouchi (1985), lering and (c) suspended load meandering river to (1) r-scale runs from top to bottom in each diageam (published with the 119Table 7.3 Characteristic luvial depositional patterns of systems tracts and their stratal boundaries associated with relative sex level changes (compiled from Nami and Leeder, 1978; Marzo et al., 1988; Van Wagoner et al., 19905 Kirshbat 1992; Shanley and McCabe, 1993; Tomgvist, 1993; Westcott, 1993) m and McCabe, Sequence boundary Lowstand systems tract Transgresive systems wack ‘Maximum flooding surface Highstand systems wet Surface related to 3 Felativeseo-lvel fl, Boundary separates change from decreasing 1 increasing rate of {generation of accommodation space, ‘Associated with 9 basinward shit in facies and potentially ‘accompanied by fluvial Relative sea-level fll causes rivers to straighten, deepen and widen ther courses resulting in incived valleys, Width of channel bet significantly smaller than the extent ofthe valley incision Sequence boundary may define an abrupt change in aces, €, change from high. to low sinvosity channel pattems, 2 change in 1a ize in response to increased stream power and sediment load, and fora change in sediment «composition Sequence boundary may separate ditforent degrees of channel sandstone amalgamation. Change in architecture reflecting change in sccommadation space Fluvial systems inked down dip to coarier srsined lowatand shorelines o bypass shelf {feed submarine fons High suieam powers, dlseharge rates and sediment loads may lead to constructive fava: Aominated deltas Depostion on the river, stream power and teansport capacity may increase leading to erosion and fluvial incision — this particularly may be the ease where base level {alls below the shelf—slope break (type 1 sequence boundary); (6) where the gradients of the shelf and river system are equivalent, the stream requires little or no adjustment to maintain equilibrium and ehe stream extends its profile accoss the shelf; (€) where the gradient of the shelf < river, a reduction in slope may reduce stream power and transport capacity, leading to deposition. SLI, inital sea level; SL2, sea level datum at lowered base level ‘Commonly, the gradient of the lower reaches of the river is less than the gradient of the coeval shelf and a relative sea-level fall will impose a steeper gradient on the graded profile of the river. In such cases stream power and dis- charge may increase with a resultant increase in the poten. tial for erosion and fluvial incision (Fig. 7.5a). The river system may attempt initially to reach a new equilibrium profile by changing its load characteristics, sinuosity and {n-channel bedform styles. However, ifthe imposed gradient is greater than can be accommodated by an increase in channel width or by a changing channel pattern, the river will tend to incise in order to reduce the slope. Rivers ofall seales commonly show a characteristic evolution in re sponse to base level falls, in which they first narrow and straighten, then deepen and finally widen (Schumm, 1981). In theory the incision is propagated upstream as a i point, and the whole river profile eventually lowers 10 121reach an equilibrium with the new base level (Begin e¢ al. 1981; Posamentier and Vail, 1988). It has been suggested that the gradients of the incised riyer are approximately double that prior to the base level fall (Salter, 1993). This change increases stream power, allowing the river to further incise unless the increased energy of the river can be dissipated by the erosion of channel banks, the widening of the river, and increasing, roughness of the channel floor or changing sediment load characteristics, such as coarsening grain size. This perhaps, explains why the width of incised valleys is often much greater than the river channels within the valley fills. Ie further suggests that wider channel belts with coarse bed- load will be a logical outcome of a base level fall. These characteristics have been noted in a number of Quaternary and ancient examples (Marzo et al. 1988; Eschard, 1989; Van Wagoner ef al., 1990). The greater the slope created by the base level fall, the greater the slope imposed on the fluvial system and the greater the degree of incision. Incision is therefore particularly significant for type 1 sequence boundaries (Posamentier and Vail, 1988; Van Wagoner, et al., 1990; Wood, 1991; Wood et al., 1991; Westcott, 1993). In practice, rivers often appear not to fully equilibrate with the lowered base level, or the incision propagates only a finite distance upstream before a subsequent base level rise induces aggradation. In the absence of significant base level fall, rivers are more likely to adjust their profiles by changing their channel pattern, discharge and load charac- teristics rather than incising major valleys into the shelf (Suter and Berryhill, 1985; Blum, 1990; Autin et al. 19945, Schumm and Ethridge, 1991; Westcott, 1993; Koss et al., 1994), Sometimes base level fall may be marked not by erosion and valley incision but by a more subtle boundary separating fluvial deposits with distinctly different flavial architectures (Shanley and McCabe, 1993; Westcott, 1993). The impact of base level rise (relative sea-level rise at the shoreline) on the fluvial system is equally complex. Base level rise may impose a lower gradient on the lower reaches of the river system, reduce stream power and discharge and decrease the capacity of the river to maintain sediment transport. In these cases the river responds by depositing sediment. Many workers have suggested that the rate of base level rise is particularly important in controlling how fluvial systems respond (Posamenticr and Vail, 1988; Posamentier ef al., 1988; Shanley and McCabe, 1993, and, references therein). Rapid base level rise may be considered analogous to damming the downstream portions of the river valley (Fig. 7.6a). In these cases the rate of rise will be greater than the rate of deposition, and the lower course of the fluvial system will be flooded. Flood-plain aggradation during shoreline transgression is limited. Upstream of the flooded area, the effect on the river channel may be limited, swith it till remaining at grade. Significant fluvial aggradation will occur mainly as the shoreline progrades basinward by deltaic extension. Signifi- 122 (@) STAGE ONE Ina rate ose > rat of deposition ever cours I Hooded ‘@ Upstream the rier chancel ia eesenaly unatlectes (@) STAGE TWO € Rate of vise slows and dais buld out atthe new (higher base level (Ful eyttoms aggradein response to imposed lower slope Fig. 7.6 Fluvial system response to a base level rise. The response of the fluvial system will be affected by the rate of relative sea-level rise and the contrast in gradient between the shelf and the alluvial system. Response of the fluvial system 0 (a) rapid base-level rise and (b) slow base-level rise andor high, sediment supply rate (modified after Posamentier and Vail, 1988; published with the permission of the Society of Economie Paleontologists and Mineralogists) cant fluvial aggradation will mainly occur as the shoreline progrades basinward by deltaic extension; when the rate of sediment supply exceeds the rate of base level rise (Fig 7.66; Posamentier and Vail, 1988; Shanley, 1991). In effect, the river system will deposit sediment in order to increase the slope and maintain grade with the upstream discharge and load. The net result may be a narrowing of the river channel belt, a change in channel style and @ decrease in sand-body connectivity. Aggradation of the flood plain during rapid base level rise is therefore @ fanction of the rate of rise and sediment supply. Coastal and marginal marine processes (ix. waves, tides and storms) may also affect the graded stream prof during transgression, In the Canterbury Plains of New Zealand, lowstand regional flood-plain deposits are pre sently being actively incised by braided river systems in an area currently undergoing transgression during a petiod of eustatic highstand (Leckie, 1994). Coastal erosion and sliff development during shoreline retreat has steepened fluvial gradients, resulting in incision, This example serves 10 further illustrate the potential difficulty in distinguishing cause and effect in ancient fluvial deposits. Changes in fluvial architecture related to downstream controls are summarised in Table 7.3 and discussed below in the contest, of the key stratal surfaces and systems tracts generated ducing a relative sea level cycle.42 Sequence boundaries and lowstand systems tracts in alluvial strata quence boundaries represent a surface along which sedi- isbypassed during.a relative sea-level fall. The bound- ' coincident with subaerial exposure and erosion, and inward shift in facies belts and coastal onlap tier and Vail, 1988; Posamentier et al., 1988; n Wagoner ct al., 1990). In alluvial successions, the cogition of sequence boundaries is complicated by lateral es changes and the difficulty of distinguishing major and abrupt basinward shifts in facies from localized 4 scour (Posamentier, 1993; Posamentier and Veiner, 1993; Westcott, 1993). Where the gradient of the elFexceeds the gradient of the fluvial system, significant ‘may occur, leading to the formation of an incised (bee also section 8.3). The fill of the incised valley be a complex function of the rate of base level rise tive to sediment supply. Low rates of sedimentation are jy to lead to rapid drowning of the fluvial system and formation of an estuary. High sediment supply rates result in thicker fluvial, deltaic and tidal deposits wand, 1989; Van Wagoner et al. 1990; Allen, 19915 mple et al, 1992; Shanley and McCabe, 1993; hanley 7 al, 19933; Richards, 1994). Valleys abandoned pastion of coarse-grained fluvial deposits on marine or ginal marine strata may record the presence of a se- quence boundary. Some caution should be applied in interpreting all cases ofstteam rejuvenation as the product of relative sea-level fall (Posamenticr, 1993). Valley incision additionally may vecur in response to: (1) increased discharge and stream power, (2) decreases in stream load, or (3) from tectonic uplift within the mid-stream and upstream portion of the drinage basin. In all three cases flavial incision results in sediment bypass and the development of localized unconformitis Sequence boundaries developed within the interiluve areas adjacent to incised valleys are represented by a sub- aerial exposure surface recording periods of sediment by- pass They may be recorded by the juxtaposition of marine shales on thin emergent pedogenic horizons, or marine shales overlain by flood-plain and overbank fines. Associ ated pedogenic horizons and coals are developed but ‘are often thin owing to the limited sediment accommo- dation available (Van Wagoner et al., 1990; and section 741). Where valley incision is less clear, such as where the ‘lope ofthe shelf is equal to or less than the fluvial gradient, the position of the sequence boundary is less easly inter” preted. In these cases recognition of sequence boundaries must come from identifying systematic changes in channel stacking patterns; vertical variations in channel style and in the degree of sandstone amalgamation. These changes may be paralleled by variations in the geain size, abrupt changes in sediment composition and variations in the scale of sedimentary structures within the channel fills (Marzo et al, 1988; Eschard, 1989; Shanley, 1991) The lowstand systems tract in the fluvial record may represent a period of renewed alluvial aggradation associ- ated with the initial period of slow relative sea-level rise, following base level fall. Channel courses are confined initially to the central axis ofthe incised valley, resulting in repeated reworking of channel and flood-plain deposits (Fig. 7.7). At this stage, differences in channel styles may be less important in affecting the gross sand-body architecture of the lower valley fill than the dominant sediment load of the fluvial system. In effect, the depositional products of the bedload-dominated, low- and high-sinuosity channel systems may be identical because the lack of accommo- dation promotes reworking and bypass of fines and the limited differentiation between channel and flood-plain deposits. Asa result, the lower valley fill may be dominated by high net: gross, multistorey multilateral sand bodies. Increasing rates of relative sea-level rise will promote increasing accommodation and an overall reduction in stream gradients. Low-sinuosity fluvial systems may re- spond by increasing their sinuosity to maintain grade whereas high-sinuosity systems may become more effect: ively partitioned into channel and floodplain components ‘without a change in channel style, ‘The complex facies character and stratal patterns of lowstand fluvial deposits are illustrated in studies of the Blackhawk Formation of the Cretaceous Mesaverde Group, Book Cliffs, Urah (Van Wagoner et al, 1990). The forma- tion represents a time interval of 6 million years and can be subdivided into three units; the Grassy, Castlegate and Desert Members. Four sequence boundaries are recognized within the formation, the first occurring at the top of the Grassy Member, a second and third within and bounding, the Desert Member, and a fourth within the Castlegate Member (Fig. 7.8). The Desert Member lowstand systems, tract consists of point-bar deposits with probable tidal influence (clay-draped lateral aceretion surfaces), coals up to 30cm thick and thin fluvial sheet sands. In places incision has resulted in coals lying directly on the Mancos, Shale marine shelf facies, and meandering channel de- posits erosively overlying lower shoreface sandstones. The sueveeding Castlegate lowstand systems tract shows a downstream evolution in fluvial channel style, reflecting a basinward reduction in stream gradient following a relative -level fall. The most proximal portions of the system are represented by braided channel deposits, 50m thick, which thin and pass down-dip into high-sinuosity channel and ‘meandering point-bar systems down-dip. The downstream sinuous channel patterns can be traced up-system to a series of narrow separate valley fills with broad interfluves over a total distance of approximately 100 km. 123Incision wTeRFLUvE © SHELA ORADIENT> RIVER GRADIENT. saarine GHALE coats! ast2 % sous 7 LOWSTAND TERRACE DePosiTS GRADED PROFILE EXTENDED ‘SHELF GRADIENT » RIVER GRADIENT = SLI re %G DEPOSITION. SHELF GRADIENT < RIVER GRADIENT Bice) 0 — © ——- © SSE © “Ly “UMITED ONAN I FLUWAL caANNEL EATER HENGE RE BOM AY TIDALLY INFLUENCED CHANNELS ABOVE SEQUENCE BOUNDARY o f Maran FLOODING = ® ! Ls ‘SEQUENCE BOUNDARY its —= + LONGITUDINAL PROFILE OF RIVER EXTENDED + NO CHANGE INALLUVIAL AGGRADATON RATE: “CHANGE IN TRANSPORT CAPACITY IN RESPoNse 10 REDUCED ALLUVIAL GRADIENTS. Owen O—e_—_( OO BED.LOAD, MIXED LOAD. DOMINATED MODERATE TO (Low SINUOSITY) HIGH SINUOSITY ELUVIAL FLUVIAL SYSTEMS SYSTEMS, 7.7 Conceptual models illustrating the potential evolution of system, which m, changes in archicecture or stream type fluctua ITS, initial transgressive surface or equivalent alluvial surfaces $B, sequence boundary or equi 7.4.3 Transgressive systems tracts and flooding surfaces in alluvial strata In shoreline and shelf suecessions, the transgressive systems tract occupies the interval between the transgressive surface and the maximum flooding surface (see Chapter 2 for definition). The identification of a transgressive systems tract in alluvial successions is less clear because neither of these stratal surfaces can be recognized in non-marine sections, Instead, the transgressive systems tract may be recognized only through changes in fluvial channel stacking pattern related to rapid base-level rise and shoreline trans- gression (Fig. 7.9) Following a type I sequence boundary, the flavial deposits 124 SUSPENDED LOAD HIGH SINUOSiTY AND ANASTOMOSING FLUVIAL SYSTEMS vel fall. In this example, the impact of reductions in accommodation space le ‘or may not include incision depending on the relationship betv re assumed to reflect changes in accommodation space in response to relative sea-level ions. No consideration has been given to the potential for stream + DEPOSITION ANDIOR POTENTIAL EVOLUTION OF CHANNEL STYLEIN RESPONSE TO REDUCED STREAM Power LOWSTAND TERRACE bePosns @ © = wy TIDALLY INFLUENCED CHANNELS ABOVE SEQUENCE BOUNDARY contrasting types of fluvial system in response to relative ls to changes in the character of the fluvial shelf and fluvial gradients. In all cases ipture or changes in upstream and mid-stream contrals, ent alluvial surface of the lowstand and transgressive systems tracts can be distinguished by the identification of the first major marine flooding surface over the interfluves to the incised valleys. In other instances, the distinction between the fluvial de posits of the lowstand and transgressive systems tracts may not be clear. This may be the case where the initial flooding and marine onlap of the sequence boundary occurs further downstream of all, or part, of the major valley incision, Where shelf and fluvial gradients are similar, the initial deposits overlying the sequence boundary represent a de- positional episode post-dating the initial transgressive surface, and true lowstand deposits will be absent, For this reason, Shanley (1991) coined the term alluvial transgressive deposits to encompass the complete fill of the7.8 Summary stratigraphy and selected ver ence boundaries in the Blackhawk Formation, Book sists) incised valley. Use of this term can avoid misapplication of sjstems tract terminology in areas where regional stratal rdationships are not apparent. The fluvial deposits of the tansgressive systems tract rellct the balance between sediment supply and the rate of tative sea-level rise. Where sediment supply is relatively {ow only a thin veneer of fluvial sediment may be preserved a the sea transgresses rapidly landwards, Where the rate oftransgression is lower, the fluvial record of a transgressive sjstems tract may include the increased partitioning of Ghinnel and flood-plain deposits in vertical section, the decreasing connectivity of channel meander-belt sands, and the increasing influence of tidal processes in the lower Aeaches of the fluvial system (Allen, 1991; Shanley, 1991; Shanley and McCabe, 1993). Reduction of flavial gradients ‘may promote an evolution in channel styles to more mixed load, meandering or anastomosing rivers, and the increasing presentation of crevasse deposits in the stratigraphic record as rivers attempt to maintain grade with rising base lesel (Ryseth, 1989; Kirschbaum and McCabe, 1992; Tomnavist, 1993). The period of maximum flooding may be represented by ‘idence for subtle or pronounced tidal influence within te fluvial system (Shanley and McCabe, 1991). Poorly drained palacosols or thick coals may occur. Relatively thick clastic and carbonate lacustrine facies reflecting rising ical sections illustrating the range of facies juxtapositions associated with ifs (after Van Wagoner, 1992). Inset stratigraphy illustrating sequence indaries associated with the Castlegate, Deser¢ and Grassy Members. Sections are 20 miles apart. Sequence boundaries resented by SB, possible sequence boundaries by PSB (published with the permission of the American Associate ‘of Petroleum water tables and incursions into poorly drained flood plains. also may be developed (Ryer, 1981; Atkinson, 1986). 7.44 Highstand systems tracts in alluvial strata ‘The highstand systems tract represents the period of de- creasing rates of relative sea-level rise. In shoreline and shelf strata the interval is bounded atits base by a maximum flooding surface and at its top by a sequence boundary. Early highstand fluvial deposits may be indistinguishable from the facies of the late transgressive systems tract. During the late highstand, reductions in flood basin accom. modation may result in the increased potential for lateral rather than vertical accretion, and the formation of laterally, interconnected and amalgamated channel and meander belt systems with poorly preserved flood-plain deposits (Shanley and McCabe, 1993; Fig. 7.10). In all cases the uvial depositional record of the highstand will be affected significantly by the rate of sediment supply to the system, the character of the fluvial system during base level rise, and the extent to which the highstand deposits are eroded during a subsequent relative sea-level fall, The broad links between a relative sea level cycle and its impact on fluvial and shoreface architectural patterns are schematically represented in Fig. 7.11 (from Shanley and McCabe, 1994), 125INTERRLUVE ‘UP-DIP INCISED VALLEY THOK CONS AND FLOOD nace cons SHELF GRADIENT > RIVER GRADIENT. I Move = - - — HoaL nFLUENOE ~ [| =F / TS ool FS wt ae ae Ya “HOALLY INFLUENCED 8 "SEQUENCE BOUNDARY GRADED PROFILE EXTENDED SHELF GRADIENT = RIVER GRADIENT ar =o a Om Oe sii © we © = —$——————r—— Ys “UMITED CHANGE N/FLUMAL ARCMITECTURE = LONGITUDINAL PROFILE OF nuenecrenneo RIVER GRADIENT SHELF GRADIENT = RIVER GRADIENT SEQUENCE BOUNDARY ef ae ——_§_— SLE re an Amen ——$_ — Se ae ~ew| |» J ess @ THK COns gS os. a ze urs. — asi ff Ss SHELF GRADIENT < RIVER GRADIENT g SS ° so —e—EE Oe eee wee “LIMITED CHANGE IN FLUVIAL ARCHITECTURE + CHANGE IN ALLUVIAL AGGRADATION RATE. MES 88 d Ria wes + STABILIZATION OF TRANSPORT CAPAGITYIN » INGREASED DEPOSITION RESPONSE TO DECREASING HATE OF ANDIOR POTENTIAL GENERATION OPACCONMODATION SPACE EVOLUTION OF CHANNEL site vey: Owes ©—=aw © Ow ‘SB: SEQUENCE BOUNDARY ee OR EQUIVALENT BEDL0D MIXED LOAD SUSPENDED LOAD TIDALLY INFLUENCED ATLOVIAU SURFACE BounareD MODERATE TO -HIGHSINUGSITY. CHANNELS ABOVE GOWSINUOSITy) —-HIGH-SINUOSITY _—-ANDANASTOMOSING SEQUENCE BOUNDARY MFS* MAXIMUM FLOODING RUVIAUSYSTeMS — FLUVIANSySTES — FLUVIAL SYSTEWS HAF awa, 50m Fig. 7.10 Conceprual models of evolving alluvial styles during highstand. Period characterized by a reduction in the rate of ‘ation of accommodation space, In this example, the impact of decre ing accommodation space leads to changes in the sharacter ofthe fluvial system and the decreased potential for alluvial aggradation, depending on the relationship between shelf aud fluvial gradients. Rates of sediment flux are assumed to exceed rates of relative sea-level rise during the late highstand. 88, sequence boundary or equivalent alluvial surface; MFS, maximum flooding surface or equivalent alluvial surface sea-level. However, systematic changes in the architecture of allvial successions clearly reflect changes in accommo- dation however formed, and the basic principles of sequence stratigraphic models still apply. 7.5.1 Eocene Castissent Formation, South Pyrenees, Spain The Eocene Castissent Formation of the Tremp-Graus Basin illustrates fluvial and deltaic deposition in a. small thnast-top basin in the Southern Pyrenees Foreland (Marzo # al, 1988). In this area, fluvial and deltaic deposition throughout the Eocene was strongly controlled by the ‘emplacement of Pyrenean thrust sheets (Seguret, 1972; Niiman and Nio, 1975; Ori and Friend, 1984). The Castis- sent formation comprises a series of coarse-grained multi- lateral, multi-storey fluvial sandstones, sandwiched between two pronounced marine to brackish mudstone intervals of Ypresian age (Niiman and Nio, 1975). The contact between the fluvial sandstones and the underlying brackish-marine mudstones is unconformable and represents a sequence boundary (Marzo et al, 1988). Detailed mapping of the sequence boundary shows that the Castissent Formation forms an incised valley-fll and upper deltaic plain complex dominated by three multilateral and maltistorey sandstone bodies (Marzo etal., 1988). These sand bodies are separated by thicker developments of finer grained flood:plain and brackish water deposits reflecting periods of enhanced flood-plain aggradation and sustained marine inundation (Fig, 7.12). Lateral variations in sand-body architecture and fluvial 127“STRATIGRAPHIC ARCHITECTURE SLOW BASE LEVEL RISE TO STL STAND: ISOLATED RIBBONS TO LATERALLY DAMALGAMATED MEANDER BELTS ep PALETE ITIL EERIE BASE LEVEL AISE: FORMATION OF TNDALLY INFLUENCED FLUVIAL DEPOSITS TER LER LE LLLE-ih ‘STIL STAND-ASE LEVEL RISE: AMALGAMATED, HIGH NET ‘FOOAOSS FLUVIAL DEPOSITS. BASE LEVEL FALL: VALLEY INCISION AND FORMATION OF TERRACE DEPOSTS ~“ Fig. 7-11 Summary digram ills RELATIVESEA- SHORELINE ARCHITECTURE. HIGHSTAND: AGGRADATION TO 6 PROGRADATION \ y | RSL CURVE ~ ~ 4 “TRANSGHESSIVE: RETROGRADATIONAL = p aaa o 7 | PROGRADATION ‘Sequence Boundary ing the relationship between shoreface and fluvial architecture as a function of base level ‘change (after Shanley and McCabe, 1993). Detailed correlations suggest the timing of th ‘ised-valley fill occurred after the initial transgressive surface. For this reason Shaaley and McCabe (1993) regard the valley-fill as “alluvial-transgressive deposits style are best illustrated in the lowest sand-body complex, (Complex A of Marzo et al., 1988). The upstream valley- fill is dominated by a braided network of channels showing, a high degree of vertical and lateral amalgamation to form a sheet sandstone complex. Correlative flood-plain deposits, form widespread red beds with caliche soils, suggesting that most of the fines bypassed the alluvial area and the low rates of sedimentation on the adjacent flood plains pro- moted oxidation and soil formation. The valley-fll sand- stones can be traced down-system into mixed-load meandering channel deposits. This down-dip change in channel style is accompanied by reduced incision, a decrease in channel intereonnectedness and the increasing, preser Fig. 7.12. (opposite.) Su vation of flood-plain deposits, due to high rates of flood: plain aggradation, Lateral changes in fluvial architecture in the Castissent Formation are therefore interpreted to reflect reduced fluvial gradients from the hinterland to the margins of the marine basin, Systematic vertical changes in alluvial sand-body archi tecture also occur within the Castissent Formation, and are interpreted to reflect relative sea-level fluctuations (sand- bodies A1-A3, BI-B2 in Fig. 7.12). These changes are particularly well developed in the proximal valley-fill of the alluvial system, where they form repeated erosional~ aggradational cycles (Marzo et al, 1988). Each cycle cot ‘ences with a phase of valley incision into the underlying ary diagram illustrating the geometry and corselation of sand bodies within the Castissent Formation, Eocene, Pyrenean Basin (after Marzo et al, 1988). Note the evolution of fluvial channel styles from coarse-gr wed sheet-sand bodies generated by low sinuosity systems to more tabular or ribbon sand bodies of meandering channel patterns. Individual channel cycles are separated by periods of flood-plain aggradation, palaeosol formation and down-dip equivalent marine incursions. Inset figure illustrates the main depositional facies and stratigraphic context of the Castissent Formation. Formation forms as incised-valley fill sandwiched between two marine flooding events and theit up-dip equivalents. Key: 1, Castissent Formation (in black); 2, fluvial and upper delta plains 3, lower delta plain; 4, fan-delta; 5, delta front; 6, slope and curbidites (Hecho Group) 128129 vise ousflood plain of channel deposits. Deposition within the valley is accomplished by vertical aggradation and lateral switching of braided channel networks resulting in highly connected sandstone bodies. The valley-fll deposits evolve from braided to more mixed-load, high-sinuosity channel systems and associated flood-plain deposits. This evolution in channel style is accompanied by a decrease in sand-body interconnectedness and increasing preservation of overbank and floodplain deposits in the succession. This erosional— aggradational cycle is repeated three times within each sandstone complex, with successively later phases of incision cutting down to the previous channel deposits to form progressively wider valley margins. Evidence for sustained, periodic flooding of the flood plain is manifested in the formation of hydromorphic soils and brackish water marl Multiple phases of alluvial aggradation and degradation indicate the operation of higher frequency allocyclic con- trols. The upward change in alluvial architecture can be interpreted in terms of a lowstand incised-valley fill, and downstream flood plain, overlain by fluvial channel and. floodplain deposits within transgressive and highstand system tracts, 7.5.2 Triassic, lishak Formation, Prudhoe Bay Field, North Slope, Alaska, USA Coarse-grained alluvial and fluvio-deltaic deposits of the Triassic Ivishak Formation form the principal reservoir facies of the Prudhoe Bay Field, North Slope, Alaska. The field is the largest oil and gas field in the USA, with original place reserves of about 22 billion barrels of oil and 42 trillion cubic feet of gas (Morgridge and Smith, 1972; Jones and Speers, 1976; McGowen and Bloch, 1985; ‘Atkinson et al., 1988). The Ivishak Formation forms part of the Sadlerochit Group, a clastic wedge of Permo- Triassic age that unconformably overlies carbonates of the Mississippian and Pennsylvanian Lisburne Group. The Ivishak Formation is sandwiched between shelf mudstones ‘of the Kavik Formation and shelfal and phosphatic car- bonates of the Shublik Formation (Fig. 7.13). The contact with the Kavik Formation is considered conformable ‘whereas the upper boundary with the Shublik Formation is sharp and associated with a phosphatic and pyritic pebbly lag. This boundary is interpreted as a disconformity de- veloped as a result of shoreface reworking of the Ivishak Formation during transgression. The Ivishak Formation varies from 400 to 700 t thick across the oilfield and is generally subdivided into two large-scale depositional cycles; a lower coarsening upward, flavio-deltaic and flavial interval of largely progradational character, dominated by mudstones, siltstones, sandstones and conglomerates, and an upper aggradational to retro- gradational cycle dominated by finer grained fluvial sand- stones and shales (Atkinson et al., 1988, 1990). Previous studies have suggested that the formation was deposited by 1 prograding fluvio-deltaic complex related to a braided- river-dominated fan delta or coastal braid plain (Jamieson 