Embry Sequence Stratigraphy

81

Transcript of Embry Sequence Stratigraphy

Page 1: Embry Sequence Stratigraphy
Page 2: Embry Sequence Stratigraphy

Published online by the Canadian Society of Petroleum Geologists, October, 2009. The document can be accessed at http://www.cspg.org Suggested reference is: Embry, A.F., 2009, Practical Sequence Stratigraphy. Canadian Society of Petroleum Geologists, Online at www.cspg.org, 79 p. The fifteen chapters were originally published as separate articles in fifteen issues of the CSPG monthly publication, The Reservoir, May, 2008 – September, 2009.

Page 3: Embry Sequence Stratigraphy

Practical Sequence Stratigraphy

Contents

1) Introduction ...............................................................................................................................................32) Historical Development of the discipline: The First 200 Years (1788-1988) ................................53) Historical Development of the discipline: The Last 20 Years (1998-2008) ...................................94) The Material-based Surfaces of Sequence Stratigraphy, Part 1:

Subaerial Unconformity and Regressive Surface of Marine Erosion ......................................... 135) The Material-based Surfaces of Sequence Stratigraphy, Part 2:

Shoreline Ravinement and Maximum Regressive Surface ........................................................... 176) The Material-based Surfaces of Sequence Stratigraphy, Part 3:

Maximum Flooding Surface and Slope Onlap Surface .................................................................. 237) The Base-Level Change Model for Material-based, Sequence Stratigraphic Surfaces ............. 298) The Time-based Surfaces of Sequence Stratigraphy ........................................................................ 359) The Units of Sequence Stratigraphy, Part 1:

Material-based Sequences .................................................................................................................. 4110) The Units of Sequence Stratigraphy, Part 2:

Time-based Depositional Sequences............................................................................................... 4711) The Units of Sequence Stratigraphy, Part 3:

Systems Tracts ....................................................................................................................................... 5112) The Units of Sequence Stratigraphy, Part 4:

Parasequences ...................................................................................................................................... 5713) Sequence Stratigraphy Hierarchy ...................................................................................................... 6114) Correlation ........................................................................................................................................... 6715) Tectonics vs. Eustasy and Applications to Petroleum Exploration ............................................. 73

Practical Sequence Stratigraphy was originally published as a fifteen-part series in theCanadian Society of Petroleum Geologists’ monthly magazine, The Reservoir, betweenMay 2008 and September 2009.

The CSPG thanks Dr. Ashton Embry for his work in making this series possible.

Page 4: Embry Sequence Stratigraphy

Acknowledgements Above all, I have to acknowledge the major contributions of Hollis Hedberg, the father of modern stratigraphic practice. Over a time span of 40 years (late 1930s - late 1970s) he established the guiding principles for stratigraphic classification and nomenclature. His concepts and advice are just as relevant today as they were when they were formulated. I would also like to thank Ben McKenzie, the editor of the CSPG Reservoir. He encouraged me to produce this series of articles on sequence stratigraphy and he carefully edited each one. He also did an excellent job of producing this compilation. Thanks also go to Heather Tyminski, the CSPG Communications Coordinator, who looked after the layout of each article and was always available for consultation. I have discussed sequence stratigraphic concepts with many individuals over the past 40 years and I am thankful for having the opportunity to do so. Erik Johannessen of StatoilHydro and Benoit Beauchamp of the University of Calgary have been especially helpful and they have consistently taken me to task when I lapsed into sloppy thinking. Many of the concepts in these articles are the result of our many conversations and debates. I would like to thank my employer, the Geological Survey of Canada, which fostered and funded my research and allowed the publication of these articles. I must also thank my colleague, Dave Sargent, who expertly drafted all the diagrams and improved the design of many.

Page 5: Embry Sequence Stratigraphy

3

Practical Sequence Stratigraphy 1Practical Sequence Stratigraphy 1Practical Sequence Stratigraphy 1Practical Sequence Stratigraphy 1Practical Sequence Stratigraphy 1Introduction by Ashton Embry

This is the first in a series of articles on oneof my favorite subjects : sequencestratigraphy. I have called the series practicalsequence strat igraphy because I ’ l l beemphasizing the application of the disciplinerather than dwelling on theoretical models.Each article will cover one main topic and Ihope by the end of the series anyone whohas had the fortitude to read all the articleswi l l have a good idea what sequencestratigraphy is and how it can be used tohelp find petroleum.

During the last 30 years , sequencestratigraphy has been discussed in dozensof books and thousands of scientific papers.It also has become the most commonly usedstratigraphic discipline for developing acorrelation framework within a sedimentarybasin because of the low costs associatedwith such an analys is as wel l as i tsapplicability in many cases to a well log andseismic data base in addition to cores andoutcrop. Despite such populari ty,considerable confus ion and var iousmisconceptions are associated with themethods and terminology (e .g . , unitdefinition) for sequence stratigraphy. This isunfortunate because sequence stratigraphycan be an excellent foundation for faciesanalysis and consequent interpretations ofpaleogeographic evolution and depositionalhistory of portions of a sedimentary basin.

I became involved in developing sequencestratigraphic methodology because I found Icould not apply the methods andterminology proposed by Exxon scientistsalmost 20 years ago. As a stratigrapher forthe Geological Survey of Canada, my mainfocus is on the descr ipt ion andinterpretation of the Mesozoic successionof the Canadian Arctic Archipelago. Sequenceanalysis is an essential part of such workand I found it frustrating that I could notapply the proposed Exxonian methods andterminology in a rigourous scientific manner.Furthermore, when I went through theliterature in an attempt to see how otherswere applying the Exxonian methods, I foundthat the applications were either seriouslyflawed or did not really employ the Exxonianmethods. This led me to develop methodsand terminology which, above all, wereguided by objectivity and reproducibility. Ialso made sure that such methods andterminology could be used in diversegeologica l sett ings , from outcrop to

subsurface, and from undisturbed basin fillsto tectonically disrupted areas with onlyfragmentary records. Finally, I also addressedthe issue of data type because it is essentialthat any proposed methods and terms canbe used equally well with outcrop sections,mechanical well logs supported by chipsamples and scattered core, seismic data, orany combination of these data types. In theseefforts I was assisted by colleagues at theGSC, especially Benoit Beauchamp and JimDixon, who also experienced the sameproblems as I did when it came to theapplication of sequence stratigraphy toregional stratigraphic successions. I alsoreceived a great deal of help and feedbackfrom my fr iend Er ik Johannessen ofStatoi lHydro, who saw the problemsstemming from Exxonian sequencestratigraphy from the perspective of apetroleum explorationist.

This series of articles will summarize theterminology and methods which I and mycolleagues have found most useful in oursequence strat igraphic studies . Thismethodology has many features in commonwith the Exxon work but i t a lso hassignificant differences. I hope to demonstratethat sequence stratigraphy, when properlyutilized, provides a very reliable way tocorrelate stratigraphic cross-sections withaccuracy and precision. The preparation ofsuch cross-sections is a fundamental activityin the exploration for stratigraphicallytrapped oil and gas and the use of sequencestratigraphy significantly enhances the chanceof success of any stratigraphic petroleumprospect. Sequence stratigraphy also allowsa stratigraphic succession to be put into atime framework which in turn allows thedepositional history and paleogeographicevolution to be interpreted against abackground of base-level changes. Suchinterpretations provide the predictiveaspect of sequence stratigraphy.

Below, I discuss how sequence stratigraphyis best viewed as a separate stratigraphicdiscipline rather than some all encompassingdiscipline which integrates data from allsources.

Stratigraphy and StratigraphicStratigraphy and StratigraphicStratigraphy and StratigraphicStratigraphy and StratigraphicStratigraphy and StratigraphicDisciplinesDisciplinesDisciplinesDisciplinesDisciplinesStratigraphy is the scientific discipline thatstudies layered rocks (strata) that obeySteno’s Law of Superposition (younger strata

overl ie older strata) . The Law ofSuperposition allows a relative temporalordering of stratigraphic units and surfacesat any location, and correlation of suchentities between different localities permitsa relative ordering of strata to be assembledfor the entire Earth. Stratigraphy includesrecognizing and interpreting the physical,biological, and chemical properties of strataand defining a variety of stratigraphicsurfaces and units on the basis of verticalchanges in these properties.

Each stratigraphic discipline focuses on aspecific property of strata for unit definition,description, and interpretation. Verticalchanges in that specific property of the strataallow the recognition and delineation ofstratigraphic surfaces within that disciplineand these are used both to define theboundaries of the units and to provide standalone correlation surfaces. The stratigraphicdisciplines of lithostratigraphy (changes inlithology) and biostratigraphy (changes infossil content) have dominated stratigraphicanalysis since the time of William Smith.However, over the past 50 years, otherproperties of strata have been used to definenew stratigraphic disciplines, each with itsown specific category of stratigraphic unitsand surfaces. The “late comers” which havebeen adopted are magnetostratigraphy(changes in magnet ic propert ies) ,chemostratigraphy (changes in chemicalproperties), and sequence stratigraphy(changes in depositional trend).

For each strat igraphic disc ip l ine, therecognized changes in the specific propertywhich characterizes that discipline arecorrelated (matched on the basis of similarcharacter and stratigraphic position) fromone locality to the next and become theboundaries of a series of units. Changes inthe various rock properties often can beextended over wide areas so as to allowthe definition of a set of regional units.Furthermore, it is useful to determine thet ime relat ionships of a strat igraphicsuccession. To accomplish this, stratigraphicboundaries which are used for correlationhave to be evaluated in terms of theirrelationship to time. Each stratigraphicsurface represents an episode of changewhich occurred over a discrete interval oft ime and thus each has a degree ofdiachroneity over its extent. To undertake achronostratigraphic analysis (i.e., to put the

Page 6: Embry Sequence Stratigraphy

4

succession into a time framework), eachcorrelated surface has to be evaluated interms of how close it approximates a timesurface.

Surfaces that have low diachroneity, that is,they developed over a short time interval,are the closest approximation to timesurfaces we have and they have the mostutility for the construction of stratigraphiccross sections and time frameworks. Suchboundaries were classically determined bybiostratigraphy with rare contributions fromlithostratigraphy (e.g., bentonites). Morerecent ly, magnetostrat igraphy andchemostratigraphy have been employed tocontribute to the construct ion of anapproximate time framework. The mainproblems with using these stratigraphicdisciplines in petroleum geology are thatthey are very time consuming, require highlytrained specialists, and often involve largecosts. Furthermore, they also require rocksamples from either outcrop or core thatare rarely available for most subsurfacestudies. All these constraints have greatlylimited the application of these types ofstratigraphic analysis in day-today petroleumexploration. As discussed below, sequencestratigraphy does not have the drawbacksand constraints that severely limit the useof the other stratigraphic disciplines forbuilding an approximate time correlationframework for subsurface successions.

Sequence Stratigraphy - The recognizableproperty change of strata that al lowssequence stratigraphic surfaces to be definedand delineated, and provides the rationaleupward trend to a fining-upward one andvice-versa, and a change from a shallowing-upward trend to a deepening-upward oneand vice-versa. Such changes are based onrelat ive ly object ive observat ions andinterpretations and they are the main onesused to def ine speci f ic sequencestratigraphic surfaces. The latter two changesin trend are often used to interpret a changefrom a regressive trend to a transgressivetrend and vice versa. Much more interpretivechanges in depositional trend – the changefrom base-level fall to base-level rise andvice-versa – are also sometimes used insequence strat igraphy but, as wi l l bediscussed, these are very difficult to applyto many datasets.

These changes in depositional trend areused to define and delineate specific typesof sequence stratigraphic surfaces (e.g.,subaerial unconformity for the change fromsedimentation to subaerial erosion) andthese surfaces in turn are used forcorrelation and for defining the specific unitsof sequence stratigraphy (e.g., a sequence).

Given the above, we can say “SequenceStratigraphy consists of:

1) the recognition and correlation ofstratigraphic surfaces which representchanges in depositional trends in therock record and

2) the description and interpretation ofresulting, genetic stratigraphic unitsbound by those surfaces.”

Each surface of sequence stratigraphy ischaracterized by a specific combination ofphysical characteristics, which are based on:

1) sedimentological criteria of thesurface itself and the strata above andbelow it, and

2) geometric relationships between thesurface and strata above and below it.

Thus the types of data available for a sequencestratigraphic analysis must allow the faciesof the succession to be reasonablyinterpreted and the stratal geometries to bedetermined. Data from other stratigraphicdisciplines such as biostratigraphy andchemostratigraphy can also contribute tosurface recognition (e.g., help determinestratal geometries) but cannot be used forsurface characterization.

For each stratigraphic discipline, it is useful,but not essential, to have a solid theoreticalfoundation which links the generation of thevarious surfaces in that disc ip l ine tophenomena which occur on our planet. Forexample, surfaces in b iostrat igraphyrepresent changes in fossil content that aredue mainly to a combination of evolutionand shifting environments of deposition. Itmust be noted that b iostrat igraphyf lourished long before the theory ofevolution was developed. Most sequencestratigraphic surfaces were recognized in therock record and used for correlation longbefore a theory was developed to explaintheir existence. Eventually it was postulatedthat these sequence stratigraphic surfacesare generated by the interact ion ofsedimentation with relative changes in baselevel and this theoretical model is widelyaccepted today. In the next article I willdescribe the historical development of boththe empir ica l observat ions and thetheoretical underpinnings which have led tothe current state of sequence stratigraphy.

Page 7: Embry Sequence Stratigraphy

5

Practical Sequence Stratigraphy IIPractical Sequence Stratigraphy IIPractical Sequence Stratigraphy IIPractical Sequence Stratigraphy IIPractical Sequence Stratigraphy IIHistorical Development of the Discipline:The First 200 Years (1788-1988) by Ashton Embry

In my in i t ia l art ic le in th is ser ies , Iemphasized that sequence stratigraphy is oneof a number of stratigraphic disciplines witheach discipline characterized by a specifictype of stratigraphic surface used forcorrelation and unit definition. I definedsequence stratigraphy as 1) the recognitionand correlation of stratigraphic surfaceswhich represent changes in depositionaltrends in the rock record and 2) thedescription and interpretation of resulting,genetic stratigraphic units bound by thosesurfaces. Sequence stratigraphic thought hastraveled a long and bumpy road to arrive atour current understanding of the disciplineand that succinct definition.

In this article and the following one, I willdescr ibe the history of sequencestratigraphic analysis from its first vestiges,which were part of the birth of moderngeology, to its current state, which is vibrantbut burdened by invisible surfaces, anoverblown jargon, and quest ionablemethodologies.

Early WorkEarly WorkEarly WorkEarly WorkEarly WorkSequence stratigraphy has been slowlyevolving ever since the late 1700s whenJames Hutton, the father of modern geology,f irst recognized an unconformity as aspecific type of stratigraphic surface andrealized that it represented a substantialt ime gap. From that t ime onward,unconformities were seen as very usefulstratigraphic surfaces for correlation andbounding strat igraphic units and forunraveling geological history. Because anunconformity represents a change indepositional trend, it is one of the mainsurfaces employed in sequence stratigraphy.Thus, i t can be sa id that sequencestratigraphy began at the moment Huttonconceptualized an unconformity.

During the 1800s, debate began on whetherunconformities were generated by a rise ofthe land surface (tectonics) or by a fall insea level (eustasy). By the end of the centurythe tectonics interpretation was favouredand unconformities, and the episodes ofdiastrophism they represented, were seenas the key to global correlations. In the firsttwo decades of the 20th century, keyrelat ionships associated withunconformities were recognized. Grabau

(1906) described stratigraphic truncationbelow the unconformity and stratigraphiconlap above it. Barrell (1917) provided thef irst deduct ive model for sequencestratigraphy when he introduced the conceptof base level, an abstract surface which actsas the cei l ing for sedimentat ion, andproposed that cycles of base-level rise andfall produced repeated unconformities in thestratigraphic record. Notably, he also defineda diastem which, in contrast to anunconformity, is a stratigraphic surface whichrepresents an insigni f icant gap in thestratigraphic record. Unfortunately Barrellwas struck down by the Spanish Flu soonafter h is paper on base level andunconformities was published, and his “aheadof their time” concepts lay in limbo for along time.

In the 1930s small-scale, unconformitybounded units were recognized in theCarboniferous strata of the mid-continentand were called cyclothems (Weller, 1930;Wanless and Shepard, 1936). We now knowthat these cyclothems were generated bynumerous eustatic rises and fall in sea levelrelated to the waxing and waning of theGondwana glaciers. However, at the time oftheir recognition, there was fierce debateas to whether cyclothems were the productof tectonics or eustasy.

Sloss and WheelerSloss and WheelerSloss and WheelerSloss and WheelerSloss and WheelerSequence stratigraphy began as a specificstratigraphic discipline almost 60 years agowhen Sloss et al. (1949) coined the termsequence for a stratigraphic unit boundedby large-magnitude, regional unconformitieswhich spanned most of North America.Krumbein and Sloss (1951, p. 380-381)elaborated on the concept of a sequencewhich they characterized as a “major tectoniccycle.” It was not until the early 1960s thatSloss (1963) fully developed the concept andnamed six sequences which occurredthroughout North America. Sloss (1963)interpreted that the unconformities whichbound his sequences were generated byrepeated episodes of continent-wide,tectonic uplift.

After Sloss et al. (1949) gave us the conceptof a sequence, Harry Wheeler published aseries of papers (Wheeler and Murray, 1957;Wheeler 1958, 1959, 1964a, 1964b) which

used theoretical deduction to provide afoundat ion for the development ofunconformities and consequent sequences.The main parameters in Wheeler’s model,like that of Barrell (1917), were sedimentsupply and rising and falling base level (base-level transit cycles). Wheeler (1958, 1959)provided real-world examples ofunconformity-bounded sequences tosupport his model. In most cases, therecognized unconformities were of smallermagnitude than the cont inent-wideunconformities of Sloss (1963) and many ofthe unconformities of Wheeler (1958, 1959)disappeared in a basinward direction. Asillustrated by Wheeler (1958, Fig. 3), whereone of the bounding unconformit iesdisappeared, that specific sequence was nolonger recognizable. Thus to Wheeler (1958),a sequence was a unit bounded byunconformities over its entire extent.

The result of defining a sequence as a unitbounded entirely by unconformities was thatmost sequences occurred only on the flanksof a basin where major breaks in thestratigraphic record were common andreadily recognized. Nomenclatural problemsoccurred as unconformities appeared anddisappeared along depositional strike andbasinward and new sequences had to berecognized at every place this happened(Figure 2.1 , page 6) . Furthermore,unconformity bounded sequences had verylimited value for subdividing the morecentral successions of a basin where breaksin the record were absent or very subtle.

In summary, by the mid 1960s, sequencestratigraphy was characterized by twoseparate approaches, one of data-drivenempiricism as exemplified by the work ofSloss (1963) and the other of theoreticaldeduction as used by Wheeler (1958).Notably, both approaches came to a similarplace, that of a sequence being a unitbounded by subaeria l unconformit iesgenerated by base-level fall (tectonic upliftor eustatic fall).

The pre-modern era in sequence historycame to a close in the mid 1960s with thepublication of Kansas Geological SurveyBul let in 169 (Merriam, 1964) whichsummarized the concepts on cycl icsedimentat ion and unconformity

Page 8: Embry Sequence Stratigraphy

6

development up to that date. After this,interest in sequence stratigraphy waned asthe focus of sedimentary geology switchedto process sedimentology and facies models.In the mid 1970s a few new terms wereintroduced. Frazier (1974) named unitbounded marine starvation surfaces (today’smaximum flooding surfaces) a depositionalcomplex and Chang (1976) renamed asequence as defined by Sloss et al. (1949) asa synthem. Neither of these suggestions wasembraced by the stratigraphic communityand sequence stratigraphy remained in thecloset until Exxon researchers publishedtheir revolutionary concepts and methods.

PPPPPeteteteteter Ver Ver Ver Ver Vail anail anail anail anail and Sd Sd Sd Sd Seismic Deismic Deismic Deismic Deismic DatatatatataaaaaInterest in sequence stratigraphy was revivedin 1977 with the publication of AAPGMemoir 26 on Seismic Stratigraphy (Payton,1977). In this watershed publication, PeterVail and his colleagues from Exxon usedregional seismic lines as their primary database and demonstrated that the sedimentaryrecord consists of a series of units that arebound mainly by unconformities (Vail et al.,1977) . This was accompl ished on thereasonable assumptions that many seismicreflectors parallel stratal surfaces and thatunconformities coincided with seismicref lectors at which other ref lectors

terminated due to truncation, toplap, onlap,or downlap. In essence, Vail et al. (1977) usedseismic data to delineate unconformities byway of the seismically imaged, geometricrelationships of the strata.

A number of the Exxon researchers ,including Peter Vail, had been graduatestudents under Larry Sloss and it is notsurprising that the seismically determined,unconformity related units were termed“depositional sequences” by the Exxonscientists. On the basin flanks, a sequence-bounding reflector was characterized bytruncation below and onlap above andappeared to be similar to the unconformityused by Sloss et al. (1949) and Wheeler(1958) to bound a sequence ( i .e . , anunconformity formed mainly by subaerialerosion). Of critical importance was theobservat ion that the ref lector thatencompassed this truncation unconformityon the basin flanks could be traced into themore central areas of the basin where it haddifferent character. In some areas thereflector exhibited no evidence of missingsection and these portions were termed thecorrelative conformity part of the sequenceboundary. More commonly, the reflectorexhibited unconformable relationshipscharacterized by either marine onlap or bymarine downlap. Thus a sequence boundary,as delineated with seismic data, seemed tobe a composite boundary characterized by atruncation unconformity on the basin flanksand by marine unconformities and stretchesof correlative conformities in the morecentral portions of the basin (Figure 2.2).On the basis of these observations, Mitchumet al. (1977) proposed a new definition of asequence – “a stratigraphic unit composedof a relatively conformable succession ofstrata bounded at its top and bottom byunconformities or correlative conformities.”

The new def in i t ion essent ia l lyrevolutionized sequence stratigraphy. Withit, the stratigraphic succession of a givenbasin could be subdivided into a series ofsequences which could be recognized overmost or all of a basin (Figure 2.3). Theproblems that had prevented the acceptanceof the unconformity-only boundedsequences of Sloss (1963) and Wheeler(1958) were thus resolved and new life wasbreathed into sequence strat igraphy.Overall, the Exxon seismic data clearlydemonstrated that sequence boundaries arekey, regional correlation horizons and thatsequences are the most practical units touse for stratigraphic subdivision if one wantsto describe and interpret the depositionalhistory of a stratigraphic succession. Oneof the most innovative aspects of the Vail etal. (1977) sequence boundary is that it is acomposite of different types of stratigraphic

Figure 2.1. The occurrence of 10 unconformities on the basin margin allows 9 “unconformity-only” sequencesto be defined. Because each unconformity extends a different distance into the basin, a new series of“unconformityonly” sequences has to be defined each time an unconformity disappears. Such a chaoticnomenclatural system and the lack of subdivision over much of the basin ensured that the “unconformity-only” sequences of Wheeler (1958) were not widely adopted.

Figure 2.2. A seismically delineated sequence boundary (red line) exhibits of a variety of specific stratigraphicrelationships along its length indicating that different types of surfaces comprise the boundary. On the basinflank the boundary is characterized by truncation whereas further basinward it exhibits marine onlap.Toward the basin centre the boundary is a correlative conformity. Seismic line from Quaternary succession ofthe Gulf of Mexico, Desoto Canyon area (modified from Posamentier, 2003).

Page 9: Embry Sequence Stratigraphy

7

surfaces rather than one specific type ofsurface. It is this composite nature of asequence boundary which allows sequencesto the correlated over large areas of a basinand is the key to the great utility of such aboundary. One problem associated with theseismically delineated, composite sequenceboundary was the uncertainty of the specificnature of the surface types which compriseda sequence boundary. Such uncertainty stemsmainly from the poor vertical resolution ofthe seismic data which was available at thetime Vail et al. (1977) were doing theirstudies . In most instances, indiv idualreflectors comprised 20-30 metres of strataand thus the seismic data were not adequateto resolve the necessary deta i ls toconfidently identify all the specific types ofstratigraphic surfaces which were generatingthe reflectors that were designated assequence boundaries on seismic sections.On the bas is of truncat ion/onlaprelationships, a reasonable interpretationwas that subaerial unconformities formedthe sequence boundaries on the basin flanks.However, it was very uncertain what specifictypes of stratigraphic surfaces formed theseismica l ly determined, marineunconformit ies and the correlat iveconformities of the sequence boundariesfarther basinward. Furthermore, in somecases, such as for the “downlap surface” ora toplap unconformity, it was uncertainwhether or not the seismically imagedunconformity (apparent truncation ofreflectors) was a real unconformity or wasan artifact of the low seismic resolution(merging rather than terminating strata). Thisuncertainty regarding the specific types ofstratigraphic surfaces which comprise asequence boundary is still causing significantproblems today.

The Rise of Sea LevelThe Rise of Sea LevelThe Rise of Sea LevelThe Rise of Sea LevelThe Rise of Sea LevelIn addition to providing a new methodologyand definition for sequence delineation, Vailet al. (1977) interpreted that the multitudeof sequence boundaries they recognized onseismic data from many parts of the worldwere generated primarily by eustatic sealevel changes. This interpretation stood incontrast to that of Sloss (1963), who hadalways emphasized tectonics as the primedriver of sequence boundary generation. Asnoted earlier, the debate of tectonics versuseustasy for unconformity generation beganin the 19th century and it continues today.Importantly, the interpretation that eustasywas the driving force behind sequencegeneration led to the development of adeductive model for sequence generationwhich combined sinusoidal eustatic sea levelchange with a constant sediment supply anda basinward increasing subsidence rate. Thismodel reproduced a number of thestratigraphic relationships seen on the

seismic sect ions such as bas in f lankunconformities associated with truncationand basin centre, condensed horizonsassociated with downlap. Because of this,the model was embraced by the Exxonscientists and became the centre piece oftheir next watershed publ icat ions onsequence stratigraphy (Wilgus et al., 1988).These papers , and the models andinterpretations they advocated, formed thefoundation for new sequence stratigraphicterminology and methods which could beapplied to the rock record of well logs andoutcrops as well as to seismic. In the nextarticle in this series, I will discuss thisrevolutionary model which took sequencestratigraphy from correlating low-resolutionseismic to interpreting high-resolution welllog and outcrop stratigraphic sections. I willalso put into context all the terminologyand disagreements the Exxon sequencemodel has spawned over the past 20 yearsand I’ll describe the alternative models andmethods which have arisen during this time.