130 et ali, 1980; Melvin and Knight, 1984; McGowen and Bloch, 1985; Lawton et al., 1987). Within the Prudoe Bay field the Ivishak reservoir has historically been subdivided into four zones on lithostrat: ‘graphic and petrophysical grounds (zones 1—4, Atkinson cet al., 1988). Recent studies demonstrate the complex stratigraphic nature of the Ivishak Formation (Richards etal. 1994). The Sadlerochit Group can be subdivided into seven sequences based on detailed facies analysis of core data, sequence stratigraphic correlations of over 1300 wells and parallel evaluation of dip-meter and three-dimensional seismic data, Sequence boundaries are recognized by abrupt ‘changes of facies associations in cores, systematic variations in the internal sedimentary structures, thickness and the dogecc of amalgamation of channel-fil deposits. Four orders of sequences are recognized and referred ro here as first order, second-order, etc. In all cases, sequence boundaries can be tied to the maximum progradation point of lowet corder sequences. Note that the term order is used here for ‘ease of discussion, and no inferences should be made concerning the duration of the sequence as defined in section 2.2.4 (Fig. 7-13). First-order sequence At the largest scale, the Sadlerochit Group forms part of a first-order lowstand delta and fluvial braid-plain complex developed in response to a major relative sea-level fall. The base of the sequence conforms to the sequence boundary at the top of the Lisburne Group. The top of the sequence is represented by the major sequence boundary separating the Triassic Sag River Formation from the overlying Jurassic Kingak Formation (Hubbard et al., 1987). The earliest fluvio-deltaic deposits are predominantly progradational in character, suggesting sediment supply exceeded the rate of creation of accommodation. Overlying alluvial deposits dominate the remainder of the Ivishak Formation to forma thick, aggeadational interval of fluvial and flood-plain deposits. These topset deposits show an evolution with time from mixed-load fluvial and deltaic systems, to coarser conglomerate and sand-dominated fluvial systems. Gross changes in flavial channel style and stacking pattern reflect an increasing rate of creation of accommodation over time consistent with a relative sea-level rise, The end of Ivishak deposition is marked by a gradual fining of alluvial deposits and the eventual reworking of fluvial sandstones during marine transgression, leading to eventual establishment of Shublik Formation carbonate and Sag River Formation clastic shelf facies (Fig. 7.13). Second-order sequences ‘The first order sequence can be divided further into two second-order sequences by 4 major sequence boundary defined as the “6000” surface. The 6000 sequence boundary isa seismically mappable event (within the limits of seismic resolution), and marks a major change in the sand : shale‘Saisie ‘Sag River 3 ‘Shublik es i PRUDHOE BAY FORMATION [ste 3 i 3 ivesHaK ‘Tango “Areapraied tera chan S im | Romeo: = = sae cer aon g a en: ee Loot a 7 Fig, 7.13 Reservoir sedimentology, stratigraphy and equity subdivisions of the Ivshak Formation, Sadlerochit Group, Prudhoe Bay Feld, North Slope. The stratigraphic subdivision recognizes a hierarchy of key strata surfaces including (a) sequence boundaries (1000 series surfaces), (b) parasequence set boundaries or their up-dip equivalents (100 sevies surfaces), and (c) Parasequenee boundaries (intra-100-series surfaces) (the authors gratefully acknowledge the permission of the Pradhoe Bay Owners in the publication of this work by the Joint BPX—Arco—Exxon RAZOR team) ratio of the succession. It also defines the upper boundary toa suite of early Triassic faults, which are confined to the hase of the Prudhoe Bay reservoir. Palacocurrent analysis from core-calibrated dip-meter data also shows that the sequence boundary marks a major switch in the palacoflow. direction of lvishak river systems from NWSE below the 6000 surface to N—S above. The lower second-order se- quence displays a systematic change in architectural style, from predominantly progradational ro aggeadational to ddeltaic and fluvio-deltaic deposits consistent with increasing, fates of relative sea-level rise. The succession above the 6000 surface is fluvially dominated, and it is unclear whether the initial deposits above the sequence boundary represent the depositional products of flavial systems de- ‘cloped within lowstand or alluvial transgressive systems tracts. Succeeding fluvial deposits show an overall aggra- dational to retrogradational character consistent with «constant or declining sediment flux associated with increasing secommodation. Thitd-order sequences Three third-order sequences are recognized within the suc- cession based upon the recognition of basinward shifts in facies belts and changes in the stacking patterns in both marginal marine and fluvial deposits. Only the upper two sequences will be discussed here (Fig. 7.14). Sequence 2 ‘The sequence is defined at its base by the 8000 surface and at its top by the second-order 6000 sequence boundary. The earliest depositional units of the sequence comprise mixed-load to sandy deltaie systems (7400-7100), which prograded from northwest to southeast into the Kavik sea, ‘The earliest deltaic lobes are progradational in character, whereas later deltaic episodes show increased evidence of aggradation and the increasing preservation of fluvial topset deposits. These changes in deltaic style, combined with the increasing thickness and continuity of deltaic and flood- plain shales favour increasing sediment accommodation 131‘ i (uu) Lov anaisie anseauDsrV om ccc en eel LL LIL ila hese si Se a ae a =o Li, cae ‘dddldllldss waawo es 132“following a period of lowstand. The maximum progradation pointof the sequence corresponds to a third-order sequence ‘Foundary (7000 surface) separating fluvio-deltaic from alluvial deposits. The boundary is recognized in core data within the southern portion of the field by the abrupt serical juxtaposition of distributary channel facies on distal mouth-bar and lower delta-front deposits. In. she forthwest of the field the same boundary is defined by 4 change in channel stacking patter from distributary shannel and delta-plain channels to sandy fluvial systems. The boundary can be traced throughout the field and corresponds to a shift in gamma-ray log response and a thange in petrophysical properties. In the eastern portion of the Field, relatively thick flood- plain and lacustrine shale complexes dominate the upper part of the sequence, where they: form laterally extensive imtra-reservoir seals. These shales contain poor to moder- ately well-developed pedogenie horizons, which can be ‘correlated between wells (Atkinson ef af., 1988). Within the northeastern parts of the field, channel-fill deposits associated with lacustrine shales form isolated fining ap- ward units with low width-to-depth ratios (<1 :200), These channel types evolved from amalgamated tabular sand bodies within lower parts of the sequence. The overall width to depth characteristics of these channel systems together sith their close association with thick lacustrine and flood-plain deposits favours an overall evolution from moderate sinuosity to anastomosing channel styles associ- ated with very low alluvial gradients and periodic flood- plan flooding. Towards the southeast and central portions Of the field these isolated channel sand bodies give way to ‘more tabular, amalgamated channel storeys with higher wwidth-to-depth ratios (>1:200) associated with flood- plain rather than lacustrine shales. These facies compare favourably with the deposits of amalgamated sandy 10 nixed-load, moderate-sinuosity channel systems developed within the centtal part of the sequence. Both types of channel deposits can be traced to the northwest, where they give way to mulistorey fluvial channel sand bodies. The increasing development of thick lacustrine and flood- plain shales in che east of the field is paralleled by a decrease in the degece of channel sand-body amalgamation in the west. This may reflect increasing rates of creation of accommodation through time, The scale and extent of the shales towards the top of sequence 2 may further su that they represent the flavial equivalents of a maximum flooding surface in marine strata. Rates of alluvial aggradation decreased during the final stages of sequence development, resulting, in the west of the field, in an upward increase in the connectivity and amalgamation of channel sand bodies to form well- developed fluvial sand-sheets. This change in architectural style is less transitional in the eastern part of the field, and isolated channel sand-bodies are abruptly overlain by stacked, multistorey channel sand-bodies dominated by incompiete fining upward intervals and a lack of flood- plain-abandonmene fines (Atkinson ef al, 19883 Richards €t aly 1994), The major change in stacking pattern and degree of sandstone amalgamation marks the 6000 sequence boundary. Sequence 3 ‘The final third-order sequence commences with the 6000 sequence boundary and terminates in close proximity to the Pre-Shublik unconformity. The exact position of the ‘upper boundary of the sequence is unclear in view of the flavially dominated nature of the succession and the evi- dence for additional sequence boundaries within the suc- cession, The sequence exhibits an overall progradational to aggradational style. Sandy and mixed sand and gravelly fluvial deposits dominate the base of the sequence, where they form a well-developed sheet sandstone covering the entire extent of the field, The deposits give way abruptly to coarse-grained fluvial channel conglomerates and con- glomeratic sandstones. The junction betwee: sglomerates and underlying sandstones has been interpreted traditionally as a sequence boundary based on the abrupt change in grain size (Atkinson ef al., 1988). Although this clearly represents a period of fluvial incision, the significance of the surface may have been overstated in the past because similar coarse-grained facies occur within fluvial deposits underlying the boundary. This suggests that the magnitude of facies dislocation is limited and that it represents a higher order boundary. The coarse-grained fluvial deposits, are dominated by erosive-based, multistorey conglomerate bodies up to 10 ft thick, characterized by normally graded, horizontally bedded and cross-stratified conglomerates, Overbank fines are rarely preserved within the section, ‘The upper part of the sequence is dominated by finer grained, fluvial-channel sandstones, which form a broadly Uupward-fining suecession related to increasing sediment accommodation. These deposits reflect the products of sandy, distal coastal plain environments (Atkinson et al. 1988). Within the western and caster part of the field, relatively thick shales occur as fine-grained flood plain and lacustrine intervals of similar character to the shales of, sequence 2. The widespread navure and thickness of these deposits may reflect periods of reduced alluvial gradients and enhanced flood-plain aggradation associated with rising, base level. The top of the sequence may pass abruptly into the carbonates of the Shublik Formation or be overlain by thin, transgressive shelf sands of the Eileen Formation. In summary, the stratigeaphic record of the Ivishak Formation illustrates the complex response of alluvial and fluvio-deltaic systems to changes in sediment accom modation, Recognition of sequence boundaries and the up- dip equivalents of flooding surfaces is complicated by the lateral and vertical variability of flavial and deltaic deposits within the succession, Changes in sediment accommodation are reflected at a variety of scales, from the evolving patterns of channel sand-body amalgamation and shale development to changes in channel pattern and style as the fluvial depositional systems attempted to achieve equilibrium. 133CHAPTER EIGHT Paralic Successions 8.1 Introduction 8.2 Parale depositional systems 844 Climate 82.1 Deltas 8.2.2 Delta physiography 8.5 The sequence stratigraphy of distinct 8.2.3 Deltas and sedimentary processes 8.2.4 Coastal plan to shorcline—shelf 8.25 Estuaries paralic systems 85.2 Thestrat 85.3. The st 8.4.2 Accommodation space 843 Grain size Western Canaan Basin 8.8.1 Steatigeaphie traps 88.2 Seals 88.3 Analogues and flow units 8.5.1 The stratigraphy of coastal plain to shoreline—shelf systems ray of delat systems igraphy of escuar 8.9 Paras systems ata seismic scale 4.9.1 Scismicscale models 8.9.2 Scismic resolution 8.933 Seismic facies 8.3 Sequences in paralie successions systems 83.1 Sequence boundaries and valley incision 8.6 Correlation procedure 8.10 Variations in paralic systems within 2 8.3.2 Imerfluve sequence boundaries 83.3 The eranspressve surface 83.4 Forced regressions 853.5 The masimum flooding surface 8.4 Parasequences in paralic successions BALL Sedimentary processes 8.7 An exampl 8.1 Introduction Paralic successions include a wide range of environments — deltas, coastal plains, shoreline—shelf systems and estu- aries — each deposited at or near to sea-level (Table 8.1). ‘As such they are extremely sensitive to changes in relative sea-level and are therefore particularly suitable for high- resolution sequence stratigraphic analysi The chapter commences with a summary of different paralic depositional systems, and a discussion of hhigh- resolution sequence stratigraphy. It proceeds by showing, {i) how stratigraphic signatures vary in distinct paralic environments; and (ii) how sequence stratigraphy can be applied to paralic petroleum reservoirs. The chapter con- cludes with an analysis of the seismic expression of paralic successions, and how paralic successions may vary during a sea-level cycle 8.2 Paralic depositional systems 8.2.1 Deltas ‘A delta is the prism of sediment that accumulates where a river enters a standing body of water. Deltas comprise a 134 8.6.1 Depostional environments 8.6.2 Parasequence correlation 8.63 Progradation, aggradation, and retrogradation 8.6.4 Sequence boundaries the Viking Formation, searlevel cycle 8.1041 Shelf processes 1.10.2 Spatial changes in processes 8.10.3 Sequence stratigraphic framework 8.11 Summary ‘Table 8.1. Environments and subenvironments encountered in paralic successions Deltas Coat pain to shoreline shelf systems Incied valleys Subenvironments Distributary channels Distributary mouth bars Crevasse channels Crevasse splays Lagoons Lakes Chenier: Beaches Shoreline-shalfcompleses Tidal ridges Tidal channels Flood and ebb tidal deltas Bertie islenes Bay-head deltas Tidal fats, ete«rial portion and a subaqueous portion. The subaerial tion, the delta plain, can be commonly divided into two the upper delta plain, which is fluvially dominated, the lower delta plain, which is marine influenced, ticularly by tides (Fig. 8.1). The following two sections uss variations in delta physiography, and the influence sedimentary processes on deltaic systems. 8.2.2 Delta physiography Shelf-edge deltas (or deep-water deltas) Shelf-edge deltas are located at the shelf—slope break. They pass directly into an extensive coeval slope (with an angle of 2=5°) and deep-water sedimentation system (Fig. 8.2a). Shelf-edge deltas commonly exhibit a range of syn- sedimentary deformational features, such as growth faults, slides and mud diapirs. These features are gravity driven, ‘Sseatovel ‘She ta = hel, si me 1008 m Fig. 81 Delta terminology (after Coleman and Prior, 1982) 135aind result from the rapid accumulation of a thick pile of soft sediment fronted by an unconstrained slope. Shelf deltas (or shoal-water deltas) Shelf deltas develop in shallow water depths, commonly 30-70m, landwards of the shelf=slope break. The sub- aqueous portion of shelf deltas comprises a. relatively coarse-grained, steep part (angles of 1-2"), known as the delta front, and a fringing muddy, lower angle section (angles <0.5°), known as the prodelta (Fig. 8.26). Shelf deltas lack large coeval slope and deep water systems, and, in general, are devoid of large-scale soft-sediment deformation features. Gilbert deltas Gilbert deltas are coarse-grained fan-deltas characterized by steep delta foresets (> 20°) that are dominated by sedi- ment gravity flow processes (Fig, 8.2c; Colella, 19885 Braga et al., 1990). Foresets range from subseismic to scismic in scale (up to 700 ms Ori, 1987). Gilbert deltas are common in rift and strike-slip settings, where subsidence and uplift combine to generate the requisite basin-margin ‘water depth and alluvial fan eatchment areas. If the water depth is 100 shallow, rapid progeadation will result and steep forescts may not develop. IF the water is too deep, sediment may not accumulate as a subaerial fan, but pass. directly t0 the basin floor to accumulate as a submarine (@) SHELF EDGE DELTA (by SHELF DELTA ‘Ate deta (@GRBERT DELTA 220" 136 Drowned deta plain deposit, Faultinduced slumping may characterize the early portions of Gilbert deltas while silty foresets up to 10m thick tend to increase in thickness and abundance basin: wards. The foresets are commonly inversely graded, 8.2.3 Deltas and sedimentary processes Fluvially dominated deltas Deltas with a strong fluvial signature in the delta front result from high sediment input and relatively low-energy shelf processes. They are commonly characterized by @ sheet-like sediment body comprising coalesced mouth bars The core of each mouth bar, adjacent to the distribucary channel, is likely to be sandy. However, depending upon the calibre of the supplied load, sediment deposited away from the channel axis is likely to be increasingly mud prone. Asa result, a finger-like sandstone isopach develops in the lower delea plain and shallow delta front (Coleman and Prior, 1982). The delta plain of flavially dominated deltas takes on the character of the fluvial system. Three main types are recognized: rivers, braidplains and alluvial fans, which feed river deltas, braid deltas and fan deltas respectively (Orton, 1988). Wave- and storm-dominated deltas, Deltas characterized by strong wave energy have relatively straight coastlines, and shore-parallel sandstone isopachs Sesovel Signicant coeval deep water deposts rectly fectrom the dota Sealovel cooval deep-water Seposts the product ot slope degrading Fig. 8.2 Delta types: (a) shelf edge, (b) shelf, and (c) Gilbert deltas,skewed in the direction of prevailing longshore currents. Fairweather waves move sediment onshore and provide an effective barrier to its offshore transport. Storms redistrib- ute the sand on to the inner shelf to increase the shore- perpendicular width of the coastal sands. Tide-dominated deltas Strong tidal currents produce numerous coeval channels at the delta front and result ina finger-like irregular sandstone isopach (Coleman and Prior, 1982). Depending on tidal strength, a tidal signacure may penetrate deep into the delta plain, favouring the development of tidally influenced lagoons, tidal flats and creeks. The strongest tidal currents are likely to be in the upper portion of the delta front and in the lowermost delta plain, where daily reworking can produce clean-swept sandstones. ‘The subaqueous delea may be extensive and grade into adjacent tidally influenced shelfal areas, providing one of the few fairweather mec! anisms for offshore sand transport. 8.24 Coastal plain to shoreline~shelf systems Goastal plain to shoreline—shelf systems lack major rivers. They receive the majority of their coarse sediment through longshore and along-shelf transport, largely from coeval deltas. Shelves are shallow, gently sloping, open-marine areas characterized by wave, storm, tidal and, rarely, oceanic, currents. Shelves pass landwards into the shoreline, which comprises the shoreface and the beach. The shoreface is asteep, narrow zone (commonly less than 1km wide), characterized by structures produced by shoaling waves (Shepard, 1960; Bernard and LeBlane, 1965; Friedman and Sanders, 1978). The beach extend from the mean low- tide mark to a supratidal zone that is only periodically influenced by marine processes, for example during storms. The beach may be backed by aeolian dunes. Shoreline shelf systems pass landwards into the coastal plain, which iscommonly fronted by a lagoon, Small rivers that debouch into the lagoons form bay-head deltas. 8.2.5 Estuaries Estuaries are drowned river valleys (Dalrymple et al., 1992}. They are characterized by sediment input from both fluvial and marine sources. At their heads, estuaries are characterized by fluvial input. Where fluvial input is strong, els become ively strong, fluvial ch progressively tidally influenced and merge downstream ito tidal channels. Estuary mouths range from tidal to vwave-dominated, allowing estuaries to be divided into two road groups (Dalrymple et al, 1992). In wave-dominated estuaries, along-shore and onshore sand transport results in the accumulation of a sandy plug. at the head of the estuary (Fig, 83a). The sand plug commonly comprises two elements, (i) a barrier with associ- ated washover deposits, and (i) a tidal inlet that breaches the barrier allowing tidal exchange with the estuary and forming a flood tidal delta. Landwards of the sand plug, the central portion of the estuary is a low-energy zone characterized by muddy facies. Together, the sand plug, the central-basin muds and the bay-head delta form a tripartite subdivision of wave-dominated estuaries In contrast, strong tides ensure an active exchange be tween the estuary and the open sea, preventing the develop- ment of bay-head deltas, a muddy central basin ot a discrete sand plug (Fig. 8.3b). Instead, in tide-dominated estuaries, a series of tidal sand bars ornamented by dunes, ate likely to pass up-system into sandflats, and then to tide- influenced and finally fully fluvial channels. Although a ‘maddy central basin is not developed, an analogous zone of relatively low energy is recognized (Dalrymple et al., 1992). It occurs landwards of the strongly flaring tide. dominated sandy estuary and is characterized by a systematic change in channel types ‘straight~meandering— straight’, reflecting changes in the balance between fluvial and tidal process. At the seaward end an outer straight section is tide dominated, with net headward sediment transport, At the landward end an inner straight portion is fluvially dominated with net seaward transport. In between, a zone of net bed-load convergence is characterized by tight ‘meanders and fine grain sizes. 8.3. Sequences in paralic successions 8.3.1 Sequence boundaries and valley in In paralie successions sequence boundaries are recorded by basinwards or ‘downwards’ shifts in facies belts, Two important signatures are associated with base level fall: (i) valley incision and (ii) forced regression, Valley incision The response of fluvial systems to relative sea-level fall is discussed in detail in Chapter 7. Ifthe new river course is steeper than che equilibrium river profile, the river will firstly straighten its coarse and then incise to form a valley Incised vall eys are common features in paralic successions (Van Wagoner et al, 1990). They are important because they represent unequivocal evidence ofa sequence boundary and also because they can form stratigraphic traps for hydrocarbons. ‘The recognition of incised valleys With limited well or outcrop data it is often difficult to distinguish amongst (i) individual channels (e.g, fluvial or tidal distributary channels); (i) mattistorey channel sand- stone bodies, and (iii) incised valleys. Both valleys and channels ate sharp based and commonly filled by fining, By@ FLOODING SURFACE “ Eos muna cuanne y covthiac ——RAMENENT ; osraurany 1 ose ; wreroista- ) PF BouNary 3 Soar ea & © fopfemmmnn | icc) £ ESD] comamecnoure ruvat SEED camer J * sel] cena easin ruova ALLUVALCRANNEL Fig, 83 (a) Wavedominated estuary upwards successions. Similarly, both valleys and multi storey channel sandstones may comprise two or more stacked-channel sandstone bodies. However, some key features can allow these distinct sediment bodies to be distinguished: 1 Valley fills are commonly wider and thicker than chan- 138 PE S SS ™ Key te section Bay-hoad delta Barrer snoretace, fad food ta [Aviat dopasita Contral basin rer snot aaine © +5) prognapwe MARINE vue [2/_Shonerace The, REGAZE TRANSGRESSIVE SHELF at FOS swmggune nels. On average, valleys have width: thickness ratios in the order of 1: 1000 or more, whereas channels are nat- rower, with width :thickness ratios of 1: 100 (Reynolds, 1994b; Table 8.2). 2 Valleys can be inferred where deep incisions (deeper than the thickness of individual channels) can be recognized,» loading sta 5 sana 4 vagy ‘ o 5 LE, wen serace 1 a ron Au F sue [2S] eaten Fuuvian FS auuuna, ‘Trangreasive shelf deposits dal sand bars UFR sanctiats Tisal meanders Eg] Inner steaight tidal ~ fava EERE] Atuvial ceposts Key to loge uaine x) SNUB TM Fader TRANSGRESSIVE ¥ fe GO TEENS Ta el ee ee ‘ “= Tidal lamination : UFR plane bed ; meer =e er nue PF eeurtaco iii tegmety a Seems ig, 83 (b) Tide-dominated estuary (after Dalrymple e¢ al, 1992). TMF, time of maximum flooding; RS, ravinement surfaces Sh, sequence boundary; FS, flooding surface; UFR, upper flow regime Commonly, these incisions take the form of terraces. 3 Valley incision is often heralded by the generation of accommodation space at progressively slower rates, and followed by a phase during which accommodation is gen- “trated progressively more rapidly. Such progressive changes ate lkely to be recorded in the rock record, leaving signals that indicate proximity to a sequence. boundary. For example, where the rate of generation of accommodation space is decreasing, the thickness of parasequences and the volume of crevasse splay deposits are likely to progressively, ‘decrease, whereas channel connectivity is likely to increase. 4 The facies that fill incised valleys record a basinwards, shift in facies belts. Two types of valley fill are common in paralic successions: estuarine valley-fills, and_valley-fills 139“Table 8.2 Summary statistics of paral sandstones (after Reynolds, 1994b) ‘wits tm) Lenaths Gn Thickness (md —— —— — numberof Mean Range Moan Range Mean Range datapoints Incied valleys 9850 «3000-500 303 215281 Fluvial channels 750 1400-60 9 2025 6 Distiutary channels S20 3800-20 73 04 268 rexase channels © 400-5 2a 17-02 4 Shoreline-shelf sandstones Undifferentiated 25000 106 000-1600 93200 90000-47000 191 4927 er Highstand systems tact 16400 43000-16000 36 “ranugressive systoms tact 7200 20000-3300 5 Distibutary mouth bars 2900 14000-1100 6500 ‘9600-2400 97 aie 26 Tidal idgee 400 29300-750 17500 61000-3250 m2 p93 3 Ceovase splays 800 ‘700-20 5600 1700-160 14 12-03, a Flood tidal delta complex 6200 13700-1700 12300 25700-2800 67 218 3 Lower tidal fat 1000 1350-400 45 9-20 1" Tidal creeks 810 1550-160 52 181 5 Tat inlet 1850 2550-700, 2300 48300 48 37 3 Estuary mouth shoal 2400 2300-1700, 3750 700-2200 10 10-35 4 Chenier 3650 ‘400-900, 21900 38600-29005 8 7-46 2 All sands 50501080005 35300 190000. 160 7 7 6a that evolve from fluvial to estuarine fills. Marine mudstones may also fill incised valleys (Fig. 8.4). 5 As a result of river rejuvenation, incised valleys com- ‘monly contain the coarsest sediment available locally 6 Hag etal. (1988) suggest that phases of eustatic sea-level fall are unlikely to exceed 100m in magnitude. Therefore, valleys produced by eustatic sea-level fall are unlikely to be deeper than 100m. This may help to distinguish valleys from slope canyons and gales. Incised-valley patterns Incised valleys in the rock record have highly variable pattems, Regional studies typically reveal patterns compar able co modern tributive drainage patterns (e.g. Dolson et al, 1991). By contrast, detailed mapping over relatively restricted areas commonly shows complex geometries characterized by short cuts between adjacent valleys, vary- ing valley widths and a great range in the spacing between adjacent valleys (eg. Jennette et al, 1992). As with all fluvial systems, valleys are, in general, aligned down the palacoslope. Where the basin floor is exposed during sea- level fall, the orientation of incised valleys may be perpen- dicular to flavial and distributary systems in the preceding hhighstand systems tract 8.3.2. Interfluve sequence boundaries Valleys are restricted features separated by interfluves (Fig. 8.4). During falling sea-level the interflaves are expos subaerially and subject to erosion and pedogenesis. With 140 relative sea-level rise the valley aggrades and the interfluves are progressively onlapped. In paralic successions, valleys are incised into the pre-existing coastal or delta plain, and, therefore, the interfluves are likely to be flat topped. When 1 valley has filled to the level of the old coastal plain, the next increment of relative sea-level rise will flood the interfluves generating a proportionately large volume of accommodation space, and causing the shoreline ro trans: ‘gress rapidly. Asa result, many interfluves are characterized by a sharp erosion surface, which is the product of ravine- ment, and ate overlain by a thin transgressive lag, In many ways this signature resembles a simple marine flooding surface. However, intesfluve sequence boundaries ‘may: (i) be underlain by well-developed palacosols indicating prolonged exposure; and (ii) be capped by anomalously ‘coarse lags, comprising clasts that were introduced to the basin by rivers during lowstand, but are not present in the underlying highstand successions. 