ReferencesReferencesReferencesReferencesReferencesBarrell, J. 1917. Rhythms and the measurementsof geologic time. GSA Bulletin, v. 28, p. 745-904.

Chang, K. 1975. Unconformity -boundedstratigraphic units. GSA Bulletin, v. 86, p. 1544-1552.

Frazier, D. 1974. Depositional episodes: theirrelationship to the Quaternary stratigraphicframework in the northwestern portion of theGulf Basin. Bureau of Economic Geology, Universityof Texas, Geological Circular 74-1, 26 p.

Grabau, A. 1906. Types of sedimentary overlap.

GSA Bulletin, v. 17, p. 567-636.

Krumbein, W. and Sloss, L. 1951. Stratigraphyand sedimentation. W. M. Freeman and Co. SanFrancisco, 495 p.

Merriam, D. (ed.) 1964. Symposium on cyclicsedimentation. Kansas Geological Survey, Bulletin169 (2 volumes), 636 p.

Mitchum, R, Vail, P., and Thompson, S. 1977. Seismicstratigraphy and global changes in sea level, part2: the depositional sequence as the basic unit forstratigraphic analysis. in: Payton, C. (ed.) Seismicstrat igraphy: appl icat ion to hydrocarbonexploration. AAPG Memoir 26, p. 53-62.

Payton, C . (ed.) 1977. Seismic stratigraphy:applications to hydrocarbon exploration. AAPGMemoir 26, 516 p.

Posamentier, H. 2003. A linked shelf-edge deltaand slope-channel turbidite system: 3d seismiccase study from the eastern Gulf of Mexico. in:Roberts, H., Rosen, N, Fillon, R., and Anderson, J.(eds.), Shelf margin deltas and linked down slopepetroleum systems, Proceedings of the 23rdGCSSEM conference, p. 115-134.

Sloss, L. 1963. Sequences in the cratonic interiorof North America. GSA Bulletin, v. 74, p. 93-113.Sloss, L., Krumbein, W., and Dapples, E. 1949.Integrated facies analysis. in: Longwell, C. (ed.).Sedimentary facies in geologic history. GeologicalSociety America, Memoir 39, p. 91-124.

Vail, P. et al. 1977. Seismic stratigraphy and globalchanges in sea level. in: Payton, C. (ed.). Seismicstratigraphy: applications to hydrocarbonexploration, AAPG Memoir 26, p. 49-212.

Figure 2.3. The same 10 unconformities as shown on Figure 2.1 are present on the basin flank and the same9 depositional sequences are delineated. With Mitchum et al.’s (1977) addition of the “correlative conformity”to the definition of a sequence boundary, the same 9 sequences can be extended over the entire basin. Thisresolved the nomenclatural nightmare of “unconformity-only” sequences and the lack of subdivision in thecentral portions of a basin.

Page 10: Embry Sequence Stratigraphy

8

Weller, J.M. 1930. Cyclic sedimentation of thePennsylvanian period and its significance. Journalof Geology, v. 38, p. 97-135.

Wheeler, H.E. 1958. Time stratigraphy. AAPGBulletin, v. 42, p.1208-1218.

Wheeler, H.E. 1959. Stratigraphic units in timeand space.American Journal Science, v. 257, p.692-706.

Wheeler, H.E. 1964a. Base level, lithospheresurface and time stratigraphy. GSA Bulletin. v.75, p. 599-610.

Wheeler, H.E. 1964b. Base level transit cycle. in:Merriam, D.F. (ed.) Symposium on c yc l icsedimentation: Kansas Geological Survey, Bulletin169, p. 623-629.

Wheeler, H.E. and Murray, H. 1957. Base levelcontrol patterns in cyclothemic sedimentation.AAPG Bulletin, v. 41, p. 1985-2011.

Wanless, H. and Shepard, F. 1936. Sea level andclimatic changes related to late Paleozoic cycles.GSA Bulletin, v. 47, p. 1177-1206.

Wilgus , C . , Hast ings , B .S . , Kendal l , C .G . ,Posamentier, H.W., Ross, C.A., and Van Wagoner,J.C . 1988. Sea level changes: an integratedapproach. SEPM Spec. Pub. 42, 407 p.

Page 11: Embry Sequence Stratigraphy

9

Practical Sequence Stratigraphy IIIPractical Sequence Stratigraphy IIIPractical Sequence Stratigraphy IIIPractical Sequence Stratigraphy IIIPractical Sequence Stratigraphy IIIHistorical Development of the Discipline:The Last 20 Years (1988-2008) by Ashton Embry

In the last article in this series, I looked atthe first 200 years (1788 to 1988) of thedevelopment of sequence stratigraphy as auseful strat igraphic disc ipl ine forcorrelation, mapping, and interpretingdepositional history. By 1988 we had arevised definition of a sequence which wasa strat igraphic unit bounded byunconformities or correlative conformities(Mitchum et al, 1977). Because this definitionwas based mainly on observations fromseismic data, there was considerableconfusion as to what specific types ofstrat igraphic surfaces const ituted asequence boundary, especia l ly thecorrelative conformities. The veil was liftedin 1988.

The Exxon Sequence ModelThe Exxon Sequence ModelThe Exxon Sequence ModelThe Exxon Sequence ModelThe Exxon Sequence ModelIn 1988, the first, comprehensive sequencestratigraphic model was described in aseries of papers authored by researchersfrom Exxon Corporation. These papersappeared in SEPM Special Publication 42 -Sea level Changes: An Integrated Approach(Wilgus et al., 1988) and they presentedExxon’s methods, models, classificationsystems, and terminology for sequencestratigraphy. These papers also made clearhow Exxon scient ists del ineated andcorrelated a sequence boundary from basinedge to basin centre. The Exxon work wasbased on a combination of theoreticalmodeling and empirical observations fromseismic records, well-log cross-sections,and outcrop data.

The paper by Mac Jervey (Jervey, 1988)presented a quantitative, theoretical modelfor sequence development and it greatlyexpanded the concepts on the interactionof sedimentation and base-level changewhich had been first explored by Barrell(1917) and Wheeler (1958). Jervey’s modelused sinusoidal sea level change, subsidencewhich increased basinward, and a constantsediment supply as its input parameters. Themodel predicted that, during a cycle of base-level rise and fall (see Jervey, 1988, Fig. 9),three different sedimentary units would besequentially developed and that these wouldconstitute a sequence. These were an initial,progradational (regressive) unit depositedduring initial slow base-level rise, a middle,retrogradat ional (transgressive) unit

deposited during fast base-level rise and anupper, progradational (regressive) unitdeposited as base-level rise slowed andduring the subsequent interval of base-levelfall.

Depositional Sequence BoundariesDepositional Sequence BoundariesDepositional Sequence BoundariesDepositional Sequence BoundariesDepositional Sequence BoundariesOn the basis of Jervey’s concepts and thestrat igraphic geometries observed onregional seismic profiles, Exxon scientists(Baum and Vail, 1988; Posamentier et al., 1988;Posamentier and Vail, 1988) developed atheoretical sequence stratigraphic model fora shelf/slope/basin depositional setting(Figure 3.1). In this model, a depositionalsequence is bound by subaeria lunconformities on the basin margin and bycorrelative surfaces farther basinward.

Two types of deposit ional sequenceboundaries, originally defined by Vail and Todd(1981), were included in the Exxon model. AType 1 sequence boundary encompassed amajor subaeria l unconformity whichextended from the basin edge, past the shelfmargin and onto the upper slope. Basinward,the boundary was called a correlativeconformity (see Baum and Vail, 1988, Fig. 1)and was extended along the base of theturbidite facies which occupied the basinfloor and onlapped the lower slope (Figure3.1).

A Type 2 sequence boundary comprised arelatively minor subaerial unconformitywhich did not reach the shelf edge. It wasconfined mainly to the proximal portion ofthe shelf often within nonmarine strata(Posamentier and Vail, 1988, Fig. 18). Thebasinward extension of the Type 2 sequenceboundary (correlative conformity) was alonga chronostratigraphic surface equal to thetime of start base-level r ise (t ime ofmaximum rate of eustatic fall) (Baum and Vail,1988, Fig. 1; Van Wagoner et al., 1988, Fig. 4;Posamentier et al., 1988, Fig 6).

Systems TractsSystems TractsSystems TractsSystems TractsSystems TractsThe depositional sequence was divided intothree component units which were termedsystems tracts. These approximated the threeunits that Jervey (1988) had deduced as beingpart of a sequence that develops during asinusoidal, base-level rise/fall cycle. Thelower unit was called the lowstand systemstract (LST) and it consisted of a basal unit ofturbidites overlain by a progradational wedgewhich onlapped the upper slope portion ofthe sequence bounding unconformity. TheLST was bound by the sequence boundarybelow and the transgressive surface above(Figure 3.1). The transgressive surface, asdefined by the Exxon workers, marks thechange from progradational sedimentationbelow to retrogradational sedimentation

Figure 3.1. The Exxon depositional sequence model of 1988. The lower boundary is a Type 1 sequenceboundary (SB1) and it coincides with a subaerial unconformity on the shelf and upper slope and with thebase of submarine fan deposits in the basin. The upper boundary is a Type 2 boundary (SB2) and it coincideswith a subaerial unconformity on the shelf and with a time surface (clinoform) equivalent to the start base-level rise farther basinward.

The transgressive surface (TS) and maximum flooding surface (mfs) occur within the sequence and allow itto be subdivided into three systems tracts – lowstand (LSW), transgressive (TST) and highstand (HST). Thesystems tract which directly overlies a Type 2 sequence boundary is called a shelf margin systems tract(SMW). Modified from Baum and Vail (1988, Fig.1).

Page 12: Embry Sequence Stratigraphy

10

above. The LST was interpreted to havedeveloped during most of base-level fall andthe early part of rise.

The middle unit was called the transgressivesystems tract (TST) and it consisted ofretrogradat ional sediments thatoverstepped the LST and onlapped the shelfalportion of the subaerial unconformity. TheTST was bound by the transgressive surfacebelow and the maximum flooding surface(MFS) above (Figure 3.1, page 9). The MFSwas defined as the surface of sequencestratigraphy that marks the change fromretrogradational sedimentation below toprogradational sedimentation above. Thetransgressive systems tract was interpretedto have developed during high rates of base-level rise.

The upper systems tract was termed thehighstand systems tract (HST) and it consistedof progradational sediments which werecapped by a subaerial unconformity (sequenceboundary on the basin flanks and by thecorrelative surfaces farther basinward (Figure3.1, page 9). The HST was interpreted to havedeveloped during the waning stage of base-level rise and the early portion of base-levelfall. The wedge of strata above the Type 2sequence boundary and below the nexttransgressive surface was called the shelfmargin systems tract (SMW) (Figure 3.1 page 9).

Van Wagoner et al., (1988) applied the sameterminology to s i l ic ic last ic sedimentsdeposited in a ramp setting (see their Fig. 3).In this case the sequence boundary wasextended basinward from the terminationof the unconformity along the facies contactbetween shallow water sandstones aboveand marine shales below and then into theoffshore shales. Van Wagoner et al., (1988)also def ined a smal l -scale sequencestratigraphic unit termed a parasequence. Itwas defined as a relatively conformablesuccession of genetically related beds orbedsets bound by marine flooding surfaces.A marine flooding surface was defined as asurface which separates older from youngerstrata across which there is evidence of anabrupt deepening.

There can be no doubt that the Exxonsequence stratigraphic work was a verysignificant and important contribution tosedimentary geology because it transformedsedimentology from a static discipline whichfocused on facies models to a dynamicdiscipline in which facies developed in aframework of shifting base level. It alsoprovided a comprehensive classificationsystem for sequence stratigraphy andelucidated the linkage between changing baselevel and the development of specific surfacesof sequence stratigraphy and the units they

enclosed. Thus it is not surprising the modelwas enthusiast ical ly embraced by theindustrial and academic sedimentary geologycommunities.

The Exxon sequence model , andaccompanying methods and classificationsystems were the product of a combinationof theoretical deduction and empiricalobservations. Most of the stratigraphicsurfaces employed in their work werematerial-based surfaces which were definedon the basis of physical criteria. Howeverthe model also included an abstract timesurface equated with the start of base-levelrise. As will be discussed in future articles,the inclusion of a chronostratigraphicsurface has caused some problems. Their1988 sequence model has a few otherinconsistencies and these also wil l bediscussed in subsequent articles.

Genetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceThe next contribut ion to sequencestrat igraphic c lass i f icat ion came withGalloway’s (1989) proposal that a sequencebe bound by maximum flooding surfaces(“downlap surfaces”) , the prominentstratigraphic surface at the top of the TST ofExxon (Figure 3.2). Such a sequence was acompletely different stratigraphic entity fromthe depositional sequence of the Exxonmodel, although it did fit Mitchum et al.’s(1977) general definition of a sequencebecause the distal portion of a MFS is oftenan unconformity produced mainly by

sediment starvation. The conformable ,proximal portion of the MFS is a suitablecorrelative conformity of the sequenceboundary. Gal loway (1989) named asequence bounded by MFSs a geneticstratigraphic sequence (GSS).

In contrast to the Exxon depositionalsequence, which was in part based onJervey’s (1988) deductive model, Galloway’sgenetic stratigraphic sequence was purelyan empirical construct based on his extensivework on the Tertiary strata of the Gulf Coastand the observation that MFSs are usuallythe most readily recognizable sequencestratigraphic surfaces in marine successions.

Revision of the Exxon ModelRevision of the Exxon ModelRevision of the Exxon ModelRevision of the Exxon ModelRevision of the Exxon ModelHunt and Tucker (1992) were the first authorsto modify the original 1988 Exxon sequencemodel and they focused on the placement ofthe Type 1 depositional sequence boundary.In the Exxon model, strata deposited duringbase-level fa l l were placed below theunconformable sequence boundary on thebasin f lanks and above the sequenceboundary in more basinward localities. Huntand Tucker (1992) correctly asserted that thedepositional sequence boundary in the basinmust lie on top, rather than below, the stratadeposited during fall (i.e., the submarine fanturbidites) to ensure a single, through-goingsequence boundary is delineated. NotablyJervey (1988, Fig. 20), in his deductive modelof sequence development, also put theturbidite fac ies below the sequence

Figure 3.2. This schematic cross-section illustrates the boundaries of both the genetic stratigraphic sequence(GSS) of Galloway (1989) (MFS boundary) and the T-R sequence of Embry (1993) (composite SU/SR-U/MRSboundary). Both of these sequence types are based solely on empirical observations and this contrasts withthe 1988 Exxon depositional sequence model which is substantially derived from theoretical deduction.Embry (1993) subdivided a T-R sequence into a transgressive systems tract (TST) and a regressive systemstract (RST) on the basis of the enclosed MFS. These systems tracts can be readily applied to a GSS.

Page 13: Embry Sequence Stratigraphy

11

boundary rather than above it. In theirmodification of the Exxon model, Hunt andTucker (1992) proposed that the basinwardextension of the sequence boundary be alongthe time surface at the start of base-levelrise which they cal led the correlativeconformity (CC) (Figure 3.3).

They also added a fourth systems tract inthe uppermost portion of a sequence andcalled it the forced regression systems tract(FRST). Their FRST was bound above by thesequence boundary (SU on the flank, CC inthe basin) and below by the newly defined,“basal surface of forced regression” (BSFR)which equated to a time surface equal to thetime of start base-level fall (Figure 3.3).

With such a def in it ion, the FRSTencompassed all the strata deposited duringbase-level fall. The LST of Hunt and Tucker(1992) was restricted to the strata betweenthe CC below and the transgressive surfaceabove and represented the progradationalstrata deposited during the initial phase ofslow base-level rise that occurred in theJervey model (Figure 3.3). Thus the LST ofHunt and Tucker (1992) was equivalent toonly part of the LST of the Exxon Type 1sequence but was entirely equivalent to theSMW of the Exxon Type 2 sequence.

To complicate matters even more, Nummedalet al., (1993) referred all strata depositedduring basin level fall as the falling stagesystems tract (FSST). The four systems tract,sequence model of Hunt and Tucker (1992)was elaborated on and clearly illustrated byHelland-Hansen and Gjelberg (1994) whoably demonstrated the theoretical logic ofsuch a classification system.

TTTTT-R S-R S-R S-R S-R SeeeeeqqqqquuuuuenenenenenccccceeeeeIn 1993, due to my inability to objectivelyapply the Exxon sequence stratigraphicmethods and classification system to verywell exposed strata of the nine km-thickMesozoic succession of the Sverdrup Basinof Arctic Canada, I suggested anotherpossible combination of surfaces whichwould satisfy the basic definition of asequence boundary (Embry, 1993; Embry andJohannessen, 1993).

Following the work of Wheeler (1958) andExxon (Posamentier et al., 1988), a subaerialunconformity (SU) was used as a sequenceboundary on the basin flank with the provisothat in many settings the SU is partially ortotally replaced by a shoreline ravinement(SR-U) which represents the landwardportion of the transgressive surface of Exxonworkers. The sequence boundary wasextended basinward from the terminationof the basin flank unconformity (SU/SR-U)along the maximum regressive surface (MRS)

which represents the basinward,conformable portion of the transgressivesurface (Figure 3.2).

Such a boundary was theoretically reasonablebecause the landward termination of themaximum regressive surface joins thebasinward termination of the shorelineravinement. Thus the composite of the SU,SR-U and MRS forms one, single through-going sequence boundary which can berecognized with objectivity from basin edgeto basin centre. The unit bound by this newlydefined sequence boundary was called a T-Rsequence because the sequence boundaryseparated strata deposited during regressionbelow from transgressive strata above.

The T-R sequence was divided into twosystems tracts: a transgressive systems tract(TST) bounded by the sequence boundarybelow and the MFS above and a regressivesystems tract (RST) bounded by the MFSbelow and the sequence boundary above(Figure 3.2). Notably, like Galloway’s (1989)GSS, a T-R sequence was entirely an empiricalconstruct based on observat ions ofsubsurface sections and extensive, very wellexposed outcrops.

Another DepositionalAnother DepositionalAnother DepositionalAnother DepositionalAnother DepositionalSequence BoundarySequence BoundarySequence BoundarySequence BoundarySequence BoundaryAnother proposal for defining a depositionalsequence boundary was made by

Posamentier and Allen (1999). They suggestedusing only a port ion of the subaeria lunconformity as the sequence boundary andthen extending the boundary basinward alongthe time surface at the start of fall; the BSFRas defined by Hunt and Tucker (1992)(Posamentier and Allen, 1999, Fig. 2.31)(Figure 3.4, page 12).

As illustrated by Posamentier and Allen(1999), the juncture between the SU and theBSFR occurs well landward of the basinwardtermination of the SU (Figure 3.4, page 12).They subdivided such a sequence into threesystems tracts – LST, TST, and HST – and thesewere defined essentially in the same way asthose used for the Exxon Type 1 sequence ofPosamentier and Vail (1988). Posamentier andAllen (1999) also suggested that the conceptof a Type 2 sequence (boundary) beabandoned.

SummarySummarySummarySummarySummaryIn summary, over the past 20 years, differentmodels for sequence boundary delineationand for the subdivision of a sequence intosystems tracts have been proposed. This hasresulted in considerable confusion andmiscommunication as different authors applydifferent sequence models and terminologyin their study areas. In some cases the sameterm is used for different entities (e.g., theLST of Posamentier and Allen (1999) versusthe LST of Hunt and Tucker (1992)). In other

Figure 3.3. The depositional sequence and component systems tracts of Hunt and Tucker (1992). Hunt andTucker (1992) advocated for the utilization of two time surfaces – the basal surface of forced regression(BSFR) (equals start base-level fall) and the correlative conformity (CC) (equals start base-level rise) fordelineating sequence and systems tract boundaries. Their depositional sequence boundary is a composite ofthe subaerial unconformity on the basin flank and the correlative conformity time surface farther basinward.Note that the forced regressive systems tract (FRST), which was defined by Hunt and Tucker (1992), is boundby both time surfaces and represents all the strata deposited during base-level fall. The other systems tractsare lowstand (LST), transgressive (TS) and highstand (HST).

Page 14: Embry Sequence Stratigraphy

12

cases different terms are used for the sameentity (e.g., the FRST of Hunt and Tucker(1992) and the FSST of Nummedal et al.,(1993) for all strata deposited during base-level fall). Such nomenclatural chaos is notconducive for effective communication.

Most importantly, the different proposals forsequence models and classification systemshave not been comprehensively comparedso as to determine the relative pros and consof each one. Such a review would helpworkers select the most appropriatesequence stratigraphic classification systemfor their studies.

In following articles in this series, I will buildsequence strat igraphic methods andclassification systems from the bottom up. Iwill also review a variety of classificationssystems with regards to applicability to realworld situations encountered by petroleumgeologists. In the next few articles we willlook at the various surfaces of sequencestratigraphy which have been defined andeach will be evaluated in terms of theirusefulness for correlation and for boundingsequence stratigraphic units.

ReferencesReferencesReferencesReferencesReferencesBarrell, J.1917. Rhythms and the measurementsof geologic time. GSA Bulletin, v. 28, p. 745-904.

Baum, G. and Vail, P. 1988. Sequence stratigraphicconcepts applied to Paleogene outcrops, Gulf andAtlantic Basins. In: Sea level changes: an integratedapproach. C.Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner(eds.). SEPM Special Publication 42, p. 309-327.

Embry, A.F. 1993. Transgressive-regressive (TR)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embry, A.F. and Johannessen, E.P. 1993. TRsequence stratigraphy, facies analysis andreservoir distribution in the uppermost Triassic-Lower Jurassic succession, western Sverdrup Basin,Arctic Canada. In: Arctic Geology and PetroleumPotential. T. Vorren, E. Bergsager, O.A. Dahl-Stamnes,E. Holter, B. Johansen, E. Lie, and T.B. Lund (eds.).NPF Special Publication 2, p.121-146.

Gal loway, W. 1989. Genet ic strat igraphicsequences in basin analysis I: architecture andgenesis of flooding surface bounded depositionalunits. AAPG Bulletin, v. 73, p. 125-142.

Helland-Hansen, W., and Gjelberg, J. 1994.Conceptual basis and variability in sequencestratigraphy: a dif ferent perspective. SedimentaryGeology, v. 92, p. 1-52.

Hunt , D. and Tucker, M. 1992. Strandedparasequences and the forced regressive wedgesystems tract: deposition during base-level fall.Sedimentary Geology, v. 81, p. 1-9.

Jervey, M. 1988. Quantitative geological modelingof siliciclastic rock sequences and their seismicexpression, In: Sea level changes: an integratedapproach. C.Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner(eds.). SEPM Special Publication 42, p. 47-69.

Mitchum, R, Vail, P., and Thompson, S. 1977. Seismicstratigraphy and global changes in sea level, part

2: the depositional sequence as the basic unit forstratigraphic analysis, In: Seismic stratigraphy:application to hydrocarbon exploration. C. Payton(ed). AAPG Memoir 26, p. 53-62.

Nummedal, D., Riley, G., and Templet, P. 1993.High resolut ion sequence architecture: achronostratigraphic model based on equilibriumprof ile studies. In: Sequence stratigraphy andfac ies assoc iat ion . H.Posament ier ,C.Summerhayes, B. Haq, and G. Allen (eds).International Association of Sedimentologists,Special Publication 18, p. 55-68.

Posamentier, H. Jervey, M., and Vail, P. 1988.Eustatic controls on clastic deposition I: conceptualframework. In: Sea level changes: an integratedapproach. C.Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner(eds.). SEPM Special Publication 42, p. 109-124.

Posamentier, H. and Vail, P. 1988. Eustatic controlson clastic deposition II- sequence and systemstract models. In: Sea level changes: an integratedapproach. C.Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner(eds.). SEPM Special Publication 42, p. 125-154.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. SEPM Concepts in Sedimentologyand Paleontology, # 7, 210 pp.

Vail, P. and Todd, R. 1981. Northern North SeaJurassic unconformities, chronostratigraphy andsea-level changes from seismic stratigraphy. In:Petroleum Geology of the Continental Shelf ofNorthwest Europe. L. Illing and G. Hobson (eds.).Heyden and Son, Ltd., London, p. 216-236.

Van Wagoner, J.C., Posamentier, H.W., Mitchum,R.M., Vail, P.R., Sarg, J.F., Loutit, T.S., and Hardenbol,J. 1988. An overview of the fundamentals ofsequence stratigraphy and key def initions. In:Sea level changes: an integrated approach.C .Wilgus, B.S. Hastings, C .G. Kendall, H.W.Posamentier, C.A. Ross, and J.C. Van Wagoner(eds.). SEPM Special Publication 42, p. 39-46.

Wheeler, H.E. 1958. Time stratigraphy. AAPGBulletin, v. 42, p.1208-1218.

Wilgus , C . , Hast ings , B .S . , Kendal l , C .G . ,Posamentier, H.W., Ross, C.A., and Van Wagoner,J.C. (eds). 1988. Sea level changes: an integratedapproach. SEPM Spec. Pub. 42, 407 p.

Figure 3.4. The depositional sequence boundary of Posamentier and Allen (1999). This proposed sequenceboundary consists of only part of the subaerial unconformity (SU) and is extended basinward along the basalsurface of forced regression (the time surface at the start of base-level fall). Note that a substantial portion ofthe subaerial unconformity lies within the sequence. The sequence is divided into the same three systemstracts as recognized for the 1988 Type 1 sequence model of Exxon.

Page 15: Embry Sequence Stratigraphy

13

Practical Sequence Stratigraphy IVPractical Sequence Stratigraphy IVPractical Sequence Stratigraphy IVPractical Sequence Stratigraphy IVPractical Sequence Stratigraphy IVThe Material-based Surfaces of Sequence Stratigraphy,Part 1: Subaerial Unconformity and Regressive Surfaceof Marine Erosion by Ashton Embry

The fundamental building blocks of sequencestratigraphy are the various sequencestratigraphic surfaces that are defined andused for correlation and for bounding units.As discussed in the first article in this series,sequence stratigraphic surfaces representchanges in depositional trend and thisdistinguishes them from surfaces of otherstratigraphic disciplines which representchanges in different observable propertiesof strata.