8.3.3. The transgressive surface The transgressive surface is. the first significant marine flooding surface across the shelf within a sequence (Van Wagoner et alu, 1988). The surface defines the top of the owstand systems tract and the base of the transgressive systems tract. Commonly, the transgressive surface is co- incident with the interfluve sequence boundary described above. It will also cap valley-fill deposits (Fig. 8.4), In delta plain and coastal plain successions it is often difficult to recognize the landwards equivalent of the trans- gressive surface. As a result, although valley fills andValley led by Estuarine Fit ® marine mudstone valoy vay mores: Incised valleys » / ie 2 sents ¢ 7 $ ° 70 1 ® 2 5 8 4 Kor Prades sear sntone het mctne cy end deta — sequence today ene — Tretrnogrnae sre [EB 20209 noumstot Fig, 84 Variable expression of sequence boundaries in paralic successions: (a) a plan view and (b) and (c) representative cross- sections (after Van Wagoner et ul, 1990) Frooding surface maximum flooding surfaces ean be recognized, it may be impossible to divide intervening successions into lowstand, and transgressive systems tracts (Fig. 8.5). Forced regressions are basinward movements of the shore- line, caused by relative sea-level fall. They are inde- pendent of sediment supply (Posamentier et al, 1992; 8.3.4 Forced regressions 141LUTHOSTRATIGRAPHY HEATER ForMariON : : Fig. 85 (continued opposite) Wireline and associated graphic log of the Brent Formation in well 211/18a-7 in the Thistle fel, North Sea, A sequenee boundary is inferred at the base of the Etive Formation, where a multistorey succession of fluvial channel sands lie above upper shoreface sands, and a maximum flooding Surface is recognized in the middle of the Ness Formation, between retrogradational and forestepping sets of bay-head deltas. However, there is no clear transgressive surface. Asa result the intervening succession cannot be divided into lowstand and transgressive tracts (from Reynolds, 1995). Lithology: C, clay; S, silty F, fine; M, medium; C, coarse Posamentier and James, 1993) and are expressed by basin: wards or ‘downwards’ shifts in facies belts. Typically, shoreline sands sharply overlie outer shelf mudstones. The sharp contact reflects the re-establishment of a sea-floor profile in equilibrium with shoreface and inner shelf pro- 142. ‘cesses, prior to sand deposition and renewed shoreline progradation. Two types of lowstand shoreline are recog- nized: (i) attached shorefaces that overlie sands of the ‘underlying highstand systems tract, and (ii) detached shore- faces that are isolated in offshore shales (Fig. 8.6; AinsworthBig. 85 (continued) and Pattison, 1994; Fitzsimmons, 1994). In addition, two ‘expressions of the sequence boundary are recognized. In the first the sequence boundary rapidly dies out basinwards imo a correlative conformity (e.g. Plint, 1988; Posamentier and Chamberlain, 1993). In the second the sequence bound- ary remains a sharp erosive surface for 10—20km basin- ‘wards ofthe lowstand shoreline, and is always overlain by a distinctive, commonly gutter cast, sand (Fitzsimmons, 1994). Key InrpiatonFacles s\.,. Sach Be arate sesso ss en rowel «| (teem ees a Eo s ‘Swatey facies £ bp Hurmoce eos taties | For #) fa elle i i=. Sent feet nto oat | s, sty sandstone: ‘sanostone se ‘itetone FS) andes 86 8B ——— sequence botany MES — — ~ toma tooang snes Fs Foding sacs To date, forced regressions have been recognized largely in wave- and storm-dominated successions, where they produce shore-parallel sand isopachs (e.g. Bergman and Walker, 1988). However, sea-level fall also occurs in suc cessions dominated by tides or semi-permanent currents. Reynolds (1994a) argues that a series of extensive undu- lating erosion surfaces in the Viking Formation of Alberta, Canada, were cut subaqueously by tidal scour following sea-level fall (section 8.6). 18Highstand sys tract a ~ Highstand stems act Highstana sylems tract, Na. Detached lawstandshorstaca ‘eneated in olshore mudstone Sheta sits ane muss Ta] Peal its “255 shoretace to inner she sande Sequenes boundary = covtetatve contri, Fig. 8.6 Two types of lowstand shoreface: (a) attached and (b) detached, and (c) a set of lowstand shorefaces, of forced regeessive set’, Note that tis is only one example of a forced regressive set. The sand bodies could be spaced more closely or connected (after Ainsworth and Pattison, 1994; Fitzsimmons, 1994) 8.3.5 The maximum flooding surface In paralic successions, the maximum flooding surface is, coeval with the most landwards position of the shoreline. ‘Commonly, the maximum flooding surface is underlain by a retrogeadational parasequence set and overlain by a progradational parasequence set. In many cases, however, itis difficult to recognize a single, discrete surface. Instead, ‘maximum flooding zone’ is recognized. Within this zone, two or three discrete surfaces may be candidates for a ‘maximum flooding surface’. Basinwards of the shoreline, maximum flooding zones are represented by outershelf sits and mudstones. Candi 144 dates for the maximum flooding surface are marked: (i) by the finest grained deposits; ii) by evidence for condensation, such as firmgrounds; or (ii) by outer shelf carbonates (Fig. 4.3). High gamma-ray values, and high organic matter content reflect anoxic bottom water. These are common during transgression and in the lower portion of maximum flooding zones. (On the delta plain, maximum flooding zones are charac- terized by tidal influence in distributary channels (Shanley and McCabe, 1993), pronounced channel erevassing, lake expansion (Atkinson, 1983) and generally werter palaeosols. In lagoonal successions, discrete maximum flooding sur faces may be recognizable between backstepping and fore-uepping bay-head delta parasequences (Fig. 8.5). 16 Parasequence sets Farasequence sets record the net movement of the shoreline e two oF more parasequences. Their wireline log ex- ession, and relation to key surfaces and systems tracts, been illustrated in previous chapters. In core and in atcrop, progradational and retrogradational parasequence n§ are expressed as systematic changes in sedimentary 1s, ichnofacies and grain size (Figs 4.3 and 8.7). 84 Parasequences in paralic successions detailed character of parasequences is. extremely and is controlled by a range of interconnected sedimentary process; (ii) the calibre of the plicd sediment; (iii) accommodation space; and (iv) linate. Each of these factors is discussed below. The following section, section 8.5, shows how they affect the tailed statigeaphy of coastal plain to shoreline shelf stems, deltas and estuaries. J Sedimentary processes general, as fluvial, wave and storm-dominated shore- lins prograde they each produce a coarsening succession. jowever, the lateral and dip extent of the associated bodies, and the detailed sedimentary structures, depending on the dominant process (Fig, 8.85 Table 2), Shoaling waves redistribute sediment alongshore and straighten the plan-form of shorelines, As a result, ive: and storm-dominated shorelines produce tabular, hore-parallel sandstone bodies with a spatially uniform tratgraphic signature. Shoaling waves also generate an hore-directed current that prevents offshore sand trans- By contrast, storms can transport sand to the inner hell, so that storm-dominated sand bodies can extend inher offshore than wave-dominated sand bodies (Fig. .8a,b). In fluvial- and tide-dominated settings the sand le ratio, and the grain size will vary across the shoreline ortion of a parasequence, reflecting its composition by a amber of coalesced mouth bars (Fig. 8.8c). 34.2 Accommodation space Ryo models relate parasequences in shoreline successions delta plain and coastal plain successions. Van Wagoner d al, (1990) show coastal plain mudstones developing. gontemporaneously with shoreline progradation (Fig. 8.9). By contrast, Devine (1991) shows strandplain and lagoonal successions developing, respectively, during (i) progradation, sand (i) aggradation and transgression (Fig. 8.10) Both models are supported by detailed field-work, they ‘ach reflec distinct rates of generation of accommodation space. In Devine’s model no accommodation space is gen- crated during regression, resulting in toplap, whereas sig: nificant accommodation space is generated during barrier island aggradation and transgression, resulting in a di matic expansion ofthe lagoon. By contrast, accommodation space is generated throughout strandplain regression in the model of Van Wagoner ef al. (1990), allowing the syn- chronous deposition of coastal plain deposits. Partitioning of sand-body types into systems tracts Analysis of coastal plain and deltaic sand bodies shows, (Table 8.3) that the overwhelming majority of sand bodies, deposited landwards of the shoreline (distributary channels, crevasse splays, tidal ereeks, and flood tidal deltas) occur in transgressive systems tracts (Reynolds, 199%4b). The data indicate that accommodation space in the delta plain and coastal plain is generated largely during transgression, and that certain sand-body types are favoured when base level is rising rapidly. For example, flood tidal deltas develop in lagoons, which are favoured during transgression. Similarly, base level rise may also favour crevassing. Sand-body dimensions Recent studies have recognized the cyclic stacking of sequences into sequence sets and composite sequences (Mitchum and Van Wagoner, 1991; Jones and Milton, 1994). The low-frequency changes in accommodation space and sediment supply that generate these patterns also Table 8.3 Data to show the preferential partitioning of sand bodies into particular systems tracts. LST, lowstand systems tract; TST, transgressive systems tract; HST, highstand systems teact; SMST, shelf-margin systems tract; %, the percentage of a given sand-body type in a given systems tracts N, the total number of sandstone bodies of a given type for which the systems tract is known systems vad % " Valleys sr 100 a7 Tidal reeks 1st 100 3 Tidal fats sr 100 4 Shoreline shelf ust 5 “8 1st ” sr o Distibutery channels Ast 2 ® 1st 6 ust 8 ‘crevasse splays 1st 100 6 Flood vidal deltas 1s 92 2 st 8ovpins Fuipooy wince *SgWy soojuns Suipooy “sq “(PFGE “SpIousoy Uo 1-9 19g) Yoneaoadionay sup suzoddns usddews jeuosox — Kaepunog aouanbos aanyaowwl Ue aq ew 29E}ANS Zaypooy pasys ays aexp aaLHpUH sass220Id aeauowipas ur a8ueqp oyp pue Be] ayp Jo ssoupry oy “SouDNRS KseAUOUIPs pu 921s urEAH UI aHUCYD SHEUIESP v ypur ‘SswuNWopaid [O66 | “S>IYSY suas) Suyppaq,ssou9 aye9s-sunp doa sa speayor nq “ouonbssesed say D4 O14 BuiKpoxo ay aseq sayy “spurs paunes-tumpsus pasoneds 910u q PUP aIUOD suospUes Ut ssv>x29p pulods v Aq PayzeUE st 39e}2Ns Supooy puodas y -oouanboseaed vu epesoid pur Surmojjeys axe>yput‘spaq-ssox> passousnqun jo uonvasssoud ay Uf aSeOI>U UE Yala Joypolfoa “SaBuEY> say “StHEO, Avoasig feONION H jeauoztaoy wos Sluey> © pur tuaiUOD auorspueS UF IseD!OUI Spray & SI DO} spuemdp] “SUILIB PUES wIMIpaus poxoNIeDs 44 pazuNDEayD pur poregimrorg s| 39";1ns Supooy aq “29epI0s Fupooy as1y a4p sYseu! aUDIUOD auoIspuLs ul asea;o9p admAqe UY“ sfuons ve soieaiauad 2100 ap Jo oseq aq, Sumas ooys-puvs jepa v ul pousodap ‘saouanbasesed aajduros ona smo4s EpLUL “CHIDgTY VOREUIOY HULA ayp UL 2409 Y Z'g “SH te og 8 7 7 | z S 3 = 8 age 5 7 3 3 a ‘STORM DOMINATED = “IDE DOMINATED | H 3 z zg 2 & 2 8 FI — 7 > ne PROGRADATIONAL PARSEQUENCE SET ge B 2 SSS Be | 8 7 ——___| ee | @ 8 e e gee B Eee a5 ital Bs 7 3288 8 Ye‘300 m 10-20 kn (ay wave-pomnaTeD A. Shorelace SHORELINE SHELF . Inner shalt iste shel (b) STORM-DOMINATED SHORELINE SHELF (@ DELTAS ‘SEDIMENTATION Fig. 88 Variations in sand distribution within parasequences sharacterized by distinet sedimentary processes. Wave-dominated (a) and stotm-dominated (b) parasequ are ch leaning and coarsening signatures that extend in three dimensions. They differ in that the sand belt extends much further ftom the shoreline in storm-dominated successions. By contrast, deltaic parasequences (c) are characterized by mnarked spatial changes in sandstone acterized by simple 147(@) LIMOSTRATIGRAPHIC MODEL COASTAL PLAIN MUDSTONES AND SANDSTONES ST TRTRE comarun \ [Sf a 7 ai { { 15 | | | COAL I Litt l Fig. 8.9 Lithostratigraphic and chronosteatigraphi models of parasequence development based on the work of Van Wagoner etal. (1990), Coastal plain mudstones and sandstones are shown to be coeval with shoreline progradation. For discussion see text. influence sandstone body dimensions on a parasequence scale, For example, the thickness of shoreline shelf sands decreases upwards through highstand sequence sets, Simi larly, the dip extent of shoreline—shelf sands tends to be twice as extensive during highstand systems tracts as in transgressive systems tracts (Table 8.2). 8.4.3 Grain size Coarse sediment (gravel and sand) favours: (i) low-sinuosity channels in the delta and coastal plain; (ii) well-drained delta and coastal plains that lack lakes; (iii) the rapid deposition of river load at river mouths; (iv) a steep ‘reflec- tive’ shoreline thar receives the full effect of wave energy’ and (v) a steep delta front characterized by mass flow processes (up to around 25° in Gilbert deltas. By contrast, fine sediment (silt and mud) favours: li) high-sinuosity channels; (i) poorly drained delta and 148 coastal plains characterized by lakes; (i) sediment dispersal beyond river mouths ay a result of buoyaney processes (iv) a low-angle shoreline that attenuates and dissipates wave energy, and forms a sandy shoreline and inner shelf zone; and (x) a low-angle delta front (around 1°). ‘The dominant grain size of deltas reflects the nature of the catchment area: the area and physiogeography of the catchment; the prevailing climate; and the bedrock ith ology. These factors, in combination with tectonic activity and relative sea-level fluctuations, may change through time to generate fluctuations in sediment flux and calibre For further discussion on the influence of grain-size, see Orton (1988) and Orton and Reading (1993). 8.44 Climate ‘On the delta and coastal plain climate influences the nature of lacustrine deposits, palacosols and the development(@) LTHOSTRATIGRAPHIC MODEL ‘Subset aus Regressive deposts ‘Standplain Haws coasts isin Fig. 8.10 (continued overleaf) Lithostratigraphic and chronostratigraphic models of parasequence development based on the ‘work of Devine (1991). Coastal plain mudstones and sandstones are shown to be coeval with transgression. For discussion see text 149(0) (ou prasoqunces shown D)-@) = 12m thick sequences Fig. 8.10 (continued) 150(b) RETROGRADATIONAL PARASEQUENCE SET Tidal strm or oceanic cureniseworkbaror Sand int shel andelone tS | nner ent ue snat Pema sce tees Fein See tee ey. Bed | aaa rate | Fig. 8:11 (a) (continued overleaf) A progradational parasequence set in a storm-dominated shoreline —shelf succession. The work of Devine (1991) and Van Wagoner et al. (1990) has been combined to show coastal plain and lagoonal successions deteloping during progradation and transgression. The map shows the extent of shoreface sands in each parasequence 1~4. The short ogs (A) and (B) illustrate spatial variations in parasequence set signature that result from lohate parasequences. A flooding, surface lag is more likely if the underlying shoreface has progeaded past a given location, as only then can shoreface erosion Jwinnow and concentrate coarse material (based on Hamblin and Walker, 1979; McCrory and Walker, 1986; Plint and Walker, 1987; Eyles and Walker, 19885 Plint e¢ al, 1988). (b) A retrogradational parasequence set in a storm-dominated shoreline—shelf succession, During transgression, the shoreline may aggrade and keep pace with relative sea-level rise, so that significant accommodation space is generated in the coastal and alluvial plain. With an inerease in the rate of relative sea-level rise, ot a drop insediment supply, the shoreline will jump landwards to a new position where these two variables are in balance (for a review see Eliot, 1986b) of coals, evaporites and lacustrine carbonates. The wind regime has an important effect on the efficacy of storm processes, while the prevailing winds determine the extent 8.5 The sequence stratigraphy of distinct paralic systems of aeolian (coastal) dune fields, the direction of longshore The fundamental controls outlined above affect the devel- sediment transport, and the presence of semi-permanent opment of all paralic suecessions. However, they combine shelf currents. Climate also influences the development of in different ways in different paralic systems. In many ways shelf carbonates. coastal plain t0 shoreline—shelf systems are the simplest. These are discussed and illustrated in detail below. Sub- sequent sections show how stratigraphy differs in deltaic and estuarine successions. 1st‘Washove tans, tod tial tas ana goons may be ‘moomtart but are bly 1 be Dimensions tm) Wm) length (km Patison 1968 0-6 0-5 60 Bergan and Walker 1@88 0-10 O-5. 10 Legos on ‘Shorelace attached detached ines 1Omhigh;2~3 km wide; 265 km spar; 10s km ing: average 22 “Io coast Stopes of afew dogrees or less Transgressie tag No vous verte Fig. 8.11 (c) A transgressive parasequence set in a storm-dominated shoreline—shelf succession. Surface 1 is cut by shoreface erosion. The sand body, 2, is the product of short-lived progradation, All evidence for subaerial exposure is removed by ravinement, producing surface 3. Ancient examples suggest that the preserved sediment bodies are linear and shore- parallel (Bergman and Walker, 1988; Pattison, 1988). Modern shelves suggest the presence of numerous shore-oblique ridges (¢ Rine et al., 1986) and ragged patches of sand above transgressed si 8.5.1. The stratigraphy of coastal plain to shoreline~shelf systems Storm-dominated parasequences In general, the lower part of storm-dominated para- sequences (shelf and lower shoreface) comprises a series of shatp-based beds 5-30. thick that become thicker and. progressively amalgamated upwards (Fig. 81a). The beds are often hummocky cross-stratifed, with bioturbated, ‘wave-rippled tops. The upper shoreface may be: i) storm dominated and characterized by swaley facies (McCrory and Walker, 1986); (i) wave dominated, with or without 152 successions, W, width; t, thickness bars; or (ii) cut by cross-bedded tidal channel units, Barred shorefaces are characterized by erosively based, trough cross-beds, deposited in longshore troughs and/or rip channels. Non-barred shorefaces are characterized by wave ripples, onshore and offshore directed cross-beds, and planar laminae (for a review see Elliott, 1986b). It is critical, but not always easy, to distinguish channels gener- ated by tidal inlets and longshore troughs, from incised: valley fils. Coastal plain successions are strongly influenced by climate, and are characterized by small fluvial systems, Lagoonal deposits are generally shale prone, but may also ‘comprise sandy facies deposited in storm washovers, flooddeltas, or bay-head deltas (Plint and Walker, 1987; ine, 1991). In shelfal locations the stratigraphic signature of pro- tional, storm-dominated prarasequence sets is generally comprising a stack of cleaning and coarsening, equences that become coarser and sandicr upwards Bnet progradation of the shoreline (Fig. 8.1 La). An ction is suggested by derailed mapping in the Cardium jon (Eyles and Walker, 1988), which has revealed a te form ro individual parasequenees that produces an Il progradational, but variable, parascquence-set sig- In proximal locations the balance between sediment wpply and relative sea-level rise will control both the sequence and the parasequence-set signature (section 4.2). The nature of retrogradational, storm dominated para- ence sets depends on the balance between sediment ply and relative sea-level rise. Figure 8.11b shows one bination of these variables — the shoreline~shelf el- is of the parasequences could be stacked more closely separated entirely. Transgressive, storm-dominated parasequence sets com- tse a scries of back-stepping parasequences, the sandy tions of which are neither connected nor overlain (Fig. Me). Many transgressive parasequence sets are charac- ized by ravinement, removing all evidence for subacrial sposure, and by shelfal processes reworking the stranded cline deposits. Transgressive parasequence sets. are ne forms of retrogeadational parasequence sets. Forced regressive, storm-dominated sets are generated hen sea-level falls in a series of more or less discrete steps ig. 8.6c). As discussed in section 8.3.4, the base of each package is characterized by a downwards shift in ies that records sea-level fall. If sea-level fall is associated ith river incision, then each sand package is a high- gency sequence. The sand bodies themnsel 12 tetizel by a progeadational, coarsening signature. Forced igresive sets can have a geometry closely similar to Aansgressive parasequence sets (compare Fig. 8.6c with Fig. $.11c). The shoreline position is likely to be fixed by shell iregularities, generated, for example, by deep-seate: faults, Wave-dominated parasequences The stratigraphic architecture of wave-dominated para Sequences is closely comparable to that of storm-dominated Parasequences. A key difference is that the predominance ‘of shoaling waves commonly limits the distance from the beach to the edge of the sand belt to less than Ikm. AS a result, dip-extensive wave-dominated sands can be gener. ated only by progradation. As in storm-dominated para Sequences, wave-dominated shoreline successions generally soarsen, and increase in bed thickness upwards. Again, the upper shoreface may be barred, non-barred, or cut by tidal shannels. Conglomeratic wave-dominated shorelines ate steeper and have their own distinctive suite of structures (Fig. 8.125 Massari and Parea, 1988; Hart and Plint, 1989). Tide-dominated parasequences Tide-dominated shoreline—shelf systems pass laterally into estuaries and deltas, and landwards into tidal flats (Fig, 8.13). If the tidal flats are richly fed with sediment by alongshore and along-shelf tidal currents, they can form the central portion of prograding coastal plain to shoreline shelf suecessions. Subsidal flats and the lower portions of intertidal flats tend to be sandy, and to pass landwards into muddy and then vegetated, intertidal and supratidal flats Consequently, prograding tidal flats generate fining up. wards successions. These in turn may be cut by fining upwards channel fils, the deposits of complex channel systems which dissect the flats (for a review sce Elliott, 1986a,b). Progradational, tidal successions are illustrated in Fig, 8.13. Analogous transgressive tidal deposits are thinner, developing during periods when rivers are drowned, form- ing estuaries, and little sediment is transported to the shelF.. In such cases sandy shelf deposits tend to be sourced by. tidal scour and reworked from the underlying succession. The combined effects of tidal scour and shoreface erosion ‘may generate complex erosional topography on the flooding surface, Deposits that overlie the surface may include sand sheets and sand ridges. Sand sheets may fine or coarsen ‘upwards depending on sand-sheer movement. The internal structure of tidal sand ridges is poorly known, but is thought to be dominated by dune cross-bedding. For a detailed discussion sce Stride (1982). 8.5.2 The stratigraphy of deltaic systems Much of the stratigraphic variability observed in deltas is controlled by interplay amongst: sedimentary processes, accommodation space, sediment supply, climate and grain size. Accommodation space and sediment supply play a role similar in deltaic systems to thar discussed for coastal plain to shoreline—shelf systems. Following Orton (1988) and Orton and Reading (1993) the importance of gr and sedimentary process can be shown by extending, the ternary process classification of Galloway (1975) into a prismatic form (Fig. 8.14). Their model can be refined by differentiating amongst shelf deltas, shelf-edge deltas and Gilbert deltas. Space precludes illustration of this extra 1 km) these clinoforms are generally not resolved and shelf deltas are expressed as parallel seismic reflections. By contrast, the slope systems that form the delta front of shelFedge deltas may be several hundreds of metres high 71“gag] Poorly) ——Pemesbity (m0) ——Gaavay API Uiiegy Dep Opargeardirecion Sacra ‘Sequree sree oom ne) ge so_ntpreaton Saigaty T i Delta plain i oer : | 7 | : ali i i 3 Is ‘ard onder x iit cmon ie = + — UI owe |S Somnus nna Rw sero Fig. 8.23 Graphic log, porosity and permeability data showing the lose relationship between key sequence stratigraphic surfaces and reservoir quality in the Scalby Formation, Yorkshire. For example, an order of magnitude dectcase in permeability occurs, ‘heross the lowermost flooding surface, anda similar change occurs across the fourth-order sequence boundary (after Eschard et aly 1991) imnl clearly resolvable. Shelf-edge deltas also exhibit a ge of syn-sedimentary deformational features that can resolved seismically; growth faults, mud diapirs, se nt slides, canyons and gullies. ingle clastic ramps owangle clastic ramps develop in those basins that lack a re-eistng slope, and are characterized by strong shelfal 6s, which redistribute sediment supplied via deltas nd prevent the growth of a shelf—slope system (Fig. {24b). On regional seismic lines they are expressed as monotonous parallel reflections (e.g. Sangree and Widmier, 97). Redisplaying conventional seismic lines at reduced jontal scales (squashing) or increased vertical scales, (stetching), commonly by a factor of two to four times, reveal subtle dips and stratigraphic pinch-outs, ot mrevals of shelf-delta sedimentation if the resolution is Wide-self model uith (1970) suggested that progeading shorelines may d wide, prograding and aggrading shelf—slope systems 8.24). Although Asquith’s original correlations have questioned (Van Wagoner et al., 1990), other (@) DROWNED SHELF Key: 5 revs 824 Basinscale models of paraic 3 Consatan ‘Actve deta. or coasa-laln-chorene shi system ES workers have shown similar geometries (eg. McCave, 1985; Lawrence et al., 19905 S. Sturrock, pers. comm., 1991). The model is important as it indicates that not all pro- grading siliciclastic slopes are fed directly by deltas — the shoreline may be some distance from the offlap break. The decision as to which of these three models is appro- priate is not straightforward, because the seismic expression Of wide-shelf systems is comparable with that of shelfedge deltas, both in large-scale form (e.g. McMillen and Winn, 1989) and in the potential for generating syn-sedimentary deformational features on the slope. In general though, wwide-shelf—slope systems tend to be muddy and to have sigmoidal clinoforms, whereas strongly prograding slopes, with toplapping clinoforms are more likely to be fed by shelf-edge deltas. 8.9.2 Seismic resolution The seismic expression of a single sandstone body isolated in shale is dependent primarily on the relationship between the thickness of the sandstone and the wavelength of the seismic data (see section 3.1.2). For example, at burial depths of 1-2km. Sandstones with 20% porosity may have a velocity of 2500 ms~', With a good quality seismic source of 50H, distinct reflections can be seen from the top and the base of sandstones that are thicker than 25 m. Seaterel Sentero! (J strand mus Nearshoreandinner shelf sands 173However, beds thinner than 6.25 m cannot be distinguished. Beds 25~12.5 m thick will be characterized by constructive interference of reflections from their top and base, and beds 12,5~6.25 m thick will be resolvable only by changes in amplitude (Meckel and Nath, 19771. Paralic sandstones have mean thicknesses that range from 5 to 30m (Table 8.2). Most are likely to be less than, 15m in thickness, and many are characterized by thickness changes. As a result, individual sands may be expressed by two rellections, interference, and/or by amplitude changes. In addition, as paralic successions typically comprise a number of closely superposed sandstone bodies, inter~ ference of reflections from individual sandstones is likely. In such situations it may be difficult to determine both the number of sandstones and their distribution (Sheriff, 1985). Seismic modelling can be used to help constrain inter- pretations (Meckel and Nath, 1977). Horizontal seismic resolution is a function of the wave- length ofthe seismic data, and the nature ofthe twordimen- sional or three-dimensional data set. For two-dimensional daca, seismic migration gives an approximate resolution of 1/4 of the seismic wavelength along the line. Using the parameters stated in the example above this would give a horizontal resolution of 12.5m, Most paralic sandstone bodies are much wider than this and should be detectable if their thickness is appropriate (Table 8.2). However, ‘unmigrated energy from out of the plane of the line may reduce horizontal resolution. In addition the resolution of two-dimensional seismic data is severely limited by the spacing of the individual seismic lines. By contrast, three dimensional data migration is performed in three dimen- sions, giving a resolution of 1/4 of the wavelength in all directions. In addition, three-dimensional data sets are data volumes, There are no line spacing problems. As a result, with good quality data sets, individual sandstones ‘commonly can be tracked on time slices or horizon slices — even thin sands expressed by subtle amplitude changes. 8.9.3 Seismic facies As a result of the problems of vertical resolution and interference of reflections generated by superposed sand. bodies, there are few examples where discrete paralie sand- stone bodies have been imaged on conventional two- dimensional seismic data. However, several studies reveal 4 consistent relationship between paralic depositional environments and seismic facies (Sangre and Widmier, 1977; Bouvier et al., 1989; LeBlanc et al., 1989; Van Vliet and Schwander, 1989). Four key facies are apparent Fig, 8.25). 1 High continuity, high amplitude. Continuous high- amplitude reflections are indicative of extensive uniform shales interbedded with uniform sands, silts, limestones, or coals. Lithological contrasts of this kind are commonly generated at parasequence, parasequence set and sequence boundaries. 174 2 Low amplitude. Low-amplitude reflections indicate cither uniform lithologies or thin beds that cannot be resolved. They are typical of offshore marine shales. 3 Low continuity and variable amplitude facies. Low con- tinuity and variable amplitude facies are characteristic of fluvial, delta plain and coastal plain successions, which comprise sands of variable size, shape and orientation, scattered in shales. 4 Low-relief mounded facies. In addition to being charac. terized by subtle clinoforms, Sangree and Widmier (1977) argue that shelf deltas may be recognizable by having a broad gently mounded extemal form. 8.10 Variations in paralic systems within a sea-level cycle Sequence stratigraphy provides a series of models that allow the temporal and spatial position of broad paralc ‘environments to be predicted (Posamentier and Vail, 1988) Posamentier et al, 1988; Van Wagoner et al, 1990). An understanding of shelf processes, delta position and gra size can be used to refine these models, and to inerease thei predictive power. 8.10.1 Shelf processes The interaction between fluvial and shelf processes is ent cal in determining the spatial distribution of sands within paralic successions. The capacity of shelf processes 10 redistribute sand is a function of a variety of controls, some of which are unrelated to any sequence stratigraphic frame- work. For example, for a basin to develop significant tides it must be freely connected t0 a major ocean: the Mediterranean is essentially tideless because itis to0 small to generate significant tides, and the straights of Gibraltar are too narrow to allow significant oceanic tidal waves to center. Nevertheless, several studies have revealed systematic relationships between parasequence stacking pattern and basin process. Where a relationship is apparent, pro: gradational parasequence sets tend to be more fluvially dominated, whereas retrogradational and transgressive para: sequence sets tend to be more tide and wave dominated (eg, Weise 1980; Bhattacharya 1988, 1991; Pulham, 1989; Tyler ef al., 1991). The decrease in the effect of fluvial processes in backstepping parasequence sets can be related ro a rising base level that drowns and causes the lower courses of rivers to be sluggish. By contrast, shelf width appears to be a major control on marine processs Tides ‘Maximum tidal ranges occur when the shelf is wide enough to allow the shelf sea to resonate with the open ocean tide (Howarth, 1982). This occurs when the shelf has a width of around 300km, i.c. one-quarter of the wavelength of the ocean tidal wave, Most modern shelves have a widthfess than 300km and exhibit an approximately linear relationship between shelf width and tidal range, with tidal range increasing as the critical wavelength is ap- proached. Restricted gulfs are often sites of amplified tidal range as they represent areas of increased shelf width. In detail, the basin gcometry required for resonance is del cately balanced (Defant, 1961) so that small changes may alter the tidal regime radically (Johnson and Belderson, 1969). Thié is illustrated by the Bay of Fundy, where changes in basin physiography and sea-level over the last ‘4000 years have resulted in the tidal regime changing from being microtidal (<2m tides) to its present macrotidal state. The implication is that temporal changes in the strength of tidal processes are to be expected, particularly when major changes in basin configuration occur, as, for example, across sequence boundaries. Waves Coastlines that face the open ocean and the prevailing winds will be more wave dominated than lee-facing coast lines in restricted basins. As shelf width increases, shoaling, oceanic waves (ie. waves that touch the sea bed) are increasingly attenuated and damped by bottom friction. ‘Therefore, wider shelves may be less wave-dominated than equivalent steeper, narrow shelves. 8.10.2 Spatial changes in processes Individual shelves are not characterized by a single sedi- mentary process: different processes dominate in different regions. For example, in the modem North Sea the tidal ridges and sand sheets off the English Coast (Stride, 1982) and the Dutch estuaries (Visser, 1980) are well known areas of tide-dominated sedimentation. Yet within the same basin, the Helgoland Bight off Germany isa classic example of astorm-dominated shoreline—shelf (Aigner and Reineck, 1982) 8.10.3. Sequence stratigraphic framework ‘The relationships between shelf width and sedimentary processes can be combined into a sequence stratigraphic framework, because sequence models suggest a systematic relationship between shelf width and systems tract. The relationships are most clear during the development of a type 1 sequence boundary in a basin with a pronounced shelf=slope break. In the highstand systems tract, the shelf narrows and the system becomes increasingly progradational in response to a decrease in the rate of rise of relative sea-level. The progradational signature favours flavially dominated deltas that may become increasingly wave-influenced as the shelf ‘width decreases, and the effect of wave damping diminishes. In the lowstand systems tract the rivers initially incise, 176 the shelf is by-passed, and sediment is transported directly to the deep basin, The lowstand prograding wedge, located at the shelf edge and fronted by a distinct slope system, i essentially a shelf-edge delta, located down-dip of the incised valley. The deltaic trunk river forms a point se: ‘ment source, fixed at the head of the incised valley, and a strongly prograding shelf-margin delta results. The low rate of generation of shelfal accommodation space favours a sandy shoreline system, Whereas the strong progeadation, favours a fluvial signature, and the narrow shelf a high degree of wave influence. ‘An increased rate of relative sea-level rise progressively slows the rate of progradation and may flood the incised valleys so that they take on an estuarine or tide-dominated shelf-delta character. As the interfluves are flooded she rate fat which accommodation space is generated increases abruptly, initiating the development of the transgressive “systems tract. A steady rise in base level causes the rivers to be sluggish in their lower portions, suppressing 4 fluvial signature. By contrast waves have more time to rework deltaic deposits and the increasing shelf width favours stronger tidal currents, resulting in wave- and. tide dominated deltas. Lateral to deltaic depocentres, the shore: line is likely to be characterized by wave-dominated barrier shorelines, which front large lagoonal areas that can be ct by tidal inlets, depending on the tidal range. ‘The culmination of the transgression is represented by the maximum flooding surface. In shelfal settings this i represented by a condensed horizon, in the lower del plain and coastal plain it may be recorded by the maximum landwards penetration of a tidal signature (Shanley and McCabe, 1993). 8.11 Summary High-resolution sequence stratigraphy is emerging a6 a powerful tool in the analysis and prediction of paralie successions. It has clear applications in deciphering funda mental stratigraphic controls (namely sediment supply and relative sea-level change) and delineating, genetic units both in the subsurface and at outerop. In the petroleum industry itprovides a rigorous framework for subsurface correlation, and can help in the delineation of reservoir flow units. In addition, it can provide a means for classifying analogue data, and it can highlight stratigraphic traps. Successful high-resolution sequence stratigraphic studies are founded on facies analysis, an approach based on the study of continuous, uninterrupted successions. The ad: vance of sequence stratigraphy has laid increased impor ance on small-scale facies dislocations, and grain-sive changes. It may be time to revisit classic facies models with this in mind. For example, is progradation a steady process or are small-scale facies breaks that could be attributed erroneously to sea-level fall a common feature? In contrast with core-based and_wireline-log-basedstudies, seismic studies are at present relatively crude. This isparticularly true for studies based on petroleum industry ‘two-dimensional seismic, where major stratigraphic sur- fhces, coal prone intervals, and seismic facies may be distinguishable, but key features such parasequences and valley fills cannot be recognized with ease. Inthe future, the increased resolution of three-dimensional_ seit ‘offers an opportunity for integrating high-reso quence stratigraphic concepts with seismic data. ic data suoyssavong oereat | 8120001)CHAPTER NINE Deep-marine Clastic Systems Introduction 9.3 Fan development d 93.1 Lows 9.2 Deep-marine elastic systems — depositional processes and classification 9.2.1 Depositional processes 9.2.2 Classification of deep-marine clastic systems 934 ad models 93.2 Basinloor versus slope fan 9.3.3. Lowstand systems tracts and deep- effet of sediment calibre on lowstands 9.4 Fan development daring highstand and tcansgression 9.4.1 Highstand fan systems 9.4.2 Fan development during Iransgressions 9.5 Conclusions the evolution of the lowstand 9.1 Introduction Sequence stratigraphy provides a powerful tool for strati- graphic analysis of deep-marine clastic systems when it is combined with an appreciation of the variability in pro- cesses and depositional products of deep marine settings. This combined approach based on sequence stratigraphy and sedimentary processes has advanced from early studies of the late 1970s, which emphasized simple model-driven interpretations of de scismie data (e.g. Mitchum ef al., 1977a,b; Vail et al.y 1977a,b; Muti, 1985; Posamentier and Vail, 1988; Van Wagoner et al., 1990; Walker, 1992a,b; Posamentier and Weimer, 1993) This chapter first reviews the range of sediment transport mechanisms in deep marine settings and the controls on coarse clastic deposition within a basinal setting. The prin- ciple large-scale deep-marine clastic depositional systems and the controls on their development are discussed. This information is used to develop a suite of sequence strati- graphic models, which show the variety of depositional systems that may develop within a given systems tract. ic systems based upon 9.2. Deep-marine clastic systems — depositional processes and classification 9.2.1 Depositional processes The erosion, transport and deposition of sediment in deep- urine clastic settings are controlled largely by sediment gravity flow processes; flows in which sedimentfluid mix- tures move under the influence of gravity (Middleton and Hampton, 1973, 1976; Lowe, 1979, 1982; Middleton and Southard, 1984). Sediment gravity flows form a broad 178 group of genetically related processes, ranging from slumps and slides associated with the downslope translation of cohesive material such as silt and mud, to fully turbulent turbidity currents Four basic types of sediment gravity flow are recognized, and include the following (Table 9.1) 1 Turbidity currents, in which the sediment is supported by the upward component of fiuid turbulence generated by, the density contrast between a sediment= fluid mixture and the surrounding ambient fluid 2 Fluidized/liquidized flows in which the sediment is sup ported by the upward movement of the escaping pore fluids. 3 Grain flows in which the particles are supported by the Alispersive pressure of colliding grains. 4 Cohesive flows in which particles are supported by matrix density and strength, ‘Turbidity currents and cohesive flows are considered to be effective agents of sediment transport, and are the principal mechanisms by which turbidite facies* develop within basinal settings (Middleton and Hampton, 1973, 1976; Lowe, 1982; Pickering et al., 1986; Postma, 1986; Pickering et al.. 1989). The remaining types of sediment gravity flow are regarded as transient phenomena; occurring, between the initiation of sediment movement by slides and slumps, and the final stages of sediment and fluid transport by fully turbulent turbidity currents (Lowe, 1979, 1982; Postma, 1986). Many authors draw the additional distinction between low- and high-density turbidity current flows (e.g. Middleton “The (0972) to refer tal sshments deposited by sediment gravy Hows and not it enti fais sw ere nee conten oF Mt and Ris Lachle9.1 Characteristic types of sediment gravity flow (modified after Nardin et al, (1979) and Cook et al, (1982) Internal f ——_mechanical Transport mechani and ‘Acoustic record Sedimentary structures tnsport behaviour dominant sesiment support characteristics, and bed geometry q] Freefall and roling single blocks along Strong hurnmocky bottom return, Grain supported framework, variable steep slopes hhyperbolae and side echoes common. Weak, matrix, diorganized. May be elongate chaotic internal return: structureless parallel to slope and narrow Perpendicular te slope ‘Shear failure along discrete shear planes Internal reflectors continuous and often Bedding may be undeformed and perale subparallel to underying beds Slide may undeformed; abrupt teminatons. Strata of to underlying beds ar deformed especialy behave elastically at top; plas ‘glide blocks may be uncomformable or at base and margies where deo flow and thin lateral margins subparallel to underying sediment conglomerate can be generated Hurmmocky, sight convex-up top, base subparalel to underlying beds; 103 to 1000s wide and long Basic Shear failure along discrete concave-up Internal reflectors continuous and Bedding may be undeformed. Upper shear plames accompanied by rotation of undeformed for short distances with and lower contacts often deformed, slide. May move elastically or elastically and deformation at toe and along base. Internal bedding at angular dicordance to plastically Concaveup falue lane at head and enclosing strata. Sige variable subparalel to adjacent bedding a toe, Surface usually hummocky 1a low Shear distributed throughout the sediment Seaclor reflectors may be hyperbolic, lasts matrixsupported; clasts may Pastic mass. Clsts supported above base or bed by regular, oF smoath. Commonly acoustically exhibit random fabri throughout the bed Cohesive strength of mad matrix and cast _wansparent with few or no internal or oriented subparallel, especially at buoyancy. Can be intisted and move long reflectors. Mounded or lens shaped with base and top of flow units; inverse grading distances along very lowangle slopes blunt termination at head, May be chactic possible. lat sie and matrix content internally \atiable. Occur as sheet to channel shaped bodies em? to sveral 105m thik and 105 to 1000s (7) long: widths variable Cohesioniess sediment supported by Massive; ast Aas parallel to flow dlspenive presure, Usvally requires steep and imbricate upstream, inverse grading slopes for initiation and sustained may occur near base downsiope movement Cohesionless sediment supported by __Inlvidual flow deposits very thin; may not ‘upward displacement of fluid (dlatance) as be resolvable with present seismic reftction loosely packed structure collapses settles techniques, Repested flows may produce 3 into a tight packed texture. Requires sequence of thin, even reflector slopes = Dewatering structures, sandstone —_———— ykes, lame and load structures, convolute red fiow Fluid Cohesionless sediment supported by bedding, homogenized sediment ‘upward motion of exaping pore fluid. Thin (© 10 susurs pen ium sour ‘auenny susure Kian use spas aim tiguep pment as eS Has SUD fpRIN Ou pena moj pare) say eu) puesaune olen seu wentery 20 Bupyoniny pi IY ANE poes9 es yous e001 350 3500 espn aieepon ot wo e2pon ras ews aieepon ——SePOHL es yus-aiepon, st oto ses toes ot paseo ween sree Sao ves 05-02 si-or ee est se-0e Aon on 6 satrap sespon 6H emporio Imoy swap dos ews soe sxevspon, erpou 3611 som as w=) es) c er sw) tu - iw 9 pues pues a pesrpne Pow pes pursue (6 unseses pesiogonnsKapos 76 Iae.pe pues yen pe pues suevoaat sueodunion ——yurvoduy tow Aaenonog yemonns —imonag—adiaens deinen yempnag anna sempnas 006 hon poo vn oor aaeepon po0 209) poo paeBAba SHON soo P05 poo SPOR Fo°9 p00 von sno sieRpow-W6si on a ‘nor iesepou-45st vou sn wor sysopou- ys way va sa 4 s3ev va > aw sew 8 © sevexpnu foand wisp ex fepou un sun suoupues penne syn suo sun pony due pues neuuey pup souaucsd a paeuuop ssf perce oor ps ae90, ‘agoq sue -oWwZOpp rero4 ‘aim suurp ——Knauoat pues eget ‘eae, pues pawn ans pore seyoioy 360) Soon. 2-0 ssowanns sqeves Sis oeveNKubh SEHR ANH ayeN AONE oe os ss yeipsmoy —spupou poe spun ysis spine peyton simp ‘0p porew spunowsn0e4> punow moe Spam wuotonaney sna, owe stom yrumbnniiey easy suv espa aS sou sug duns e018 soya pamueD 2901 RUFF aE SG SEO PARED sot esa ‘seam pu ‘os pir sous ‘atpan suum seao]-puvey> seen-puumD ———sAEPaM auon sda) uue see PIURD wo) ws) sn on 9) ‘ase a sw 0 pein prs pa pr en puesipre, pus psn Pow es wei ewe Iq pazyoBi said atop BIRO {a wei ieuop&@ pRNOBRID sdues BuNeUATE SIMONA {a si RUB KG panHEBAID we 2a wa pas (best Spreqpny pe Suypeay s2ye) sushs ausepp aueaH-doop yo sousUaR NEED aEpNSGAS 6 IGE,Gravelrich systems Gravel-tich submarine fan and ramp systems are generally small in scale and commonly form che down-dip. marine equivalents of coarse-grained fan-delta, alluvial cone and braid plain settings (Fig. 9.5, Tables 9.2 and 9.3). Gravel- Fich slope aprons are derived from reworking, mass-wasting and catastrophic submarine rock fall avalanches associated With high-angle slopes and scarps (Fig. 9.5a). They com ‘monly occur adjacent to relict or active submarine fault scarps. Sediment is often poorly sorted prior to deposition. The depositional products of these systems reflect a wide satiety of mass flow processes and include non-channelized shaotic boulder/cobble beds, and intraformational ro- tational slumps of fine grained interbedded sandstone and mudstones with exotic clasts. These coarse-grained facies commonly interbed with turbidite sandstones and mud- stones (Surlyk, 1978; Ineson, 1989). Ephemeral input from pslip alluvial fans and fan-delta systems may liberate ‘sni-grade material to the apron, resulting in the formation of more sand-rich facies fringing the distal edges of the system. In seismic data the slope apron shows little coherent acoustic character and forms wedge-shaped aggradational packages which thicken towards relict fault scarps. Grayelrich submarine fan and ramp systems typically form broad, wedge-shaped sediment bodies, characterized in their proximal parts by conglomerates and sandstones, derived from rock-fall and debris flows (facies A and F of Motti and Ricci-Lucchi, 1972; Surlyk, 1978; Kessler and Moorhouse, 1984; MacDonald, 1986; Ineson, 1989; Ferrentinos et al., 19885 Piper et al., 1990) as well as thick bedded and erosively based gravel- and sand-tich high. density turbidites (Fig. 9.5b,c). Sand-rich facies are best developed within the medial parts of the system, where they occur as stacked, massive or graded sandstones charac- terized by dewatering and fluid-escape structures typical of high-density turbidites (facies B, Mutti and Ricci-Lucchi, 1972; Lowe, 1982}. The distal margins of the systems pass abruptly into thin sandstone turbidite beds and inter- brlded, hemipclagic shales (Ineson, 1989; facies D, Mui and Ricci-Luechi, 1972; Prior and Bornhold, 1989). The fon and ramp systems show wedged-shaped seismic ge- metres with little or no coherent internal seismic character. The wedge commonly thickens into the hanging-wall areas, adjacent to up-dip transfer segments. Rapid lateral and Vertical facies variations are the dominant depositional motif of gravel-rich fans and ramps. This facies va is illustrated by the internal architecture of gravel-riel submarine ramps from the Jurassic syn-rift successions of the Wollaston Foreland, East Greenland (Surlyk, 1978) and the South and North Brae fields of the North Sea Basin (Harms e¢ al, 19813 Stow et al., 1982) Sand-rich systems ‘The term sand-rich denotes those submarine fan, slope- apron andl ramp systems that exhibit a sand content exceed- ing 70%, when measured throughout the system (Fig. 9.6). This sand content defines the lower limit for a major change in the seismic character and facies architecture of deep marine turbidite systems (Reading and Richards, 1994; Richards et al., 1996). Sand-rich systems form rela tively small-scale features (generally 1—5Okm radius) sourced from the incision oF failure of relict sand-rich shelves or by direct canyon access to littoral drift cells (Reading and Richards, 1994), Sand-rich slope aprons are derived from the reworking and mass wasting of relic shelf—slope systems, and form isolated wedge-shaped sand bodies paralleling the basin margins (Fig. 9.6a). The slope apron is limited in basinward extent, reflecting the local nature of sediment source areas to the turbidite system. Coarse-grained high-density tur- bidites dominate the proximal facies of the system, whereas more distal areas show an increasing basinward inter- bedding of hemipelagic mudstones (Kumar and Slatt, 1984). ‘The systems form lobate to aggradational wedged-shaped seismic packages that bank up against basin-margin slopes. The apron is internally seismically opaque with rare devel. ‘opment of clinoform goometries Sand-tich fan and ramp systems are dominated by up- dip channel sand bodies, which pass down-dip into channel ined fan lobes (Nelson and Nilsen, 1984; Busby-Spera, 1985; Kleverlaan, 1989; Reading and Richards, 1994). Studies of outcrop and subsurface examples demonstrate similar high sand : shale ratios throughout the extent of the fan and camp system (Fig, 9.6b,c). The mid-fan and medial ramp comprise elongate, channelized lobe systems, which coalesce to form a broad sand-sheet (Link and Nilsen, 1980; Link and Welton, 1982; Chann and Dott, 1983, Busby Spera, 1985; Heller and Dickinson, 1985; Guardado et al., 1989; Kleverlaan, 1989). The systems are largely dominated by massive to internally graded high-density turbidites, with only subordinate thin-bedded turbidites forming fan-abandonment and fringe deposits. Sand-body continuity and connectivity are typically very good, although more widespread shales may occur, reflecting lobe abandonment intervals (e.g. MeGovney and Radovitch, 1985), Mid-fan deposits are dominated by abrupt changes from shale into thick, amalgamated sandstones. Cleaning upward, coarsening upward trends are noticeably absent, whereas the transition from fan/ramp to basin plain is generally abrupr. On seismic data the fan and ramp systems occur a single seismic package, with poor to well-developed mounding. The scismic envelope defines the full extent of the turbidite system. Internal reflectors may be subhorizon- tal to low-angle inclined with rare clinoform geometries. Examples of sand-tich ramps and fans include the Tortonian sandy fans of the Tabernas Basin (Kleverlaan, 1989), the Eocene Matilja and Rocks Sandstone (Link and Nilsen, 1980; Link and Welton, 1982), and the Jurassic Magnus and Miller systems of the early post-rift fill of the North 185uNe-souRceD COALESCING co Smee Se es UR iN SUING ANS WA SAGE SRY LA SERS Rumble 186 Fig. 9.5. Summary block diagram sgeavelrich, deep-marine clastic systems (after Reading and Richa 1994)CANYON FED BY ACTIVE (@)NEARSHORE LITTORAL DRIFT (OF RELICT SHELF SANDS MULTIPLE RIVERS/SANDY DELTA (0 AND COASTAL SYSTEMS. Fig. 9.6 Summary block diagrams of sani-rich, deep-marine clastic systems (after Reading and Richards, 1994)Sea and Palacogene Frigg, Cod, Balder, Andrew and Forth, Fans, also of the North Sea Basin (Heretier et al., 1979; Kessler ef al., 1980; De'Ath and Schuyleman, 19815 Sarg, and Skjold, 1982; McGovney and Radovitch, 1985). Mixed sand—mud systems The term mixed sand-mud is used here to denote fan and related systems that exhibit sandstone contents between 30% and 70% throughout the system (Fig, 9.7). These systems commonly are derived from large mixed-load delta, shoreline and coastal plain provenances. Mixed sand~mud slope aprons, fans and ramps form moderate-scale features (10-350km radius) and account for a major portion, of the ancient deep-marine clastic record (Reading and Richards, 1994). Mixed sand—mud slope aprons are characterized by a wide variety of mass flow processes, resulting in complex, irregular and often disorganized lithofacies distributions (Fig. 9.7a). The apron system is dominated largely by slump packages composed of deformed hemipelagic shales and contorted thin-bedded turbidites, slide blocks and chutes infilled with slope-related mudstones or thick- and thin-bedded, discontinuous turbidites and debris flows (Hill, 1984; Nelson and Maldanado, 1988; Alonso and Maldanado, 1990). Gullies and constructional channel sys- tems may traverse the disrupted slope-apron surface and pass laterally and basinward into more stable areas, where laminated mudstones and sandy mudstones predominate. Local development of turbidites within these gullied and channelled areas may lead to isolated channel-fills of hetero- sgencous turbidites and sandstones with interbedded mud- stones. The systems commonly form well-developed mounded scismie packages dominated internally by hum: mocky to chaotic reflectors (Nardin et al 1979). The mounds are located in close proximity co the base of the submarine slope owing to sediment input to the apron from mass wasting and failure. Detailed mapping of the incernal seismic facies commonly reveals a chaotic seismic character. ‘Mixed sand~mud submarine fans and ramps are domi nated by two main architectural elements; a channel levee system and down-dip depositional lobes (Fig. 9.7b,¢5 Normark, 1978; Walker, 1978; Normark ef ala. 19795 Droz and Bellaiche, 1985). The dominance of one or either of these architectural elements in seismic data is controlled by the grain sizeof the system. In general, increases in mud content within the fan or ramp result in a reduction in the relief and acoustic expression of lobes and the increasing dominance of channel—levee systems. Channel—Ievee systems form the conduits through which sediment is distributed to the main area of the fan, These may be mud-filed, reflecting rapid fan abandonment, or contain a central core of coarse-grained, highly hetero: geneous channel-fil deposits, flanked by levee siltstones and mudstones (Walker, 1978; Winn and Dott, 1979; 188 1985; Tyler et al., 1984; Mutti et al. 1985a; Weuller and James, 1989; Schuppers, 1992). Channel fill facies vary from sandy conglomerates and pebbly_ sandstones with thick bedded, high-density turbidites to fine-grained thin- bedded turbidites and hemipelagic mudstones. Individual sandstone beds are commonly lenticular and erosively based. In high-resolution and multichannel seismic data, the channel system may display a poorly to well-developed wedee shaped or ‘gull wing’ geometry dominated by diffuse or well-defined low-angle reflectors, which downlap away from the apex of the channel axis. Constructional lobes form the down-dip equivalents of the channel—levee system (Fig. 9.7b,c). They comprise overlapping, layered sand bodies characterized by complex, and variable sand-body geometries and architecture. Sand: shale ratios are highest within the core or apex of the lobe, bat decrease markedly towards the lobe margins. The lobe may be dominated by massive, thick-bedded high-density turbidites (Kleverlaan, 1989; Kulpece and Van Geuns, 1990) or display more classic turbidites and interbedded hhemipelagic shales (Stevens Fan, MacPherson, 1978; Webb, 1981; Marnosa-Arenacea Fan, Ricci-Lucchi and Valmori, 1980) with poorly developed cleaning upward log cycles or coarsening upward grain-size trends. Such fan systems display a high degree of variability at the small to medium .. On seismic data, mid-fan depositional lobes and associated channel systems show poor to well-developed hummocky acoustic character. These features may occur as isolated mounds developed a significant distance from the submarine slope. The depositional character and facies of ‘mixed sand—mud fans and ramps are illustrated in studies by Murti (1972, 1985, 1992), Walker (1978, 1985}, Kleverlaan (1989) and Pickering (1981). Subsurface examples include the Forties, Nelson and Gannet fan and ramp systems, North Sea Palacogene (Armstrong et aly 1987; Kulpece and Van Geuns, 1990), the Miocene Yowlumne field, San Joaquin Basin, California (Berg and Royo, 1990) and the Permian Spraeberry Trend of Wes Texas (Tyler and Gholston, 1988). Mud-rich systems Mud-rich turbidite systems contain fess than 30% sand (Fig. 9.8). They form a common feature of basins with mature drainage patterns, large source areas and river and deltaic systems dominated by fine-grained suspended load (Reading and Richards, 1994). Mud-rich systems ate volumetrically the most important deep-water clastic sys tems occurring in the world’s oceans today (Damuth and Kumar, 1975; Kolla et al., 1980; McHargue and Webb, 1986; Kolla and Coumes, 1987; Weimer, 1990), and form large-scale slope aprons, fans and ramps 50~3000 km in radius (Reading and Richards, 1994) off many continental margins. Mud-dominated slope aprons are broadly similar in character to their sandier, mixed sand—mud counterparts) MEDIUM GRADIENT [GOASTAL PLAIN, DELTAS, (CHENIERS, GARAIER IStaNDs ss w (CHENIERS, BARRIERS IMuD/sANo-RICH SUBSURFACE o 10-1004 | Fig. 9.7 Summary block diagrams of mised sand—miud, deep-marine clastic systems (after Reading and Richards, 1984) 189* MARSH aor MUD-RICH MERGER EZ © Lowmuooy ‘COASTAL PLAIN, PAL > FEEDING CCHANNEL-LEVEES : i INTHE Fig. 9.8 Summary block diagrams o HETEROGENEOUS Y SUBSURFACE HETEROGENEO mudé-rich, deep-marine clastic systems LOBE SANDS AND SULTS: (after Reading and Richards, 1994) 190ig, 9.8). The slope margin is characterized by erosional les and corational slumps generated by sediment load- fand foundering together with curvilinear extensional nits associated with the downslope translation of sub- rine slides (Gorsline and Emery, 1959; Nardin et al 1079), The base of the mud-rich apron may be characterized bysile and mud-dominated lobes, debris flows and uneven topography generated by the compressional coes of slides, ‘The nature of depositional processes within the system least a complex and irregular distribution of fine-grained lthofacis. irregularly distributed sandstone turbidite pack- fais may develop where retrogressive slumping exhumes tele coarser grained shelf and upper slope deposits. Slide flcs located along the margins of the slope apron may fom conduits for basinward sediment transport in shelf. lige delta systems, Slope aprons show a dominance of tnounded seismic packages characterized by hummocky £0 thaotic reflectors, which may pass basinward into dis- continuous parallel reflectors (Mud-rich fan and ramp systems are typically large sys- tems (up to thousands of kilometres) dominated by well- dheloped channel—levee systems (Kolla ef aly 19843 Manley and Flood, 1988; Weimer, 1990; Fig. 9.8). Sands are poorly developed within the system and are confined (0 arelly restricted, highly heterogencous channeb-fill