Before describing various surfaces in detail,a few general i t ies about surfaces arerequired. First of all, there are two distinctlydifferent types of sequence stratigraphicsurfaces in use today – material-based andtime-based.

A material-based surface is defined on thebasis of observable physical characteristicswhich include 1) the physical properties ofthe surface and of overlying and underlyingstrata and 2) the geometrical relationshipsbetween the surface and the underlying andoverlying strata.

A time-based surface in sequencestratigraphy is defined on the basis of aninterpreted, site-specific event related to achange in either the direction of shorelinemovement (e.g., landward movement toseaward movement) or the direction of base-level change (e.g., falling base level to risingbase level).

Surfaces are also described in terms of theirrelationship to the interpreted time gapacross the surface. A surface across whichthere is a large, significant time gap asevidenced by the missing stratigraphicsurfaces (e.g., truncation, onlap) is called anunconformity. If the time gap is very minorand is inferred mainly on the basis of ascoured and/or abrupt contact rather thanon missing surfaces across the contact, thesurface is called a diastem. If there is noinferred time loss across the surface, it isreferred to as a conformity . Notably,different portions of a single surface typecan exhibit different relationships to time(e.g., one portion can be conformable and

another portion unconformable with yetanother portion being diastemic). Finally,surfaces are often interpreted in terms oftheir relationship to time over parts or allof their extent, that is, the relationshipbetween the surface and time surfaces. If agiven surface is conformable and the sameage over its entire extent, it is a time surface.However, no material-based, conformablesurface is equivalent to a time surface becausethe generation of such a surface is alwaysdependent in part on sedimentation rate. Thisfactor always varies in space and time,ensuring all conformable, material-basedsurfaces will develop over an interval of timeand will always exhibit some diachroneity(i.e., time surfaces will pass through them).

Surfaces which develop over an extendedinterval of time such that time surfaces crossthem at a high angle are classed as beinghighly diachronous. Those which developover a relatively short time interval such thattime lines cross them at a low angle arereferred to as having low diachroneity. Insome cases, time surfaces do not cross asurface but rather terminate against it (e.g.,truncation, onlap) (Figure 4.1). Such a surfaceis either an unconformity or a diastem and isreferred to as a time barrier. Wherever asurface is a time barrier, all strata below itare entirely older than all strata above it.

It must be noted that some unconformitiesor diastems are diachronous and timesurfaces pass through them (offset) ratherthan terminating against them. Once again, asingle surface can exhibit more than onerelationship time over its extent (e.g., a highlydiachronous diastem over one portion anda time barrier unconformity over another)

The six, material-based surfaces of sequencestratigraphy (Embry, 1995, 2001) in commonuse for correlation and/or as a unit boundaryare:

1) Subaerial unconformity,2) Regressive surface of marine erosion,3) Shoreline ravinement,4) Maximum regressive surface,5) Maximum flooding surface, and6) Slope onlap surface.

Importantly, each of these surfaces ischaracterized by a combination of observableattributes that allow it to be distinguishedfrom other stratigraphic surfaces and allowfor its recognition by objective criteria. Inthis article, the first two of these surfacesare described and interpreted as to theirorigin, their relationship to time, and theirpotential usefulness for correlation andbounding a sequence stratigraphic unit. Theremaining material-based surfaces, as well

Figure 4.1. Time surfaces terminate at the unconformity that is a time barrier. The time surfaces below theunconformity are truncated and they onlap on top of the unconformity. All the strata below the unconformityare entirely older than all the strata above.

Page 16: Embry Sequence Stratigraphy

14

as time-based surfaces, will be discussed insubsequent articles.

Subaerial Unconformity (SU)Subaerial Unconformity (SU)Subaerial Unconformity (SU)Subaerial Unconformity (SU)Subaerial Unconformity (SU)The subaerial unconformity is an important,sequence stratigraphic surface and was thesurface used to empirically define a sequencein the first place (Sloss et al., 1949). It wasfirst recognized through observation over200 years ago and James Hutton’s discovery/ recognition of the Silurian / Devoniansubaerial unconformity at Siccar Point,Scotland is legendary. The defining attributesof a subaerial unconformity are an erosivesurface or weathering zone (e.g., paleosol,karst) overlain by nonmarine/brackish marinestrata, and the demonstrat ion that itrepresents a s igni f icant gap in thestratigraphic record (Figure 4.2). Any typeof strata can lie below. Shanmugan (1988)elaborates on the physical characteristics ofa subaerial unconformity.

It is worth emphasizing that nonmarine tobrackish strata are required to overlie asubaerial unconformity. When marine strataoverl ie strata that had been formerlyexposed and eroded, the surface marking thecontact is not a subaerial unconformity. Thereis little doubt that an SU once overlay theeroded strata but it is no longer presenthaving been eroded during the passage ofmarine waters over it. Most often theremaining unconformable surface is ashorel ine ravinement a lthough othersurfaces can potentially erode through andthus replace a subaerial unconformity as thesurface marking a major gap in the succession.

The occurrence of a significant stratigraphicgap across a subaerial unconformity is criticalfor its recognition because this establishesthe unconformable nature of the surface.Impor tant ly, this a l lows a subaeria lunconformity to be distinguished fromsubaerial diastems which are scouredcontacts at the base of fluvial channel strataand which are much more common in therecord. Such subaerial diastems originatethrough channel migration on a flood plainand are highly diachronous, diastemicsurfaces which harbour only a very minortime gap at any locality.

To demonstrate the presence of a significanttime gap beneath an SU, it is usually necessaryto show that truncated strata lie below thesurface . The occurrence of onlappingnonmarine strata above the surface addsfurther support to such an interpretation.These stratigraphic relationships are mostreadily seen on seismic data integrated withfacies data from wells (Vail et al., 1977)although sometimes such seismical lydetermined relationships are not real andare an artifact of the seismic parameters

(Cartwright et a l . , 1993;Schlager, 2005; Janson et al.,2007).

The geometrical relationships(truncation, onlap) which helpto delineate an SU can also oftenbe determined on crosssections of well log and/oroutcrop data (Figure 4.3). Datafrom other strat igraphicdiscipl ines, especia l lybiostratigraphy, can be useful indemonstrating the occurrenceof a substantial time gap acrossa suspected subaeria lunconformity.

Barrell (1917) and Wheeler(1958) related the origin of asubaerial unconformity to themovement of base level, whichis the conceptual surface ofequilibrium between erosionand deposition. Deposition canpotentially occur where baselevel occurs above the surfaceof the Earth and erosion will occur in areaswhere it lies below the Earth’s surface. Asubaerial unconformity is interpreted toform by subaerial erosional processes,especially those connected to fluvial and/orchemical erosion, during a time of base-levelfall (Jervey, 1988). As base level falls beneaththe Earth’s surface, subaerial erosion cutsdown to that level.

Wheeler (1958) and Jervey (1988) also

showed that a subaerial unconformityadvances basinward during the entire timeof base-level fall and reaches its maximumbasinward extent at the end of base-levelfall. It continues to form during subsequentbase-level rise as it retreats landward and isonlapped by nonmarine to brackish sediment.

In regards to its relation to time surfaces, asubaerial unconformity is commonly anapproximate time barrier and time surfaces,

Figure 4.2. Outcrop of Lower Cretaceous strata on east central Axel Heiberg Island. The surface labeled SUis an abrupt scour surface beneath fluvial channel strata. The surface truncates strata below and is onlappedby fluvial strata above. It has all the characteristics of a subaerial unconformity and is interpreted as such.Regional correlations indicate major truncation below this surface confirming its interpretation as an SU.

Figure 4.3. An SU is delineated on this cross section on the basis ofthe truncation of the underlying strata of the Deer Bay Formationand the occurrence of fluvial strata directly above the unconformity.

Page 17: Embry Sequence Stratigraphy

15

Figure 4.4. An SU is often an approximate time barrier because fluvial stratadeposited during base-level fall can be preserved in incised valleys. Suchstrata (deposited at time T2) are the same age as deltaic strata depositeddown dip and which underlie the SU. In this case some strata on top of theinterpreted SU are older than some strata below it and it is not a perfect timebarrier.

Figure 4.5. An RSME is interpreted at the base of the sandstone because it isan abrupt contact which is underlain by offshore shelf strata which coarsen-upward (see sonic log) and overlain by coarsening-upward shoreface strata(after Plint, 1988).

Figure 4.6. The RSME forms as a scour zone on the inner shelf in front of theshoreface during base-level fall. It migrates basinward during the entire inter-val of base-level fall and is downlapped by shoreface strata which are in turncapped by an SU.

Figure 4.7. The RSME is a highly diachronous surface and time surfaces passthrough it at a high angle. The time surfaces are offset across the RSME withshoreface strata above the RSME being the same age as offshore shelf stratabelow it. The RSME is not a time barrier.

for the most part, do not pass through it. In other words, almost allstrata below the surface are older than almost all those above.

There are definitely exceptions to this and these can be associatedwith migrating uplifts (Winker, 2002). Also, some fluvial strata, especiallyat the base of incised valley fills, may have been deposited during base-level fall (Suter et al, 1987; Galloway and Sylvia, 2002; Blum and Aslan,2006) and are thus older than some of the down-dip strata below theunconformity (Figure 4.4). In this case, the actual subaerial unconformitywould lie on top of fluvial strata deposited during fall. However, such asurface would likely be very hard to identify within a succession offluvial strata and the basal contact of the fluvial strata is best interpretedas the SU unless compelling data dictate otherwise (Suter et al., 1987).

The time-barrier aspect of a subaerial unconformity makes it animportant surface for correlation and for bounding genetic units. Thissurface has been given other names besides subaerial unconformitysuch as lowstand unconformity (Schlager, 1992), regressive surface offluvial erosion (Plint and Nummedal, 2000), and fluvial entrenchment /incision surface (Galloway and Sylvia, 2002). However, the termsubaerial unconformity has the widest acceptance and is the one Iwould recommend for this surface.

Regressive Surface of Marine Erosion (RSME)Regressive Surface of Marine Erosion (RSME)Regressive Surface of Marine Erosion (RSME)Regressive Surface of Marine Erosion (RSME)Regressive Surface of Marine Erosion (RSME)The regressive surface of marine erosion was first empiricallyrecognized and named by Plint (1988) mainly on the basis of studies onCretaceous strata in Alberta. Its defining characteristics include being asharp, scoured surface and having offshore marine strata (usually midto outer shelf) that coarsen upwards below the surface and

coarseningand shallowing-upward shoreface strata above the surface(Figure 4.5). The underlying shelf strata are variably truncated and theoverlying shoreface strata downlap onto the RSME. Sometimes aGlossifungites trace fossil assemblage is associated with the RSME(MacEachern et al., 1992). The surface occurs within an overall regressivesuccession but is considered a change in depositional trend fromdeposition to nondeposition and back to deposition.

Plint (1988) interpreted the RSME formed during a time of base-levelfall when the inner part of the marine shelf in front of the steepershoreface is sometimes eroded. This area of inner shelf erosion, whichcan be up to a few tens of kilometres wide, moves seaward during theentire interval of base-level fall and is progressively covered byprograding shoreface deposits (Figure 4.6). Thus the RSME canpotentially be quite widespread both along strike and down dip.However, it should be noted that in many cases an RSME either has apatchy distribution or does not form at all due to variations in seafloor slope, sedimentation rate, and base-level fall rate (Naish andKamp, 1997; Hampson, 2000; Bhattacharya and Willis, 2001).

In most cases, erosion beneath the RSME is minor and localized andthus it is almost always a diastem and not an unconformity (Gallowayand Sylvia, 2002). Sometimes local truncation of strata beneath an RSMEcan be demonstrated but this requires very close control. However,the potential for more substantial erosion exists and, in a few examples,it has been shown to be an unconformity where it has eroded througha subaerial unconformity (Bradshaw and Nelson, 2004; Cantalamessaand Celma, 2004).

Because the regressive surface of marine erosion migrates basinwardduring the entire time of base-level fall, it is a highly diachronous surfaceand time lines pass through it (offset) at a high angle (Embry, 2002)(Figure 4.7). It is not an approximate time barrier like a subaerial

Page 18: Embry Sequence Stratigraphy

16

unconformity except in the few instanceswhere it has eroded through a subaerialunconformity (e.g., Cantalamessa and Celma,2004). Because it is most often a highlydiachronous, diastemic surface and has a verypatchy distribution, the RSME is not suitablefor use as a bounding surface for sequencestratigraphic units or for being part of acorrelation framework. However, it isimportant to recognize such a surface whenit is present and to use it as part of faciesanalysis inside the established sequencestrat igraphic correlat ion framework.Galloway and Sylvia (2002) referred to thissurface as the regressive ravinement surface.The term regressive surface of marineerosion is most commonly used and isrecommended.

ReferencesReferencesReferencesReferencesReferencesBarrell, J. 1917. Rhythms and the measurementsof geologic time. Bulletin of the Geological Societyof America, v. 28, p. 745-904.

Bhattacharya, J. and Willis, B. 2001. Lowstanddeltas in the Frontier Formation, Powder RiverBasin, Wyoming: implications for sequencestratigraphic models. American Association ofPetroleum Geologists Bulletin, v. 85, p. 261-294.

Blum, M. and Aslan, A. 2006. Signatures of climateversus sea level change within incised valley fillsuccessions: Quaternary examples for the TexasGulf Coast. Sedimentary Geology, v. 190, p. 177-211.

Bradshaw, B. and Nelson, C. 2004. Anatomy andorigin of autochthonous late Pleistocene forcedregressive deposits, east Coromandel inner shelf,New Zealand: implications for the developmentand definition of the regressive systems tract. NewZealand Journal of geology and Geophysics, v.47, p. 81-97.

Cantalamessa, G. and Di Celma, C . 2004.Sequence response to syndepositional regionaluplift: insights from high-resolution sequencestratigraphy of late early Pleistocene strata,Periadriatic Basin, central Italy. SedimentaryGeology, v. 164, p. 283-309.

Cartwright, J., Haddock, R., and Pinheiro, R. 1993.The lateral extent of sequence boundaries. In:Williams, G. and Dobb, A. (eds.) Tectonics andsequence stratigraphy. Geological Society LondonSpecial Publication 71, p. 15-34.

Embry, A.F. 1995. Sequence boundaries andsequence hierarchies: problems and proposals.In: Steel, R.J., Felt, F.L., Johannessen, E.P. andMathieu, C. (eds.) Sequence stratigraphy on thenorthwest European marg in . NorwegianPetroleum Society (NPF) Special Publication 5, p.1-11.

Embry, A. 2001. The six surfaces of sequence

stratigraphy. American Association of PetroleumGeologists Hedberg Conference on sequencestratigraphic and allostratigraphic principles andconcepts, Dallas. Abstract volume, p. 26-27.

Embry, A.F. 2002. Transgressive-Regressive (T-R)Sequence Stratigraphy. In: Armentrout, J andRosen, N. (eds.). Sequence stratigraphic modelsfor exploration and production. Society forSedimentary Geology (SEPM), Gulf Coast SEPMConference Proceedings, Houston, p. 151-172.

Galloway, W.E. and Sylvia, D.A. 2002. The manyfaces of erosion: theory meets data in sequencestratigraphic analysis. In: Armentrout, J and Rosen,N. (eds). Sequence stratigraphic models forexplorat ion and product ion . Soc iety forSedimentary Geology (SEPM), Gulf Coast SEPMConference Proceedings, Houston, p. 99-111.

Hampson, G. 2000. Discontinuity surfaces,clinoforms and facies architecture in a wavedominated, shoreface-shelf parasequence. Journalof Sedimentary Research, v. 70, p. 325-340.

Janson, X, Eberli, G., Bonnaffe, F., Gaumet, F., andde Casanove, V. 2007. Seismic expression of aprograding carbonate margin, Mut Basin, Turkey.American Association of Petroleum GeologistsBulletin, v. 91. p. 685-713.

Jervey, M. 1988. Quantitative geological modelingof siliciclastic rock sequences and their seismicexpression. In: Wilgus, C., Hastings, B.S., Kendall,C .G., Posamentier, H.W., Ross, C.A., and VanWagoner, J.C . (eds .). Sea level changes: anintegrated approach. Society for SedimentaryGeology (SEPM) Special Publication 42, p.47-69.

Maceachern, J., Raychaudhuri, I., and Pemberton,G. 1992. Stratigraphic applications of theGloss i fung i tes ichnofac ies de l ineat ingdiscontinuities in the rock record. In: Pemberton,G. (ed.), Applications of Ichnology to PetroleumExploration. Society for Sedimentary Geology(SEPM) Core Workshop Notes 17, p. 169-198.

Naish , T. and Kamp, P. 1997. Sequencestratigraphy of 6th order (41 k.y.) Pliocene –Pleistocene cyclothems, Wanganui Basin, NewZealand: a case for the regressive systems tract.Geological Society of America Bulletin, v. 109, p.979-999.

Plint, A. 1988., Sharp-based shoreface sequencesand offshore bars in the Cardium Formation ofAlberta: their relationship to relative changes insea level. In: Wilgus, C., Hastings, B.S., Kendall, C.G.,Posamentier, H.W., Ross, C.A., and Van Wagoner,J.C . (eds.). Sea level changes: an integratedapproach: Society for Sedimentary Geology(SEPM) Special Publication 42, p. 357-370.

Plint, A. and Nummedal, D. 2000. The falling stagesystems tract: recognition and importance insequence stratigraphic analysis. In: Hunt, D. and

Gawthorpe, R. (eds.). Sedimentary responses toforced regressions. Geological Society of London,Special Publication 172, p. 1-17.

Schlager, W. 1992. Sedimentology and sequencestratigraphy of reefs and carbonate platforms.American Association of Petroleum GeologistsContinuing Education Course Notes Series # 34.71 p.

Schlager, W. 2005. Carbonate sedimentology andsequence stratigraphy. Society for SedimentaryGeology (SEPM) Concepts in Sedimentology andPaleontology 8, 200 p.

Page 19: Embry Sequence Stratigraphy

17

Practical Sequence Stratigraphy VPractical Sequence Stratigraphy VPractical Sequence Stratigraphy VPractical Sequence Stratigraphy VPractical Sequence Stratigraphy VThe Material-based Surfaces of Sequence Stratigraphy,Part 2: Shoreline Ravinement and Maximum RegressiveSurface by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionAs discussed in the last installment of thisseries, s ix materia l-based surfaces ofsequence stratigraphy have been empiricallyrecognized over the past 200 years. Eachsurface represents a specific change indepositional trend which can be recognizedon the basis of observat ional data.Collectively, these surfaces are the basicbuilding blocks of sequence stratigraphy andallow high resolution correlations, definitionand del ineat ion of speci f ic sequencestratigraphic units, and interpretations ofdepositional history in terms of base-levelchange. The two surfaces discussed in my lastarticle, the subaerial unconformity and theregressive surface of marine erosion, formedprimarily during base-level fall. The twosurfaces which are discussed in this article,shorel ine ravinement and maximumregressive surface, form at the start of, andduring, base-level rise.

As will be discussed, both these surfacespotentially have great utility in sequencestratigraphic analyses. Furthermore, asmateria l -based surfaces, they can beidenti f ied on the basis of physicalcharacteristics which include the nature ofthe surface itself, the nature of underlyingand overlying strata, and the geometricalrelationships between the surface andsurfaces in underlying and overlying strata.The relationship of the surfaces to eitherbase-level change or to a change in shorelinedirection has no role in their definition andcharacterization. However, the origin of eachsurface is interpreted in terms of theinteraction of sedimentation with base-levelchange.

Shoreline Ravinement (SR)Shoreline Ravinement (SR)Shoreline Ravinement (SR)Shoreline Ravinement (SR)Shoreline Ravinement (SR)A stratigraphic surface referred to herein asa shoreline ravinement has been empiricallyrecognized for a long t ime. Excel lentdescriptions of the surface and its mode oforigin were given by Stamp (1921), Bruun(1962), and Swift (1975). The characteristicattributes of a shoreline ravinement whichallow its recognition are an abrupt, scouredcontact overlain by estuarine or marinestrata which fine and deepen upwards.Underlying strata can vary from non-marine

to fully marine. As a scoured contact, itrepresents a change in trend from depositionto non-deposition and, as will be discussedin more detail, it can vary along its extentfrom being a minor diastem to being a majorunconformity.

The origin of a shoreline ravinement surfacewas determined by early workers on thebasis of observat ions a long modernshorelines which are transgressing (i.e.,moving landward). Because the slope of thealluvial plain is commonly less than that ofthe shoreface, erosion carves out a newshoreface profile as the shoreline moveslandward during transgression. One or moresuch erosional surfaces form as wave and/ortidal processes erode previously depositedshoreface, beach, brackish, and non-marinesediment. The eroded sediment is depositedboth landward and seaward of the shoreline(Figure 5.1). When both tidal and waveprocesses are acting in a given area, both atidal shoreline ravinement and a waveshoreline ravinement can form (Dalrympleet al., 1994; Zaitlin et al., 1994), although inmost cases only a wave shoreline ravinementis preserved.

The SR begins to form at the start oftransgression which occurs when rate of

base-level rise exceeds the sedimentationrate at the shoreline. This often occurs verysoon after the start of base-level rise alongmost of the shorel ine where thesedimentation rate is low to moderate(Embry, 2002). The SR stops being generatedat the end of transgression which can occurat anytime during base-level rise dependingon the interaction of the rate of base-levelrise with the rate of sediment supply. Becauseit develops over the ent ire t ime oftransgression, a shoreline ravinement isoften considered to be diachronous (e.g.,Nummedal et al., 1993). However, over itsextent, it can either be a diastem or anunconformity and thus can exhibit twodifferent relationships with regards to time(Figure 5.2, page 18). It is important todetermine which parts of a given shorelineravinement are unconformable(unconformable shoreline ravinement, SR-U) and which parts are diastemic (diastemicshoreline ravinement, SR-D).

A diastemic port ion of a shorel ineravinement (SR-D) has the def in ingcharacteristics of an SR as described aboveand is further characterized by the presenceof penecontemporaneous, nonmarine strataunderlying the surface and the preservationof the previously developed subaerial

Figure 5.1. The shoreline ravinement (SR) forms by shoreface erosion during transgression. It migrateslandward during the entire interval of transgression and thus can form a widespread surface. In this examplethe shoreline ravinement surface has cut down only partially through the succession of non-marine/brackishstrata that were deposited contemporaneously with the formation of the SR. In this case the SR is a highlydiachronous diastem.

Page 20: Embry Sequence Stratigraphy

18

unconformity (Figures 5.1, page 17; 5.2; 5.3;5.4). At any given locality, there is only a veryminor time gap across a diastemic shorelineravinement and overal l i t is a highlydiachronous surface with time lines cuttingit at a high angle and somewhat offset (Figure5.5).

Figure 5.2. The two different time relationships of a shoreline ravinement surface. A shoreline ravinementsurface is a highly diachronous diastem (SR-D) when it has not eroded the underlying subaerial nconformity.However, when it has eroded the underlying subaerial unconformity (SU), it is an unconformity and a timebarrier (SR-U) with all strata below being older than all strata above the surface.

Figure 5.3. In this outcrop of Early Cretaceous strata from eastern Axel Heiberg Island, a subaerial unconformity(SU) is present beneath the white-weathering fluvial sandstone. A shoreline ravinement occurs at a scouredcontact which occurs at the base of a thin, marine sandstone which fines upward into marine shale andsiltstone of mid-shelf origin. The strata between the SU and the SR are fluvial in origin. In this case the SR is adiastemic shoreline ravinement (SR-D) which is highly diachronous.

Figure 5.4. A subsurface section of Early Cretaceous strata with accompany-ing gamma ray/sonic logs. A shoreline ravinement separates non-marine stratabelow from marine strata above. The subaerial unconformity which formedduring the preceding base-level fall is preserved and the SR is a highlydiachronous, diastemic shoreline ravinement (SR-D).

Figure 5.5. The time relationships of a diastemic shoreline ravinement (SR-D) whichoverlies penecontemporaneous fluvial-brackish strata. Time surfaces cut the SR-Dat a high angle and are offset across the surface. Thus the SR-D is a highly diachronoussurface.

In contrast , a port ion of a shorel ineravinement that has removed both thepenecontemporaneous, non-marine stratathat were deposited behind the shorefaceas it moved landward, and the subaerialunconformity that had formed during thepreceding base-level fall and regression

(Figure 5.2), is an unconformity and not adiastem. With the removal of the subaerialunconformity, the shoreline ravinement takeson the time relationships of the subaerialunconformity and becomes a time barrierthat represents a significant gap in thestratigraphic record. All strata below anunconformable shoreline ravinement areolder than all strata on top of it (Figure 5.6).

The SR-U has the defining characteristics ofan SR and an additional characteristic is that,in most cases, the underlying strata aremarine rather than non-marine (Figures 5.7,5.8). However, the key characteristic whichallows the confident recognition of an SRUis that the strata below are regionallytruncated and the marine strata above oftenonlap (Figure 5.9, page 20). Such relationshipsare often clearly imaged on seismic data(Suter et al., 1987) or are determined bycorrelations on well log and outcrop crosssections. Notably the SR-U illustrated on thelog of Figure 5.8 would be hard to identify ifit could not be demonstrated that truncationwas occurring at the stratigraphic level.

Many major unconformit ies in thestratigraphic record, including some of thoseused by Sloss (1963) to define his continent-wide sequences, are unconformableshoreline ravinements rather than subaerialunconformities (e.g., the major, base Norianunconformity illustrated in Figure 5.9, page20). An unconformable shoreline ravinementcan be differentiated from a subaerialunconformity by the presence of marinestrata directly above the surface . Thischaracteristic contrasts with that of an SUwhich has fluvial/brackish strata directlyoverlying the surface . When estuarinedeposits directly overlie an unconformity, itis sometimes difficult to decide if the SU hasbeen preserved or has been eroded byestuarine (tidal?) currents (i.e., surface is anSR-U).