sand~ Sones within the axis of the channel—levee system {fmperato and Nilsen, 1990; Weimer, 1990). The outer fan tres is dominated by thin, distal sheet-sands (Pickering, 1983; Nelson et al., 1992; Twitchell et al., 1992). Mu fich fans and ramps are generally readily distinguished on fesmic data because of their large scale and the character- fic development of extensive channel—levee systems ffeliche et al, 1981; Kolla ef al 1984; McHargue and Webb, 1986; Kolla and Coumes, 19875 Damuth et al, 1988), The upper or inner fan area is dominated by chaotic and mounded seismic reflector packages, reflecting slope failure and slumping, separated by more organized channel— levee systems. These latter features show’ a characteristic “pill wing’ seismic goometry with an orientation perpen iclar to the channel axis. The mid-fan area displays 1g geometries, reflecting, channel— Tevees with limited seismic relief. The outer or lower fan gystem is manifest seismically as more continuous, parallel to subparallel reflector sets (Damuth, 1980; Droz, 1983: 0'Connell, 1986). smaller scale, gull wi 9.3 Fan development during lowstands 9.3.1 Lowstand models ver the past ewo decades, sequence stratigraphic models have been used widely to predict and locate submarine fan reservoirs in frontier and mature basin areas. These models lssume that the growth and deposition of a turbidite system are intimately ted to a cycle of eustatie or relative sea-level change. Early sequence stratigraphic models envisaged fan development to be related to eustatie se Tevel fall, where base level fell below the offlap break, causing, entrenchment and erosion of fluvial systems into a subaerially exposed shelf, and the bypass of coarse-grained Ulastic sediment into the basin (Fig. 9.9, Mitchum et al.. 1977agbs Vail and Todd, 1981; Posamentier and Vail, 1988; Posamentier ef af., 1988; and see Chapter 2). The implications of this model were clears sequence boundaries generated by eustatic sea-level fall should be associated ‘with submarine fan complexes in the basin. The model ‘vent further in suggesting that the timing of fan deposition Could be predicted from the eustatic sea-level curve (Vail et aly, 1977a,b,¢; Shanmugam and Moiola, 1985). Barly Iowstand fan models largely ignored the role played by tectonics, sediment type and sedimentation rate in influencing, the growth, architecture and evolution of these systems through time (Shanmugam et al, 19853 Kolla and Macurda, 1988; Posamentier and Vail, 19885 Posamentier et al., 1988; Galloway, 1989; Reading and Richards, 1994). For the Quaternary stratigraphy record there is some evidence to support a eustatic control on fan development. The Bengal and Indus fans, ewo of the largest modemn fans, both show an abrupt decrease in tertigenous clastic input following the Holocene sea-level rise. A similar picture is seen in recent deep-sea drilling evidence from the Mississippi and Amazon fans, which show comparable changes in Sedimentation patterns from Holocene pelagic deposition (G—30em 10000 years") t0 rapid turbidity current depo- sition [1200 cm 100000 years") during the glacial lowstand (Fig, 9.10). This pactern is also repeated in the Astoria fan system (Shanmagam et al, 1985). Tn the geological record, most of the hydrocarbon-bearing fons are interpreted to relate to periods of relative sea-level fall, if not global eustatic lowstands (Fig. 9.11). However, in many of these cases proving correlation of fan develop- rent with custatic lowstand events is difficult. It is clear that base level fall must have a profound effect on the potential for erosion of continental, marginal marine and hallow marine depositional systems. Whilst fan develop- tent may occur at any time, the volumes of clastic sedi- ‘ents delivered to a basin for fan development during hhighstand or transgression are unlikely to match the poren- al volumes of elastic input liberated during periods of relative sea-level fall 9.3.2. Basin-floor versus slope fan systems Models for deep-water, lowstand depositional systems ree- dgnize two major types of fan. These comprise a ‘basin oor or lowstand fan and a ‘slope fan’ forming part of the lowstand wedge (Mitchum et al, 1977a,b; Posamenticr and Vail, 1988). These terms require some clarification in the light of the classification scheme above. “The lowstand or basin-floor fan is considered to form a 191 és anse5 ouseurdoog | 6 191d049uuoneurjaxa poqieasp 40) 3893 995 “(1861 “PPOU, PUE HEA -PYe) fPA9]-e98 UL s|fey sHEISNa ZuLNP wouidojsarp avy 404 sfopout SEI GG Beg Wass adors ani wonaq/JT3HS 3H1-40 NOLWaMEGENERALIZED [ia] PELAGICMARL AND OOZE | GREY SILTY CLAY [GRADED Sano BEDS Anazonand other deep-sea fans and adjacent areas during the last Ericson Damuth and Kumar (1975), Dau mmoresand-prone system in response to the increased stream ‘apacity and enhanced alluvial gradients associated with the early phases of relative sea-level fall (Van Wagoner ‘tal, 1990). During this period, the shelf system is largely bypassed and the major locus of coarse sand deposition is shifted 10 the basin (Posamentier and Vail, 1988; Van Wagoner et al, 1990). This increase in sediment calibre and load may be enhanced further by the reworking and tmirenchment of incised valleys on the shelf. This scaward shift in deposition associated with the relative sea-level fall may be recognized as a downshift in coastal onlap in seismic data, change inshelfal and shoreline parasequence stacking patterns and the abrupt change from hemipelagic mmodstones to clean sandstones in logs penetrating the hasinal succession (Fig. 9.12; Van Wagoner ef al., 19905 Vail and Wornardt, 1990; Posamentier and Erskine, 1991). The slope fan is developed as relative sea-level stabilizes and later starts to rise. During this period, progressive reductions in alluvial gradients associated with the increas- ing shelfal accommodation result ina decrease in the volume and calibre of sediment supplied to the basin. This Fig 9.10 Generalized diagram showing the relationship between glacio-eustatic flict unrOLoGY crsn GENE GO apts oF sepmnraon RATES {ON MODERN FANS (onv10 ya ee oN (ete) DEEP-SEA i 3 2 — rane rsna—e) FANS AMAZON MISSISSIPFI_INDUS ASTORIA z "| [rsuscoercemon| ats | a7 | sis | om 1 P I 0 | TuRgory ” Y se wisconsin | een a | | t80 | >to erosion x | | et @ [ranacosrosmon| 345 | 370 | 7 7 19" caciaL. w 10" nF stacianion InveRoLsciATiON tions and sediment lithology on the 130 000 years (from Shanmugam ef al., 1985), Age boundaries, tic zones, sea-level curve and lithologic log are modified from Damuth (1977). Data from other areas based on ith (1977), DSDP Leg 96 Scientific Party (1994) and Ewing et af. (1958) change in the character of basin depositional systems is ac companied by coastal onlap and the progressive landward shift in deposition associated with inereasing shelfal accom- modation and the progressive infilling of up-dip incised valleys. The relative sea-level rise may be recognized by a change in shelfal and shoreline parasequence stacking pat~ tems and the transition on logs from clean sandstones of the lowstand fan into fine-grained, interbedded sandstones and mudstones reflecting levee channel and overbank di posits of the slope fan (Fig. 9.13; Posamentier and Vail, 1988; Vail and Wornardt, 1990; Van Wagoner et aly 1990; Posamentier and Erskine, 1991). Seismic expression Basin-floor fan and slope fan are terms used widely in the literature. Studies suggest that these two fan types ean be recognized in both seismic and wireline-log data based on the simple criteria outlined above (Posamentier and Vail 1988; Vail and Wornarde, 1990; Posamentier and Erskine, 1991; Sangree ef al, 1991). Furthermore, each fan type is 193(CHANGES OF SEA-LEVEL (m) epochs | AISING FALLING TURBIOITE RESERVOIRS +30 ° 300 — ° L [ PEST - { MISSISSIPPI CANYON (LA, USA) | poo FE z oe 2 REPETTO FM, VENTURAFLD (CA, USA) PSSSEEE S STEVENS SANDSTONE (CA, USA) : 4 PUENTE Fa, WILMINGTON FLO (CA, USA) PRESENT SeMLevEL | GEOLOGIC TIME (ny) L ‘5 PUCHKIRGHEN FM, MOLASSE BASIN AUSTRIA) GHACKBERRY FO. (TX, USA) “7YORKUN CHANNEL (TX, USA) SFRIGG FLD (UKCNORWAY BORDER, N'SEA) ‘9 FORTIES FLD (UK, NSEA) 1OMONTROSE FLD (UK, N SEA) $1.00 FAN (NORWEGIAN N SEA) S2BALDER FLD (NORWEGIAN N'SEA) CCHANGES IN 'SEALEVEL’ REFLECT INTERPRETED EXTENT OF COASTAL ONLAP Fig. 9.11 Apparent correlation of hydrocasbon-bearing submarine canyon and fan deposits with periods of low sea-level (after Shanmuga and Moiola, 1982) considered to exhibit a different sand body architecture, sand :shale ratio and facies. Basinloor and slope ns are considered to occur at different times during the relative sea-level eycle and are regarded therefore as separate, mutually exclusive entities. Mitchum (1985) has argued that good seismic evidence cexists ina number of examples to show a distinct difference in the timing of basin-lloor and slope fan development. This interpretation has been questioned in numerous paper in the segimentological literature, which suggest that basin- floor and slope fans may form the components of the same fan system, rather than being trae fans in their own right (Walker, 1978; Muttiand Normark, 19875 Walker, 19921 Kolla, 1993). To date, the evidence in favour of contem- poraneity between basin-floor and slope fan systems has. been equivocal because of the lack of stratigeaph ution in outerop data needed to resolve the timing of fan development. ‘The terms basin-floor fan and slope fan are misleading, and inappropriate for describing fans for two main reasons. Firstly, they refer to the seismic expression of a specific 194 architectural element ofa fan system, rather than represent= ing fans in their own right. Secondly, the terms assume that the spectrum of decp-marine clastic systems can be rep resented by two single depositional models (Kolla, 1993; Reading and Richards, 1994). “The seismic expression of a fan system is largely controlled by the acoustic properties of the dominant grain size, and lithology of the fan and its contrast with hemipelagie basinal sediment. As discussed earlier, the lithological character of the fan is controlled by sediment source area composition (Reading, 1991; Reading and Richards, 1994; Richards and Bowman, in press). Furthermore, coarse grained fan systems are composed of a different set of architectural clements to their finer grained counterparts, (Fig. 9.4). Sand-rich fan systems are expressed as a single seismic envelope (mound) developed within a basin-floor location, where the boundaries of the envelope define the whole fan system and its component architectural elements (Figs 9.6 and 9.14). This type of fan system would be referred 10 as a basin-floor fan. In contrast, mixed sand-mud fans contain architecturalreunvescatevercunve PM an shed rom til syst by sev! oweng San evry cat proceses pe fai mar pla Seman rl a theo of revous HST FALLING: eam profile = z.. A aaa ol : nelaiveses-oel-1| on a e-1= a < = hava vara a EARLY RISE. Incised valley = ‘Sediment slumped away P* 2 | foming writes uta cangon \ Tepbasin oor oratpe sea cena (causes vy snp a ridycorens) —c. °S coaencevongany Ute Das oor los chaaacTERisTics SPorGR SPorGR toss Loss SLOPE FAN | SLOPE FAN 2 SONDENSED snow SECTION Bas : FLG0R © . Fn 5 en | 2 > sequence seeeee! —eeeene—. SEGHBARY ustiTsT (se) BASIN FLOOR FAN ATTRIBUTES SEC GiaMAcTERIsTICs Latoralpinchout reflection geemory high-ampitude feleetons onlapring the basin margin; internal bidirectional Aovnlap of on tirecton abuts against downitvown side ota ‘you aut Eitle extemal mounding overving Ip | unconformity and below iowstand Slope ‘anor a condensed section Fig 9.12. Key characteristics and log responses of a basin-loor fan system (after Vail and Wornardt, 1990; Kolla, ‘SCHEMATIC CROSS SECTION 1° dip strike section UPPER BOUNDARY estan ‘ Hemipelagic sale or channel ‘Querban apron facies above Douncary «+ Sharp Boundary wih minimal ransiion INTERVAL + Turbiltes sands = amalgams ‘ands ~ Amalgamated massive turbidte ‘anos with shale breaks = Winer erosional surfaces Major erosional surface at top of fan May bs remnant fan mounds + Redeposited massive shingled sande “Contourte sends in separate mounds 1d massive trbidtes SEQUENCE BOUNDARY Massive sand above hemipelagic shale 1 Sharp boundary, limited orosion at base ‘SEDIMENTARY Los, GRanAcTeRIsTiCs CHARACTERISTICS ‘Sheet to shest-moundod, basinal turpidites, Massive amalgamated {@ facies, Mutt and Fic! Lucchi, 1972) to un-amoigemeted sande with ‘shale breaks (C, D facies) from central tomarginal aroas of basin, Deposited by unchanrelled and some channelled (wihout significant overbanks) turbidity currents; sometimes eroded by bottom currents. Corresponds mainly 12 type! but also typeI systems of Mut, (1985) Blooky log shapes vath sharp, flat base nd 1p 1993). 8P, spontaneous potentials GR, y-ray HST, highstand systems tract; TST, transgressive systems tract 195RELATIVE SEALEVEL CURVE Plan ew of channolevee lobe Lo CHARACTERISTICS spopor ay Lowsrano Proamavins cuannet, Steric senewaTic cRoss secrion ‘Stream profile Stream profile -2 Early tage of owstand data (stumping) Broxinal slope far (Canyon il) Sequence ‘oun overbank racic Sum LUePER BOUNDARY + Boma trian so + Ping-opwa dpi og patern {Fount abundane oak + Ser ite swat tten + Mutvsorey ean i 10 chan) {rie sndstanes tng upast Imudstone an ine-graned tubes asi Lowen BOUNDARY Foon + Henipalage sae wih foul Faner) "Sti tsar eeevee— SEQUENCE din SEQUENCE | «bouncy conforma, Le on $8.0 FF [SLOPE FAN ATTRIBUTES Ghanvaltil” Relave sea-ievel 2 ‘Sandisit —__ surace Pine-grained eee Tunes SS “op slope fan Distal ‘channel ‘Top basin floor stachea an surlace eves | hanna ‘Overbani | Distal turbos rode wunites ‘pre “Channa ‘Tsbidtes splay Sand Sstachos lobes Seismic cnanacrenionics 7 Gham ost ann ath nw ae Sorel imngh ama ‘etecone nee ovrarks > {enine cannes aral preneat smefneen 2 chee nda dicots (Sem staetne moneocsy Up plato paral sr asin nm moda oH Siu stro) Fig. 9.13 Key characteristies and log responses boundary; BFF, basin-floor fan 196 106 CHARACTERISTICS ‘Sonrcterearwints Soh tsk, ha minnng ns teny stapes ky aps wh ning pat haar n oar ‘faurchomea dune bate a to vers ‘poy corte, mai ln Sone we ryacana bt nao opens 2 Chole aks: tay ee ome by sat ‘anniv comes. ahs ‘apes ar sve posted Oy ‘nasinasing iaoveg sos 2: My exit agai comsning ra ‘rormany es 4 Shay nines SEDIMENTARY. HAAR TERISTICS ‘Chametovrnuinrcharet oe rte ‘oycorttel eae sans © ae tosone cence) ander sone aes ‘ttn cereal ovarian yore sacs ocravtober tbe lenin oF ena ‘wars tictreor compen exh ome Niveratle smarty chrocorsice eta allow war aye ccs momperaneoucy wih eee aerotee 1s of a slope-fan system (after Vail and Wornardr, 1990; Kolla, 1993) SB, sequencedlements that mimic the seismic expression of both basi: floor and slope fans (Figs 9.7 and 9.15; Normark, 1978; Droz and Bellaiche, 1985). The levee and channel-fll tlements may not be imaged seismically where their litho- logical characteris similar to the background hemipelagic sediments. Where sufficient lithological and acoustic differ- ences exist, the channel—levee system is imaged as a short, stubby g° seismic geometry dominated internally by poor to well-developed downlapping reflector sets, which pass laterally into a channol-fill characterized by either hummocky clinoforms (sand-flled) or parallel reflector sets {mudsilled channel). Down-dip, the constructional lobes ofthe fan are commonly expressed as a series of overlapping ‘ormutually exclusive constructional mounds characterized Jnernally by clinoforms or parallel reflector sets. In this ‘ase the depositional lobes of the fan form constructional ‘mounds located within the basin (basin-floor mounds) that can be traced up-dip into the channel—levee system. In conventional terminology the channellevee complex ‘would be referred 0 as a ‘slope fan’, whereas the basinal, time equivalent facies would be represented by ‘basin-floor’ lobes. For more mud-rich fan systems, the dominant architec- tural elements ofthe fan are large-scale (> 10 km) channel— levee systems, within an overall gull-wing geometry (Figs 98 and 9.16). The features show characteristic wedged. shaped seismic geometries, dominated internally by well developed downlapping reflectors oriented at right angles to the main axis of the channel systems (MeHargue and Webb, 1986; Kolla and Coumes, 19875 Weimer, 1990). The acoustic response of mud-rich systems therefore is, characteristic of a slope fan Basin-floor fans and slope fans om wireline logs ‘The application of wireline log analysis to basinal environ- iments, and che main criteria for recognizing major strat graphic surfaces in the basin are outlined in section 4.4.5 Criteria also have been developed to recognize ‘basin-floor” and ‘slope-fans" on the basis of core-calibrated wireline log response (Pacht et al., 1990; Vail and Wornadt, 19905 Vail ct al, 19915 Figs 9.12 and 9.13). Sequence boundaries are considered fo occur atthe base of log patterns indicating an abrupt increase in silt and/or sand overlying, hemipelagic shales. Sequence boundaries at the base of basin-floor fans ate easy to define by the abrupt juxtaposition of a blocky “box car’ sand-log pattern over ‘rail road” log patterns typical of hemipelagic shales (Vail and Wornardt, 1990; Vail et al, 1991). Where the slope fan rests directly on older sequences and no basin-floor fan exists, the sequence houndary is more difficule to identify and is placed at the firs significant increase in silt content above the hemipelagic shales. The sequence boundary separates the railroad track pattern of the underlying hemipelagic shale section from the crescent-shaped log pattern of the coarsening/fining Upward packages of the slope fan (Vail and Wornardt, 1990). ‘These models should be used with some caution, and they require calibration to core and seismic data before final interpretations are made (Kolla, 1993). Without cali- bration, log-based models for fan prediction are imprecise. More recent studies illustrate the relationship beeween log, character and the depositional elements of the fan system, based upon the integeation of core, wireline log and seismic data, These studies show that sand-rich fan systems show log responses characteristic of ‘basin-floor fans’, whereas, log patterns for mud-rich fans compare favourably with ‘slope fans’. This apparently simple relationship is invali dated in mixed sand—mud fans, where the system may show both *basin-loor fan’ and ‘slope fan’ log signatures Fig, 9.17). 9.3.3 Lowstand systems tracts and deep-marine clastic systems ‘Three principal models have been documented in the litera ture for lowstand systems tracts developed within shelf— slope, ramp and extensional-basin-margin settings (Mitchum, et al, 1990; Vail and Wornadt, 1990) In basins with a pronounced shelf—slope break, the Jowstand systems tract deposited in a basin is commonly separated into a lowstand fan unit and a lowstand prograd- ing wedge (sce section 2.4.3 and Fig. 9.18). The high sand: mud ratio lowstand fan unit is followed by a finer grained lowstand wedge dominated by slope deposits that have been described variously as slope fan, slope-front-fll, wedge and submarine fans. These slope-fan systems are characterized by active leveed channel deposits at the apex of the canyon mouth. In ramp margins, the lowstand systems tract is considered to form a tworpart wedge (see section 2.4.7 and Fig 9.18b). The lowstand wedge therefore comprises up-dip incised valley deposits, which may be traced down- tropical to temperate carbonate facies inthe northern Pacific (from Peas. CORAL-ALGAL REEF BRYOZOAN-ALGAL BIOSLASTIC DEBRIS Schlanger, 1981) clearly is not the whole story. In fact, the most rapid, but short-lived, sea-level rises (corresponding, to fifth-order, or xen higher, sea-level fluctuations discussed in Chapter 2) fan deliver the knock-out blow to a carbonate system, panicularly if this rapid sea-level rise occurs just as the system is beginning to develop, or ‘start-up’ (Neumann and Macintyre, 1985). As these authors point out, there is a sigmoidal growth pattern to organic carbonate systems, where the creation of new living space (or ‘accommodation’, (Chapter 2) is at first more rapid than the ability of the new population ro fill che space. Following this start up or lag” phase, the population growth may exceed the rate of accommodation creation inthe ‘catch up’ phase, and finally the ability of the population co grow will be limited by the creation of accommodation in the ‘keep up’ phase (Fig. 10.5). Figure 10.6 illustrates the start-up, catch-up and eep-up phases for an idealized isolated carbonate platform experiencing, transgression. ‘Once the carbonate-producing system has survived the start-up phase, it will grow vigorously if environmental conditions remain favourable, keeping pace with rising se level, which can lead to the development of strongly aggra- ational margins. This contrasts with siliciclastic systems, which will tend to backstep rather than accrete vertically during transgression. If, however, the rate of sea-level rise in the early startup phase is so rapid that the photic zone pulls away from the earliest start-up population prior to the catch-up phase, the carbonate system will drown, or 223PERIODS BIOHERMS MAJOR SKELETAL ELEMENTS a - CORALS TeRTiany 1 4 Rudists + Bryozoe 1o0-] CRETACEOUS RUDETS atoporeida K 4 conats Sponges JURASSIC = stromatoporoids eto Gas ~——Sronatoporoids | TRIASSIC ERNE ——— Coral Sponges 7 “TuBPAIES ————Coral Sponges 4 é Spang Tap Stara ——~ e PERMIAN 2 sco 8 CARBONIFEROUS Rees g 3 $ Devonian StROMATOPOROIDS Corals 4004 SILURIAN 4 STROMATOPOROIDS ‘Biyozoa onpovician. SPONGES Skeletal signe 500) ‘Skeletal algae ‘CAMBRIAN RRCHREOCYATHIDS {SKELETAL ALGAE Fig. 10.3 Simplified stratigraphic column indicating the major framework-building organisms through time (from James, 1983) step back to a shallower water position. As Schlager (1992) ‘comments, losing this race for the photic zone can extinguish carbonate platforms in the short term, but as Fig. 10.4 illustrates, longer term third- and second-order processes are unlikely to be able to drown healthy platforms. Some cther mechanism must be responsible for massive platform carbonate extinctions. 10.2.2 ‘Highstand” shedding ‘The *keep up’ phase of Neumann and Macintyre (1985) mentioned above is a period of platform development during which the production of carbonate exeeeds the space available (0 accommodate it (Fig. 10.6). Under such circumstances, produced carbonate may be redeposited elsewhere, in deeper water, on tidal flats and as acolianites.. The Bahamas carbonate platform at present is such an overproducing system, Neumann and Land (1975) were able to demonstrate that the production of carbonate from algal breakdown in a Bahamian lagoon far exceeded the mass deposited in situ. The excess production was probably redeposited on neighbouring tidal flats, and in deeper water. Redeposition into deep water of huge quantities of lagoonally produced carbonate also has been demonstrated 214 by two deep cores taken through a progeading carbonate ‘margin off the Bahamas. The seismically defined progeading clinoforms comprise in excess of 90% pellets, which were produced at water depths of less than 5~10m on the flat topped Great Bahama Bank. The importance of the shallow- water carbonate factory as a supplier for deep-water sedi ‘mentation also has been demonstrated unambiguously by other workers in the Bahamas (Droxler et al.,19833 Reijmer et aly, 1988), the Great Barrier Reef (Davies et al., 1989) and the Indian Ocean. Furthermore, ancient calciturbidite successions demonstrably deposited in deep water contain a very high proportion of grains that can have originated ‘only in shallow water (Reijmer et al., 1991). These obser vations suggest that deeper water carbonate accumulation can proceed to any great degree only when a carbonate system is healthy and actively overproducing sediment, and this situation will apply only when carbonate platform tops are flooded 10 a few tens of metres, such as the Bahamas today. Redeposition of carbonate from platform top tothe slope and the basin is referred to as ‘highstand shedding’. The term is a litle unfortunate in a strict sequence strat staphic sense because a carbonate platform will also shed sediment during transgression (Driscoll et al., 1991) and sea-level fall, but all other factors being equal, a platform"FREQUENCY OF PLATFORM ACCUNULATION RATES n=65 Holocene production rates and sea- 10" fev rise from Schlager, 1981, 1992) UW oF SIGMOIDAL GROWTH Tine Fig. 10.5 Population growth versus time, indicating ‘start-up’, ‘catch-up’ and ‘keep-up’ phases of Neumann and Macintyre (1985) = OP 3 g 3 A zg gat § e £ 3 i Sop 2 ° 10" 10 0 oF vot Coxon — Holocene sea-level 8 Long er subsidence z Recent cas asic Abs cco rs <5 sgzmipepatirms| ——_logna rats 0-207 2 Sai 8 Fig 104 Growth cates of carbonate Bahamas Re comes 3 gysens compared with relative seas ccm Ine ris. Solid bars represent the of al record, open bars represent ee : Ee 107 10 10% 10° Fates in mm/1000 years (Bubnof Units) will tend to shed most sediment during highstand as the rate of creation of accommodation declines (Chapter 2), and the platform top becomes partially by-passed by over- produced carbonate. In general, carbonate systems are likely to be less productive during lowstand, because the area of shallow-water carbonate production will be reduced. ‘The areal reduction of production may be enormous, par- ticularly on steep-sided platforms such as the Bahamas, (Fig. 10.7). Highstand shedding in carbonates provides ‘major contrast with classic sliciclastc sequence stratigraphy, which predicts that most sediment is shed into the basin during lowstand (Chapter 2). This contrast is illustrated on Fig. 10.8, which shows the highest and most variable basinal sedimentation rates of silicielastie systems occurring during glaciations when sea-levels were low, and highest and most variable basinal sedimentation rates for carbonate systems occurring during interglacials when sea-levels were high. In both eases the variability of sedimentation rate is, 21s‘STARTUP Sea-level foods 1} antecoden nigh Seat! t] and carbonate : produeton ated deepwater siliastios a ISS, TRANS. SERRE ED : Gneseve SvSTeMS ‘caTCH UP year | See seve t | oducton racks ‘Sestevel || fing sea-level ‘and bus aggroatoral rargin SARA noche ‘KEEP UP’ ® Catborateprotucton 4] srseds toate HIGHSTAND eaters t] creraton of Systems sccommodaton space TRACT artoate le shed ‘the plator op LRN tothe ope and basin LILIA EROS sequence soe ‘SUBAERIAL EXPOSURE inet cabonaia producton pe {erminsod, part Lowstano from minor eng SYSTEMS reefa(a). Platform TRACT tops larstedin tui ate) PLATFORM DROWNING nonmetal deterioration may. PLATFORM, ‘cause platiorm to DROWNING. cease recicton (DROWNING fcc Crowned | UNCONFORMITY tzom rae e 10.6 Schematic model for an ecard cowniaoped isolated carbonate platform, showin ‘by prograding 7 Piafom idealized systems tract geometries and platform drowning probably due to turbidite-controlled sedimentation, rather than a uniform, continuous rain of pelagic sediment (see the carbonate ooze sedimentation rates for comparison). Another important point is that during low-stands, car- bonate systems will not experience the same type of physical erosion as siliciclastic provenances. During subaerial ex- posure, carbonates will tend to be eroded chemically, result- ing in dissolution of grains and cements, and reprecipitation Of dissolved material (see further discussion in section 10.2.4}. Relatively small quantities of eroded carbonate will be generated during lowstand and the most likely source of lowstand carbonate is sediment production from growing carbonate, such as fringing reefs (Fig. 10.7), or 216 areally smaller platforms that have down-stepped during sea-level fall. ‘Although highstand shedding is the main way of con: tributing carbonate sediment to the basin from steep-sided platforms, on lower angle ramps and rimmed shelves, lowstand basinal carbonate deposition can be significant. Possibly the best-constrained example has been demon- strated by Jacquin et al. (1991) in the Cretaceous of the Vercors, France, where lowstand carbonate fans shed from downstepping rimmed shelves are developed. The lowstand fans have a different grain composition from the highstand shedded material.Fig. 10.7 Cartoon of highstand progradation compared with Yolumetrcally smaller lowstand wedge, Bahama platform, Temiaey (after Eberli and Ginsburg, 1987) 102.3 Carbonate platform drowning, As wall as starting, catching and keeping up, carbonate sysems will also give up. Rapid, but short-lived sea-level tics may outpace carbonate platform growth (section, 102.1), but once the rate of sea-level rise slows down, all ther things being equal, the carbonate platform may be restored to productive health. However, there are several ‘vironmental factors independent of sea-level change that ay seriously impede carbonate development. As well as the appropriate light and temperature requirements, there are at least five additional controls that need to be con- sidered; nutrient supply, clastic input, oxygenation, salinity and predation. Table 10.1 lists the major causes of carbon- ate platform drowning, with examples from modern car bonate systems and the geological record. For more detail Niantic > Inergiacial ‘Sedimentation rates in mm per 1000 years Fig, 10.8 Highstand and lowstand basinal sedimentation rates for carbonate and silicielastic systems {from Droxler and. Schlager, 1985) ‘on mechanisms of platform drowning the reader is referred to Schlager (1992). All the mechanisms described in Table 10.1, individually, ‘or in combination, or in concert with rapid sea-level rise, can cause a carbonate platform to drown. In the fossil record, the term ‘drowning unconformity’ has been coitied (Schlager, 1989), and is shown schematically in Fig, 10.6. This ‘unconformity’ can have developed only when the carbonate platform top was flooded, ie. within the trans- {gressive or highstand systems tract. Drowning unconform- ities have several distinctive properties, which, in detail, differentiate them from sequence boundaries. Firstly, there is usually a very rapid change from shallow-water carbonate to deep shelf, slope or basinal deposits. Observations by the author of cores taken from the top of a Miocene carbonate platform, offshore central Vietnam, show a change from shallow-water platform carbonate sedimen- tation, through a drowning ‘sequence’ of deeper water Foraminifera, with evidence of marine condensation, in- cluding glauconite and phosphates, and finally into basinal raudstones. This transition takes place over several toa few tens of metres, but represents several million years worth of sedimentation. A similar succession has been described from the top of the Liuhua carbonate platform, Pearl River Mouth Basin, offshore China (Erlich et al., 1990). Here, the transition is more gradual, with the drowning unit 217‘Table 10.1 uses of carbonate platform drowning ‘marion supy lenient enironment ate most vourabl for erganic carbonate gronth arid reels Inhigh rier settings its apparent that corbonateseretng, framenor-bulding coal are replace by feshy alge, sponges ost cara Mosern examples ice the East aa Sea, Indonesia, whee coral are saree bow 5m and are replaced green algae. The ack of ree inerpreted tobe the esto upueling rutintch waters, which stint the growth of coral competitors bees and Phipps, 1888) Sic nut lay partes in suspension te water column wil inhibit ight penetration, Sntcay eucng or preventing reef rowah. nation, coral po} are, and many predecessor framework bling forms wer, unable to ope wih day partes during feeding, Ccurstantal evidence of carbonate deme by clastic ‘put nthe geclogia werd i demonstrated by seovnrd stepping Macene Carbonate sytem fem south to north aro the Laconia Province, ofshore Sarawak, Malays, aay trom northward prograding dita pte (ping, 1969; Figure 129}, Note that changes water saint ornutrent content esocated with a sliedastc depositional stem such as deta, may be the prime cause of carbonate production demise Sali vriaions Changes of salty may exert a dramatic effect on carbonate producti, paniculata of reas. in the Hlocene, 02d, shallow Lagos on fst topped Carbonate platforms acted highly viable sales. Te seaward ebb fw of thes mimi bank wnt’ Hed theres, which aso became overvheles by Seward anspor of carbonate ies rom the sgoons Neuman and Mace, 1985) ‘exygenavon {Wen ongen & depleted, ees wil die, However, the est evidence for masve caronate demise esting fram oxygen decency is cicumstantal Schlager and Php (1950) have used evidence for Ceteceous oceric anata and carpe this with the frequen af Cretaceous carbonate platform deme to nda @ posible causal rk Predation ‘The importance of predation on carbonate communities weltocumented ‘rom meer day examples. Te ‘Crown of Thos stash blight, wich tiveatened to wige ut lari tracts of the Great Baer Reet s won, However, iti difcut, not impossible, to demonstrate desl the effect of predation inthe fol record, anf curentyunknowm on the scale of an entire carbonate platform heralded by the appearance of a red algal boundstone awhich passes laterally into thodolith beds of low faunal diversity. This red algal interval was probably deposited in several tens of metres of water depth, too deep for prolife coral production. The final expiry of carbonate sedimen: tation is shown by a deepwater condensed section, whichis in turn succeeded by basinal muds. Drowning unconform- ites are also recognizable on seismic data. Figure 10.10 shows a seismic line from the Atlantic margin of the USA, where a flat-topped Early Cretaceous carbonate platforms downlapped by Late Cretaccous to Tertiary siliciclastis “The latter interval must have been deposited on top of the carbonate when sufficient accommodation had been gener- ated on top of the drowned platform to produce the iciclastic slope. ‘An intermediate version of complete platform drowning is backstepping or retrogradation, This may take place on the scale of a single platform, where the platform margin retreats away from incoming siliciclastcs, such as in the Devonian build-ups of westem Canada (Stoakes and Wendt, 1988; Scaturo ef al., 1989), or on the scale of a carbonate ‘province’ such as the Luconia shoals, described in Table 10.1. Backstepping of carbonate margins may precede ‘complete platform drowning, such asin the case of Liuhus, where the margin effectively shrank to the red algal bound- stone described above, before its final demise (Fig. 10.11), In short, the most important point abour platform drowning is that it can be controlled entirely by environ- imental factors; sudden rapid rises in sea-level may outpace carbonate production temporarily, but a deterioration in the environment of the carbonate factory through any of the mechanisms discussed above is a more likely cause of large-scale platform demise. Drowning will take place only when the platform top is flooded, within the transgressive or highstand systems tracts. 10.2.4 Carbonate platform exposure Subaerial exposure is an obvious alternative method of a, _ MAGNETIC. BASEMENT Fig. 10.9 Tertiary carbonates of the Luconia Province, offshore Sarawak. The carbonates young towards the north, as they backstep away from northward directed siliciclastic input (from Epting, 1989) 218WILMINGTON PLATFORM (©) Nw Silicictasties Crot.— Tort. Platform Late Jur. TE, Cret, MILE 7 Ih = = — Drowning unconformity —— bownlapping reflections 40 mat : + Courtesy Shell Oi Co. Elevated platform margin with drowning unconformity on seaward side Fig. 10.10 Drowning unconformity on the Wilmington Platform, offshore USA (seismic profile courtesy of Shell Oil Co.; from Schlager (1992)) Killing a carbonate platform. Because shallow-marine carbonates build at, or close to, sea-level, they are highly susceptible to frequent, and in some cases, long-lived subacrial exposure. The effect of exposure on carbonates is firstly to turn-off large areas of the carbonate factory and then to erode chemically the exposed carbonate. Depo- Sitional carbonate mineralogies are highly unstable, and will experience alteration in the presence of meteoric water, which is acidified by atmospheric and soil carbon dioxide, aswell as humic and other biologically derived acids present in soils. Modern marine carbonate is composed chiefly of aragonite, a preferred precipitate of many marine organ- isms, including corals and molluscs, and high-magnesium, calcite, precipitated by echinoderms, benthic Foraminifera and red algae. Dissolution of these grains, and reprecipi- tation as more stable low-magn during meteoric diagenesis, leading to a redistribution of porosity and the formation of zones of considerably, enhanced porosity and permeability. Where meteorie dia ‘genesis is very intensive, extensive cavern systems, fissures 219Fig. 10.11 Red algal mou Mouth Basin, South China Si (from Erlich et al, 1990) and subterranean streams may form, giving ise to karst (Trudgill, 1985). Karst may provide major secondary porosity and permeability in carbonate reservoirs, and its recognition and prediction is an important practical con- sequence of identifying carbonate platform exposure (Warson, 1982; Loucks and Anderson, 1985; Craig, 1988). Nor all subacrial exposure will result in meteoric dissol- ution or karst formation. In more arid climates, where rainfall is limited, exposed carbonates will suffer limited dissolution; instead, carbonate precipitation may give rise to the formation of ealerete crusts, needle fibre calcite and thizocretions, the last formed by preferential cementation around plant roots. For a more detailed discussion of subaerial-exposure-related diagenesis, the reader is referred to Longman (1980), James and Choquette (1988) and Tucker and Wright (1990), ‘The most important point from the discussion of earbon- ate exposure is that most of the processes acting on carbon ates during lowstand are chemical. In general, the physical products of carbonate exposure are volumetrically small, although collapse around cave systems and along major fissures can result in carbonate debris cones, and can even control the shape of carbonate margins (Back e¢ al., 1984). Ifa carbonate platform system is attached to a siliciclastic source area, lowstand may result in siliciclastic transport, across the exposed carbonate edifice and into the basin. Examples of this include the Bone Spring Formation of the Permian of New Mexico (Saller et al., 1989), where sand- stones onlap the carbonate slope, and the Carboniferous of ‘Anglesey, where fluvial channels incise into shelf carbonates, (Walkden and Davies, 1983). In climates that are too arid, to sustain major fluvial systems, evaporites may be precipi- tated close to the surface of the exposed carbonate, and, ‘where arid conditions are combined with a restricted basin, subaqueous evaporites may form, onlapping the base of the carbonate slope (Tucker, 1991). On isolated carbonate platforms, remote from a source of siliciclastics, the products of lowstand are likely to be meteoric diagenesis, ‘with associated porosity redistribution, and in sitw carbon- 220 backstepping and preceding final drowning on top of the Liuhua carbonate platform, Pearl River ate deposition as fringing reefs, shown schematically on Fig. 10.6. Feinging rcefs also may develop on attached platforms, but may be affected adversely by silicilasie input of evaporative conditions. Note that the volume of in situ carbonate produced during lowstand will be highly dependent on the slope of the carbonate margin. This is discussed more fully in the following section. 10.3 Carbonate slopes, platform classification and facies belts 10. 1 Slopes and platform classification ‘The geometry of a carbonate platform will exert a major control on its response to sea-level change. For example, on steep sided, isolated platforms the area available for in situ owstand carbonate sediment production will be very small, restricted, at least initially, to a fringing reef around the platform (Fig. 10.12). By contrast, gently sloping carbonate ramps will maintain large areas for carbonate production during sea-level fall. Bevween these two extremes lie 2 range of carbonate platform types defined on two main criteria, (i) the slope angle and presence or absence of a slope break, and (ii) whether land-attached or isolated. The classification scheme to be used here follows that of Tucker and Wright (1990), with minor modifications. Carbonate ramps are defined as low-angle (<1*) homoclinal dipping surfaces, without a significant break in slope (Ahr, 1973), although Read (1985) has described many variants on this theme, Rimmed shelves show a break in slope, and have slope angles between 1° and vertical. Escarpment margins are a subset of rimmed shelves, with the margin having angles of 35° to vertical. These platforms types may be land-attached or detached (isolated). Detached platforms ‘or ‘atolls’ generally maintain steeper slopes, although such platforms occasionally may develop ramp margins. ‘Although the above platform categories are convenient for classification purposes, in reality there is a continuum between the different margin types, which has a direct,(Woden Bahamas) (Neogene Bahamas) Lonstane wedge Fig. 10.12. The importance of platform slope angle and rometry on the volume of in site lowstand carbonate provluction, Steep-sided platforms will permit only a relat sinall area of carbonate production during sea-level fall, inhereas broader ramps can maintain much larger areas; fearing on unravelling their sequence stratigraphy (Schlager, 1992). The platform margins of Fig. 10.13 show net accumulation on the Blake transect, bypass of the Exuma transect, and net erosion of the Bahamas escarpment. All of these geometries can be produced during platform growth, finan identical sea-level regime, yet the bypass margin has given rise to a sediment package onlapping the base of slope, and the erosional margin shows truncation of the ‘margin and accumulation of a mounded package on the basin loor (Schlager and Camber, 1986). It would be an ‘unwise seismic stratigrapher who attempted to relate such geometries purely to sea-level change! Net crosion and a ‘mounded basinal package docs not necessarily correspond tolowstand. Purdy (pers. comm., 1992) has suggested that erosion of the margin and development of the steepest slopes are more likely during transgression and eatly high- stand, when a carbonate system is aggrading and the steepening flanks, subjected oan overlying load of accumu- lating carbonate, collapse into the basin The angle of a depositional carbonate margin will also be controlled by the type of sediment supplied to the slope | and the sedimentary process both on and off the platform. Kenter (1990) has shown that coarse-grained sediments, can maintain much higher angles of repose, and hence steeper slopes, than finer grained deposits. He was also able to show that cohesionless coarse-grained sediments could build slopes up to 35°, whereas muddy cohesive sediments tended to maintain low slope angles characterized by repeated massive slumping (Fig. 10.14). The type of sedi- EROSION Bahamas escarpment 225 NEUTRAL (Gand bypassing) Exuma taneort 112 ACCRETION et ee Fig, 10.13 Erosion, bypass and accretionary carbonate margins (Irom Schlager and Camber, 1986). ODP, Ocean Drilling Program ment supplied to a slope will be dependent on the type of carbonate factory, and on the orientation of the slope with respect to current regimes. For example, the carbonate platforms of the Italian Dolomites show slope angles up 0 30°, some of the steepest examples on Fig. 10.14, because of gravel-sized particles of calcisponge supplied from the carbonate factory. This contrasts with the low slope angles evident from seismic sections across the muddy northern flanks of the modern Little Bahama Bank, The effect of current activity on carbonate platforms will also influence sediment supply t0 the slope and basin, On windward margins, wave activity will tend to push most of the fine sediment on to the platform, leaving the coarser fraction 10 accumulate on the windward slope. Leeward margins will tend to accumulate finer sediment and, in general, build shallower slopes (Hine et al, 19815 Tucker, 1985). 10.3.2. Facies belts on carbonate platforms Using the general platform classification described above, idealized facies belts for ramps, rimmed shelves and escarp: ment margins are outlined below. Although carbonate- producing organisms have evolved through geological time, 21Slope angle in degres ROMA | wuo.sueroRrED Dominant oximent fabric Key: O wer-ocumentatod examples 1 Examples lacking prose contol en geometry A. Planks stabized by orgie ramebulsing or cementation shallow-water carbonates have tended to develop broadly similar facies belts through time. Carbonate ramps Ramps are low-angle seaward dipping surfaces, with no continuous elevated rim or clear break in slope. The tran- sition between low energy, lagoonal facies through high- energy grainstone or reefal facies to deeper water facies is thus very diffuse, and ramp sedimentation may be domi nated by basinal processes (storms and tidal currents) over wide areas. The generally uniform dip on a ramp surface is maintained largely by sediment redistribution by hydraulic processes across its broad but shallow expanse. Deeper ramp facies commonly contain sediment of shallow-water origin transported offshore during storm surges (Aigner, 1984; Wright, 1986). This ‘regrading” process is important in allowing the ramp to maintain its inherently unstable geometry. Ramps are often seen to develop into rimmed shelves as the differential build-up potential of the more paces that of adjacent deeper and productive margin shallower water f The idealized facies belts of a land-attached ramp are 2 Fig. 10.14 Sediment composition versus slope angle on carbonate margins (from Kentet, 1990) illustrated on Fig. 10.15. A broad area of peritidal carbon. ates, controlled largely by climatic and hydraulic setting, comprising evaporites and tidal flat deposits is developed adjacent to the ramp topography. Evaporites and evaporiti structures, such as tepees, laminated crusts and enterolithic, folds (Shinn, 1983; Scoffin, 1987; Tucker and Wright, 1990) are characteristic of tidal flat deposition in arid areas, but in more humid climates evaporites may be absent, replaced by soil and palacokarsts developed upon subaerially exposed beach and lagoon deposits, such as in wrrasneir | cangowwre PERT Stara ‘Below FWWB Waveragitaied | Low energy Salon) Bee} Oavsand Oa matonsaoe leentasnsle | Faia | stoas”” | ipa ta Fairwocther wave base Fig. 10.15 Carbonate ramp facies model (after Read, 1982), FIWWE, fair-weather wave baseCarboniferous of South Wales (Wright, 1986). Seaward the tidal flat environment, lagoonal sediments pass into wave-dominated ramp margin, a diffuse belt of bio- ic and/or oolitic shoals, up to several kilometres in jth, Reefs also may develop in this area, although most ‘eeh in the modern Arabian Gulf ramp system tend to ‘ccupy submarine swells and areas between active ooid sand sedimentation (Purser, 1973). No slope environment “such exists on a ramp, and the deeper ramp environment Ischaracterized by muddy facies, often with well-developed ‘toa beds of originally shallow platform sediment devel- ‘ped above storm wave base. Rinmed shelves Carbonate rimmed shelves havea distinct slope, with angles ‘of upto 35°, separated from a broad, shallow platiorm top. swith carbonate ramps, the landward portion of modern timmed shelves may comprise evaporitic or mangrove! imirobial peritidal facies, depending on climate. The major Uifferences between rimmed shelves and ramps are (i) the thinner margin of a rimmed shelf, often only several hun ted metres or ess in width, and (i) the presence of a well- defined slope on which slope sedimentation processes act. atlized facies belts for a land-attached rimmed shelf would thus be intertidal and lagoonal deposits passing seaward into a well-defined margin, comprising. recfal build-ups and/or oolitic and bioclastic sand shoals (Fig. 10.16), The lagoon itself may be subject to wave and tidal agitation where gaps in the margin exist. Alternatively, ‘only occasional agitation may affect the lagoon and it may become hyper- or hyposaline depending on the climate and runoff, On rimmed shelves, the margin isa very turbulent zone, where storm and oceanic waves and tidal currents impinge on the sea-bed. This zone is much better defined than on a ramp, where shallowing towards the margin is, ‘more gradual and wave energy is dissipated more gently ‘over a wider area. Scaward of the rimmed shelf-margin, proximal slope facies may develop, comprising large reef talus blocks. These pass downslope into debris flow deposits, and turbidites, which fine distally. The deeper rimmed shelf facies may consist of very distal turbidites, interbedded swith basinal mudstones of pelagic or hemipelagic origin. Escarpment margins Escarpment margins can be thought of as rimmed shelves ‘without a significant accretionary slope facies (the “escarp- ment by-pass" type of Read, 1982), with proximal lagoonal, inter- and supratidal facies belts, as described for rimmed shelves in che preceding section. On such margins, at slope angles in excess of 35~40", most sediment will tend to bypass the upper slope and accumulate on the lower slope or in the basin (Fig. 10.13). A steep shelf-edge diff is developed through rapid aggradation, which may be partly eroded by gullies down which shallow-water carbonate debris passes. Alternatively, the cliff may be highly eroded, cautting back into older carbonates. A good example of the latter are erosional shelf-margins off Belize, where Pleis- tocene limestones are being exposed along the erosional scarp (James and Ginsburg, 1979). The sediment accumu- lating at che base of the escarpment margin is usually coarse-grained and disorganized, and may form a con- tinuous apron of peri-platform talus, particularly where Faisw—] cmewsex | peer sneur” | Fonesioee”] onsnwc, | wawowe5 [auELr Goon] nEsrncTER. [EVPOATES 1 sae | Rie | SURB | MOSER | grey Gincassrise | SW Expas - Siidbumon | SEEPS” | hunks moeon.7s | a EA ee o SLA ee g 99 COR i Fence | mes, somos [Eterece [noua | Fenws, Ehimares | Ey Saar fia | Siete te seat, ora” 5 ee ue sums [28 Fig 10.16 Standard facies belts on a rimmed shelt (after Wilson, 1975) 223the shelf margin is highly erosional. Deeper into the basin, more organized proximal carbonate turbidite facies may develop, which pass basinward into distal curbidites inter~ bedded with pelagic and argillaceous carbonate (Fig. 10.17). Isolated carbonate platforms Isolated platforms are areas of shallow-water carbonate accumulation surrounded by deep water. Modern isolated platforms range from Pacific atolls a few kilometres across to massive areas of carbonate production such as the Bahama Banks, 700km from north to south, Isolated platforms may develop any of the three categories of margin described above, although steeper rimmed shelf or escarp- ment margins tend to be more common, The major differ~ ence between large and small isolated platforms is the area of overall carbonate productivity. In both eases, the margin usually will have roughly the same dimensions, but on larger platforms the extensive carbonate production in the lagoon will provide a massive additional source of sediment to shed into deeper water. 10.4 Sequence stratigraphic models for carbonate platforms 10.4.1 Introduction The foregoing sections hopefully have convinced the reader that carbonate systems may behave quite differently from siliciclastics, and that a modification of the standard silici clastic sequence stratigraphic model is necessary to encom pass carbonate idiosyncrasies. To this end, a seties of general sequence stratigraphic models, with examples from outcrop and the subsurface, are outlined in the following sections. This section also draws on previously published pe ules, ' Lower slope slide sear, 34 blocks, debries 33 bros, z Upper lope Base of ‘slope apron Basin too: pelagic and hemipetagic turpis | BASE OF SLOPE APRON 24 carbonate sequence stratigraphic models; Sarg (1988), Calver et al, (1990), Tucker and Wright (1990), Hunt and Tucker (1991) and Tucker (1991). Each carbonate platform type will be presented in an idealized sea-level cycle of transgression, highstand and lowstand or drowning, using the facies models described in section 10.3. Climate also exerts an influence on the sei mentology of systems tracts, particularly lowstand, and the important differences between arid and humid lowstand systems tracts will be presented. It is important also to bear, in mind that for all these models, only the gross systems- tract geometries will be discussed, ie. those visible on exploration seismic data (Chapter 3) or at mountainside outcrop scale (Bosellini, 1984). The finer scale cycliciies observed on a several to a few tens of mette scale will be the subject of section 10.5. 10.4.2 Ramps Once a ramp system (Fig, 10.18) is established, rising sea level in the transgressive leg of an idealized sea-level cycle ‘will commonly cause the ramp to step landward because of its low carbonate productivity. Both the coastal onlap and downlap points will migrate landward, with a progressive superimposition of deeper-water on shallow-water facies ‘Atthis time, the deeper ramp may become sediment starved and as water depth increases over the deep ramp, the potential for significant accumulation of organic mater increases. An example of petroleum source-rock accumt- lation during ramp transgression may be the distal facies of the Jurassic Smackover ramp system in the Gulf of Mexico (Moore, 1984; Sassen, 1988), which forms a major source- rock interval. More proximal facies of a ramp transgressive systems tract are shown by the Jurassic Lincolnshire Lime: stone Formation in eastern England (Ashton, 19805 Emery Fig. 10.17 Simplified facies model for a carbonate ‘escarpment margin, showing bypass of the upper slope and sediment accumulation at the toe of slope and in the basin (after Mullins, 1983)and Dickson, 1991), Here, the carbonate ramp was initiated ‘on estuarine muds and silts as pre-ramp topography was inilled during transgression and clastic facies belts were pushed landward. The carbonate succession, only up to 440m thick in total, comprises oolitic and bioclastic shoal facies overlying back-barrier and lagoonal facies, indicating overall transgression (Fig. 10.19). Other examples of land- yard stepping ramp systems during transgression include the Lower and Middle Anisian ramps of the Lower Muschelkalk, Catalan Basin, Spain (Calvet et al, 1990), and the Upper Muschelkalk ramps of the intracratonic German Basin (Aigner, 1984), ‘As the rate of sea-level rise decreases, carbonate ramp systems will tend to prograde basinwards. In early high- stand, when accommodation is still being created landward ofthe margin, supra-, intertidal and lagoonal deposits may accumulate as topsets, but during late highstand, very little ‘6r no topset deposition will occur, and a ‘strandpl tamp may result (Tucker and Wright, 1990). The seismic expression of the highstand systems tract of carbonate ‘Supraland inertial facies Lagoon! facies jnal shoals ot sea-level 0) Deep ramp foc Assan siiciclastios Salhi evaponites See-lovel ‘Sequence boundary ramps is typically as a series of shingled clinoforms, with rarely resolvable topsets, such as the Smackover Formation (Moore, 1984), Lower Permian ramps of the Midland Basin, Texas (Mazzullo and Reid, 1989) and the Upper Devonian Nisku Formation, Alberta, Canada. In the latter example, deep ramp build-ups sitting on low-angle clino- forms are the main exploration targets. In outerop, pro- gradational oolitic facies are recognized from the Castell Coch Limestone of the Carboniferous of South Wales, (Burchette, 1987), the Muschelkalk of Europe and the Middle Jurassic of France, Spain (Aurell, 1991) and southern England, Daring falling sea-level the landward portion of the ramp will be exposed. Where clastic sediment supply is low. and the ramp system was prograding into an open shelf setting, the ramp may offlap, stepping downward such that its onlap point falls below the earlier ramp margin. Alterna. tively, if the siliciclastic sediment supply is significant, it may overwhelm carbonate production and a lowstand siliciclastic system may be established across the ramp © Sea-tevel a se al 181 Gea-tevel Drowning unconformity Fig, 10.18 Sequence stratigraphic models for ramp systems. (a) Transgressive systems tract shows landward stepping of ramp faces and sediment starvation in the basin, with the potential for the development of onganic-rich mudstones. (b) Highstand systems tact shows seaward progradation of the margin and progressive thinning of topsets. In late highstand, ropsets may not be developed and a ‘strandplain’ system (cf. Calvet et al, 1990) may result. (c) Lowstand systems tract in an arid, restricted basin, Het, the basinal facies is developed as a subaqueous evaporite wedge onlapping deeper ramp facies of the previous highstand systems tract above a sequence boundary. The exposed ramp may be the site of extensive sabkha evaporite development and may beencroached upon by aeolian silicilastics. Ina humid setting, the exposed ramp may be incised by fluvial channels and karstified, ofthe sliciclastic input is relatively low, a new lowstand ramp may nucleate below the margin of the previous ramp system. {d) Drowning of a carbonate ramp followed by siliciclastic progradation over the drowned ramp. The ramp and overlying Slicclastic wedge will be separated by a drowning unconformity 225pray log ° oor 131 ee [eal ete] | =]! ater _ ay [opment 3 Nightane ~ ig 3 propedaton Be ser eee wd e 3 3 22 om Lincolnshire Limestone | & | Fig. 10.19 Summary log and facies ees, erteroe model for the Lincolnshire Limestone 28 | | werwingon sence Formation, eastern England. The ae pecans! Jowstand system tract (LST) is 52 ‘reat Merer «. dleveloped in a clastic system, the ZB | | ssemeneneer estuarine Grantham Formation, which BB |e is succeeded by a landward stepping 8 ‘Gretna Meat transgressive systems tract (TST; the 4 Lincolnshire Limestone}, as shown by the superimposition of more distal fdotere 50. 