In carbonate rocks, a subaerial unconformity

Page 21: Embry Sequence Stratigraphy

19

which develops during an episode of base-level fall is not oftenpreserved, in part because little sediment is deposited above hightide. The shoreline ravinement which develops during the followingtransgression usually removes any thin veneer of supratidal sedimentand erodes the subaerial unconformity such that marine strata occuron both sides of the surface. Admittedly, because carbonate stratatend to be cemented very early, especially in situations of exposure,such shoreline erosion during transgression may be extremely minor.However, for consistency and clarity, I suggest use of the termunconformable shoreline ravinement rather than subaerialunconformity in situations where marine carbonate strata directlyoverlie such an unconformable surface.

In terms of utility, the unconformable portion of an SR (SR-U) is veryuseful for correlation and for bounding sequence stratigraphic unitsbecause it is a time barrier. However, the diastemic portion of an SR(SR-D) is not useful for these purposes because of its highlydiachronous nature. Like the RSME, the SR-D is correlated todelineate separate facies units within a sequence stratigraphicframework.

Figure 5.6. The time relationships of an unconformable shoreline ravinement (SR-U) which has completely eroded the penecontemporaneous fluvial-brackish strataas well as the subaerial unconformity. Time lines are truncated below the SR-U andonlap the SR-U. All strata below the SR-U are entirely older than all strata above it,making an unconformable shoreline ravinement a time barrier.

Figure 5.7. In this outcrop of Triassic strata from northern Ellesmere Island, an unconformable shorelineravinement (SR-U) occurs at the base of a thin, marine shelf limestone unit (3m) which overlies marinesiltstones of mid-shelf origin. The key characteristics of an SR illustrated here are the sharp, scoured contactand the deepening-upward, marine succession directly overlying the SR. Importantly, regional correlationsindicate that about 400 metres of strata are missing beneath the SR-U at this locality.

Figure 5.8. The subsurface succession of Jurassic marine strata from the MelvilleIsland area illustrated here contains a major unconformity as determined fromseismic data and regional correlations. The unconformity is an unconformableshoreline ravinement (SR-U) which occurs at the base of a thin, fining-upwardmarine sandstone interval. The key characteristics which lead to this interpretationinclude, a sharp contact, marine strata which fine and deepen-upward directlyoverlying the surface, marine strata below the surface, and regional truncationbelow the surface.

This distinctive surface has been given a variety of names includingravinement surface (Swift, 1975), transgressive ravinement surface(Galloway and Sylvia, 2002), transgressive surface (Van Wagoner etal., 1988), transgressive surface of erosion (Posamentier and Allen,1999), and shoreface ravinement (Embry, 2002). I prefer to use theterm shoreline ravinement for this very distinctive surface with theproviso that modifiers such as tidal and wave can be added to it. Iwould emphasize it is important to add the modifier diastemic orunconformable to any stretch of shoreline ravinement surface todifferentiate between the two very different relationships to time(highly diachronous or time barrier) that exist for a given shorelineravinement (Figure 5.2).

Maximum Regressive Surface (MRS)Maximum Regressive Surface (MRS)Maximum Regressive Surface (MRS)Maximum Regressive Surface (MRS)Maximum Regressive Surface (MRS)The maximum regressive surface has been recognized from empiricaldata for as long as fining/coarsening and deepening/ shallowing cycles(“transgressive-regressive or T-R cycles”) have been recorded in

the stratigraphic record (at least 150 years).The main characteristic for identification ofan MRS in marine clastic strata is it is aconformable horizon or diastemic surfacewhich marks a change in trend fromcoarsening-upward to fining-upward. TheMRS is never an unconformity. Over most ofits extent, the MRS also coincides with achange from shal lowing-upwards todeepening-upward and this criterion is veryhelpful, especially in shallow water facies(Figure 5.10, page 20). In deeper water, highsubsidence areas, the change from ofshallowing to deepening may not coincidewith the MRS as defined by grain size criteria(Vecsei and Duringer, 2003)

In nonmarine siliciclastic strata, the changefrom coarsening to fining is also applicablefor objectively identi fy ing an MRS. Incarbonate strata the change from shallowingupward to deepening upward is usually themost reliable and readily applicable criterionfor identifying an MRS. The change in trendfrom coarsening to fining also is applicablefor carbonates but sometimes can bemisleading.

Page 22: Embry Sequence Stratigraphy

20

For larger magnitude MRSs which separatesuccessions that contain smaller scale,sequence stratigraphic units, such coarseningand fining trends are sometimes recordedby stacking patterns of the smaller scale units(Van Wagoner et al., 1990). For example, in astacking pattern which represents acoarsening trend, each small scale unitcontains a greater proportion of coarsermaterial than the underlying one. Thus anMRS separates an overall coarsening-upwardstacking pattern (often referred to asprogradational) from a finingupward stackingpattern (often referred to asretrogradational).

The recognition of an MRS depends on theavailability of data which reflects the grainsize of the sediment (with or without smallscale units) and from which general waterdepths of the deposits can be interpretedfrom facies analysis. The MRS may occurwithin a gradational interval of facies change(conformable horizon) or it can be ratherabrupt with a scour surface marking it(diastem). On a gamma log of siliciclasticsediments, the MRS in marine strata often,but certainly not always, marks the inflectionpoint from decreasing gamma ray (gradualshift to the left) (coarsening-upward anddecreasing clay) to increasing gamma ray (ashi ft to the right) ( f ining-upward andincreasing clay) (Figure 5.11). In purecarbonate systems, gamma logs are of no helpfor MRS identification and facies data fromcore and/or cuttings are required.

It must be noted that the MRS is laterallyequivalent to the shoreline ravinement(Figure 5.12) and this relationship resultsfrom the fact that both surfaces begin to begenerated at the start of transgression (seebelow). Furthermore it may be difficult todistinguish an MRS from an unconformableshoreline ravinement (SR-U) because bothcan separate coarsening-upward marinestrata below from fining upward marine strataabove and both can be scoured contacts. Forexample an MRS might have been interpretedin the succession illustrated on Figure 8 atthe top of the thin, transgressive sandstonewhich overlies the SR-U.

The key criterion for distinguishing thesetwo different surfaces is that an SR-U is anunconformity with truncation below andonlap above whereas the MRS is either aconformity or diastem which is notassociated with truncation or onlap. Thusregional data in the form of cross-sectionsand/or seismic data are usually requiredwhen uncertainty exists.

Given the coincidence of the start of finingand the start of deepening in shallow faciesat the MRS, it is reasonable to interpret that

Figure 5.9. This outcrop of Carboniferous and Late Triassic strata on northeastern Ellesmere Island containstwo major unconformities, both of which are unconformable shoreline ravinements (SR-U). The lower oneplaces Norian strata on tilted Carboniferous strata (time gap of about 90 MA). A thin unit of shallow marinesandstone which fines and deepens upward directly overlies the unconformity leaving no doubt that it is anSR-U rather than an SU. The upper unconformity is at the base of Rhaetian strata (Late Triassic) and, on theright side, a marine sandstone can be seen onlapping the SR-U towards the left. egional correlations indicatesubstantial truncation of Norian strata beneath the unconformable shoreline ravinement (SR-U) at the baseof the Rhaetian.

Figure 5.10. In the outcropping succession of Early and Middle Triassic strata on northern Ellesmere Island,a maximum regressive surface (MRS) has been delineated near the top of a white-weathering, shorefacesandstone unit. Beneath the MRS, the strata coarsen- and shallow-upward. On top of the MRS, the strata fine-and deepen-upward. At this locality, the MRS is a conformable horizon.

Page 23: Embry Sequence Stratigraphy

21

an MRS is generated at or close to the startof transgression. Transgression begins whenthe rate of base-level rise exceeds the rateof sediment supply at the shoreline. Finergrained sediment is then deposited at anygiven locality along an offshore transect andthe MRS is marked by the change fromcoarsening upward to fining upward.

Given that the rate of sediment supply alonga siliciclastic shoreline will substantially vary,the start of transgression occurs at differenttimes but, in most cases, transgression willbe initiated along the entire shoreline withina relatively short time interval. Furthermore,this time interval of MRS generation occursfrom the start of base-level rise (areas ofmoderate to no sediment input) to soon afterthe start of base-level rise (areas of highersediment input). Thus the MRS will besomewhat diachronous but suchdiachroneity will be minor (Figure 5.13).Empirical data from carbonate strata indicatethe same relationships of the MRS to time.Theoretically there may be exceptions tothis generality but they have not beendocumented.

This surface has been called a variety of namesincluding transgressive surface (Van Wagoneret al., 1988), conformable transgressivesurface (Embry, 1993, 1995), maximumprogradation surface (Emery and Myers,1996), and sometimes by the more generalterm, flooding surface. The more descriptiveand less ambiguous term, maximumregressive surface, which was introduced byHelland-Hansen and Gjelberg (1994), isrecommended when referring to thissurface.

The low diachroneity of the MRS as well asits ready identification in outcrop, on logs(siliciclastic), and on seismic sections makethe MRS a very useful surface for correlationand contributing to a regional, quasi-timeframework as well as for bounding sequence-stratigraphic units.

ReferencesReferencesReferencesReferencesReferencesBruun, P. 1962. Sea-level rise as a cause of shoreerosion. American Society of Civil EngineersProceedings, Journal of the Waterways andHarbors Division, v. 88, p. 117-130.

Dalrymple, R., Zaitlin, B., and Boyd, R. (eds.). 1994.Incised valley Systems: Origin and sedimentarysequences. SEPM, Special Publication 51, 391 p.

Embry, A.F. 1993. Transgressive-regressive (T-R)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embry, A.F. 1995. Sequence boundaries and

Figure 5.11. Two maximum regressive surfaces (MRS) have been delineated in this subsurface succession ofJurassic strata from the Lougheed Island area. The MRSs have been placed at the change in gamma log trendfrom decreasing gamma ray to increasing gamma ray. This change in gamma ray trend is terpreted toreflect a change from shallowing-upward (decreasing clay content) to deepening-upward (increasing claycontent).

Figure 5.12. A schematic diagram showing the interpreted relationship between an MRS and other surfacesof sequence stratigraphy. The landward termination of the maximum regressive surface (MRS) adjoins thebasinward termination of the shoreline ravinement (SR). This relationship occurs because both surfaces eginto form at the start of transgression and is an important one in terms of selecting boundaries for equencestratigraphic units.

Figure 5.13. The relationship of the maximum regressive surface (MRS) to time. The MRS will approximate atime surface perpendicular to the shoreline but will exhibit minor diachroneity along strike due to varying ratesof sediment input along the shoreline.

Page 24: Embry Sequence Stratigraphy

22

sequence hierarchies: problems and proposals,In: Steel, R.J., Felt, F.L., Johannessen, E.P., andMathieu, C. (eds). Sequence stratigraphy on thenorthwest European margin: NPF SpecialPublication 5, p. 1-11.

Embry, A.F. 2002. Transgressive-Regressive (TR)Sequence Stratigraphy, in Armentrout, J. and Rosen,N. (eds.). Sequence stratigraphic models forexploration and production: Gulf Coast SEPMConference Proceedings, Houston, p. 151-172.

Helland-Hansen, W. and Gjelberg, J. 1994.Conceptual basis and variability in sequencestratigraphy: a different perspective. SedimentaryGeology, v. 92, p. 1-52.

Helland-Hansen, W. and Gjelberg, J. 1994.Conceptual basis and variability in sequencestratigraphy: a different perspective. SedimentaryGeology, v. 92, p. 1-52.

Emery, D. and Myers, K. 1997. SequenceStratigraphy. Blackwell, London, 297 p.

Galloway, W.E. and Sylvia, D.A. 2002. The manyfaces of erosion: theory meets data in sequencestratigraphic analysis, in Armentrout, J. and Rosen,N. (eds.). Sequence stratigraphic models forexploration and production: Gulf Coast SEPMConference Proceedings, Houston, p. 99-111.

Nummedal, D., Riley, G., and Templet, P. 1993.High resolut ion sequence architecture: achronostratigraphic model based on equilibriumprofile studies. In: Posamentier, H., Summerhayes,C ., Haq, B ., and Allen, G. (eds .). Sequencestratigraphy and facies association. InternationalAssociation of Sedimentologists, Special Publication18, p. 55 – 68.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. SEPM Concepts in Sedimentologyand Paleontology, no. 7, 210 p.

Sloss, L. 1963. Sequences in the cratonic interiorof North America. GSA Bulletin, v. 74, p. 93-113.

Stamp, L. 1921. On cycles of sedimentation in theEocene strata of the Anglo-Franco-Belgian Basin.Geological Magazine, v. 58, p. 108-114.

Suter, J., Berryhill, H., and Penland, S. 1987. LateQuaternary sea level fluctuations and depositionalsequences, southwest Louisiana continental shelf.In: Nummedal, D., Pilkey, O., and Howard, J. (eds.).Sea-level changes and coastal evolution: SEPMSpecial Publication 41, p. 199–122.

Swift, D. 1975. Barrier island genesis: evidencefrom the central Atlantic shelf, eastern USA.Sedimentary Geology. 14, p. 1-43.

Van Wagoner, J.C., Posamentier, H.W., Mitchum,R.M., Vail, P.R., Sarg, J.F., Loutit, T.S., and Hardenbol,

J. 1988. An overview of the fundamentals ofsequence stratigraphy and key definitions, In:Wilgus , C . , Hast ings , B .S . , Kendal l , C .G . ,Posamentier, H.W., Ross, C.A., and Van Wagoner,J.C . (eds.). Sea level changes: an integratedapproach: SEPM Special Publication 42, p. 39-46.

Van Wagoner, J.C., Mitchum, R.M., Campion, K.M.,and Rahmanian, V.D. 1990. Siliciclastic sequencestratigraphy in well logs, cores and outcrops: AAPGMethods in Exploration, no. 7, 55 p.

Vecsei, A. and Duringer, P. 2003. Sequencestratigraphy of Middle Triassic carbonates andterrigenous deposits (Muschelkalk and lowerKeuper) in the SW Germanic Basin: maximumflooding versus maximum depth in intracratonicbasins. Sedimentary Geology, v. 160, p. 81-105.

Zaitlin, B., Dalrymple, R., and Boyd, R. 1994. Thestratigraphic organization of incised-valleyssystems associated with relative sea level change.In Dalrymple, R. and Zaitlin, B. (eds.). Incisedvalley systems: origin and sedimentary sequences:SEPM special publication 51, p. 45-60.

Page 25: Embry Sequence Stratigraphy

23

Practical Sequence Stratigraphy VIPractical Sequence Stratigraphy VIPractical Sequence Stratigraphy VIPractical Sequence Stratigraphy VIPractical Sequence Stratigraphy VIThe Material-based Surfaces of Sequence Stratigraphy,Part 3: Maximum Flooding Surface and Slope OnlapSurface by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionFour material-based surfaces of sequencestratigraphy – subaerial unconformity,regressive surface of marine erosion,shorel ine ravinement, and maximumregressive surface, were described in theprevious two articles in this series. In thisinstallment, the final two materialbasedsurfaces – maximum flooding surface andslope onlap surface – are described anddiscussed. Like the other material-basedsurfaces, each of these surfaces has a uniquecombination of physical characteristics whichallow it to be defined and delineated in avariety of stratigraphic settings and withvarious types of data.

The origin of these surfaces, like thosepreviously described, can be explained bythe interaction of sedimentation and base-level change . And, also l ike the othersurfaces, these have substantial utility forcontribut ing to an approximate t imecorrelation framework and for acting asboundaries for speci f ic sequencestratigraphic units.

Maximum Flooding Surface (MFS)Maximum Flooding Surface (MFS)Maximum Flooding Surface (MFS)Maximum Flooding Surface (MFS)Maximum Flooding Surface (MFS)The maximum flooding surface has beenrecognized on the basis of empirical data forover a century, although the specific namemaximum flooding surface has been appliedto it for only the past 20 years. Its value forcorrelating well log sections was recognizedby the 1950s and many so-called “markers”on published cross-sections would be nowdesignated as maximum flooding surfaces(e.g., Forgotson, 1957; Oliver and Cowper,1963). Frazier (1974) called such a surface a“hiatal surface” and Vail et al. (1977) calledthe seismic reflector which encompassedthis surface a downlap surface.

In marine siliciclastic strata, the MFS marksthe change in trend from a fining upwardtrend below to a coarsening upward trendabove (Embry, 2001) (Figure 6.1) . Innearshore areas, this change in trendcoincides with a change from deepening toshal lowing. Farther of fshore, thisrelationship does not hold and the deepestwater horizon sometimes can lie above theMFS. In terms of stacking pattern, the MFS is

underlain by a retrogradational patternwhich displays an overall fining upward andis overlain by a prograding one which recordsan overall coarsening upward (see VanWagoner et al., 1990).

In nonmarine, s i l ic ic last ic strata , theexpression of the MFS can be more subtle,but once again the surface is best placed atthe change in trend from a fining upward to acoarsening upward. In general, such aplacement coincides with the change fromdecreasing fluvial channel content to one ofincreasing channel content (Cross andLessenger, 1998). The MFS in nonmarine stratais sometimes associated with an absence ofclastic material, which can coincide with aprominent coal bed (Hamilton and Tadros,1994; Allen et al., 1996) or even a nonmarineto brackish water limestone.

In carbonate strata, the MFS also marks achange in trend from fining to coarseningNotably, in a shallow-water carbonate-bank

setting, the MFS will mark the horizon ofchange between deepening upward toshallowing upward and this criterion, whichemploys facies analysis, can often be morereliable than grain-size variation for itsdelineation in such a setting. In deeper water,carbonate ramp settings, the MFS marks achange from decreasing and / or finercarbonate material to increasing and / orcoarser carbonate material. In platformsettings the MFS is most easily identified onthe basis of the change from deepening toshallowing whereas on the adjacent slopeand basin areas the grain-size criterion ismore reliable.

Similar to identifying an MRS, the recognitionof an MFS usually requires the availability ofdata which reflect the grain size of thesediment and from which general waterdepths of the deposits can be interpretedfrom facies analysis. On the basin flanks, thesurface is either a minor scour surface(diastem) or conformity. In offshore areas it

Figure 6.1. A surface section of Lower Triassic strata along the northeastern coast of Ellesmere Island, about10 km north of the entrance to Hare Fiord. A maximum regressive surface (MRS) has been delineated highin a succession of shelf sandstones that coarsen and shallow upwards. The strata above the MRS fine anddeepen upward to a thin, fossil-rich, limestone bed, the top of which is delineated as a maximum floodingsurface (MFS). Above the MFS, the strata coarsen upwards as shown by the increasingly lighter colour of thesection.

Page 26: Embry Sequence Stratigraphy

24

can be an unconformity that developed mainlydue to starvation and minor scouring in bothcarbonate and clastic regimes. Notably suchan unconformity usually is not associatedwith any demonstrable truncation of stratabut rather marks a major loss of time asevidenced by paleontological data. In offshoreareas, the MFS often occurs within condensedstrata which contain numerous diastems and,in siliciclastics, may be associated with achemical deposit such as a limestone orironstone (Figure 6.2).

On a gamma log of siliciclastic sediments,the MFS is best placed, in the absence ofmore precise data (e.g., core), at the inflectionpoint from increasing gamma ray (gradual shiftto the right indicating fining-upward andincreasing clay) to decreasing gamma ray (ashift to the left indicating coarseningupwardand decreasing clay) (Figure 6.3). Where theMFS is represented by a chemical depositsuch as an ironstone or limestone bed orconcentrat ion of g lauconite, the log

expression of such lithologies can be variable(Loutit et al., 1988). In pure carbonate strata,it is not possible to use log response torecognize an MFS, and facies data from coreare mandatory. On seismic data the MFS isrepresented by a reflector often referred toas a “downlap surface.” On cross-sections,higher order MFSs often appear to downlaponto a lower order MFS (e.g., Plint et al.,2001).

Given the physical characteristics of the MFS,it has been interpreted to be generated at agiven locality mainly by a change fromdecreasing sediment supply to increasingsediment supply at that locality. Such a changein supply rate is most often associated withthe change from transgression to regression.Regression begins when the rate of sedimentsupply starts to exceed the rate of base-level rise at the shoreline and the shorelinesubsequently moves seaward. Coarsergrained sediment is then deposited at anygiven locality along an offshore transect and

Figure 6.2. In this outcrop of Middle Jurassic strata from central Axel HeibergIsland, a maximum flooding surface (MFS) is placed at the top of an ironstone bed.Note that ironstone content increases upward in the shale below the ironstone bedand that argillaceous ironstone overlies the MFS. The MFS is drawn at the horizonwith least clay influx.

Figure 6.3. Two maximum flooding surfaces (MFS) have been delineated inthis subsurface succession of Jurassic strata from the Lougheed Island area.The MFSs have been placed at the change in gamma log trend from increas-ing gamma ray to decreasing gamma ray. This change in gamma ray trendis interpreted to reflect a change from fining and deepening-upward (increas-ing clay content) to coarsening and shallowing-upward (decreasing clay con-tent).

Figure 6.4. A schematic diagram showing the interpreted relationship betweena maximum flooding surface (MFS) and other surfaces of sequence stratigra-phy. The MFS overlies the SU/SR-U/MRS surfaces and, as shown, representsthe change in trend from fining to coarsening. The surface develops close to thetime of onset of regression when the shoreline begins to move seaward andcoarser sediment arrives at a given locality on the shelf. In distal areas, theMFS can be an unconformity due to starvation and episodic scouring and itis downlapped by prograding sediment.

the MFS is marked by the change fromfiningupward to coarsening-upward (Figure6.4). Thus, the MFS is interpreted to begenerated very near the time of start ofregression.

On a regional scale, the start of regressionwill occur at slightly different times alongthe shoreline, and the MFS is generated laterin areas of lower sediment supply (Figure6.5). For example, the MFS of the last inter-glacial has already formed in high-input areasof the Gulf of Mexico but has yet to begenerated in low-sediment-input areas awayfrom the major rivers (Boyd et al., 1989). Inmost situations, an MFS is a low diachroneitysurface with maximum diachroneity beingparallel to depositional strike. Where theMFS is an unconformity, it is an approximatetime barrier.

This surface has been called a hiatal surface(Frazier, 1974), a downlap surface (Vail et al.,1977; Van Wagoner et al., 1988), maximum

Page 27: Embry Sequence Stratigraphy

25

Figure 6.5. The relationship of the maximum flooding surface (MFS) to time. TheMFS will approximate a time surface perpendicular to the shoreline but willexhibit minor diachroneity along strike due to varying rates of sediment inputalong the shoreline. It will develop earlier (i.e., be older) in areas of higher inputwhere regression begins earlier.

Figure 6.6. In this outcrop of an upper Devonian patch reef on northeast Banks Island, shelfal strata occur to the left of the reef and a basin occurs to the right. Aprominent slope onlap surface (SOS) occurs asinward of the reef and it is onlapped by prograding siliciclastics. See Embry and Klovan (1971) for a description of thegeology of this outcrop.

transgressive surface (Helland- Hansen andGjelberg, 1994) and a final transgressivesurface (Nummedal et al., 1993). I recommendthe name maximum flooding surface, whichis by far the most commonly used name, forthis surface.

The low diachroneity and occasional timebarrier property of the MFS make itpotentia l ly a very useful surface forcorrelation and building an approximate timeframework as well as for acting as a boundaryfor specific sequence stratigraphic units. Itsusefulness is greatly enhanced by the fact itcan usually be reliably identified in outcrop,well sections, and on seismic data.

Slope Onlap Surface (SOS)Slope Onlap Surface (SOS)Slope Onlap Surface (SOS)Slope Onlap Surface (SOS)Slope Onlap Surface (SOS)The slope onlap surface is a surface whichhas been recorded in the geological literaturefor a long time but which was not given aspecific name until Embry (1995) referred to

Figure 6.7. A schematic diagram showing the interpreted relationship between a slope onlap surface (SOS)and other surfaces of sequence stratigraphy for a carbonate shelf / slope / basin setting. The SOS developswhen the shelf (carbonate factory) is exposed and the slope is starved of sediment. Minor sediment inputonlaps the basal portion of the slope as a wedge of detached slope deposits. The upper slope is onlapped byshelf-derived sediment deposited during the early phase of transgression.

it as a slope onlap surface (SOS). Embry (2001) included the SOS asone of the six surfaces of sequence stratigraphy. It is a prominent,unconformable surface which is developed in slope environmentsand is characterized, above all else, by the onlap of strata onto thesurface. The strata below the SOS can be either concordant withthe SOS without any evidence of scour or erosion or can be clearlyscoured and / or truncated. In cases where the SOS is not scoured,the surface is one of starvation onto which younger beds onlap.Where there is scour and loss of section below the SOS, the surfaceis formed in part by erosion (gravity collapse, current scour)followed by onlap.

The SOS is best expressed in carbonate strata in a shelf / slope /basin physiographic setting (Figures 6.6, 6.7) and is often readilyseen in outcrop (Figure 6) and on seismic sections (Schlager, 2005).The SOS forms when carbonate production is greatly reduced dueto exposure of the platform (carbonate factory) during base-level

Page 28: Embry Sequence Stratigraphy

26

fall. When this occurs, most of the slopeis starved of sediment. Erosion by margincol lapse or by currents can createprominent scarps on the upper slope andfeed very coarse sediment down dipwhere it onlaps the basal portions of theslope. During base-level fall, the slopecan be onlapped by progradingsiliciclastics as illustrated in Figure 6 orcan remain relatively starved, receivingoccasional coarse carbonate sediment.During the following base-level rise, theplatform is transgressed, widespreadcarbonate production resumes, and theremainder of the SOS is onlapped byplatform-derived, carbonate sediment.Thus the SOS is usually onlapped bysediments deposited during bothbaselevel fall and subsequent base-level

rise and transgression. This results in themaximum regress ive surface (MRS)occurring within the onlapping slopesediments (Figure 6.7, page 25).

In shel f / s lope / basin sett ings forsiliciclastics, a slope onlap surface alsoforms when sea level reaches the shelfedge. At this time, sediment flux to theslope changes from being widelydistributed before sea level reaches theshelf edge to being areally restricted andconcentrated down submarine channelswhich develop in front of input centres.Such a concentration of sediment flowresults in much of the slope being starvedof sediment. Once again this starvedslope can remain intact or can be erodedby currents or submarine landslides. The

slope is eventually onlapped by laterallyexpanding fan deposits (Figure 6.8)followed by transgressive sedimentsdeposited during the subsequent base-level rise (Figure 6.9).

In some cases, when falling sea level doesnot reach the shelf edge, a slope onlapsurface can develop at the start oftransgression when sediment supply tothe slope is substantially reduced due tomore accommodation space for sedimentbeing available on the shelf and coastalplain. Early in transgression, the waterdepth of the shelf is shallow enough toallow most sediment to be swept off theshelf as part of the ravinement process.The shelfderived sediment onlaps theslope, forming an onlapping,transgressive wedge, which has beencalled the “healing phase wedge” byPosamentier and Allen (1993). Theseauthors provide a thorough explanationfor the formation of an SOS in such asetting. In this case the SOS is onlappedonly by transgressive sediment andusually shows no evidence of lost sectionbelow it.

Notably, an SOS in siliciclastic sedimentsis usually very difficult to recognize inoutcrop because of the dif f iculty inestablishing onlapping relationships inslope lithologies. However seismic dataoften images the SOS in siliciclastics(Figure 6.9) and good examples areprovided by Greenlee and Moore (1988)and Posamentier and Allen (1999, Figures4.92, 4.93, 4.94). An SOS can also bedelineated on detailed log cross-sections(e.g. , Posamentier and Chamberlain,1993).

A slope onlap surface is an unconformityand is a time barrier. All strata belowthe surface are older than all strataabove. In cases where there has been noremoval of strata below the SOS, theSOS can be interpreted as representinga preserved depositional slope which waspresent at the time of the initiation ofthe SOS. However, in many cases, stratabelow an SOS are truncated by the SOSwith current erosion and / or gravitycollapse having removed part of thestratigraphic record. The time span ofthe onlapping strata can be highlyvariable and often ranges from part wayinto base-level fall (regression) to theearly part of base- level r ise(transgression). In some cases, onlytransgressive strata onlap the surface.

Curiously, this distinctive surface has notbeen given any specific names despite

Figure 6.8. This schematic diagram, modified from Figure 3 of Posamentier and Vail (1988), illustrates theformation and characteristics of a slope onlap surface (SOS) in siliciclastic strata. When the shelf edge isexposed, sediment is funneled down submarine channels and a submarine fan is deposited on the basinplain. Most of the slope becomes starved of sediment at this time. Over time, the fan sediments onlap the slopeas the locus of fan sedimentation shifts and deposits build up. The upper part of the SOS is usually onlappedby sediments deposited during the subsequent transgression.

Figure 6.9. A seismically delineated slope onlap surface (SOS) with onlapping strata clearly apparent. Suchstrata appear to be deposited during transgression and would equate to the “healing phase wedge” ofPosamentier and Allen (1993). The SOS adjoins to an interpreted unconformable shoreline ravinement (SR-U) which truncates underlying deltaic strata. Seismic line from Quaternary succession of the Gulf of Mexico,Desoto Canyon area (modified from Posamentier, 2003).

Page 29: Embry Sequence Stratigraphy

27

i ts widespread recognit ion in bothcarbonate and siliciclastic shelf / slope /basin settings. Given the importance ofsuch a surface for correlat ion,establ ishing a chronostrat igraphicframework, and bounding sequencestratigraphic units, a name is clearlyrequired i f only for adequatecommunication purposes. Galloway andSylvia (2002) called slope surfaces onwhich there was significant erosion slopeentrenchment surfaces but such a namedoes not include the commonoccurrence of slope onlap unconformitieswhere there has no been no loss ofsection below the unconformity (only ontop of it). I named this surface a slopeonlap surface (Embry, 1995) and I wouldrecommend the use of this name, whichis descriptive and captures the mainfeatures of the surface. The time barrieraspect of the surface makes the SOS animportant surface for correlat ion,chronostratigraphic analysis, and forpotent ia l ly bounding sequencestratigraphic units.

This article concludes the description ofthe s ix , materia l -based surfaces ofsequence strat igraphy. As wi l l bedescribed in subsequent articles, thesesurfaces are the “workhorses” ofsequence stratigraphy and are very usefulfor bui ld ing an approximate t imecorrelation framework and for boundingmaterial-based sequence stratigraphicunits. Before describing such units andi l lustrat ing the appl icat ion of thesesurfaces for correlation, it is necessaryto discuss two time-based surfaces whichsome workers advocate as equivalentsof the six material-based surfaces. Thesehave been named the “basal surface offorced regression” and the “correlativeconformity” and they will be discussedin next month’s article.

ReferencesReferencesReferencesReferencesReferencesAllen, G., Lang, S., Musakti, O., and Chirinos,A. 1996. Application of sequence stratigraphyin continental successions: implications orMesozoic craton ic bas ins of easternAustralia. Geological Society of Australia,Mesozoic Geology of the Eastern AustralianPlate. Brisbane, September, 1996. p. 22 -27.

Boyd, R, Suter, J., and Penland, S . 1989.Sequence Stratigraphy of the MississippiDelta. Gulf Coast Association of GeologicalSocieties Transactions, v. 39, p. 331-340.

Cross, T.A. and Lessenger, M. 1998. Sedimentvolume portioning: rationale for stratigraphicmodel eva luat ion and h igh reso lut ion

stratigraphic correlation. In: Predictive highresolution sequence stratigraphy. K. Sandvik,F. Gradstein, and N.Milton (eds.). NorwegianPetroleum Society Special Publication 8,p.171- 195.

Embry, A.F. 1995. Sequence boundaries andsequence h ierarch ies : problems andproposals, In: Sequence stratigraphy on thenorthwest European margin. R.J. Steel, F.L.Felt, E.P. Johannessen, and C. Mathieu (eds).NPF Special Publication 5, p. 1-11.

Embr y, A . 2001. The s ix surfaces ofsequence strat igraphy. AAPG HedbergConference on sequence stratigraphic andallostratigraphic principles and concepts,Dallas. Abstract volume, p. 26–27. http://www.searchanddiscovery.net/ documents/a b s t r a c t s / 2 0 0 1 h e d b e r g _ d a l l a s /embry03.pdf.

Embry, A. and Klovan, E. 1971. A LateDevonian reef tract on northeastern BanksIsland, NWT. Bulletin of Canadian PetroleumGeology, v. 19, p. 730-781.

Forgotson, J. 1957. Nature, Usage, and Defin i t ion of Marker-Def ined Ver t ica l l ySegregated Rock Units. Geological Notes.AAPG Bulletin, v. 41, p. 2108-2113.

Frazier, D. 1974. Depositional episodes: theirrelationship to the Quaternary stratigraphicframework in the northwestern portion ofthe Gulf Basin. Bureau of Economic Geology,University of Texas, Geological Circular 74-1, 26 p.

Helland-Hansen, W. and Gjelberg, J. 1994.Conceptual basis and variability in sequencestrat igraphy : a d i f ferent perspect ive .Sedimentary Geology, v. 92, p. 1-52.

Galloway, W.E. and Sylvia, D.A. 2002. Themany faces of erosion: theory meets datain sequence strat igraphic analys is . In :Sequence strat igraphic models forexploration and production. J. Armentroutand N. Rosen (eds.). Gulf Coast SEPMConference Proceedings, Houston, p. 99-111.

Greenlee, S. and Moore, T. 1988. Recognitionand interpretation of depositional sequencesand calculation of sea-level changes fromstratigraphic data – offshore New Jerseyand Alabama. In: Sea level changes: anintegrated approach. C. Wilgus, B.S. Hastings,C.G. Kendall, H.W. Posamentier, C.A. Ross,and J.C. Van Wagoner (eds.). SEPM SpecialPublication 42, p. 329-353.

Hamilton, D., and Tadros, N. 1994. Utilityof coal seams as genetic strat igraphic

sequence boundaries in nonmarine basins:an example from the Gunnedah Basin. AAPGBulletin, v. 78, p. 267-286.

Loutit, T., Hardenbol, P., Vail, P., and Baum,G. 1988. Condensed sections: the key toage dating and correlation of continentalmargin sequences. In: Sea level changes: anintegrated approach. C. Wilgus, B.S. Hastings,C.G. Kendall, H.W. Posamentier, C.A. Ross,and J.C. Van Wagoner (eds.). SEPM SpecialPublication 42, p. 183-216.

Nummedal, D., Riley, G., and Templet, P.1993. High resolution sequence architecture:a chronostrat igraphic model based onequilibrium profile studies. In: Sequencestratigraphy and facies association. H.Posamentier, C. Summerhayes, B. Haq, andG. Allen (eds.). International Association ofSedimentologists, Special Publication 18, p.55-68.

Oliver, T. and Cowper, N. 1963. Depositionalenvironments of the Ireton formation,Centra l A lber ta . Bu l let in of CanadianPetroleum Geology, v. 11, p. 183-202.

Plint, G., McCarthy, P., and Faccini, U. 2001.Nonmarine sequence stratigraphy: updipexpression of sequence boundaries andsystems tracts in a h igh reso lut ionframework, Cenomanian DunveganFormation, Alberta foreland basin, Canada.AAPG Bulletin, v. 85, p. 1967-2001.

Posamentier, H. 2003. A linked shelf-edgedelta and slope-channel turbidite system: 3dseismic case study from the eastern Gulf ofMexico. In: Shelf margin deltas and linkeddown slope petroleum systems. H. Roberts,N. Rosen, R. Fillon, and J. Anderson (eds.).Proceedings of the 23rd GCSSEMconference, p. 115-134.

Posamentier, H. and Vail, P. 1988. Eustaticcontrols on clastic deposition II – sequenceand systems tract models, In: Sea levelchanges: an integrated approach. C. Wilgus,B .S . Hast ings , C .G . Kendal l , H.W.Posament ier, C .A . Ross , and J .C . VanWagoner (eds.). SEPM Special Publication42, p. 125-154.

Posamentier, H. and Chamberlain, C. 1993.Sequence stratigraphic analysis of VikingFormation lowstand beach deposits atJoarcam Field, Alberta, Canada. In: Sequencestratigraphy and facies association. H.Posamentier, C. Summerhayes, B. Haq, andG. Allen (eds.). International Association ofSedimentologists, Special Publication 18, p.469-485.

Posamentier, H. and Allen, G. 1993. Variability

Page 30: Embry Sequence Stratigraphy

28

of the sequence stratigraphic model: effectsof local basin factors. Sedimentary geology,v. 86, p. 91-109.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. SEPM Concepts in Sedimentologyand Paleontology, no. 7, 210 p.

Schlager, W. 2005. Carbonate sedimentologyand sequence stratigraphy. SEPM Conceptsin Sedimentology and Paleontology 8, 200 p.

Vail, P. et al. 1977. Seismic stratigraphy andglobal changes in sea level . In: Seismicstratigraphy: applications to hydrocarbonexploration. Payton, C. (ed.). AAPG Memoir26, p. 49-212.

Van Wagoner, J .C . , Posament ier, H.W. ,Mitchum, R.M., Vail, P.R., Sarg, J.F., Loutit, T.S.,and Hardenbol, J. 1988. An overview of thefundamentals of sequence stratigraphy andkey definitions , In: Sea level changes: anintegrated approach. C. Wilgus, B.S. Hastings,C.G. Kendall, H.W. Posamentier, C .A. Ross,and J.C. Van Wagoner (eds.). SEPM SpecialPublication 42, p. 39-46.

Van Wagoner, J.C., Mitchum, R.M., Campion,K.M., and Rahmanian, V.D. 1990. Siliciclasticsequence stratigraphy in well logs, cores andoutcrops. AAPG Methods in Exploration, no.7, 55 p.

Page 31: Embry Sequence Stratigraphy

29

Practical Sequence Stratigraphy VIIPractical Sequence Stratigraphy VIIPractical Sequence Stratigraphy VIIPractical Sequence Stratigraphy VIIPractical Sequence Stratigraphy VIIThe Base-Level Change Model for Material-based,Sequence Stratigraphic Surfaces by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionAs was described in the previous threearticles in this series, six material-basedsurfaces of sequence stratigraphy, whichrepresent either breaks in sedimentation orchanges in deposit ional trend, wereempirically and separately recognized oversome 220 years. Furthermore, the origin ofeach was independently interpreted to bedue to the interaction of base-level changeand sedimentation, as was also discussed inthe past articles. For example, almost 100years ago, Barrell (1917) postulated thatsubaerial unconformities formed by a fall inbase level.

As part of the revitalization of sequencestratigraphy by Exxon scientists, Mac Jervey(1988) demonstrated that the generation ofalmost all of these surfaces (the RSME wasnot considered) can be explained by a modelwhich involves oscillating base level with aconstant sediment supply. In this article, Idiscuss the concept of base level, the factorsthat cause base level to oscillate, and thegeneration of the six sequence stratigraphicsurfaces during one cycle of base-level riseand fall. I also touch on a point of contentionin sequence stratigraphy, which involves twovariants of the base-level change model andconsequent dif ferences in geometricalrelationships between the surfaces.

Base levelBase levelBase levelBase levelBase levelHarry Wheeler (1964) succinctly reviewedthe history of the use of the term base levelin stratigraphy and subsequently Tim Cross(Cross, 1991; Cross and Lessenger, 1998)clearly demonstrated how the concept ofbase level has direct application to sequencestratigraphy. Base level, in a stratigraphicsense, is not a real, physical surface but ratheris an abstract surface that represents asurface of equilibrium between erosion anddeposition. It can be thought of as a ceilingfor sedimentation and thus, in any area whereit lies below the Earth’s surface, no sedimentaccumulation is possible and erosion willoccur. Where base level lies above the Earth’ssurface, deposition can and usually doesoccur in the space between the Earth’ssurface and base level. The places where baselevel intersects the earth’s surface areequilibrium points between areas of erosionand areas of deposition. Such points definethe edges of a depositional basin.

Figure 7.1. Base-level change refers to the relative movement between baselevel (BL), here equated to sea level, and a datum below the sea floor. Twomain factors control base-level change – movement of the datum (uplift,subsidence) and eustatic sea level change. The space between base leveland the datum is known as accommodation space (Jervey, 1988). Changesin base level thus equate to changes in accommodation space. Modifiedfrom Figure 3.6 of Coe (2003).

In a marine area, base level is usually veryclose to sea level and intersects the seabottom only where strong currents or wavesresult in net sediment removal. This resultsin net deposition for most marine settings.In a nonmarine area, base level mostcommonly lies at or below the Earth’ssurface and these areas are thus oftensubjected to active erosion by a variety ofprocesses. However, in some terrestrialareas, base level can be above the Earth’ssurface, usually in areas of standing water,and in these situations it usually closelycoincides with lake / swamp level. Riversestablish a base level profile which meansthey aggrade or erode unti l , with theestablished water and sediment supply,sediment is not deposited or eroded by therivers. In this case, base level occurs at thebase of the riverbed until a change occurs inwater energy, sediment supply, or slope ofthe channel.

Oscillations of Base LevelOscillations of Base LevelOscillations of Base LevelOscillations of Base LevelOscillations of Base LevelBase level can be seen as a surface that isassociated with the amount of energy neededto erode sediment. The erosive energyavailable at any given point may change as aresult of eustatic or tectonic activity. Thiswi l l result in either base- level fa l l ing(increased energy) or base level rising(decreased energy) at that point. Because ofthe dynamic nature ofthe Earth, base levelrarely remains static inany given location and isusually moving upwardsor downwards relativeto a datum below thesurface of the Earth. Adatum is used ratherthan the Earth’s surfaceitsel f to ensure theconcept of base-levelchange is independentof sedimentation anderosion. Thus base-levelchanges can beenvisioned as changes inthe distance betweenbase level and the datum.The space createdbetween the datum andbase level , during aspecific interval of timehas been cal led

accommodation space (Jervey, 1988) (Figure7.1). Thus, base-level changes equate tochanges in the creation or destruction ofaccommodation space.

There are two main drivers of regional base-level change (i.e., increased or decreasedenergy over a substantial part the surface ofthe Earth). The first one is tectonics thatresults in upward or downward movementsof the reference horizon (datum). In thissituation the datum, and not base level, ismoving. Downward movement of the datumis referred to as subsidence and, in a relativesense, subsidence results in rising base leveland increased accommodation space (i.e.,more space between base level and thedatum). Conversely, upward movement of thedatum (uplift) results in base-level fall as thetwo reference horizons approach eachother and accommodation space is reduced.The second driver of regional base levelmovement is eustatic sea-level change thatrecords the movements of the surface ofthe ocean in relation to the centre of theearth (Figure 7.1). In this case, the datumremains stationary and base level, which isclosely tied to sea level, moves up or down.Thus, rising eustatic sea level equates toris ing base level and increasedaccommodation space, and falling eustatic sealevel equates to fal l ing base level and

Page 32: Embry Sequence Stratigraphy

30

decreased accommodation space.Furthermore, any reduction or increase ofvolume in the sedimentary column due tosuch phenomena as compaction, salt solution,and salt intrusion also will cause changes inbase level and the amount of accommodationspace made available.

In addition to the two main drivers ofregional base-level change, it must bementioned that the erosive energy can alsochange due to local climate variations. Forexample, when water flow is increased in ariver due to a wetter climate, e.g., a monsoonseason, the erosive energy level rises,resulting in effectively a base-level fall withconsequent downcutting and erosion by theriver. Such climate-driven, base-level changesthat are independent of tectonics and eustasyare usually local (basin margin) and short-lived and will not be discussed further.

Overall, we know that tectonics and eustasyare the main controls on regional base-levelchange. However, it is often impossible todetermine the ef fect of each factorseparately (Burton et al . , 1987). Theircombined, net effect is expressed as a changein base level. The term relative sea-levelchange is sometimes used (Van Wagoner etal., 1988) for a combination of eustatic andtectonic movements but I prefer the termbase-level change because it has priority anddoes not result in any confusion in regardsto moving sea level. Use of the term base-level change also avoids often irresolvablearguments of whether tectonics or eustasyis responsible for additions and reductionsin accommodation space and theaccompanying breaks in sedimentation andchanges in depositional trends in givensituations.

As noted by Barrell (1917), base level at anygiven locality is constantly changing due tothe interaction of the above factors. Suchchange is manifest in cycles (episodes) ofbase-level rise and fall that can occur at avariety of time scales, with a variety ofmagnitudes, and on local and regional scales.In general, high magnitude base-level changes( i .e . , large fa l ls and r ises) occur lessfrequently than do smaller magnitude ones.This is an empirical observation that has greatimportance for determining a hierarchicalarrangement of sequence stratigraphic units.This will be discussed in a later article.

Given that base level is continually changing,the key point for sequence stratigraphy isthat such oscillations in base level result in anumber of distinct sedimentary surfaces thatreflect depositional breaks and / or changesof depositional trend due to the interactionbetween changing rates of the addition orreduction of accommodation space and the

rate of sedimentation. Forexample, whenaccommodation space iseliminated in a given areadue to base-level fall belowthe surface of the earth,there is a major changefrom sedimentat ion toerosion and a break insedimentation occurs. Theinit iat ion of erosionpotentially results in theoccurrence of var ioustypes of unconformitiesthat can then be used forcorrelat ion and thedelineation of sequencestrat igraphic units . Asemphasized previously, an unconformityrepresents a s igni f icant gap in thestratigraphic record and contrasts with adiastem that records only a minor gap.

Sedimentary Breaks and Changes inSedimentary Breaks and Changes inSedimentary Breaks and Changes inSedimentary Breaks and Changes inSedimentary Breaks and Changes inDepositional TrendsDepositional TrendsDepositional TrendsDepositional TrendsDepositional TrendsDuring a cycle of base-level rise and fall,

Figure 7.2. A base-level change model for the generation of the six mate-rial-based surfaces of sequence stratigraphy. Each surface is generatedduring a specific time interval of base-level change due to the interactionof rates of change in accommodation space with sedimentation rates.Modified from Figure 1 of Embry (2002).

Figure 7.3. Schematic evolution of the five material-based sequence stratigraphic surfaces associated with aramp setting. (A) By the end of base-level fall (= start base-level rise), the subaerial unconformity (SU) reachesits maximum extent and a regressive surface of marine erosion (RSME) also migrates basinward at the baseof shoreface deposits during base-level fall. (b) Transgression begins as base level begins to rise and a shorelineravinement (SR) migrates landward, eroding portions of the SU. At the same time, finer sediment is depositedat any given marine locality and a maximum regressive surface (MRS) is generated. (C) In the late phase ofbase-level rise, regression occurs and coarser sediment reaches the marine shelf. A maximum flooding surface(MFS) is generated near the time of maximum transgression and a progradational, coarsening-upwardsuccession then begins to accumulate on the shelf.

various depositional breaks and changes ofdepositional trend are generated and suchchanges are represented by the recognized,materia l -based surfaces of sequencestratigraphy which were described in detailin previous articles (Figures 7.2, 7.3, 7.4).Here I summarize the development of thesesurfaces during one base-level cycle.

Page 33: Embry Sequence Stratigraphy

31

Also at this time, less and finer clasticmaterial will reach a given locality in themarine area and the water depth will startto increase. All these changes, which occurat or very soon after the start of base-levelrise, result in two surfaces of sequencestratigraphy.

Along the shoreline, the slope of the alluvialplain is less than that of the shoreface anderosion carves out a new shoreface profileduring transgression. The erosional surfaceis known as a shoreline ravinement (SR) andit develops during the ent ire t imetransgression occurs (Figures 7.2, 7.3B,7.4B). This erosional surface almost alwayscuts down through the basinward portionof the underlying subaerial unconformity (SU)and sometimes erodes most of the SU. Thisresults in segments of a shoreline ravinementbeing either unconformable (SU eroded) ordiastemic (SU preserved) as was describedin Part V of this series (Figure 7.3B).

Also, as base level starts to rise and finersediment starts to be deposited at any givenshelf locality due to the overall reducedsupply to the marine area, there is asignificant change from a coarsening-upwardtrend that characterized the preceding base-

Figure 7.4. Schematic evolution of the six material-based sequence stratigraphic surfaces associated with ashelf / slope/basin setting. (A) During base-level fall the subaerial unconformity (SU) migrates basinward. (b)Late in fall, when the shelf is exposed (SU at shelf edge), a slope onlap surface (SOS) is generated andturbidites begin to be deposited in the basin. At the start of base-level rise, transgression begins and amaximum regressive surface (MRS) is generated in the basinal turbiditic deposits. (C) As transgressionproceeds during base-level rise, the shelf is flooded and a shoreline ravinement is cut, removing most of the SU.Finer sediments are deposited in the basin with the horizon of finest sediment marking the maximum floodingsurface (MFS). (D) As base-level rise gives way to fall, a wedge of sediment progrades basinward and anotherSU migrates basinward.

level fall to a fining-upward one. A describedearlier, the horizon that marks this significantchange in depositional trend is known as themaximum regressive surface (MRS) (Figures7.2, 7.3B, 7.4B). The MRS also marks a changefrom shallowing-upward to deepening-upward in the shallow water areas.

Eventually, the rate of base-level rise slowsand sedimentation at the shoreline once againexceeds the rate of base-level rise. Thedevelopment of the shoreface ravinementstops and the shoreline reverses directionand begins to move seaward (regression).This is associated with increasedsedimentation to the marine basin due toless storage capacity in the non-marine areaand coarser sediment begins to be depositedat any given shelf locality. This produces achange from a fining-upward trend to acoarsening-upward one and the horizon thatmarks this change in trend is a maximumflooding surface (MFS) (Figures 7.2; 7.3B, C;7.4C, D). Notably this surface wi l lapproximate the horizon of deepest waterin nearshore areas, but in areas fartheroffshore that have higher rates of base-levelrise, the horizon of deepest water will notcoincide with the MFS but will be higher inthe section.

With the start of base- level fa l l ,accommodation space begins to be reducedand sedimentation ceases on the basin margin.Subaerial erosion advances basinward duringthe entire time of fall and this produces asubaerial unconformity (SU) that reaches itsmaximum basinward extent at the end ofbase-level fall (Figures 7.2, 7.3A, 7.4A). Theseaward movement of the shorel ine(regression), which began in the waningstages of base- level r ise, continuesthroughout base-level fall but at a faster pace.

Also, when base level starts to fall, the innerpart of the marine shelf in front of thesteeper shoreface begins to be eroded asdescribed by Plint (1988). This is due to theerosion of the inner shelf as it is replaced bythe shoreface that has a higher slope. Thisinner shelf erosion surface moves seawardduring the entire interval of base-level falland is progressively covered by progradingshoreface deposits . This results in aregressive surface of marine erosion (RSME)(Figures 7.2, 7.3A). It should be noted thatan RSME often does not form due to variableenergy levels and base-level fall rates. Also,because such localized, submarine erosionresults in only a minor gap in the stratigraphicrecord at any one locality, an RSME is a highlydiachronous diastem rather than anunconformity over its extent.

Finally, when falling sea level reaches a shelf/ slope break, sedimentation patterns aresubstantially altered. In siliciclastics, sedimentis channeled down submarine canyons andmuch of the slope becomes starved. Incarbonates, the slope becomes starved atthis time because the carbonate factory ofthe shelf is shut down due to subaerialexposure. Erosion of the slope can also occurdue to current scouring and gravity failures.The starved and / or eroded slope is graduallyonlapped during the remainder of base-levelfall and / or during the early part of base-level rise and transgression and a slope onlapsurface (SOS) is thus generated (Figures 7.2,7.4B).

When base level starts to r ise, newaccommodation space begins to be createdin areas formerly undergoing erosion. Thisresults in landward expansion of the basinmargin and progressive onlap of the subaerialunconformity by nonmarine stratathroughout the entire time of base-level rise.With rising base level, less sediment istransported to the marine portion of a basinbecause of reduced fluvial gradients andincreased sediment storage in thenonmarine area along the expanding basinmargin. In most s ituat ions, a lmostimmediately after the start of base-level rise,the shoreline ceases its seaward movementand begins to shift landward (transgression).