7 shoal facies over lagoonal and back: ialathalnForrakot fate) nals barrier sediments (after Emery and ust (Estuarine facies) Dickson, 1991). HST, highstand systems tract surface. Note that, in general, non-skeletal carbonate margins are better able to persist during phases of silici- dlastic input than reef-dominated margins. An example of lowstand siliciclastie deposition replacing carbonate ramp deposition during lowstand is from the Carboniferous of Anglesey (Walkden and Davies, 1983). In more humid climates, proximal ramp deposits that become exposed during lowstand will develop soils and karst surfaces. Ramsay (1987) has described palaeokarsts developed on Carboniferous ramp oolites of South Wales deposited during slight sea-level falls. By contrast, more arid environ- ments may develop caleretes or extensive sabkha evaporites ‘on the exposed ramp, with encroachment of aeolian facies if the climate is too dry to develop fluvial systems, Mixed silicilastic and evaporitic lowstand systems tracts on exposed carbonate ramps have been described by Calvet et al. (1990) from the Triassic of Spain. In an arid climate, basin isolation may occur during lowstand, resulting in evaporation. A lowstand evaporite wedge may onlap the distal ramp facies with cessation of carbonate production because of hypersaline conditions. The idealized sea-level cycle described above does not account for drowning on a carbonate ramp, which may 226 take place during transgression or highstand due to environ mental deterioration, Unequivocal examples of drowning (on ramps are not numerous. Perhaps the best-documented example is again from the Triassic of the Catalan Basin in Spain (Calvet and Tucker, 1988). Here, deeper ramp ‘mounds show evidence of exposure during an abrupt sea- level fall. Subsequent transgression led to the development of a stratified sea, with oxygen-depleted waters. Instead of nntinued -mud-mound growth, organic-rich carbo accumulated. Strictly speaking these Triassic mud mounds ‘were not really drowned, but rather failed to re-establish themselves in the transgression that followed sea-level fall and karstification, 10.4.3. Rimmed shelves Carbonate rimmed shelves (Fig. 10.20) may be initiated in several ways. They may start up during transgression over a non-carbonate substrate, particularly over a pre-existing, shelf break or structural high. In so doing, a barrier may become established whereby wave energy can be dissipated and lagoonal and inter-and supratidal carbonate deposition can proceed in the protected lee of the barrier. Rimmedako may be initiated over ramps. As transgression es, the productive margin may be the only element the platform able to keep up with sea-level, and a Idefined slope may become established. At this time, mal carbonate production also may be outpaced by sea-level, leaving a deepening hole in the lee of the ‘och as in the present-day Belize lagoon, More com- y observed, however, is “keep up” of the entire rimmed up to and including the margin, to give a flat-topped tate rimmed shelf overlying a low-angle ramp. Excel ‘examples ofthis transition include the Lower Permian the Midland Basin, Texas, where Wolfcampian ramps ‘overlain by rimmed shelves of the Wichita and Lower ear Fork Formations (Mazzullo and Reid, 1989 Fig. W021), In outcrop, the ramp-to-rimmedshelf transition Jserved in the Triassic of the Italian Dolonites is quite cular (Fig, 4.5). One of the key differentiating factors ween silcielastic and carbonate transgressive systems isthe ability of the carbonate to aggrade, rather than step during rapid sea-level rise. Aggradation, and even progadation, during transgression is common in rimmed heives (Pomar, 1991). However, if transgression was to fontinue uninterrupted, with no change in carbonate productivity, the rimmed shelf would be unable to keep Iuilding vertically and maintain its slope angle, and would lool into an escarpment mars During highstand the rimmed shelf can shed ove produced carbonate on to the slope and into the basin. In fcadly highstand, topsets may continue to accumulate as Ingoonal and inter- and supratidal facies. However, during Intehighstand, topsets may be poorly resolved, particularly ‘on exploration seismic data, It also may be impossible to identify imum flooding surface in the platform imerior facies because of the lack of backstepping in the tnderlying transgressive systems tract. However, seaward ‘of the aggraded rim of the transgressive systems tract it may be possible to identify the major downlap surface and hence maximum flooding surface of the rimmed shelf interval, but this is often equivocal, Examples of pro- grading rimmed shelves are common, and are particularly well shown on scismic data. Figure 10.22 shows examples from the Tertiary of the Dampier Basin, Northwest Shelf, ‘Australia, Progradation of rimmed shelves in outcrop is shown superbly by the Triassic of the Ialian Dolomites (Fig. 45). Sea-level fll ona rimmed shelf, as on a ramp, may result in a range of different lithologies, facies and geometries. Salle ef al. (1989) have shown major clastic input during lowstand in the Permian of New Mexico, where a series of fowstand sandstones interdigitate with carbonate slope deposits in the Mescalero Fscarpe Field. All the sands appear to sit above type I sequence boundaries. The importance of lowstand evaporite precipitation both in the topsets of rimmed shelves and as wedges onlapping carbon ate slope facies has been described by Tucker (1991) for the Permian of northeastern England. Here, aridity restricted significant fluvial input during lowstand and hypersalinity prevented further carbonate growth. Karst formation on rimmed shelf topsets in more humid systems is also highly significant as a porosity modifying provess. Examples include karstic carbonate reservoirs of the Gargino Peninsula, eastern Italy, where spectacular cavern and fissure systems provide the porosity, and, most significantly, the permeability for subsurface oil reservoirs such as the Rospo Mare Field (Zeza, 1975). “Type 2 sequence boundaries and associated shelf-margin wedge systems tracts also may be developed on attached rimmed shelf systems, where sea-level falls over the platform top, but sediment accommodation remains at the platform edge (Sarg, 1988, and Chapter 2). Shelf-margin wedge systems tracts are likely to be developed only if there has been differential subsidence of the top of the flat-topped carbonate platform to create a dipping surface on to which the shelf-margin wedge can onlap back on to (Fig. 10.20). Note that the shelf-margin wedge systems tract will not be developed on ramps, because ramps lack a shelf margin (Calvet et al., 1990), Drowning of rimmed shelves is better documented than for ramps. The example of the Wilmington Platform is illustrated on Fig. 10.10, where a silicilastic system overlies 4 drowned, flat-topped rimmed shelf. Another example is, the Jurassic of Jebel Bou Dahar of the Moroccan Atlas Mountains, where outcrop exposure shows a carbonate margin and slope onlapped by basinal shales, with evidence for progressive environmental deterioration on the platform top provided by a decreasing diversity and productivity of carbonate (Campbell, 1992). 10.4.4 Escarpment margins [As diseussed in the foregoing section, aggradation of a rimmed shelf can give rise to an escarpment margin (Fig. 10.23). The margin initially may be of the bypass type, but ‘oversteepening can lead to erosion and apparent back stepping of the margin as transgression proceeds (Fig 10.13). The products of platform erosion usually will be large blocks of talus derived from the margin, which will accumulate at the base of the carbonate slope, passing distally into more organized caleiturbidites derived from the platform interior. On the carbonate platform itself, it tasually will be very difficult, if not impossible, to differen tiate the aggradational transgressive systems tract from the overlying highstand systems tract. Bec: cretionary Slope is developed, the platform will be unable to prograde during highstand, and the platform top facies will continue to build vertically, albeit with thinner ropsets. Figure 10.24 shows the steep-walled margin of the Cretaceous West Florida platform, which accumulated during transgression ‘and highstand and has remained exposed since the Cre~ taccous, Erosion of the ‘margin’ has been demonstrated by dredge hauls from the wall of the platform, which reveal platform interior deposits (Corso et al., 1988).a) Supra and intertidal Lagponal facies Margin shoals andor rests Seatovel ‘ops faces Exposed Platform top Feinging © Sea-tevel © Drowning uncontormiy Fig. 10.20 Sequence stratigraphic models for rimmed shelves. (a) Transgessive systems tract (TST) of a immed shelf system showing aggradation of the margin. (b) Highstand systems tract (HST) of a rimmed shelf showing progradation and progressive thinning of topsets. (c) Type 1 sequence boundary and development ofthe lowstand systems tract on a humid riomed shel. Here, sea-level has fallen significantly below the margin ofthe immed shelf, and the exposed platform top is karsified and may become incised by fluvial channel. Siliciclastic sediment may be deposited inthe basin, onlapping the eathonate slope. Dr sitw carbonate production may take place as fringing reefs if che silicicasti input is sufficiently low and directed away from favourable cacbonate proxluction sites. () Type 2 sequence boundary and development of a shelf-margin-wedge systems tract on a rimmed shelf. Sea level has not fallen signiicandy below the previous maegia, and the platform top isnot extensively exposed to meteoric diagenesis The shelf-margin wedge expands landward and seawatd with rising sea-Level, but note that this ean occur only if ther is greater seaward subsidence ofthe or shelf as carhonate production is terminated by envi deep-water siliccla (On the basin floor, highstand and lowstand deposits can be differentiated better. Highstand shedding from the plat- form will supply platform interior material to the basin, which may accumulate as calciturbidites or settle out of suspension. These deposits may onlap and cover the under~ lying talus cones developed during the previous trans- gression. Glaser and Droxler (1991) documented thick, 228 inally flat platform top to provide a dipping surface that can be onlapped. (e) Drowning of a rimmed samental deterioration whi characterized by a drowning unconformity, which separates underlying platfor still submerged. The top of the carbonate is carbonate from onlapping and downlapping . BE, basinal facies; SF, slope facies; SB, sequence boundary; MIS, maximum flooding surface Holocene highstand wedges onlapping the base of the erosional margin of Pedro Bank, Nicaragua Rise. Note that margin failure also may oceur during highstand, as well as, during transgression, but highstand talus cones are more likely to be interbedded intimately with platform interior facies. The angular contrast arising from the contact be- ‘oween talus cones and finer grained sediment ean give rise0.5 seconds: Fig. 10.22 Scismie section showing progeadation of Tertiary carbonates of a rimmed shelf system, Dampier Basin, northwest Australia Supra "ace Lagoon facies gente le ic E\S Se, 2 A eearoest ign shoals Tabs “IER? Tho Poss silicic tans Fig. 10.23 Sequence stratigraphic models for escarpment margins (a) Transgressive systems tract (TST) oF an escarpment showing aggradation and failure of the margin, The products of margin failure onlap the ste {b) Highstand systems tract (HIST) of an esearpment margin showing continued aggradation, but a progressive thinning of topsts Inthe basin and at the base of slope, carbonate shed from the platlorm top during highstand onlaps the talus apron of the transgressive systems tract, Some talus also may be shed off the carbonate margin during highstand. (c) Lowstand systems tract (LST) on a humid escarpment-margin system. Here, sea-level has gin, and the exposed platform top is karstiied and may become incised by fluvial channels, Siliciclastic sediment may be deposited in the basin, onlapping the Carbonate slope, and failure of the margin also may result in the deposition of talus cones at the base of slope. In situ carbonate production is likely © be very minor on very steep margins. (d) Drowning of an esearpment margin, as earbonate production is Terminated by environmental deterioration while still submerged. The top of the carbonate is characterized by a drowning tinconformity, which separates underlying platform carbonate from onlapping and downlapping deep water slicilastis slope as a talus apron. llen significantly below the 230sw Fig. 10.24 Seismic profile of the West Florida Cretaceous platform margin. Although the margin shows a ‘seismic reef, labelled as margin facies, dredge hauls show that the margin facies acually comprise platform interior deposits (from Corso et al., 1988) 6 to geometries that resemble onlap on to sequence bound: aries. Grammer et al. (1990) and Grammer (1991) have described such geometries from submersible dives over slopes of Bahamian platforms, and Bosscher and Southam (1992) have generated computer simulations oF gra related “sequence boundaries’, which are clearly not sequence boundaries sensu stricto (Chapter 2). Lowstand on the platform top of escarpment margins be similar to that described for rimmed shelves. Karsti- fication and fluvial incision may occur in humid climates, vaporite precipitation may occur on the platform top fr onlap the platform margin in arid, restricted settings (Tucker, 1991). On very steep erosional margins, very little insitu carbonate deposition is possible, but on lower angle bypass slopes fringing reefs may become established. The low rates of sedimentation during lowstand on an escarp- ment margin provide an opportunity for older slope deposits to become lithified by marine cementation. Grammer and Ginsburg (1991) have dated thick marine cements from lithifed onlapping wedges off the Great Bahama Bank. Carbon-14 ages from these cements range from 11 000 to 13000 yearsBP, corresponding to the last sea-level low- stand, Ginsburg et al. (1991) also have been able to link Sequence boundaries to intervals of intense submarine cementation off the Great Bahama Bank. Drowning of escarpment margins proceeds much as described for rimmed shelves. Excellent examples of drowned steep-sided platforms can be seen in the Cre- taceous of southern Europe, where Schlager (1992) has suggested that their demise may be linked to global oceanic anoxic events, 10.4.5 Isolated platforms Isolated carbonate platforms may show one or more of the 6.25 MILES different types of margin described above, depending on the tectonic setting, windward —leewaed orientation of the platform and the platform size, To avoid repeating the systems tract geometries described above, this section will be devoted to short descriptions of published examples of sequence stratigraphy of isolated platforms. A simplified model for isolated platform sequence stratigraphy has been, ‘outlined in Fig. 10.6. Symmetrical platforms Epting (1980) has described two symmetrical isolated build up types. The ‘pinnacle’ type build-up is essentially an aggraded mound which shows continuous gradual back- stepping (Fig, 10.25). It is impossible to identify different systems tracts from seismic data alone in this example. Pinnacle build-ups are particularly common in the Tertiary of southeast Asia (Hatley, 1980; Grainge and Davies, 1985). Rudolph and Lehmann (1989) describe a giant gas accumulation from the Miocene of offshore Indonesia, the Natuna D-Alpha Field, which differentiated from a broad tunderlying carbonate platform into several discrete pin- nacles. Epting’s second category was the ‘platform type” build-up, which shows aggradation in the underlying trans- gressive systems tract, followed by progeadation during highstand, and a well-developed maximum flooding surface separating the two systems tracts (Fig, 10.26). Asymmetrical platforms The Liuhua Oilfield in the Miocene of the Pearl River ‘Mouth Basin, South China Sea, shows a strong asymmetry. The steep southwest-facing escarpment margin was fault controlled (Fig. 10.11), and built up vertically as an escarp- ‘ment margin through time prior to eventual drowning, The 21northeast-facing margin, by contrast, probably evolved from a ramp to a rimmed shelf through time, and shows evidence of aggeadational and progradational geometries (Erlich ef al., 1990). Eberli and Ginsburg (1987) contrasted the development of different platform margin types across the Bahamas. Figure 10.27 shows a windward margin to the west and a leeward margin to the east. The windward margin was reefal_and aggradational and built a steep escarpment margin during transgression and highstand. The leeward ‘margin, with an accretionary slope consisting chiefly of hhighstand shedded material and some shingled lowstand reflectors, prograded westwards, ultimately infiling the embayment between two previously isolated platforms. Thus the extensive flat-topped Great Bahama Bank seen today is a result of the coalescence of separate carbonate platforms. 10.5 Cyclicity and parasequences on carbonate platforms The discussion thus far has centred chiefly on seismic-scale expressions of carbonate sequence and_systems-tract development. Carbonates also exhibit very well-developed eyelicity on scales of decimetres to several metres. These cycles form the building blocks of the carbonate systems tracts and sequences. The stacking patterns of these carbon- ate rhythms for different systems tracts on different carbonate-margin types are illustrated schematically on Figs 10.18, 10.20 and 10.23, and although the stacking patterns of these cycles can provide some clues as to the overall behaviour of sea-level and/or carbonate pro- ductivity, individual cycles may be quite different geologi- cally, and may originate in several different ways. 232 Fig. 10.25. ‘pinnacle’ type carbonate platform from the Miocene of offshore Philippines. Ic is impossible to break ‘out different systems tracts from seismic data alone in this example, which shows continuous gradual backstepping and confused internal seismic character. The top of the platform is very rugose, This may represent the effects of karstification, or the differentiation of the platform top into several small pinnacles price to drowning Carbonate cycles and parasequences — examples Several types of carbonate cycle can be recognized. The ‘most common is the shallowing upward cycle, which com- prises a lower energy, deeper water base that may have been deposited below wave base, shallowing up into a shoal or reefal facies, The shoal facies may be overlain abruptly by a deeper water facies and a further shallowing upwards facies succession. This is a carbonate parse sequence. Examples of stacked carbonate parasequences such as this have been described from the Cambrian of the USA (Markello and Read, 1981; Montanez and Osleger, 1993), from the Lower Carboniferous of Wales by Gray (1981), and the Jurassic of southern and eastern England (Ashton, 1980). These eycles are a few to several metres thick and are inferred to have been deposited over tens of thousands of years, at least in the latter two examples, by seaward progradation of shoal facies interrupted by rapid deepening events, Shallowing upward cycles deposited in more proximal positions arc also abundantly described in the literature, However, these cycles are not parasequences under the definition used in Chapter 2, although they may link-up distally with parasequences. The Middle and Upper Triassic of the Dolomites in northern Italy is particularly notable for its merre-scale cycles (Goldhammer et al. 1990). In the Latemar Limestone of the Ladinian, the cycles are usually less than I'm thick and comprise limestone—dolomite pairs, with shallow subtidal grainstones and packstones capped by dolomite containing pisoids, tepee structures and vadose cements, indicating subaerial exposure (Fig, 10.28). Up to 500 such eycles have been described, but these thinner cycles may be disturbed by thicker horizons of disrupting tepees, reflecting longer periods of subaerial exposure. The96200 96440 96320 Time in seconds 96320 96440 96200 Time in secondsEASTERN EDGE Pati |-——strarrs “reef? WINDWARD MARGIN two-way travel time = LEEWARD MARGIN, | ve.-6x Fis) Fig. 10.27 Windward (lefe) and leeward (right) platiorm margins of the Bahamas (from Eberli and Ginsburg, 1987). See text for details absence of any tidal-flat facies in these carbonates rules out the possiblity of idal-flat progradation as a mechanism for producing the cyclicity, suggesting that an external control such as sea-level fluctuation may have caused the ehythm Other proximal shallowing upward cycles include gra stone-based cycles shallowing upwards into tufa, and tu based cycles shallowing upwards into tepees and breccias from the Early Proterozoic of the Rocknest Formation, Canada (Grotzinger, 1986), In arid settings carbonate cycles may be associated with ‘evaporite precipitation, usually where gypsum or anhydrite ‘occurs at the top of the cycle, overlying dolostones oF limestones with intertidal sedimentary structures. This type of cycle has been interpreted as the development of supra tidal sabkhas with attendant evaporite precipitation over intertidal carbonates, and again represents a shallowing upward rhythm, Such units are usually a few metres thick, and have been observed in the Jurassic Arab Formation of Abu Dhabi by the author, and in the same Formation in Saudi Arabia (Wood and Wolfe, 1969). In both areas the 24 evaporites provide major intraformational sealing units overlying. reservoir bioclastic and oolitic shoal facies, Sabkha sulphate seals and carbonate—evaporite cycles are also well described from the Permian carbonates of the Permian Basin, Texas (Handford, 1982; Craig, 1990), In more humid settings, carbonate cycles on attached platforms may include a significant siliciclastic component. The Yoredale Series of the Upper Dinantian of northern England comprise limestones deposited below wave base daring overall transgression when siliciclastic input_was more muted, overlain by mudstones and sandstones de- posited by prograding delta systems. Abandonment of the prograded delta lobes led to the development of swamp environments and the subsequent formation of coal (Leeder and Strudwick, 1987). In the Upper San Andres Formation of the Permian of New Mexico, mixed carbonate and siliciclastic cycles that contain all the elements of a sequence ‘on the seale of a few metres have been described by Sonnenfeld (1991). Here, siliciclastics represent lowstand deposition on the slope and in the basin, and carbonates\ amen anes piso lenses Unettuboc Simastone solomite, couplets eyCUC FACES. 2p doiomite crust {= | | wamvacone “ayy emer 4 7 5] sunesrinesore cout so ‘skeletal-peloidal |o2| Linedon-tont coset A (Gages i Serre oyeic Be coaiepo ace Fig. 10.28 Cyclic carbonates of the Latemar Limestone, Tlian Dolomites (after Hardie et al., 1986). See text for details represent transgressive and highstand deposits. These cycles have been termed fourth- and fifth-order sequences, because they have a greater degree of symmetry than a typical Parasequence, and show all the systems tracts of a typical sequence. Goldhammer et al. (1991) have recognized ‘fourth order’ sequences from the Carboniferous of the Paradox Basin, comprising carbonates, sliciclastics and evaporites, Wwith a mean thickness of 35m and periodicity of about 260000 years. Carbonate cycle stacking patterns and Fischer plots In recent years, there has been a revival of interest in analysing the stacking pattems of high-frequency platform- top carbonate cycles. Much of this analysis has been achieved by the use of accommodation ot ‘Fischer plots’ (Fischer, 1964; Sadler et al.. 1993). A Fischer plot is a series of sawtooth waves, with the vertical line of the satooth representing the vertical thickness of a carbonate «yee, linked by a diagonal line drawn from the top of the «yc down to the right of the plot. The bottom right of the diagonal links to the base of the next vertical line represent- ing the next carbonate cycle (Fig. 10.29). The only factual information contained in the Fischer plot is the cycle Grose agaracaion Fig, 10.29 Fischer plot of four hypothetical high-frequency carbonate platform cycles (from Sadler et al., 1993) thickness, The diagonal line subjectively represents the pas sage of time and subsidence prior to the deposition of the next cycle, and is based on two simplifying assumptions of a fixed cycle period and linear subsidence, and, as such, Fischer plots are ‘purely descriptive representations of stratigraphic sections that emphasize stacking pattems (Sadler e¢ af., 1993). Despite the simplifying assumptions, Fischer plots combined with detailed facies analysis have been used to show how platform-top carbonate-cycle stacking can help to distinguish different systems tracts and the location of sequence boundaries within otherwise ‘monotonots platform-top carbonate systems, Goldhammer etal, (1993) were able to use Fischer plots from Ordovi carbonates of the El Paso Group, West Texas, to differe tate systems tracts, Lowstand systems tracts comprised thin quartz arenites, transgressive systems tracts comprised thickening upward subtidal eycles, and highstand systems tracts comprised dolomitic peritidal cycles containing some quartz sand, Fischer plots of the stacking pattern of the West Texas cycles and other cocval Ordovician cycles in the USA ate illustrated on Fig. 10.30. Other studies that have used Fischer plots to define sequences and systems tracts include Read and Goldhammer (1988) and Montanez, and Osleger (1993). Sadler et al. (1993) provide a good review of Fisher plots, with recommendations for their construction and interpretation, Carbonate cycles — origin The origin of carbonate and mixed cycles is the subject of considerable and continuing debate (Chapter 2). Ii a vast subject in its own right, and will be considered only brielly here. Three mechanisms are usually invoked eustasy, and. two types of autocyclic mechanism. The eustat trol involves changes in sea-level that are governed by orbital perturbations (Milankovitch cyclicty, Chapter 2). 235aaa za [aera as] aaa a Za) Seckmancun Grou Znaton ‘Sr t967 Spelman (080) itary Arh, PR Key: 30 arenas Elosbper Group Zenton, Cd and Bune (45) 10. ‘Geo ate, A yf (ash hewn, van Hon. TX “Raila Mountains, OK Fischer plots for Lower Ordovician platform carbonates from several locations across the USA, correlated, biostratigraphicaly. Third-order sequence boundaries (491, 489 and 487 Ma) were correlated using high-frequency cycle stacking patterns and the distribution of quartz sand (from Goldhammer eta, 1993), SB, sequence boundary Goldhammer etal. (1987) have observed that, in detail, the cycles observed in the Triassic of the Dolomites comprise asymmetric packages of five cycles that thin upwards. This “pentacycle” has been interpreted to reflect the 20000 year precession periodicity superimposed on the 100000 year cycle, Read e7 al. (1992) and Read and Horbury (1993) have attempted to relate carbonate cyclicity and diagenesis to Milankovitch forcing, using many examples drawn from, the Phanerozoic. However, it is evident that throughout geological time the volume and extent of polar ice-caps has been highly variable. Invoking rapid and high-amplitude sea-level change for cyclic carbonate development in ‘ice house’ time is reasonable, but it becomes more difficult in, ‘greenhouse’ times. In addition, any cycles developed by an orbital forcing mechanism ought of be of great lateral extent, but cycles of limited lateral extent have been described in the literature (e.g. Pratt and James, 1986). A popular autocyelic mechanism to explain carbonate rhythms, attributable to R.N. Ginsburg, is the tidal-at progradational model (James, 1984). Here, the area of carbonate production is reduced gradually and over whelmed eventually by tidal-flat progradation. Because the carbonate factory can no longer supply material to the tidal flats, progradation ceases, and the reduced-productivity carbonate platform is unable to keep pace with continuing relative sea-level rise, subsides, and is flooded during the cnsting transgression. During transgression, the carbonate factory is re-established and the cycle recommences. This, mechanism has been used to explain intervals that cannot, be interpreted easily in terms of orbital forcing, and where tidal-lat_progradation is evident. Lankester (1993) has suggested that a similar mechanism may be applicable to the platform-top facies of the Miocene of Majorca for the highest frequency cycles, which show no evidence for sea-Jeretchange, and Goldhammer et al. (1993) have speculated that such an autocyclic mechanism may be responsible for high-frequency carbonate cycles of the Ordovician of Texas described above. An additional autocyclic mechanism pro- posed by Pratt and James (1986) is the tidal island model. Here, a carbonate platform is never exposed or flooded cmirely, but is covered with tidalflat islands, These islands aay progeade or build vertically, and may coalesce with one another. On doing so, the subtidal area for carbonate production may decrease, and the coalesced island may sulside. Hydrographic forces also may shift the site of sedimentation to a new area, and a new island may develop asthe old one subsides. This model was invoked originally toexplain the lack of lateral continuity of carbonate cycles inthe Ordovician St George Group, Newfoundland (Pratt and James, 1986). 10.6 Conclusions Carbonate sequence stratigraphy has elements in common With silciclastic systems. However, there are many features that are specific to carbonates that need to be considered ‘carefully before a sequence stratigraphic interpretation of a carbonate system is undertaken. The following check-list, summarizes the most important points. 1 Carbonate is generally produced in platform-top environ- ments. Carbonate production usually exceeds the rate of, creation of accommodation, especially during highstand, and this carbonate will bypass the platform top and be discharged into adjacentbasins, a process termed ‘highstand, shedding’. Most carbonate is shed into deeper water during, highstand in contrast to siliciclastic systems. 2 Carbonate platforms accumulate at or close to sea-level, and provide excellent depth indicators for interpreting, relative sea-level changes. Carbonate platforms also are susceptible to repeated subaerial exposure because of the water depths in which they accumulate, and abundant smallscale (a few to several metres) carbonate cycles commonly are developed in carbonate platform-top enviton- ments, Analysis of the stacking patterns of carbonate cycles, can enable a sequence stratigraphic interpretation to be made. Fischer plots may be of some use in stacking pattern analysis, 3 Organic carbonate production usually can keep pace With sea-level rise, except where the rise is very rapid or where the environment of carbonate production deterio- rates. Environmental deterioration may be caused by changes in light penetration, salinity, oxygenation, nutrient levels, siliciclastic input and predation. It is usually very difficult to identify the cause of environmental deterioration, in the rock record. 4 Organic carbonate systems may build much more ag dational margins than siiciclastic systems, especially during, transgression. Backstepping is less common in organic carbonate systems, except where environmental deterio- 5 ‘Drowning unconformities’ developed over carbonate systems may share certain geometrical characteristics with sequence boundaries. From seismic data alone it may be difficult to distinguish between these two surfaces. 