Page 34: Embry Sequence Stratigraphy

32

In general, three surfaces are formed duringbase-level fal l : subaerial unconformity,regressive surface of marine erosion, andslope onlap surface; and three surfaces areformed during base-level rise: shorelineravinement, maximum regressive surface, andmaximum flooding surface. However, it mustbe noted that, in some instances, the SOS isnot generated until after the start of base-level rise and, in many cases, an RSME is notgenerated at all. It is also worth mentioningthat the surfaces generated during the base-level rise portion of the cycle (SR, MRS, MFS)can also be generated during cycles of varyingrates of base-level rise rather than duringrisefall cycles. They can also be generated byautogenic processes which result in markedchanges in sediment supply (e.g., delta-lobeswitching) during base-level rise (Muto etal., 2007).

Overall, a model of oscillating base-levelchange with constant sediment supply canreasonably explain the occurrence of theempirical ly recognized six surfaces ofsequence strat igraphy. Because thesesurfaces form during specific time intervalsduring a base-level cycle (Figure 7.2, page30), they have a predictable spat ia lrelationship to each other, regardless ofsediment supply. Figures 7.5, 7.6, and 7.7schematically illustrate these relationships forboth a ramp and a shelf/slope/basin setting.

As will be discussed in subsequent articles,these spatial relationships are the key forpredictions with sequence stratigraphy andfor allowing sequence stratigraphic units tobe defined by using various surfaces andcombinations of surfaces as unit boundaries.

Initial Base-level rise Models and SurfaceInitial Base-level rise Models and SurfaceInitial Base-level rise Models and SurfaceInitial Base-level rise Models and SurfaceInitial Base-level rise Models and SurfaceRelationshipsRelationshipsRelationshipsRelationshipsRelationshipsThe base level / sediment supply modeldiscussed above for the generation of thesix material-based surfaces of sequencestratigraphy is characterized by relativelyhigh rates of base-level rise occurring at orvery soon after the start of base-level rise.This variant of the general base level /sediment supply model is known as the fastinitial rise model (Figure 7.8A) and I favourit because empirical data from studies of base-level changes driven by either eustasy(Shackleton, 1987) or tectonics (Gawthorpeet al., 1994) indicate that high rates of riseoccur very soon after the initiation of rise.

In this model, the maximum regressivesurface (MRS) is generated directly after thestart of base-level rise because, as describedearlier, sediment supply to the marine shelfdecreases and becomes finer at any givenlocality as base level rapidly starts to riseand transgression begins. The shorelineravinement also starts to form and move

Figure 7.5. A schematic cross-section for a ramp setting showing the geometric relationships for the fivematerialbased surfaces of sequence stratigraphy. Note a combination of SU and SR-U forms the basin flankunconformity and the basinward termination of the SR-U adjoins the landward termination of the MRS. Thisgeometrical relationship is the result of transgression beginning at or very soon after the start of base-level rise.The MFS extends from shelf to basin.

Figure 7.6. The time relationships of the five material-based surfaces of sequence stratigraphy for a rampsetting.

landward at or soon after the start of base-level rise landward because of the commonoccurrence of very low sedimentation ratesat the shoreline or significantly reduced ratesin areas of greater supply (e.g., deltaiccentres) . Consequently, the landwardtermination of the MRS joins with thebasinward termination of the shorelineravinement (SR) (Figures 7.2, 7.3A, 7.8A).Importantly, the SR removes the basinwardportion of the subaerial unconformity (SU)as it migrates landward and it often truncates

much of the SU with remnants left at thebase of incised valleys (Figures 7.3, 7.4, 7.5,7.7). Thus, in this model, the basinwardtermination of the SU joins with a portionof the SR. The importance of this relationshipwi l l be emphasized when sequencestrat igraphic units are discussed inforthcoming articles.

In support of such a model, almost allpublished, rock-based, sequence stratigraphicstudies in both carbonates and siliciclastics,

Page 35: Embry Sequence Stratigraphy

33

Figure 7.7. A schematic cross section for a shelf / slope / basin setting showing the geometric relationships forthe six material-based surfaces of sequence stratigraphy. Once again, a combination of SU and SR-U formsthe basin flank unconformity and the basinward termination of the SR-U adjoins the landward terminationof the SOS. The MRS onlaps the SOS and the MFS extends across all the physiographic elements

demonstrate that the basinward terminationof the SU joins the SR as predicted in themodel (e.g., Suter et al, 1987; Pomar, 1991;Embry, 1993; Beauchamp and Henderson,1994; Johannessen et al., 1995; Mjos et al.,1998; Plint et al., 2001; Johannessen and Steel,2005; and many others).

The other base level/sediment supply modelwhich has been proposed (Jervey, 1988) isone of very slow, initial base-level rise (slowinitial rise model) and it is illustrated inFigure 7.8B. In this case, the SR and MRS arenot generated over much of the marine areauntil well after the start of base-level rise.The reason for this is that sedimentationrates are high enough to exceed the veryslow rise rates that occur during the earlypart of base-level rise in this model. As aresult, regression and sediment coarseningon much of the shelf, which occurred duringfall, continue during early base-level rise.

Furthermore, because the SR does not startto form until well into base-level rise, it doesnot cut down through the basinward portionof the SU as it did in the fast initial rise model.Consequently, in the slow initial rise model,there is no material-based surface whichconnects to the termination of the SU (Figure7.8B). The ramificat ions for sequenceclassification of this lack of a basinwardcorrelative surface for the SU will bediscussed in later articles.

I do not favour the slow initial rise modelbecause there are no empirical data toindicate that base-level rise rates are initiallyvery slow. In fact, the opposite appears tooccur, as discussed above. Also, empiricalstratigraphic relationships established inrock-based studies indicate that thebasinward termination of the SU almostalways joins the SR, indicating the basinwardportion of the SU has been eroded by the SR.These stratigraphic relationships also negatethe viability of the slow initial rise model.The only studies, which have been offered insupport of the model, are interpretations ofseismic data not calibrated with logs andcores (e.g., Posamentier, 2003). These canbe reasonably re-interpreted so as to becompatible with the fast initial rise model.

In my next article, I’ll elaborate on the time-based approach to sequence stratigraphy andthe two time surfaces which are advocatedfor use in sequence stratigraphy.

ReferencesReferencesReferencesReferencesReferencesBarrell, J. 1917. Rhythms and the measurementsof geologic time. GSA Bulletin, v. 28, p. 745-904.

Beauchamp, B. and Henderson, C. 1994. The LowerPermian Raanes, Great Bear Cape and TrappersCove formations. Bulletin of Canadian Petroleum

Figure 7.8. Two base-level-change models for sequence stratigraphy for a ramp setting. The same fivematerial-based surfaces are generated in both models. The key difference between the two models is theresulting relationship between the subaerial unconformity and the shoreline ravinement. (A) The fast initialrise model. As shown on the left, both a tectonically-generated and a eustatically-generated base level curveare characterized by rapid initial base-level rise. This results in the shoreline ravinement cutting out thebasinward portion of the subaerial unconformity and the surface relationships shown on the right. (B) Theslow initial rise model. As shown on the left, early base-level rise is slow on a sinusoidal change curve. In thissituation, the shoreline ravinement does not erode the basinward portion of the subaerial unconformity asshown on the right and a wedge of nonmarine sediment always separates the SU from the overlying SR andMRS. Note the basinward termination of the SU does not join with another material-based surface.

Geology, v. 42, p. 562-597.

Burton, R., Kendall, C., and Lerche, I. 1987 Out ofour depth: on the impossibility of fathoming eustasyfrom the stratigraphic record. Earth ScienceReviews, v. 24, p. 237-277.

Coe, A. (ed.). 2003. The sedimentary record of

sea-level change. Cambridge University Press,New York, 287 p.

Cross, T. 1991. High resolution stratigraphiccorrelation from the perspective of base levelcycles and sediment accommodation. In J. Dolson(ed.) . Unconformity re lated hydrocarbonexploration and accumulation in clastic and

Page 36: Embry Sequence Stratigraphy

34

carbonate settings. Short course notes, RockyMountain Association of Geologists, p. 28-41.

Cross, T.A. and Lessenger, M. 1998. Sedimentvolume portioning: rationale for stratigraphicmodel eva luat ion and h igh reso lut ionstratigraphic correlation. In K. Sandvik, F.Gradstein, and N. Milton (eds.). Predictive highresolution sequence stratigraphy. NorwegianPetroleum Society Special Publication 8, p. 171-195.

Embry, A.F. 1993 Transgressive-regressive (T-R)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embry, A. 2002. Transgressive-Regressive (T-R)Sequence Stratigraphy, In J. Armentrout, and N.Rosen (eds.). Sequence stratigraphic models forexploration and production. Gulf Coast SEPMConference Proceedings, Houston, p. 151-172.

Gawthorpe, R., Fraser, A., and Collier, R. 1994.Sequence stratigraphy in active extensionalbasins: implications for the interpretation ofancient basin fills. Marine and Petroleum Geology,v. 11, p. 642- 658.

Jervey, M. 1988. Quantitative geological modelingof siliciclastic rock sequences and their seismicexpression. In C. Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner,(eds.), Sea level changes: an integrated approach:SEPM Spec. Pub. 42, p. 47-69.

Johannessen, E., Mjos, R., Renshaw, D., Dalland, A.,and Jacobsen, T. 1995. Northern limit of the Brentdelta at the Tampen Spur – a sequencestratigraphic approach for sandstone prediction.In R. Steel, V. Felt, E. Johannessen, and C. Mathieu(eds.). Sequence Stratigraphy of the NorthwestEuropean Margin. Norwegian Petroleum SocietySpecial Publication 5, p. 213-256.

Johannessen, E.J. and Steel, R.J. 2005. Shelf-marginclinoforms and prediction of deep water sands.Basin Research, v. 17, p. 521-550.

Mjos, R., Hadler-Jacobsen, F., and Johannessen, E.1998. The distal sandstone pinchout of the MesaVerde Group, San Juan basin and its relevancefor sandstone prediction of the Brent Group,northern North Sea. In K. Sandvik, F. Gradstein,and N. Milton, (eds.). Predictive high resolutionsequence stratigraphy. Norwegian PetroleumSociety Special Publication 8, p. 263-297.

Muto, T. , Steel , R. and Swenson, J . 2007.Autostratigraphy: A Framework Norm for GeneticStratigraphy. Journal of Sedimentary Research, v.77, p. 2-12.

Plint, A. 1988. Sharp-based shoreface sequencesand offshore bars in the Cardium Formation of

Alberta: their relationship to relative changes insea level. In C. Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner,(eds.). Sea level changes: an integrated approach.SEPM Special Publication 42, p. 357-370.

Plint, G., McCarthy, P., and Faccini, U. 2001.Nonmarine sequence strat igraphy: updipexpression of sequence boundaries and systemstracts in a h igh reso lut ion f ramework,Cenomanian Dunvegan Formation, Alber taforeland basin, Canada. American Association ofPetroleum Geologists Bulletin, v. 85, p. 1967-2001.

Pomar, L. 1991. Reef geometries, erosion surfacesand high frequency sea level changes, UpperMiocene reef complex, Mal lorca, Spain .Sedimentology, v. 38, p. 243-269.

Posamentier, H. 2003. A linked shelf-edge deltaand slope-channel turbidite system: 3d seismiccase study from the eastern Gulf of Mexico. In H.Roberts, N. Rosen, R. Fillon, and J. Anderson, (eds.).Shelf margin deltas and linked down slopepetroleum systems, Proceedings of the 23rdGCSSEPM conference, p. 115-134.

Shackleton, N.J. 1987. Oxygen isotopes, ice volumeand sea level. Quaternary Science Reviews, v. 6, p.183-190.

Suter, J., Berryhill, H., and Penland, S. 1987. LateQuaternary sea level fluctuations and depositionalsequences, southwest Louisiana continental shelf.In D. Nummedal, O. Pilkey, and J. Howard, (eds.).Sealevel changes and coastal evolution. SEPMSpecial Publication 41, p. 199-122.

Van Wagoner, J.C., Posamentier, H.W., Mitchum,R.M., Vail, P.R., Sarg, J.F., Loutit, T.S., and Hardenbol,J. 1988. An overview of the fundamentals ofsequence stratigraphy and key definitions. In C.Wilgus , B .S . Hastings , C .G. Kendall , H.W.Posamentier, C.A. Ross, and J.C. Van Wagoner,(eds.). Sea level changes: an integrated approach.SEPM Special Publication 42, p. 39-46.

Page 37: Embry Sequence Stratigraphy

35

Practical Sequence Stratigraphy VIIIPractical Sequence Stratigraphy VIIIPractical Sequence Stratigraphy VIIIPractical Sequence Stratigraphy VIIIPractical Sequence Stratigraphy VIIIThe Time-based Surfaces of Sequence Stratigraphy by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionIn parts four, five, and six of this series, Idescribed the six, material-based surfacesof sequence stratigraphy, which have beenrecognized and characterized over the past200 years. Notably, each of thesematerialbased surfaces is defined on thebasis of observable physical characteristicsthat include:

• the physical properties of the surfaceand of overlying and underlying strataand

• the geometrical relationships betweenthe surface and the underlying andoverlying strata.

These surfaces can be sa id to bemodelindependent because they wereempirically recognized before a model wasproposed to explain or rationalize theirexistence. The delineation and use of suchsurfaces for correlation and for definingspeci f ic sequence strat igraphic unitsconstitutes a materialbased approach tosequence stratigraphy.

Another approach to sequence stratigraphy,which is advocated by some authors (e.g.,Hunt and Tucker, 1992; Helland-Hansen andGjelberg, 1994; Posamentier and Allen, 1999;Catuneanu, 2006; Catuneanu et al., in press),is a time-based approach. In a time-basedapproach, some of the surfaces used forsequence stratigraphic analysis are definedon the basis of time rather than observablecharacterist ics and geometricalrelationships. Such an approach is indicatedby Posamentier (2001) “Crit ical to asequence stratigraphic analysis is theidentification of time synchronous surfacesthat punctuate rock successions”.

Time-based surfaces are known aschronostratigraphic surfaces and are definedon the basis of a specified event at an exactlocation. Basically, a chronostratigraphicsurface represents a depositional surfacethat existed at the moment in time whenthe specified event took place. As stated byCatuneanu (2006) “Sequence stratigraphicsurfaces are defined relative to the fourmain events of the base-level cycle”. Suchevents are related to a change in either thedirection of base-level change (e.g., fallingbase level to rising base level) or thedirection of shoreline movement (e.g.,landward movement to seaward movement).

Figure 8.1. A sinusoidal base-level change curve illustrating the timing of the four “events” that are used todefine four, time-based surfaces of sequence stratigraphy:

Start base-level rise = correlative conformity (CC)Start transgression = maximum regressive surface (MRS)Start regression = maximum flooding surface (MFS)Start base-level fall = basal surface of forced regression (BSFR)

As shown on Figure 8.1, four base level cycleevents are defined and utilized in the time-based approach, with the fundamentalunderpinning of this approach being thehypothesis that each event is associated witha specific, sequence stratigraphic surface. Thefour events and their assigned surfaces are:

• start base-level rise (1) = correlativeconformity,

• start transgression (2) = maximumregressive surface,

• start regression (3) = maximumflooding surface, and

• start base-level fall (4) = basal surfaceof forced regression.

The time-based approach differs from thematerial-based approach in two main ways:

• a different way of defining somespecific surfaces that are common toboth approaches (e.g., maximumregressive surface) and

• the addition of two new surfaces whichhave no equivalents in the material-based approach.

These two, t ime-based surfaces wereproposed (deduced) by Hunt and Tucker

(1992) on the basis of the sequencestratigraphic model of Jervey (1988) ratherthan on empirical data. In contrast to themodel-independent, material-based surfaces,these two time-based surfaces are model-dependent (i.e., “no model – no surfaces”).They are best viewed as hypotheticalsurfaces which represent two events on thebase level curve.

Old surfaces / new definitionsOld surfaces / new definitionsOld surfaces / new definitionsOld surfaces / new definitionsOld surfaces / new definitionsTwo important, material-based, sequencestratigraphic surfaces are the maximumregressive surface (MRS) and maximumflooding surface (MFS) and these surfaceswere defined and described in previousarticles. As was noted in those articles, boththe MRS and MFS were empir ica l lyrecognized many years (under differentnames) before sequence strat igraphicmethodology and models were formulatedand they are defined and delineated solelyon the basis of their physical characteristics.As part of modern day, sequence stratigraphictheory, the MRS and MFS are interpreted tohave formed due to the interplay of base-level change and sedimentation although itmust be emphasized that such interpretationsplay no role in their definition.

Page 38: Embry Sequence Stratigraphy

36

In the time-based approach, these twosurfaces are def ined on the basis ofinterpreted changes in shoreline direction.For example, Catuneanu (2006, p. 135) states“The maximum regressive surface is definedrelative to the transgressiveregressive curve,marking the change from shorel ineregression to subsequent transgression”.Similarly, Catuneanu, 2006, p. 142) states “Themaximum flooding surface is also definedrelative to the transgressive-regressivecurve, marking the end of shorel inetransgression.”

In reality, the distinction between the twomethods – the material-based definitionsbeing dependent on observablecharacteristics and the time-based onesbeing dependent on theoretical events - doesnot have a significant effect on the final result.This is because the observable characteristicsused for the material-based definition of asurface are used as proof of the occurrenceof the given event associated with thatsurface. Thus, in most cases the same horizonis picked for a given surface by bothapproaches although, as will be discussed,this is not always the case. Regardless, it isimportant to understand the profounddifference in the manner in which surfacesare defined in the two approaches as thisdifference has a significant impact with theintroduction of two new surfaces in thetime-based approach.

Two New SurfacesTwo New SurfacesTwo New SurfacesTwo New SurfacesTwo New SurfacesTwo, time-based surfaces were introducedinto sequence stratigraphy by Hunt andTucker (1992) on the basis of two theoreticalevents – start base-level fall and start base-level rise. These surfaces had not been definedbefore the modeling work of Jervey (1988).One was named the basal surface of forcedregression (BSFR) (Hunt and Tucker, 1992)and the other the correlative conformity(CC) (Helland-Hansen and Gjelberg, 1994).Subsequent books (e.g., Posamentier andAllen, 1999; Coe, 2003; Catuneanu, 2006)have advocated for the use of theseconceptual, time-based surfaces for sequencestratigraphic unit definition and correlation.

For illustrative purposes, I have added botha BSFR and a CC to model cross-sectionswhich were constructed to show therelationships of the material-based surfacesof sequence strat igraphy. The modelsrepresent three different scenarios relatedto differences in physiography and speed ofinitial base-level rise:

• ramp setting, fast initial base-level rise(Figure 8.2),

• ramp setting, slow initial base-levelrise (Figure 8.3),

• shelf / slope / basin setting, SOS (slopeonlap surface)-generated, fast initialbase-level rise (Figure 8.4).

The relationships of the two hypotheticaltime surfaces to the six material-basedsurfaces for a shelf / slope / basin settingwith a slow initial base-level rise is essentiallythe same as that shown on Figure 8.4. It mustbe emphasized that the placement of thesetime-based surfaces on these model crosssections is based on theoretical reasoningand not empirical evidence.

Basal Surface of Forced RegressionBasal Surface of Forced RegressionBasal Surface of Forced RegressionBasal Surface of Forced RegressionBasal Surface of Forced Regression(BSFR)(BSFR)(BSFR)(BSFR)(BSFR)Hunt and Tucker (1992, p. 5) defined a BSFRas “a chronostratigraphic surface separatingolder sediments…deposited during slowingrates of relative sea level rise… fromyounger sediments deposited duringbaselevel fall”. In short, it represents a timesurface generated at the start of base-levelfall. Plint and Nummedal (2000), Catuneanu(2006), and Catuneanu et al. (in press)characterize the BSFR as the clinoform(paleo-seafloor) present at the start of offlap(equals start base-level fall at the shoreline)along a given transect perpendicular to theshoreline. From a theoretical point of view,a BSFR will be truncated updip by the SU,will be offset at the RSME and then will occursomewhere within a thick, upward-coarsening succession of shelf and slopestrata. Basinward, it will approach the

underlying MFS and may downlap onto it(Figures 8.2-8.4).

Because the BSFR is a time-based surfaceand does not correspond with anymateria lbased surface of sequencestratigraphy, the obvious question becomes– “Does such a hypothetical surface have anyobservable, characteristic features thatwould al low it to be del ineated withreasonable objectivity so as to allow it to beused for correlation and bounding sequencestratigraphic units?”

This does not appear to be the case and Ibel ieve it is basical ly impossible toconvincingly recognize “the first clinoformassociated with offlap” in almost everyconceivable geological setting. As shown onFigures 2-4, such a time surface occurs withina succession of coarsening-upward strata andno sedimentological variation or change ingrain size trend has been identified ortheorized to characterize the surface andallow its recognition in such a succession.This lack of criteria for the recognition ofsuch a surface over most of a basin has beennoted by Posamentier et al. (1992), Embry(1995), Posamentier and Allen (1999), Plintand Nummedal (2000), and Catuneanu (2006)– among others. Posamentier et al. (1993, p.1695) state “This surface becomes a crypticsurface, virtually impossible to identify,where the shoreface deposits becomegradationally based”. Posamentier and Allen

Figure 8.2. A schematic cross-section for a ramp setting with a fast initial base-level rise. Five material-basedsurfaces of sequence stratigraphy (SU, RSME, SR, MRS, and MFS) and two time-based surfaces (BSFR, CC)are illustrated on the cross-section. The time-based BSFR and CC occur within the coarsening-upwardssuccession between the material-based MFS and MRS. The BSFR is truncated updip by the SU / SR-U and theCC is truncated very near the basinward end of the SR-U. Neither of these hypothetical time surfaces ismarked by any sedimentological changes and, because they are conformities, they are not distinguishable byany geometrical relationships.

Page 39: Embry Sequence Stratigraphy

37

(1999, p. 90) state “it exists only as achronohorizon, … precise identification …can be limited”. Plint and Nummedal (2000,p. 5) note that such a time surface is “difficultor impossible to recognize in outcrops orwell logs”. Catuneanu (2006, p. 129) states“the basal surface of forced regression …has no physical expression in a conformablesuccession of shallow water deposits”. Thusit appears widely accepted that the BSFR hasno characteristic physical attributes to allowits objective recognition in well exposedsections or in core.

Authors who advocate the use of theconceptual BSFR in sequence stratigraphicclassification offer two ways of delineatingsuch a surface. One is through the use ofseismic data, and authors such as Posamentierand Allen (1999) and Catuneanu (2006)suggest a BSFR can be approximated by theseismic reflector that intersects the SU(subaerial unconformity) at the start of adownward trajectory of the SU (i.e., startofflap). In theory, this has some merit, butthe main problem with such a proposal isthat subsequent erosion on the subaerialunconformity during the entire time of base-level fal l destroys such a geometricalrelationship. Consequently it is virtuallyimpossible to identify on seismic sectionsor well log sections the “clinoform whichintersects the SU at the start of offlap” exceptin extremely rare cases.

The other strategy for delineating a BSFR isto use one of the material-based surfaces ofsequence stratigraphy or, in some cases, alithostratigraphic surface (within-trend facieschange) as a “proxy” for it. Some authorshave associated a BSFR with a regressivesurface of marine erosion (RSME) (e.g.,Posamentier et al. , 1993). However, asdescribed in part 4 of this series (Embry,2008a), the RSME is a highly diachronoussurface which forms during the entire timeof base-level fall and is almost entirelyyounger than the time-based BSFR (Plint andNummedal, 2000).

Sometimes, in offshore shelf environments,sedimentat ion on the unconformableportion of an MFS is interpreted not to havebegun until after base level starts to fall (i.e.,outer shelf initially starved after the start ofregression) and that portion of the MFS issometimes labeled as a BSFR (e.g., MacNeiland Jones, 2006, Fig. 11; Catuneanu, 2006,Fig. 4.19). However, such a surface shouldbe recognized as an MFS rather than a BSFRas a BSFR is a chronostratigraphic surfaceand thus cannot be an unconformity. Intheory, the BSFR in the above-cited caseswould have downlapped onto the MFS,although even this would be very difficult todemonstrate in a real-world situation.

Figure 8.3. A schematic cross section for a ramp setting with a slow initial base-level rise. Once again the BSFRand CC occur in the coarsening-upward succession between the MFS and MRS. In this model, the landwardend of the CC adjoins the basinward termination of the SU. Unlike the fast initial rise model, the CC and MRSare separated by a substantial thickness of sediment and the SR does not erode the basinward portion of theSU. Also in this model, a time-based MRS is hypothesized to occur within nonmarine strata. As with the time-based BSFR and CC, no concrete criteria have ever been proposed for delineating a time-based MRS innonmarine strata

Figure 8.4. A schematic cross-section for a shelf / slope / basin setting showing the geometric relationships forthe six material-based surfaces of sequence stratigraphy as well as the two time-based ones. Similar to theother models, the BSFR is truncated updip by the SU / SR-U. It occurs within the coarsening-upward successionbetween the MFS and MRS and, in some cases, will downlap onto the MFS basinward. The CC occurs in thebasinal turbidites and may coincide with or lie just below the material-based MRS. It onlaps the slope onlapsurface (SOS). The BSFR and CC placement on this and the other schematic cross sections is based solely onmodel-based deduction.

Another materia l -based, sequencestratigraphic surface which is occasionallyequated to a BSFR is the slope onlap surface(SOS) (e.g., Posamentier and Allen, 1999).Once again, such a comparison isinappropriate because an SOS alwaysdevelops after the start of base-level fall. Forsiliciclastics, this will almost always be a

significant time after the start of base-levelfal l . Furthermore, an SOS is often anunconformity (Embry, 2008b).

Two commonly used proxies for a BSFRinvolve the use of highly diachronous facieschanges at the base of turbidite strata or atthe base of shallow water carbonate or

Page 40: Embry Sequence Stratigraphy

38

clastic strata (e.g., Hunt and Tucker, 1992; Plintand Nummedal, 2000; Mellere and Steel, 2000;Coe, 2003; Catuneanu, 2006, and very manyothers). The obvious pitfall in using the baseof submarine fan deposits as an equivalentof a BSFR is that it is highly unlikely the firstgravity flow deposits will coincide, or evenbe remotely close to coinciding, with thestart of base-level fall. Turbidite depositioncan be initiated any time during fall and, inmany cases, does not occur at any time duringfall (Catuneanu, 2006). The same logic appliesto the use of the highly diachronous, basalcontact of a shallow marine deposit for aBSFR (e.g., Burchette and Wright, 1992). Sucha facies contact forms throughout the entireinterval of fall as the shallow water faciesprogrades basinward over deeper waterfacies. A serious problem of trying to equatea BSFR with inappropriate materialbasedsurfaces as discussed above is that such apract ice can result in mis leading anderroneous interpretations of depositionalhistory.