6 Apparent sequence boundaries can be produced by changes in the grade of carbonate sediment accumulating ‘on depositional slopes. These ap yy be difficult to distinguish from genuine sequence boundaries, especially on seismic data. 7 Relative sea-level fall usually will result in exposure of much of the carbonate platform area, depending on the geometry of the carbonate system. Subaerial exposure will tend not to yield much carbonate debris co the basin, because carbonates will tend to be eroded chemically, rather than physically. Ifthe carbonate platform is attached, siliciclastic material may be transported into the basin at this time. 8 The effects of subacrial exposure on the carbonate plat form will be climate dependent. Humid climates will result in major dissolution and reprecipitation of carbonate, and may form karst terrains. Arid climates will result in less intense carbonate diagenesis, bur potentially more evaporite precipitation on the exposed carbonate shelf and in the baCHAPTER ELEVEN Organic-rich Facies and Hydrocarbon Source Rocks 1A Ineroduction 11.1.1 Controls on organic richness and source potential 11.1.2 Source rocks, tectonics and sea level change 1.2 Dektafeoastal plain oxganicrich Facies 1 and source rocks 11.2.1 Sequence stratigraphic significance of coals 11.1 Introduction Sequence stratigraphy provides a useful geological frame- work for considering the distribution of organic-rich facies {source-rocks*). It can allow a seismic, well or outcrop section to be subdivided into systems tracts in which the factors conteolling source-rock deposition can be considered. Understanding the sequence stratigraphic context ofa source rock is necessary in order 0 predict its lateral extent and variability. However, it should be emphasized that source- rocks cannot be predicted from stratal geometries alone because there are too many variables involved (Fig. 11.1). This chapter first explores the controls on deposition of dorganic-rich facies and to what extent these can be predicted from sequence stratigraphic analysis. It then considers in derail the sequence stratigraphic context of coals and trans- _gessive marine black shales and concludes by describing stratigraphic models for organic-tich carbonate facies. 11.1.1. Controls on organic richness and source potential Enhanced organic matter preservation is a function of many factors, the most important being the physiogeo raphy of the basin, climate, terrestrial organic productivity, marine aquatic organic productivity, oceanic circulation, sedimentation rate and water depth (Fig, 11.1). A number of these factors clearly are not predictable from systems tract analysis alone, eg. climate and oceanic circulation. + An oil prone source-tock compris sediments that are rich in organic earbon aun contain organic material ufc hydrogen wo avert ail to ‘om maturation (Tote al, 1974) 11.2.2 Geochemistry of delta plas 18 Masi lorganictich faces 11.3 Organicrich facies and systems tacts in clastic systems 11.3.1 Lovstand systems tract 2 Transgressive systems tract 3 Highstand systems tract Carhonate Source Rocks 114.1 Genetic classification scheme 11.4.2 Intereorbonace build-up 11.43 Ineeaplatform depression 1144 Uneestrcted basin margin 114.5 Deep ocean basin Conclusions Terrestrial organic productivity Terrestrial organic productivity is a primary influence on the development of coals and coaly sediments deposited in, coastal/delta-plain environments. McCabe (1984) con siders that the potential for modern peat accumulation ia complex function of climate, which controls the balance between the rates of plant production and decay. Hot, humid climates favour plant production and cool climates favour plant preservation. The nature of the plantecosystem has a strong influence on the type of organic matter pre served and hence the potential for oil- or gas-prone source: rock development (section 11.2.2). Terrestrial organic matter supply The rate of terrestrial organic matter supply to marine sediments is controlled principally by the nature of the floral ecosystem in the hinterland, the grain size of the sediment and the distance from the shoreline (Schlesinger and Mellack, 1981). Terrestrial organic productivity was negligible in pre-Devonian times and therefore little is preserved in marine sediments. Terrestrial organic matter will, because of its low density and grain size, tend to be concentrated in the fine-grained mud and silt facies of the systems tract. Terrestrial organic matter supply rates will be highest in fine-grained sediments in close proximity to wwell-vegetated deltas. Swamp and marsh areas are particu: larly important sources of terrestrial organic matter. Supply decreases exponentially with increasing distance from the shoreline and increasing water depth. However, shel-edge deltas potentially may deliver high terrestrial organic matter fluxes to the upper slope, and submarine canyons may tapClimate “eesti Basin prysoneearnhy Satanic po cen oe susely + —_ Serle! Marine aquatic organic productty Fg. 11.1 The thar most marine petroleum rocks were deposited undes Sources of terrestrial organic master in the coastal plain and ttansport it directly into deep water settings. Primary productivity Oilprone kerogen in marine source-rocks originates pre- dominantly from marine phytoplankton. The flux of algal ‘organic matter to marine sediments is a function of primary organic productivity in the photic zone and water depth (Cabert, 1987; Schwartzkopf, 1993). Zones of high organic productivity usually occur in the vicinity of continents Zones of particularly high productivity are located in areas of coastal upwelling, where shore-parallel winds result in upwelling of nutrient-rich deep water. Modern upwelling eas are concentrated on the westeen margin of modern continents (Pelet, 1987). Note, however, that palaco- productivity prediction in the geological record is difficult. Palscoclimatic models can allow prediction of prevailing wind patterns and hence upwelling areas in the geological record. The accuracy of these predictions of upwelling areas is very variable (R. Miller, pers. comm.) Water depth The flux of carbon from surface productivity to bottom sediments is strongly influenced by water depth. Degradation and recycling of organic matter in the water column (whether xic or anoxic) sharply reduces the carbon flux at shallow water depths. At 1000m water depth, the carbon flux is < 10% of the value at 100m water depth (Suess, 1980; Teter et al, 1984), As a general rule-of-thumb, the supply tors that influence the organic richness and source noxie bottom water conditions. The development of anoxic conditions is sintrolled by the factors shown. Stippled zone highlights the transgressive systems tract. potential of sediments in clastic depositional systems. Note “MES, maximum flooding surface of marine algal organic matter t0 bottom sediments is likely to decrease with increasing water depth and distance from the shoreline, owing to lower surface productivity and remineralization in the water column (Schwartzkopf, 1993). Organie matter preservation Aquatic organic matter derived from phytoplankton is the main precursor of cil-prone kerogen. Preservation of oil- Prone organic matter in sediments is greatly enhanced under anoxict bottom-water conditions (Demaison and Moore, 1980) which develop where oxygen demand from decaying organic matter exceeds supply. Anoxic conditions result from restricted vertical circulation of sea-water andi or high biological productivity. Physiogeographic restriction helps limit the warer colamn circulation, and hence the resupply of oxygen to bottom waters. It ix a favourable element for predicting anoxic environments. Physiogeographic restriction may take several forms and occur at a variety of scales. Examples include classic silled or intra-shelf basins; geographically restricted oceanic basins, such as the Gulf of Mexico and An basins in Cretaceous times; and geographically enclosed epeitic seaways, such as the Cretaceous Western Interior Seaway, USA, the Jurassic of the North Sea and the Holocene Black Sea. 4 Anoxia suse hereto cover oth teu anoxic and oxygen depleted dysarobie (<0. ml oxygen botwan waters unde which organi mater preseeation chanced 2390 ca 0 or oF 5 ‘SIVAISAS 38NOS 40°ON 4 vl oe » -----1 ---x wore 2 sal wouen0 “Z asoroa | auwino vansHas men) av SOINOLDAL ‘uh YSnans 902 gounos Uo Sone‘Water column stratification enhances the probability of oxic conditions developing by inhibiting bottom-water glenishment. A positive water balance occurs where run- from land and rainfall excceds evaporation. A less lense freshwater cap forms over more saline marine water, a restricted circulation may lead to episodically anoxic ‘ondtions in basins of limited extent (see Fleet et al.'s (1987), scussion of the Lower Jurassic of Southern England). A Iegative water balance means that evaporation exceeds supply by runoff and rainfall. Dense saline water sinks to feed the bottom waters, which may become anoxic as oxygen is used by the degradation of organic matter. This isthe model commonly used to explain anoxia in Jurassic aged intrashelf basins in the Middle Rast (eg, Droste, 1990), Stratification and hence anoxic conditions can persist only below the surface mixing layer, ic. effectively below ‘gor wave base. The surface mixing layer thickness varies ‘vith wind and tide energy in the basin, and usnally is ‘ithin the range 50-200 m. Devanic circulation Water column oxygen profiles are key factors in predicting anoxic conditions and hence source potential in deep-water sediments of the continental slope and ise. In present-day oceans, the dissolved oxygen minimum is at depths of 100-1000m because most oxidation of sinking organic Inatter occurs at these depths. However, at present the ‘oxygen minimum is rarely sufficiently intense to give anoxic ‘onltions and enhance organic matter preservation. Oxygen Contents rise below the oxygen minimum zone because deep occan water today is supplied by cold, oxygen-rich polar waters. The oxygen content of ocean bottom-water therefore decreases with increasing distance from the poles. The oxygen minimum (and hence anoxic conditions) is presently most intense in areas of high surface productivity, ie upwelling areas. Where the oxygen minimum is intense, anoxic conditions may occur on the upper slope and outer continental shelf. Although itis understood that oceanic circulation patterns inust have changed dramatically through geological time, prdiction of ancient water-column oxygen profiles is diffi ‘ult, For example, at times of high global sea-level and nore extensive continental shelves, the oxygen minimum done may have covered large areas of the continental shelf. nan ice-free Earth, sources of deep water may have been the warm and saline waters from low latitudes rather than the present-day cold and less saline polar waters. This could have resulted in an expanded oxygen minimum zone [Brass et al., 1982). Palacogeography is a pointer to past areas of intense oxygen minima. Intense minima seem to have occurred in physiogcogeaphically restricted areas of the oceans, distant from the main body of the oceans (e.g, Cretaceous Arctic Ocean, Cretaceous Gulf of Mexico). Sedimentation rate Changes in sedimentation rate and, in particular, intervals ‘of condensed sedimentation can be predicted from sequence stratigraphic analysis, Ithas been suggested that condensed sections are likely candidates for source-rock intervals (Loui, ef al,, 1988). However, sedimentation can have either a positive or negative effect on organic carbon preservation (see review in Schwarzkopf, 1993). The relationship between ‘organic preservation and sedimentation rate is complex, because it is a delicate balance between enhanced preser- vation through rapid burial and dilut nie matter by lastic material during rapid sedimentation (Schwartzkopf, 1993). Nevertheless, there are increasing numbers of case studies demonstrating the correlation between enhanced organic preservation and regional transgression and, in particular, anoxic condensed facies (Hallam and Bradshaw, 1979; Demaison and Moore, 1980; Jenkyns, 1980; Loutit ef al, 1988; Leckie et al, 1990; Palsey et al, 19913 Creaney and Passey, 1993). 11.1.2 Source-rocks, tectonics and sea-level change Figure 11.2 shows the distribution of the world’s major source rocks for oil through geological time (based on an -house compilation). Certain plate tectonic configurations, in the past were favourable for source-rock deposition. Peaks of marine source-rock development in late Devonian and in the Late Jurassic to Cretaceous coincide with con- tinental break-up and peaks in extensional activity in the ‘Wilson Cycle’ of global plate tectonic movements. These periods also coincide with peaks in the first-order sea-level, ede. ‘The vast majority of source-rocks form in extensional tectonic settings, particularly passive margins, and intra- ‘ratonic and back-arc basins. A low in marine source-rock, development occurs in the Permo-Carboniferous interval, at a time of predominantly contractual tectonic activity and widespread glaciation. Note also the influence of stratigraphic age. Gas-prone, land-plant-derived, source rocks are present only since the Devonian, Delta top oil-prone source-rocks are restricted to the late Mesozoic and Cenozoic times, coincident with the evolution of the lowering angiosperms. 11.2 Delta/coastal plain organic-rich facies and source rocks 11.2.1 Sequence stratigraphic significance of coals Coals can form in a variety of basin settings. For example, Tertiary coals in southeast Asia formed predominantly in extensional basins, whereas the widespread Upper Cretaccous coal deposits in westem USA formed in a foreland basin setting. Autochthonous coals are an import- ant component of delta/coastal plain deposits because theyrepresent vertical accumulation of sediment and hence pure aggradation of the delta/coastal plain. Controls on coal formation are many and varied, but given favourable fac- tors, such as a hot and wet climate, a stable, high ground water table and no clastic infux to the peat forming mire, coals can accumulate rapidly on the delta/coastal plain (McCabe, 1984). Recently, it has been proposed that coals ‘ean be used for regional correlation in non-marine basins (Hamilton and Tadros, 1994). Peat accumulation rates in modern Arctic climates are very slow compared with modern tropical climates, where rates of 2.5 m per 1000 years are typical and where rates up to Sm per 1000 years have been recorded (Fig. 11.3). In ancient coal deposits itis worth considering the time required for the accumulation of thick coals. After allowing for ‘compaction, a 3m coal seam deposited at 1.8m per 1000 years may have required favourable conditions to persist for perhaps 16 000 years. Rates of coal accumulation there fore can be rapid but ultimately are limited by the rate of base level (relative sea-level) rise. Modern coal swamps are unlikely to be able to keep pace with the most rapid rates of Holocene eustatic sea-level rise of 20m per 1000 years (Fairbanks, 1989), Coals accumulaté at or close to base level, and it follows thar for a thick coal to accumulate, aggradation of the delta/coastal plain must occur. In other words, the rate of coal accumulation and the rate of creation of sediment accommodation must balance. In addition the coal swamp must be effectively sheltered from clastic input in order for low-ash coals to accumulate. It is suggested that the devel- ‘opment of raised mires on the coastal plain is necessary for this 10 occur (e.g. McCabe and Shanley, 1993). Raised mires occur where the build-up of waterlogged peat elevates the surface of the mire above regional ground-water tables. In southeast Asia, raised mites tens of kilometres across can be elevated up to 7m above adjacent flood plains (Cameron et al, 1989). | o «8 os « 56 6 (Coal accumulation rate myear (10°) Fig. 11.3 Modern coal (peat) accumulation rates in different ‘geographical settings 242 The development of coals during progeadation and aba donment of a delta lobe will now be considered (Fig. 11.4}. Coals can form in local depocentres on the delta plain during active delta progradation. These coals are likely 10 be thin owing to low delta:plain aggradation rates, and discontinuous owing to erosion by active allavial channels and dilution by clastic sediment. Thick regionally extensive coals are more likely to form during abandonment of a delta lobe. The abandoned delta lobe will subside at a rate governed by subsidence and eustasy. IF the abandonment phase persists, sediment starvation can ensue and a thick coal may accumulate, Eventually, either through a drop in coal accumulation rate or an increase in the rate of relative sea-level rise, the delta plain is flooded by transgressive marine sediments and the cycle begins again. Workers in the Cretaceous Western Interior Seaway, USA have related the stratigraphic position of coals to parasequence stacking patterns (Ryer, 1984; Levey, 1985; ‘Cross, 19883 Shanley and McCabe, 1993). Coals developed, in coastal plain mires extending several tens of kilometres landward of coeval shorelines (Fig. 11.5). Coals are thickest landward of aggradational parasequence sets, i.e. where the shoreline stacks vertically. Coals associated with basin: ward and landward stepping parasequences are thinner (Cross, 1988). Ryer (1984) specifically relates the largest coal fields in Utah with extensive aggeadational stacking of parasequences (fourth-order cycles) developed at the trans: agressive and regressive maxima of third-order cycles. The inference from these observations is that thick coals corre late with periods of rapid coastal plain aggradation, How- ever, the fact that coals can form up to 45% of the coastal plain suecession in these deposits, and that low-ash coals accumulate only 4m from a clastic shoreline, requires, further thought. Shanley and McCabe (1993) suggest that, the formation of raised mires in the coastal plain can stabilize the shoreline during moderate rates of sea-level rise, inhibiting transgression and encouraging vertical stack ing of facies belts in much the same way as a carbonate platform margin might behave (Chapter 10). 11.2.2 Geochemistry of delta plain organic-rich facies ‘The controls on the oil or gas-prone nature of organic-ich delta/coastal plain sequences are still debated. It is known that some late Cretaceous and Tertiary delta/coastal plain source-rocks have sourced significant quantities of oil, but can oil-prone delta/coastal plain source rocks be predicted? This is a complex subject for which generalized rules-of thumb can sometimes be misleading (Fig. 11.6). Cil-prone terrigenous kerogen is derived from plant cuticles and resins. It is thought that only in post-Jurassie times have plants had sufficient foliage to yield significant amounts of oil-prone kerogen (there is uncertainty over late Triassic and Jurassic land plants). Pre-Jurassic land plants would yield gas-prone kerogen. Hot, wet climates are conducive t0 the production of@ Tin eecontnuose cole Progradation 1988) high foliage : wood ratio plant communities with potentially, high yields of oil-prone kerogen. Cool, wer climates with gymnosperm (coniferous) dominated plant communities, have yielded prolific resiniferous oil-prone kerogen in some Gretaceous delta/coastal plains in Australasia. Dry climates, result in predominantly gas-prone kerogen. Preservation of oil-prone terrigenous kerogen is thought to be enhanced in the brackish to saline conditions of the lower delea/coastal plain. Upper delta-plain (freshwater influenced) coalsand coaly sediments re considered mainly to be gas-prone, It sometimes may be possible to map coal distributions sing techniques such as seismic attribute and seismic facies analysis given well calibration. However, the prediction of relative oil- or gas-prone potential of delta/coastal plain sediments in a seismic package is very difficult, if not impossible. Firstly, the differentiation of lower and upper delta plain settings within the top sets of a seismic sequence isunlikely. Secondly, the relative oil- and gas-prone poten- tial of coals can vary laterally both within and between individual coal seams. Finally, there is presently no con- sensus over the relative potential of coals and associated carbonaceous shales. Koy: Sand Prone Datta Plain and Deta Front Sity Mudctonos in Prodetta ERE Oifshore Marine Mudstones, EES coal ES 11.4 Cartoons illustrating coal formation during progradation and abandonment of a delta lobe (after Allen and Mercier, 11.3. Organic-rich facies and systems tracts in clastic systems The following section attempts to integrate some of the controls on clastic source-rock development described earlier within a sequence stratigraphic framework, Source-rock development is discussed within a series of systems tracts, developed during a cycle of changing relative sea-level. The systems tract block diagrams are adapted from those pub- lished by Posamentier and Vail (1988). 11.3.1. Lowstand systems tract Significant coal accumulation on the coastal plain is unlikely in the early lowstand systems tract and terrestrial organic, matter is likely to be highly oxidized. The shelf and upper slope are considered t0 be zones of sediment bypass and hence source potential there is negligible (Fig. 11.7). Ter- restrial organic matter supply is limited to locally reworked material in basinal fan deposits. [Fanoxic conditions persist in the basin a condensed basinal source-rock facies may occur away from the sediment entry points. The lowstand, prograding wedge has some potential for coal development during the aggradational phase but the areal extent of the 243FERRON SANDSTONE, TURONIAN, UTAH, Landard-¢ + Seawara ROCK SPRINGS FORMATION, CAMPANIAN, WYOMING. Landers <—_—+ Seaward seat ‘500 o ADAVILLE FORMATON, CAMPANIAN, WYOMING andar <——____—+ Seaward Fig. 11.5 Diageammatic cross-sections illustrating the distribution of coal relative to the position of the shoreline in different stratigraphic intervals of the Cretaceous, Western Interior Seaway, USA. Shorelines stepping basinward indicate progradation, whereas shorelines stacking vertically indi aggradation and shorelines stepping landward indicate retrogradation (from Cross, 1988) ‘onganic-rich facies will be restricted to the incised valleys and associated canyons. Marine organic-rich slope and basinal facies are unlikely owing tothe rapid sedimentation rates. Although marine source-rocks could occur, overall this appears to be the least prospective systems tract for source-rock development. 244 2 Transgressive systems tract This is the most important systems tract for the develop ment of marine oil-prone source rocks. Many authors have noted the correlation between the occurrence of organic: rich facies and regional transgression (Hillam and Bradshayy 1979; Demaison and Moore, 1980; Jenkyns, 1980; Loutt etal. 1988; and see Fig. 11.2), However, not all transgressive systems tracts result in deposition of organic-rich facies The link is complex and is explored in some detail in this section, General features of the transgressive systems tract In the transgressive systems tract the shoreline retreat, landward, resulting in a progressive increase in the ge ‘graphical extent of shallow-marine shelf deposition, reaching a maximum at the maximum flooding surface. Increasing distance from the contemporary shoreline results in reduced clastic sediment supply and reduced terrestrial organic matter supply t0 the outer shelf and slope. Very low sedimentation rates on the shelf and slope can resalt in deposition of a condensed facies (Fig. 11.8) However, reduced sedimentation rates alone will not result in deposition of organic rich facies. Enhanced presee vation of organic matter in poorly oxygenated benthic environments is a common characteristic of transgressive black shales (c-g- Demaison and Moore, 1980; Oschmann, 1988; Miller, 1990; Wignall, 1991a,b. Shallow water depths on the shelf result in high organic fluxes and there: fore high oxygen demand. The higher the surface pro- ductivity, the higher is oxygen demand in the bottom ‘waters. In the surface mixing layer above storm wave base storms and tidal currents can effectively mix the water column and reoxygenate the bottom waters. It follows that the sediment—water interface will have to remain below the surface mixing layer (H) for oxygen deficient/depleted conditions to persist for significant periods of time, The ‘depth (I) will vary depending on storm and tide energy. For example, H1 may be shallower on wide shelves compared with narrow shelves owing to reduced tidal and wave activity (Hallam and Bradshaw, 1980). Development of a stratified water column and physiogeographic restriction will increase the chances of anoxic conditions and organic- Fich facies development during transgression. Models for source-rock development In the model shown (Fig. 11.9), the basin is permanently anoxic below a certain water depth, marking the base of the surface mixing layer. A distal deep-water, relatively, condensed, source-rock facies occurs in an anoxic basin ‘entre in the distal toc-sets of both lowstand and highstand systems tracts, which would be manifested as a downlap, surface on seismic sections. The transgressive systems tract results in landward translation of this facies belt until the‘AND LAKES. 2 i a Controls: eyo Be 1G imate ay ra Age SREP EE —— sare, nent, resins 7 28 cerggen cams} Si TERRESTAIAL | odrSe FACIES stun ‘Cotamer | Boro} WATER tae COLRIET > JURASSIC Fig. 11.6 Cartoon illustrating different types of source-rock setting on a typical modern delta plain from southeast Asia (based on the Klang Delta). Lower delta-plain environments are thought to be areas where the preserva likely to generate oil) is enhanced in brackish alkaline environments. Hot humid climatic regimes in the post 0-wood ratio plant communities, which, given favourable preservation, nos favourable for generating high foliage ‘kposiion of oiLprone coal source rocks, deeper water condensed source-rock facies impinges on the transgressed shelf. In this model the greatest extent of source-rock facies will coincide with the time of maximum, flooding. During subsequent highstand progradation, water depths decrease and the area of source-rock deposition gradually shrinks. Thus source rocks are not restricted to a particular systems tract but expand to a maximum during, the transgressive systems tract. This model is similar to the ‘expanding puddle’ model of Wignall (199 a), derived from the study of black shale deposits in epicontinental shelf seas, (fig. 11.10). It is less applicable to passive continental margins, where the deep ocean may be well-oxygenated. Wireline logs are particularly useful in characterizing source rocks (Passey et al., 1990; Myers and Jenkyns, 1992), Creaney and Passey (1993) noted that many marine ikprone source rocks are characterized by an initially abrupt upward increase in organic richness, against back~ ground values, and a subsequent gradual decrease in rich- ness (Fig. 11-11). They attributed this pattern to the control ‘of organic carbon contents by clastic sedimentation rate under anoxic bottom water conditions. The initially rapid ion of vik-prone (organic matter accous era are the can eesult in increase in TOC (total organic carbon) results from a rapid decrease in the rate of clastic sediment supply to the shelf during transgression. The subsequent gradual decrease in TOC reflects increasing clastic sediment supply and dilution of organic carbon during highstand progradation. In this model, the source interval would thicken from the ba: margin into a series of stacked source intervals in the basin centre. Wignall (199 1a) interprets the Toarcian black shales, ‘of England to be of this type. ‘An alternative model is where the anoxic conditions develop only in the transgressive systems tract itself (Fig. 11.12). Basinal areas will have interbedded source facies deposited in the TST, with non-source facies deposited in the HST and LST. Shelfal source-rocks are restricted to the TST. This model implies that the palacogeographic con- ditions necessary for anoxia to develop are unique to the transgressive systems tract. Leckie et al. (1990) document an example from the Cretaceous Shaftsbury Formation of Canada, where a nearshore organic-rich zone deposited during rapid transgression passes basinward to more nor- mally oxygenated sediments. In this ease, a nearshore zone 245 2 = g >oa oats Shetcusmorged Food caste input te shotsopo and teresa om. supply -ighproducthy and halo water depth tm set may est 9 ‘roc condtons i fSednertvater trace ISbabow the surece rig ayer ‘Nuine o-rone souree- oc wl only be epee ance Contos Seolop—not al vansressions give foureesoss onshetand rich facies development (after Posamentier and Vail, 1988) af high productivity is assumed to be localized in the transgeessve systems tract, Ie-should be clear from the above that although there is a general relationship between source rocks and transgression, the stratigraphy of organic facies is complex in detail. This complexity is shown by the work of Curiale et al. (1991) and Palsey etal, (191) in the Cretaceous Western Interior Seaway. Curiale et al, (1991) show maximum organic richness occurring in the early highstand systems tract above the condensed section associated with maximum susan nconession tek 356 suruce | Te xannsreanscression \ seowsi2 | 2 fener Lan Fig. 11.9 Chronosteatigraphic implications of a source-cock mo basin at times of both maximum transgression and regression. Ti Marine i-prone sourca-ocks Condtone develop fig. 11.8 Schematic cartoon of a transgressive systems tract on a shelf-break margin adapted to show the potential for or COALS AND COAL SEDIMENTS IN COASTAL PLAIN 1 As highstand tract 1 Distbulon may be 2ographicaly fore restricted ‘dope anoxic NEARSHORE ee ‘TaansGReSsIve Pave sent SANDS. vt High energy, oor source Condensed ofshore Fetental mai feces ganic- flooding in the Cenomanian~Turonian interval they studied. Palsey ef al. (1991), in contrast, show that in the overlying, Coniacian strata, maximum organic richness occurs in the transgressive systems tract below the condensed section associated with maximum flooding. Organic richness will vary from the proximal to distal portions of the systems tract and should not be considered to be uniform. A common problem for the petroleum, geologist is to extrapolate a source-rock proven in the shelfal portion of a transgressive systems tract into the (CHRONOSTRATIGRAPHIC IMPLICATIONS NeARSHORE ‘COASTALIDELTA PLAIN (DWAR. + BASINWARD the rich odel where organic-rich rocks are developed in a distal facies in ‘ansgression serves only to spread the area covered by orgs facies landward. MFS, maximum flooding surface; TST, transgressive systems tractte) tonsran na ue tte 7" Steet = TL ‘deposition 0) TaANsaRESSION eninp ute ek seta