Given the above arguments, a BSFR is bestseen as a purely deductive construct (i.e.,hypothet ica l surface) which has nocharacteristic physical attributes to allow itsrecognition in well exposed strata, in core,and on almost all seismic lines. Despite theseissues, the BSFR has been proposed as botha sequence boundary (Posamentier and Allen,1999) and a systems tract boundary (Huntand Tucker, 1992; Plint and Nummedal, 2000;Catuneanu, 2006). The pract ica l i ty ofemploying a “cryptic”, time-based surface asa unit boundary wi l l be discussed inforthcoming articles that look at howsequence stratigraphic units are defined.

Correlative Conformity (CC)Correlative Conformity (CC)Correlative Conformity (CC)Correlative Conformity (CC)Correlative Conformity (CC)Hunt and Tucker (1992, p. 6) characterized acorrelat ive conformity, as “truly achronostratigraphic surface” equivalent tothe depositional surface (clinoform) at theend of base-level fall (i.e., start base-levelrise). It represents the sea floor at themoment in time when base-level fall givesway to base-level rise. Like the BSFR, a CC ismodel-dependent and had not beendescribed as a distinct surface before theJervey (1988) model for explaining the originand geometries of sequence stratigraphicsurfaces was published. Hunt and Tucker(1992) did not provide any specific criteriawhich would allow the recognition of a CCexcept in areas of submarine fan deposition.Hel land-Hansen and Gjelberg (1994),Helland-Hansen and Martinsen (1996), andCatuneanu (2006) have elaborated on thissurface and advocated for its use in sequencestratigraphic classification.

From a theoretical point of view, the CCjoins the basinward end of the subaerial

unconformity (SU) in a ramp setting for theslow initial rise model (described in part 7of this series) (F igure 8.3, page 37) .Basinward, it occurs within a coarsening-upward succession situated between the MFSbelow and the MRS above. In a ramp settingfor the fast initial rise model, the CC will betruncated at the end of the unconformableshoreline ravinement (SR-U) (Figure 8.2,page 36). In a shelf / slope / basin model,where an SOS develops, and for either slowor fast initial base-level rise, the CC willtheoretically occur in a succession of basinalturbidites and will onlap the SOS (Figure 8.4,page 37).

To my knowledge, no one has ever publishedany observable criteria for recognizing thecorrelative conformity over most of a basin.This is not surpris ing g iven that nosedimentary break or change insedimentation style or trend occurs overmuch of the marine area at the start of base-level rise, especially when base-level risesslowly at the start (Figure 8.3, page 37).

This lack of observable characteristics isrecognized by Catuneanu (2006, p. 122) whostates “The main problem relates to thedifficulty of recognizing it in most outcropsect ions, core or wirel ine logs.” AsCatuneanu (2006) explains, the correlativeconformity “develops within a conformableprograding package (coarsening upwardtrends below and above); lacking anylithofacies and grading contrasts”. The mainproblem associated with the correlativeconformity is also enunciated by Plint andNummedal (2000, p. 5) who succinctly state“From a practical point of view, this marinesurface will be difficult to impossible toidentify.”

Catuneanu (2006) and Catuneanu et al. (inpress) suggest that seismic data offer thebest opportunity to identify and correlate aCC. A CC can be approximated by abasinward seismic reflector which joins witha more landward reflector that encompassesthe SU and / or the SR-U. Catuneanu (2006)interprets such a seismic-based CC in hisFigure 4.17. As shown in Figure 8.2 (page36), the MRS and the CC will theoreticallya lmost coincide when the start oftransgression occurs very soon after startof base- level r ise and perhaps moreimportant ly, the MRS adjoins to thebasinward end of the unconformity. In thiscase the seismic ref lector whichencompasses a theoretical CC will alsoencompass a material-based MRS. Thequestion remains if a seismically recognized,time-based CC for a ramp setting is inactuality a material-based MRS. I suspect it isin most, if not all cases, but we need studiesinvolving core and seismic to resolve the

question of whether or not a CC is a realsurface which has physical properties thatcan generate a seismic reflector. The othermaterial-based surface which is sometimeslabeled as a CC on seismic is the slope onlapsurface (SOS). The reason for such aportrayal is shown in Figure 8.4 (page 37),which i l lustrates that the landwardtermination of the SOS adjoins the basinwardtermination of the basin flank unconformity(SU or SU / SR-U). Thus the same seismicreflector that encompasses the SU / SR-U onthe basin flank encompasses the SOS fartherbasinward.

Hunt and Tucker (1992) suggested that thechange from a coarsening-upward successionof turbidites to a fining-upward successionmight approximate such a boundary and thishas theoretical support (Catuneanu, 2006).However, the material-based maximumregressive surface would also be placed atsuch a horizon of change in depositional trend(coarsening trend changing to a fining trend).Notably, Catuneanu (2006) and Catuneanuet al. (in press) would not put the time-basedMRS at this horizon, but rather would placeit strat igraphical ly higher at an oftenunrecognizable (“cryptic”) horizon withinshaly turbidites. The position of this horizondepends on a specific sequence stratigraphicmodel.

This significant difference in the placementof the MRS in deep water strata highlightsthe essential difference between the twoapproaches to surface defin it ion. Thematerial-based approach uses an MRS withdefined, observable criteria whereas thetime-based approach uses a theoretical,model-dependent, indefinite horizon for theMRS.

In summary, the correlative conformity,although it has theoretical appeal, is a time-based, sequence-stratigraphic surface lackingdefining characteristics which would allowsuch a surface to be recognized withreasonable scientific objectivity (i.e., withempirical observations) in most data sets.Despite these formidable problems, the CChas been proposed as both a sequence andsystems tract boundary (Hunt and Tucker,1992; Plint and Nummedal, 2000; Catuneanu,2006). The practicality of such usage will bediscussed in future articles in this series.

With this article, all the various specific typesof sequence stratigraphic surfaces which havebeen recognized / proposed, including bothmaterial-based ones and time-based ones,have been described. Such surfaces providethe means for defining a variety of specifictypes of sequence stratigraphic units.Material-based sequence stratigraphic unitsare defined by various combinations of

Page 41: Embry Sequence Stratigraphy

39

bounding, material-based surfaces. Timebasedsequence stratigraphic units employ thetime-based surfaces discussed above, inaddition to material-based surfaces, fordefining unit boundaries. The existence ofboth material-based units and time-basedunits has been a major source of confusionfor those wanting to employ sequencestratigraphic units in their studies and tocommunicate their findings. In the nextarticle, I will describe and evaluate thepract ica l i ty of the di f ferent types ofsequences, both material-based and timebased, which have been proposed for use. Insubsequent articles, I’ll tackle systems tracts,followed by parasequences.

ReferencesReferencesReferencesReferencesReferencesBurchette, T. and Wright, V.P. 1992. Carbonate rampdepositional deposits. Sedimentary Geology, v.79,p. 3-57.

Catuneanu, O. 2006. Principles of SequenceStratigraphy. Elsevier, New York, 375 p.

Catuneanu, O. et al. In press . Towards theStandardization of Sequence Stratigraphy. EarthScience Reviews.

Coe, A. (ed.) 2003. The sedimentary record ofsea-level change. Cambridge University Press,New York, 287 p.

Cross, T. 1991. High resolution stratigraphiccorrelation from the perspective of base levelcycles and sediment accommodation. In:Unconformity related hydrocarbon explorationand accumulation in clastic and carbonatesettings. J. Dolson (ed.). Short course notes, RockyMountain Association of Geologists, p. 28-41.

Embry, A.F. 1995. Sequence boundaries andsequence hierarchies: problems and proposals.In: Sequence stratigraphy on the northwestEuropean margin. R. J. Steel, F. L. Felt, E.P.Johannessen, and C. Mathieu (eds.). NPF SpecialPublication 5, p. 1-11.

Embr y, A .F. 2008a. Pract ica l SequenceStratigraphy IV: The Material-based Surfaces ofSequence Stratigraphy, Part 1: SubaerialUnconformity and Regressive Surface of MarineErosion. Canadian Society of Petroleum Geologists,The Reservoir, v. 35, issue 8, p 37-41.

Embr y, A .F. 2008b. Pract ica l SequenceStratigraphy VI: The Material-based Surfaces ofSequence Stratigraphy, Part 3: Maximum FloodingSurface and Slope Onlap Surface. CanadianSociety of Petroleum Geologists, The Reservoir, v.35, issue 10, p 36-41.

Helland-Hansen, W. and Gjelberg, J. 1994.Conceptual basis and variability in sequencestratigraphy: a different perspective. SedimentaryGeology, v. 92, p. 1-52.

Helland-Hansen W. and Martinsen, O.J. 1996.Shoreline trajectories and sequences: descriptionof variable depositional-dip scenarios: Journal ofSedimentary Research, v. 66, p. 670-688.

Hunt , D. and Tucker, M. 1992. Strandedparasequences and the forced regressive wedgesystems tract: deposition during base-level fall.Sedimentary Geology, v. 81, p. 1-9.

Jervey, M. 1988. Quantitative geological modelingof siliciclastic rock sequences and their seismicexpression, In: Sea level changes: an integratedapproach. C. Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C.Van Wagoner(eds.). SEPM Special Publication 42, p.47-69.

MacNeil, A. and Jones , B. 2005. Sequencestratigraphy of a Late Devonian rampsituatedreef system in the Western Canadian SedimentaryBasin: Dynamic responses to sea level changeand regressive reef development. Sedimentology,v. 53, p. 321-359.

Mellere, D. and Steel, R. 2000. Style contrastbetween forced regressive and lowstand/transgressive wedges in the Campanian of north-central Wyoming (Hatfield Member of theHaystack Mountains Formation). In: Sedimentaryresponses to forced regressions. D. Hunt, and R.Gawthorpe (eds.). Geological Society of London,Special Publication 172, p. 141-162.

Plint, A. and Nummedal, D. 2000. The falling stagesystems tract: recognition and importance insequence stratigraphic analysis. In: Sedimentaryresponses to forced regressions. D. Hunt, and R.Gawthorpe (eds.). Geological Society of London,Special Publication 172, p.1-17.

Posamentier, H. 2001. Sequence stratigraphy:Balancing the theoretical and the pragmatic(abstract). Canadian Society of PetroleumGeologists, The Reservoir, v. 28, issue 11, p. 14.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. SEPM Concepts in Sedimentologyand Paleontology, no. 7, 210 p.

Posamentier, H., Allen, G., James, D., and Tesson, M.1992. Forced regress ion in a sequencestratigraphic framework: concepts, examples andexploration significance. AAPG Bulletin, v. 76, p.1687-1709.

Page 42: Embry Sequence Stratigraphy

40

Page 43: Embry Sequence Stratigraphy

41

Practical Sequence Stratigraphy IXPractical Sequence Stratigraphy IXPractical Sequence Stratigraphy IXPractical Sequence Stratigraphy IXPractical Sequence Stratigraphy IXThe Units of Sequence Stratigraphy,Part 1: Material-based Sequences by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionOver the past 50 years, three different,general types of sequence stratigraphic unitshave been introduced: sequence (Sloss etal., 1949), systems tract (Brown and Fisher,1977), and parasequence (Van Wagoner etal., 1988). Specific types of sequences andsystems tracts have also been defined. Eachspecific type of sequence stratigraphic unitis pr imari ly def ined by the sequencestratigraphic surfaces which bound it. In thisarticle and the next, I will describe theevolution of sequence boundary definition,discuss the two specific sequence typeswhich have become popular, and illustratethe various types of materia l-based,sequence boundaries which have beenintroduced into the literature over the past20 years. The next article will look at time-based sequence boundaries and summarizeal l the di f ferent types of sequenceboundaries which have been proposed. Thefollowing two articles will look at systemstracts and parasequences, respectively.

Evolving Definition of a SequenceEvolving Definition of a SequenceEvolving Definition of a SequenceEvolving Definition of a SequenceEvolving Definition of a SequenceBoundaryBoundaryBoundaryBoundaryBoundaryIn the beginning – As was described in myearlier articles that dealt with the historicaldevelopment of sequence stratigraphy(Embry, 2008a, b), a sequence was firstdefined as a stratigraphic unit bound by large-scale, regional unconformities (Sloss et al.,1949). Wheeler (1958) retained this overalldefinition but included units bound bysmaller-scale unconformities. Although aparticular type of bounding unconformity wasnot specified by either Sloss et al. (1949) orWheeler (1958), applications of this conceptin the 1950s and 60s used either subaerialunconformities or unconformable shorelineravinements as the bounding unconformitiesof a sequence (e.g., Wheeler, 1958; Sloss,1963). Because these types ofunconformities are, for the most part,confined to the flanks of a basin, and asequence boundary was restricted to theunconformity, most sequence boundariesand their enclosed sequences could not becorrelated over much of the central portionsof a basin (see Figure 1 in Embry, 2008a).This greatly limited the practical applicationof sequences for subdiv iding thestratigraphic succession of a basin and such

Figure 9.1. A diagrammatic representation of a sequence boundary as a generic surface. A sequenceboundary is defined as a specific type of unconformity (red unconformity) and its correlative surfaces (blueconformity, green unconformity, and brown conformity). The correlative surfaces must adjoin to the end ofthe defining unconformity and join together so as to form one continuous sequence boundary. A specific typeof sequence has the same combination of surfaces for both its base and top.

a stratigraphic methodology was not widelyapplied until 1977.

New definitions – The 1977 watershedpublication, Seismic Stratigraphy - AAPGMemoir 26 (Payton, 1977), contained a seriesof articles on sequence stratigraphy byExxon scientists. A key observation was thata seismic reflector that encompassed a basinflank unconformity similar to that used byWheeler (1958) for bounding a sequence(i.e., characterized by truncation) could befollowed basinward where it encompassedsubmarine unconformities and conformities(Vail et al., 1977). On this basis, the Exxonresearchers modified the definition of asequence from a unit bounded byunconformit ies to one “bounded byunconformit ies or their correlat iveconformities” (Mitchum et al., 1977) and theycalled such a unit a depositional sequence.This, in effect, defined a sequence boundaryas a combination of surfaces rather than onespecific type of surface as Sloss et al. (1949)and Wheeler (1958) had done . Mostimportantly, such a modification allowedsequence boundaries to potentially becorrelated across an entire basin and thisgreatly expanded the application of sequenceboundaries for correlation and subdividingthe stratigraphic succession of a basin.

In 1988, Exxon scientists modified thedef in it ion of a deposit ional sequenceboundary to a subaerial unconformity andcorrelative conformities (Van Wagoner et al.,1988, p. 41), thus making it a much morespecific unit. At the same time, Galloway

(1989) defined a very different type ofsequence boundary which he termed agenetic stratigraphic sequence boundary, andit consisted solely of a maximum flooding isoften unconformable, such a proposedsequence boundary fit the Mitchum et al.(1977) general definition of a sequenceboundary but was clearly much different fromthe depositional sequence boundary of VanWagoner et al. (1988).

Generic definition – In light of the fact thattwo specific types of sequences have beendefined in the literature, a suitable, genericdefinition of a sequence is required. To fulfillthis need, Embry et al. (2007) defined asequence as “a stratigraphic unit bound by aspeci f ic type of unconformity and itscorrelative surfaces”. This definition resultsin a sequence being a general unit and specifictypes of sequences can be defined and namedon the basis of di f ferent types ofunconformities.

Correlative surfaces are an important partof the generic definition and are essentialfor allowing a sequence boundary to beextended over all or most of a basin.Correlat ive surfaces are sequencestratigraphic surfaces which join with theend(s) of the defining unconformity and witheach other so as to form one continuoussequence boundary (Figure 9.1). Correlativesurfaces can be unconformities, diastems. orconformities and, for maximum utility forsubsequent facies analysis in a sequencestratigraphic framework, they preferablyhave low diachroneity or are time barriers.

Page 44: Embry Sequence Stratigraphy

42

Figure 9.2. The boundaries (MFS) of a genetic stratigraphic sequence are shown in red on this sequencemodel characterized by a ramp setting with a fast initial base-level rise rate. The boundaries of such asequence are the same for all sequence models.

As wil l be demonstrated, the currentcontroversies concerning sequenceboundaries centre on correlative surfaces.

As noted earlier, two different, specificsequence types have been defined in theliterature and are in use today. These are thegenetic stratigraphic sequence of Galloway(1989) and the depositional sequence ofMitchum et al. (1977) / Van Wagoner et al.(1988) and they are described below. Otherspecific sequence types may be defined inthe future.

Genetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceGenetic Stratigraphic SequenceA genetic stratigraphic sequence was definedby Galloway (1989) and the unconformableportion of the maximum flooding surface(MFS) is the specific type of unconformitywhich defines this sequence type . Thecorrelative surfaces which compose theremainder of this type of sequence boundaryare the diastemic and conformable portionsof the MFS. Given that the MFS is a readilyrecognizable, material-based surface, such asequence type can be delineated in mostcases with objective analysis.

Consequently, a genetic strat igraphicsequence is classified as a material-basedsequence type and it is a very straight forwardand uncomplicated type of sequence. Itsboundaries are illustrated on a ramp, fast-initial-rise sequence model in Figure 9.2.Notably, such boundaries can be delineated,without modif icat ion, on any type ofsequence model with either a ramp or shelf/ slope / basin physiography and with eithera slow initial base-level rise or a fast initialrise. As will be seen, this “one boundary fits

all models” situation is not the case with adepositional sequence, and such simplicityis one of the attractive features of a geneticstratigraphic sequence.

The one serious drawback of a geneticstratigraphic sequence is that it commonlyencloses a subaerial unconformity or anunconformable shoreline ravinement on theflanks of a basin (Figure 9.2). Given that amajor time gap can be associated with suchsurfaces, not to mention a notable structuraldiscordance, such a sequence is really two,very different genetic units on the basinflanks. However, the MFS is usually the mostreadily recognizable and objective sequencesurface on both logs and seismic sections inthe offshore and deep marine areas wheresubaerial unconformities are absent. In theseareas, a genetic stratigraphic sequence clearlyhas great value for mapping andcommunication.

Depositional SequenceDepositional SequenceDepositional SequenceDepositional SequenceDepositional SequenceIntroduction – A depositional sequence wasintroduced by Mitchum et al. (1977) and thedefinition was refined by Van Wagoner et al.(1988). The specific type of unconformity thatdefines this sequence type is a subaerialunconformity. There is no doubt of the needand ut i l i ty of designat ing a subaeria lunconformity as the defining unconformityfor a depositional sequence boundarybecause of the time gap and significantdepositional and tectonic changes which areoften associated with such a surface. It wasthese properties of a subaerial unconformitythat led Sloss et al. (1949) to put such a surfaceon the boundaries of a stratigraphic unitrather than within it, thus giving birth to the

concept of a sequence as a stratigraphic unit.

Multiple depositional sequenceboundaries – Unl ike the genet icstratigraphic sequence boundary whichencompasses only a single surface type (MFS),eight different combinations of a subaerialunconformity with correlative surfaces havebeen put forward in the l iterature toconstitute a depositional sequence boundary.Thus, it is not hard to understand why thereis considerable confusion and controversywhen it comes to how one delineates andcorrelates a depositional sequence boundary.

This wide variety of depositional sequenceboundaries results from two main sources.One is the existence of both a materialbasedapproach and a time-based one to sequencestratigraphic classif ication as has beendescribed in previous articles. Materialbased,depositional sequence boundaries use onlymaterial-based, sequence stratigraphicsurfaces as correlative surfaces whereast ime-based, deposit ional sequenceboundaries also use the two time-basedsurfaces in this capacity. Another source ofsuch diversity is the existence of differentsequence models which include specificcombinations of either a ramp or shelf / slope/ basin physiography with either a slow initialbase-level rise rate or a fast initial base-levelrise rate. Some of these models weredescribed in previous articles and are usedherein to illustrate the different ways inwhich a depositional sequence boundary hasbeen delineated. Siliciclastic sediments areused in the models but the same stratigraphicsurfaces with the same relationships to eachother would occur on the models if carbonatesediments were used instead.

Valid correlative surfaces – As a preface tothe description and evaluation of eachproposed depositional sequence boundary,it is imperative to briefly review the criteriafor what constitutes a valid correlativesurface for a subaerial unconformity. First ofall, any designated correlative surface mustbe a sequence stratigraphic surface andrepresent either a break in deposition or achange in depositional trend. Just l ikemagnetostratigraphic surfaces would not besuitable for bounding a biostratigraphic unit,only sequence stratigraphic surfaces can beused as part of the boundary of a sequencestratigraphic unit.

Secondly, because the entire, preservedsubaerial unconformity must be part of agiven depositional sequence boundary, oneof the correlative surfaces has to join withthe basinward termination of the definingsubaerial unconformity. All of the correlativesurfaces then have to join with each otherso as to form one continuous boundary (see

Page 45: Embry Sequence Stratigraphy

43

Figure 9.1, page 41). Furthermore, given thatthe subaerial unconformity reaches itsmaximum basinward extent at the end ofbase-level fall (Jervey, 1988), correlativesurfaces must develop at or soon after thestart of base-level rise so as to fulfill thissecond criteria (Figure 9.3). As illustrated inFigure 9.3, surfaces which develop wellbefore or well after the start of base-levelr ise wi l l not join with the basinwardtermination of the subaerial unconformityand thus would not form a continuousboundary which includes all of the subaerialunconformity.

With these fundamental concepts andconstraints in mind, the various proposalsfor a depositional sequence boundary canbe evaluated as to their validity and utility.Material-based, deposit ional sequenceboundaries will be examined first, followedby proposed, time-based boundaries.

Material-based, Depositional SequenceMaterial-based, Depositional SequenceMaterial-based, Depositional SequenceMaterial-based, Depositional SequenceMaterial-based, Depositional SequenceBoundariesBoundariesBoundariesBoundariesBoundariesEarly depositional sequence boundaries– The first, material-based, depositionalsequence boundaries were proposed andillustrated by Van Wagoner et al. (1988). Theyillustrated the boundary on two differentsequence models – a shelf / slope / basinwith a slow initial rise rate (Figure 9.4A)and a ramp with a slow initial rise rate (Figure9.5A, page 44).

On their shelf / slope / basin model, thesubaerial unconformity occurs on the shelfwhere it is overlain by nonmarine strata. Thebasinward termination of the SU joins with aslope onlap surface, which in turn eventuallyjoins with the facies contact at the base ofthe submarine fan (Figure 9.4A). A slightlydifferent rendition of this model, based onexquis ite c l i f f exposures in Svalbard(Johannessen and Steel, 2005) (Figure 9.4B),better illustrates the Van Wagoner et al.(1988) depositional sequence boundary (SU/ SOS / facies change).

The one major flaw which invalidates such acombination of surfaces for a depositionalsequence boundary is the inclusion of thefacies boundary at the base of the submarinefan strata. This is a lithostratigraphic surfacerather than a sequence stratigraphic one andis not a valid correlative surface. As will besubsequently shown, a minor alteration tosequence boundary placement on such amodel allows a valid depositional sequenceboundary to be drawn.

The Van Wagoner et al. (1988) depositionalsequence boundary for a ramp setting with aslow initial rise is more problematic. Asshown on Figure 9.5A (page 44), thecorrelative surfaces employed for such a

Figure 9.3. A depositional sequence boundary includes all of a given subaerial unconformity by definition.One of the correlative surfaces must join with the basinward termination of the subaerial unconformity toensure a continuous, through-going, sequence boundary. Because the SU reaches its basinward extent at theend of base-level fall, the correlative surfaces must be developed at or soon after the start of base-level rise toensure one joins with the end of the SU. Surfaces developed well before or after the start of rise do not join withthe end of the SU and a continuous boundary is not possible.

Figure 9.4A. The Van Wagoner et al. (1988) depositional sequence boundary for a shelf / slope setting consistsof a subaerial unconformity (SU) on the shelf, a slope onlap surface (SOS) on the slope and the facies changeat the base of the submarine fan deposits in the basin (modified from a portion of Figure 2 of Van Wagoneret al., 1988).

Figure 9.4B. A sequence model for a shelf / slope / basin / setting with a slow initial base-level rise based onexposures of Eocene strata on Svalbard (Johannessen and Steel, 2005). The Van Wagoner et al. (1988)depositional sequence boundary is outlined in red.

boundary include the facies change at thebase of the shallow water sandstone and anon-descript surface within the shel f

mudstone facies. This non-descript surfacemay or may not represent an attempt to placethe boundary on a time surface equal to the

Page 46: Embry Sequence Stratigraphy

44

start of base-level rise (CC). Unfortunatelythis portion of the boundary is not discussedin their text. This combination of surfaceshas no validity for a depositional sequenceboundary because it includes alithostratigraphic surface (facies change) anda completely unknown and uncharacterizedsurface inside the mudstone facies.

Subsequent studies of the sequencestratigraphy of ramp successions oftenattempted to apply the Van Wagoner et al.(1988) boundary. In such analyses, the baseof a shal low water unit is used as acorrelative surface of the SU despite theirbeing no justification for it joining with thetermination of the subaerial unconformityand the fact that it is not a surface ofsequence stratigraphy. Figure 9.5B illustratesan example of such an invalid depositionalsequence boundary which was proposed ina major review of carbonate ramps byBurchette and Wright (1992). Notably, theliterature is replete with examples of suchinappropriate boundary placement for bothcarbonates and siliciclastics.

Ramp setting – Stratigraphic relationshipsin a ramp setting tend to be simpler thanthose in a shelf / slope / basin setting andthis generalization applies to sequencestratigraphy. Figure 9.6 illustrates a sequencemodel characterized by a ramp physiographyand a fast initial base-level rise. As shown, avalid depositional sequence boundary can bereadily identified on such a model. Thesubaeria l unconformity is truncatedbasinward by a shoreline ravinement (SR-U)which in turn joins to a maximum regressivesurface (MRS) farther basinward. Thus theSR-U and MRS are correlative surfaces of theSU in this model and all three together forma continuous, deposit ional sequenceboundary from basin edge to basin centre.

Such stratigraphic relationships are createdby a fast initial base-level rise rate such thattransgression occurs soon after the start ofbase-level rise and the shoreline ravinement(SR) cuts out the basinward portion of thesubaerial unconformity (SU). Notably, a largeamount of empirical data for both siliciclasticand carbonate successions confirm thecommon existence of these stratigraphicrelationships (references in Embry, in press)and substantiates the validity and utility ofsuch a depositional sequence boundary (SU /SR-U / MRS).

F igure 9.7 i l lustrates the sequencestratigraphic relationships for a sequencemodel which combines a ramp with a slowinitial base-level rise rate. In this case,transgression occurs significantly later thanthe start of base-level rise and the shorelineravinement does not truncate the basinward

Figure 9.5A. The Van Wagoner et al. (1988) depositional sequence boundary for a ramp setting with a slowinitial base-level rise rate consists of a subaerial unconformity (SU), a facies change at the base of shallowwater sandstones (yellow) and an unknown surface within the shelf mudstones (grey). This unknown surfacemay represent the time surface at the start of base-level rise (correlative conformity, CC) (modified from aportion of Figure 3 of van Wagoner et al., 1988).Figure 9.5B. Proposed placement of a depositional sequence boundary for a carbonate ramp setting (modi-fied from Burchette and Wright, 1992, Figure 13). The authors followed Van Wagoner et al. (1988) andextended the depositional sequence boundary along the facies change at the base of the shallow watercarbonates. Such a boundary is invalid because the facies change is a lithostratigraphic surface, not asequence stratigraphic one.

Figure 9.6. The boundaries of a material-based, depositional sequence are shown in red on this sequencemodel characterized by a ramp setting with a fast initial base-level rise rate. Due to the fast initial rise, theshoreline ravinement (SR-U) truncates the basinward portion of the subaerial unconformity (SU) andbecomes a correlative surface. The basinward termination of the shoreline ravinement joins the landwardtermination of the maximum regressive surface (MRS). Thus a continuous, depositional sequence boundaryconsists of a SU, a SR-U, and a MRS.

Page 47: Embry Sequence Stratigraphy

45

Figure 9.7. The boundaries of a material-based, depositional sequence are shown in red on this sequencemodel characterized by a ramp setting with a slow initial base-level rise rate. The boundaries consist only ofthe subaerial unconformities that are restricted to the basin flanks. There are no material-based, correlativesurfaces, which would allow the sequence boundaries to be extended into the basin.

Figure 9.8. The boundaries of a material-based, depositional sequence are shown in red on this sequencemodel characterized by a shelf / slope / basin setting in which slope onlap surfaces (SOS) form. The correlativesurfaces of the subaerial unconformity are an unconformable shoreline ravinement, a slope onlap surface,and a maximum regressive surface. This combination of surfaces forms a viable depositional sequenceboundary that has great utility.

portion of the subaerial unconformity. Thenet result is that the SR and MRS are notcorrelative surfaces (i.e., do not join with)of the SU and, as illustrated on Figure 9.7,there are no correlative, material-basedsurfaces for an SU in such a model. Thedepositional sequence boundary is limitedto the SU and cannot be extended fartherbasinward than the termination of the SU.Such a depositional sequence boundary isvalid but of limited utility. Of interest, noconvincing examples of such stratigraphicrelat ionships have ever been wel ldocumented in the literature but there is nodoubt that they are theoretically possibleand likely await discovery.

Shelf / slope / basin setting – The sequencestratigraphic relationships for a sequencemodel with a shelf / slope / basin physiographyare illustrated in Figure 9.8. The lowersequence boundary was generated underconditions of fast initial rise whereas theupper boundary represents slow initial base-level rise. For both boundaries, base levelfell to the shelf edge such that a slope onlapsurface (SOS) was generated. Notably, thekey stratigraphic relationships are the samefor both boundaries despite the differencein initial rise rate.

As illustrated on Figure 9.8, the definingsubaerial unconformity (SU) is truncatedbasinward by the shoreline ravinement (SR-U). The SR-U then joins the SOS at the shelfedge and a maximum regressive surface,which occurs within the submarine fandeposits, onlaps the SOS. Thus the SR-U, SOS,and MRS are all correlative surfaces of theSU and, in combination, allow a depositionalsequence boundary to be delineated frombasin margin to the deep basin. Such adepositional sequence boundary (SU / SR-U /SOS / MRS) represents a modification of thatproposed by Van Wagoner et al. (1988) whichis illustrated in Figure 9.4A (page 43). Theone change is that, within the basin, the MRSnear the top of the submarine fan deposits,rather than the highly diachronous, facieschange at the base of the submarine fan strata,is used as the correlative surface.

In shelf / slope / basin settings in which anSOS is not developed, a deposit ionalsequence boundary is readily drawn alongthe SU / SR-U on the basin flank and alongthe correlative MRS which is developed onthe outer shelf, slope and basin.

A sequence is best defined as a generic unitthat is bound by a speci f ic type ofunconformity and its correlative surfaces.Two specific types of sequences have beenrecognized and defined so far – a geneticstratigraphic sequence (part MFS, definingunconformity) and a depositional sequence(SU, defining unconformity). The genetic

strat igraphic sequence has the sameboundaries for all sequence models and thebounding surfaces are always material-based.Numerous combinations of material-basedand time-based surfaces have been proposedfor a depositional sequence boundary.

For a material-based, depositional sequenceboundary in a ramp sett ing, the onlycombination of surfaces which is valid andhas widespread utility consists of a subaerialunconformity, an unconformable shorelineravinement, and a maximum regressive

surface. A similar combination of surfaces,with or without the addition of a slope onlapsurface, is valid and has great utility for adepositional sequence boundary for a shelf /slope / basin setting.

The next article will examine the time-basedboundaries which have been proposed for adepositional sequence in both ramp and shelf/ slope / basin settings.

ReferencesReferencesReferencesReferencesReferencesBrown, L. and Fisher, W. 1977. Seismic-stratigraphic

Page 48: Embry Sequence Stratigraphy

46

interpretation of depositional systems: Examplesfrom the Brazilian rift and pull-apart basins. In:Seismic stratigraphy: application to hydrocarbonexploration. C. Payton, (ed.). American Associationof Petroleum Geologists Memoir 26, p.213-248.

Burchette, T. and Wright, V.P. 1992. Carbonate rampdepositional deposits. Sedimentary Geology, v. 79,p. 3-57.

Embr y, A . F. 2008a. Pract ica l SequenceStratigraphy II: Historical Development of theDiscipline: The First 200 Years (1788-1988).Canadian Society of Petroleum Geologists, TheReservoir, v. 35, issue 6, p. 35-40.

Embr y, A . F. , 2008b. Pract ica l SequenceStratigraphy III: Historical Development of theDiscipline: The Last 20 Years (1988-2008).Canadian Society of Petroleum Geologists, TheReservoir, v. 35, issue 7, p. 24-29.

Embry, A. F. 2009. Practical Sequence StratigraphyVIII: The Time-based Surfaces of SequenceStratigraphy. Canadian Society of PetroleumGeologists, The Reservoir, v. 36, issue 1, p. 27-33.

Embry, A. F. In press. Correlating SiliciclasticSuccessions with Sequence Stratigraphy. In:Application of Modern Stratigraphic Techniques:Theory and Case Histories. K. Ratcliffe and B.Zaitlin (eds.). Society of Economic Paleontologistsand Mineralogists, Special Publication.

Embr y, A. , Johannessen, E. , Owen, D, andBeauchamp, B. 2007. Recommendations forsequence stratigraphic surfaces and units(abstract). Arctic Conference Days, Abstract Book.Tromso, Norway.

Gal loway, W. 1989. Genet ic strat igraphicsequences in basin analysis I: architecture andgenesis of flooding surface bounded depositionalunits. American Association of Petroleum GeologistsBulletin, v. 73, p. 125-142.

Jervey, M. 1988. Quantitative geological modelingof siliciclastic rock sequences and their seismicexpression. In: Sea level changes: an integratedapproach. C. Wilgus, B.S. Hastings, C.G. Kendall,H.W. Posamentier, C.A. Ross, and J.C. Van Wagoner,(eds.). Society of Economic Paleontologists andMineralogists, Special Publication 42, p. 47-69.

Johannessen, E. J. and Steel, R. J. 2005. Shelf-marginclinoforms and prediction of deep water sands.Basin Research, v. 17, p. 521-550.

Mitchum, R, Vail, P., and Thompson, S. 1977. Seismicstratigraphy and global changes in sea level, part2: the depositional sequence as the basic unit forstratigraphic analysis, In: Seismic stratigraphy:application to hydrocarbon exploration. Payton,C . (ed.). American Association of PetroleumGeologists Memoir 26, p. 53-62.

Payton, C . (ed.). 1977. Seismic stratigraphy:appl icat ions to hydrocarbon explorat ion:American Association of Petroleum GeologistsMemoir 26, 516 p.

Sloss, L. 1963. Sequences in the cratonic interiorof North America. Geological Society of AmericaBulletin, v. 74, p. 93-113.

Sloss, L., Krumbein, W., and Dapples, E. 1949.Integrated facies analysis. In: Sedimentary faciesin geologic history. Longwell, C. (ed.). GeologicalSociety America, Memoir 39, p. 91-124.

Vail, P. et al. 1977. Seismic stratigraphy and globalchanges in sea level. In: Seismic stratigraphy:applications to hydrocarbon exploration. Payton,C . (ed.). American Association of PetroleumGeologists Memoir 26, p. 49-212.

Van Wagoner, J. C., Posamentier, H. W., Mitchum,R. M., Vail, P. R., Sarg, J. F., Loutit, T. S., and Hardenbol,J. 1988. An overview of the fundamentals ofsequence stratigraphy and key definitions. In: Sealevel changes: an integrated approach. C. Wilgus,B.S. Hastings, C.G. Kendall, H.W. Posamentier, C.A.Ross, and J.C. Van Wagoner, (eds.). Society ofEconomic Paleontologists and Mineralogists,Special Publication 42, p. 39-46.

Wheeler, H.E. 1964. Base level, lithosphere surfaceand time stratigraphy. Geological Society ofAmerica Bulletin. v. 75, p. 599-610.

Page 49: Embry Sequence Stratigraphy

47

Practical Sequence Stratigraphy XPractical Sequence Stratigraphy XPractical Sequence Stratigraphy XPractical Sequence Stratigraphy XPractical Sequence Stratigraphy XThe Units of Sequence Stratigraphy,Part 2: Time-based Depositional Sequences by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionAs described in the last article on material-based sequences (Embry, 2009b), a sequenceis best defined generically as “a stratigraphicunit bound by a specific type of unconformityand its correlative surfaces.” Two specifictypes of sequences have been defined in thel i terature – the genetic strat igraphicsequence of Galloway (1989) (part of amaximum flooding surface for definingunconformity) and the depositional sequenceof Mitchum et al. (1977) and Van Wagoner etal . (1988) (subaerial unconformity fordefining unconformity).

The boundaries of a genetic stratigraphicsequence are always material-based for allsequence models and consist of maximumflooding surfaces. However, proposedboundaries for a depositional sequence aremuch more diverse. The proposed, material-based boundaries were described andevaluated in the last article (Embry, 2009b).The proposed boundaries for a depositionalsequence which include timebased surfacesas correlative surfaces are described andevaluated herein. These t ime-based,deposit ional sequences are somewhatcontroversial as to their validity and utility.

Time-based, Depositional SequenceTime-based, Depositional SequenceTime-based, Depositional SequenceTime-based, Depositional SequenceTime-based, Depositional SequenceBoundariesBoundariesBoundariesBoundariesBoundariesIn the time-based approach to sequencestratigraphy, two time-based surfaces arerecognized as valid surfaces of sequencestratigraphy. These surfaces were introducedby Hunt and Tucker (1992) and are: 1) thebasal surface of forced regression (BSFR),which equates to the t ime surface(depositional surface) at the start of base-level fall and 2) the correlative conformity(CC), which equates to the time surface(depositional surface) at the start of base-level rise. These time-based surfaces werediscussed in detail in a previous article inthis series (Embry, 2009a).

Employing the Correlative ConformityEmploying the Correlative ConformityEmploying the Correlative ConformityEmploying the Correlative ConformityEmploying the Correlative ConformityOne proposed, time-based, depositionalsequence boundary uses the correlativeconformity (CC) as a key correlative surfaceof a subaerial unconformity so as to extendthe sequence boundary well into the basin(e.g., Hunt and Tucker, 1992; Helland-Hansenand Gjelberg, 1994). In a sequence model

with a ramp physiography and a fast, initialrise, the correlative conformity joins thebasinward end of the shoreline ravinement(SR-U) which in turn truncates the subaerialunconformity (SU) as previously discussed(Figure 10.1, see also Figure 2 in Helland-Hansen and Gjelberg, 1994). Thus, the CC isan acceptable correlative surface of an SU inthis model and such a depositional sequenceboundary (SU/SR-U/CC) is theoretically valid.

Figure 10.2 (page 48) illustrates a time-based,depositional sequence boundary using a CCas a correlative surface for a sequence modelwith ramp physiography and a slow initialrise (see also Figure 1 in Helland-Hansenand Gjelberg, 1994). As was previouslydiscussed in Embry (2009b), there are nomaterial-based, correlative surfaces for theSU in such a model. However, in the time-based approach, the CC provides such acorrelative surface because it adjoins thebasinward termination of the SU as shownin Figure 10.2 (page 48). Once again, such adepositional sequence boundary (SU/CC) istheoretically valid.

In a shelf/slope basin model, the CC closelycoincides with the MRS as was discussed inEmbry (2009a). As illustrated in Figure 10.3(page 48), a continuous boundary consistingof an SU, an SR-U, an SOS, and a CC can bedelineated on such a sequence model. Thus,such a time-based, depositional sequenceboundary would also be theoretically valid.

Although depositional sequence boundarieswhich employ a CC as a correlative surfaceare theoretically valid, the practical utility ofsuch boundaries is debatable. The reason forthis is that no published studies havedemonstrated how a CC can be delineatedand correlated in well exposed strata or onclosely spaced well logs with abundant core(see Embry, 2009a for a detailed discussion).Unconstrained interpretations of a CC onseismic data have been of fered (e.g . ,Catuneanu et al., in press) but these havenot been corroborated by rock-based dataand remain questionable. As discussed byEmbry (2009a), such seismic reflectors, whichare interpreted to encompass timebasedCCs, may actually be harbouring material-

Figure 10.1. The boundaries of one proposed, time-based, depositional sequence are shown in red on thissequence model characterized by a ramp setting with a slow initial base-level rise rate. The correlativeconformity (CC) is the only theoretically possible correlative surface of the subaerial unconformity for such amodel. Unfortunately a CC has no physical characteristics which allow its delineation.

Page 50: Embry Sequence Stratigraphy

48

based MRSs. Overall, much more research isnecessary before a depositional sequenceboundary which employs a CC can beaccepted as having practical utility.

Employing the Basal Surface of ForcedEmploying the Basal Surface of ForcedEmploying the Basal Surface of ForcedEmploying the Basal Surface of ForcedEmploying the Basal Surface of ForcedRegressionRegressionRegressionRegressionRegressionAnother time-based depositional sequenceboundary which has been proposed involvesthe use of the basal surface of forcedregression as a correlative surface of an SU(Posamentier and Allen, 1999; Coe, 2003).This sequence boundary comprises the samecombination of surfaces (SU/BSFR) for allsequence models. Such a depositionalsequence boundary is shown for a rampsetting with a slow initial base-level rise(Figure 10.4).

Because the BSFR is developed long beforethe start of base-level rise, it intersects theSU landward of the basinward terminationof the SU (Figure 10.4). It is also slightly offsetby the regressive surface of marine erosion(RSME) if such a surface is developed. As wasdiscussed by Embry (2009a), the BSFR has nophysical characteristics making it of dubiousvalue for comprising part of a sequenceboundary. However, more importantly, theBSFR cannot be considered to be a validcorrelative surface of an SU because it doesnot join with the end of the SU as shown inFigure 10.4. The use of such a sequenceboundary results in much of the subaerialunconformity being inside the proposedsequence rather than on its boundaries(Figure 10.4), an inappropriate relationshipfor a depositional sequence. This wouldsuggest that such a proposed depositionalsequence boundary (SU/BSFR) be rejectedas a possible option.

SummarySummarySummarySummarySummaryBy defining a sequence as a generic unit whichis bound by a specific type of unconformityand its correlative surfaces, two specifictypes of sequences are recognized – adeposit ional sequence (SU, def in ingunconformity) and genetic stratigraphicsequence (part MFS, defining unconformity).Numerous combinations of material-basedand timebased surfaces have been proposedfor a depositional sequence boundary.

In a time-based approach, a correlativeconformity (CC) which represents a timesurface (depositional surface) at the start ofbase-level rise is advocated for use as acorrelat ive surface for extending theboundary well into the basin. Although theCC is a theoretically valid correlative surface,its use as part of a sequence boundary iscompromised by a lack of physicalcharacteristics that would allow a CC to bedelineated and correlated with reasonableobjectivity.

Figure 10.3. The boundaries of a time-based, depositional sequence are shown in red on this sequence modelcharacterized by a shelf/slope/basin setting with a fast initial base-level rise rate. The correlative surfaces ofthe SU are the shoreline ravinement (SR-U), most of the slope onlap surface (SOS) and the correlativeconformity (CC) (time surface at start base-level rise).

Another proposed, time-based, depositionalsequence uses a basal surface of forcedregression (time surface at start base-levelfall) as a major part of the sequence boundary.The largest objection to such a proposal isthat the BSFR is not a correlative surface ofan SU because it is truncated by the SU farlandward of the basinward termination ofthe SU. Such a proposed boundary is notcompatible with the established definitionof a depositional sequence.

The main proposed material-based and time-based sequence boundaries are summarized

in Figure 10.5 and most are for a depositionalsequence boundary. The only boundarieswhich have widespread utility are material-based ones and include the MFS of the geneticstratigraphic sequence and the combined SU/SR-U/MRS, with or without an SOS, for thedepositional sequence. All other proposedboundaries use an inappropriate correlativesurface (e.g., BSFR, facies change) or includea correlat ive surface that cannot berecognized in most situations (e.g., CC).

The next article will examine systems tractswhich are component stratigraphic units of

Figure 10.2. The boundaries of a time-based, depositional sequence are shown in red on this sequence modelcharacterized by a ramp setting with a fast initial base-level rise rate. The correlative surfaces of the SU arethe shoreline ravinement (SR-U) and the correlative conformity (time surface at start base-level rise).

Page 51: Embry Sequence Stratigraphy

49

Figure 10.4. The boundaries of another proposal fora time-based, depositional sequence are shown inred on this sequence model characterized by aramp setting with a slow initial base-level rise rate. Inthis case, the basal surface of forced regression (BSFR)(time surface at start base-level fall) is used as theprimary correlative surface. Such a proposal is notreasonable because, as illustrated, the BSFR doesnot join the end of the subaerial unconformity.

Figure 10.5. A summary of the various combina-tions of surfaces for the different types of sequenceboundaries which have been proposed.

Page 52: Embry Sequence Stratigraphy

50

a sequence. Once again, both material-basedand time-based systems tracts have beendefined. The main ones in each approach willbe discussed and appraised as to their validityand utility for mapping and communication.

ReferencesReferencesReferencesReferencesReferencesCatuneanu, O., et al., in press. Towards theStandardization of Sequence Stratigraphy. EarthScience Reviews.

Coe, A. (ed.). 2003. The sedimentary record ofsea-level change. Cambridge University Press,New York, 287 p.

Embry, A.F. 1993. Transgressive-regressive (T-R)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embr y, A .F. 2009a. Pract ica l SequenceStratigraphy VIII: The Time-based Surfaces ofSequence Stratigraphy. Canadian Society ofPetroleum Geologists, The Reservoir, v. 36, issue 1,p. 27-33.

Embr y, A .F. 2009b. Pract ica l SequenceStrat igraphy IX: Part 1 Mater ia l -basedSequences. Canadian Society of PetroleumGeologists, The Reservoir, v. 36, issue 2, p. 23-29.

Gal loway, W. 1989. Genet ic strat igraphicsequences in basin analysis I: architecture andgenesis of flooding surface bounded depositionalunits. American Association of Petroleum GeologistsBulletin, v. 73, p. 125-142.

Helland-Hansen, W. and Gjelberg, J., 1994,Conceptual basis and variability in sequencestratigraphy: a different perspective. SedimentaryGeology, v. 92, p. 1-52.

Hunt , D. and Tucker, M. 1992. Strandedparasequences and the forced regressive wedgesystems tract: deposition during base-level fall.Sedimentary Geology, v. 81, p. 1-9.

Mitchum, R, Vail, P., and Thompson, S. 1977. Seismicstratigraphy and global changes in sea level, part2: the depositional sequence as the basic unit forstratigraphic analysis, In: Seismic stratigraphy:application to hydrocarbon exploration. C. Payton,(ed.). American Association of Petroleum GeologistsMemoir 26, p. 53-62.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. Society of Economic Paleontologistsand Mineralogists, Concepts in Sedimentology andPaleontology, no. 7, 210 p.

Van Wagoner, J.C., Posamentier, H.W., Mitchum,R.M., Vail, P.R., Sarg, J.F., Loutit, T.S., and Hardenbol,J. 1988. An overview of the fundamentals of

sequence stratigraphy and key definitions, In: Sealevel changes: an integrated approach. C. Wilgus,B.S. Hastings, C.G. Kendall, H.W. Posamentier, C.A.Ross, and J.C. Van Wagoner, (eds.). Society ofEconomic Paleontologists and Mineralogists,Special Publication 42, p. 39-46.

Page 53: Embry Sequence Stratigraphy

51

Practical Sequence Stratigraphy XIPractical Sequence Stratigraphy XIPractical Sequence Stratigraphy XIPractical Sequence Stratigraphy XIPractical Sequence Stratigraphy XIThe Units of Sequence Stratigraphy,Part 3: Systems Tracts by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionThe sequence is the primary unit of sequencestratigraphy and, as was discussed in theprevious two articles (Embry, 2009a, b), twospecific types of sequences have been defined.Both a depositional sequence and a geneticstratigraphic sequence can be subdivided intocomponent units that are called systemstracts. Like a specific sequence type, a definedsystems tract must be bound by specific,recognizable sequence stratigraphic surfacesif it is to have validity and utility.

Van Wagoner et al. (1988) and Posamentierand Vai l (1988) advanced sequencestratigraphy with the innovation that asequence can be subdivided into componentunits on the basis of sequence stratigraphicsurfaces that occur within a sequence. Thisenhances mapping and communication andadds to the resolution capability of sequencestrat igraphy. They referred to suchcomponent units of a sequence as systemstracts, a unit originally defined by Brown andFisher (1977), as “a l inkage ofcontemporaneous depositional systems”.Such a definition does not make clear whattype of surfaces bound systems tracts.Furthermore, such a definition implies thatsystems tracts are chronostratigraphic unitsand have time surfaces and / or time barriersfor boundaries. Van Wagoner et al. (1988, p.39) adopted the Brown and Fisher (1977)definition and noted that systems tracts “aredefined by their position within the sequenceand by the stacking patterns of parasequencesets and parasequences”. This methodologyalso is somewhat problematic becausesequence stratigraphic units are primarilydefined by their bounding surfaces ratherthan internal properties such as “stackingpatterns”.

A s impler and more straightforwarddefinition is proposed for a systems tract.Embry et al. (2007) defined a systems tractas “a component unit of a sequence which isbound by sequence-stratigraphic surfaces”.Such a definition is generic, leaves no doubtas to a systems tract being a sequencestratigraphic unit, and allows specific typesof systems tracts to be defined. The definitionalso honours the Brown and Fisher (1977)original definition and makes it clear thatsequence stratigraphic surfaces rather than

time surfaces form the boundaries. It also iscompatible with the Van Wagoner et al. (1988,1990) usage in that various sequencestratigraphic surfaces are often delineatedon the basis of a change in stacking patternas discussed in previous articles in thisseries. Importantly, this definition alsocovers the common situation where stackingpatterns are not evident. F inal ly, theproposed def in it ion emphasizes theboundaries of the unit and it can be readilyapplied for subdividing any specific type ofsequence including ones which may beproposed in the future.

Like other sequence stratigraphic units, asystems tract is defined by its boundingsurfaces. A specific type of systems tract canbe defined by key sequence stratigraphicsurfaces and their correlative surfaces whichform its lower and upper boundaries. It isemphasized that i t is the boundingstratigraphic surfaces which define a giventype of systems tract and not thecharacteristics of the strata within thesystems tract. Of course, the characteristicsof the strata, such as stacking patterns ofsmaller-scale units and grain-size trends inthe strata, substantially contribute to thedelineation of the various bounding surfacesand thus indirectly contribute to thedelineation of a given systems tract.

Similar to sequence boundaries, the specificbounding surfaces that have been proposedfor systems tracts include both material-based surfaces and time-based surfaces. Thisresults in the existence of both material-based systems tracts, which have onlymaterial-based surfaces for boundaries, andtime-based systems tracts, which have at leastone time-based surface as a part of one orboth of its boundaries The two differentapproaches to systems tract definition aredescribed below.

Material-based Systems TractsMaterial-based Systems TractsMaterial-based Systems TractsMaterial-based Systems TractsMaterial-based Systems TractsTwo different, material-based systems tractclassification schemes have been proposed.One defined three specific systems tractsfor a depositional sequence and the otheronly two.

Three Systems Tracts: Van Wagoner et al.(1988) and Posamentier et a l . (1988)

subdivided a depositional sequence intothree specific systems tracts. As describedin Part 9 of this series (Embry, 2009a), thedefining boundary of a depositional sequenceas proposed by these authors is acombination of a subaerial unconformity andan unconformable shoreline ravinement (SU/ SR-U) on the shelf, a slope onlap surface(SOS) along the slope, and a facies change atthe base of the turbidites in the basin. Theydefined three systems tracts within such adepositional sequence (Figure 11.1, page 52)on the basis of two sequence stratigraphicsurfaces – a transgressive surface and amaximum flooding surface – within it. Incurrent terminology, a “transgressivesurface” is equivalent to a combinedmaximum regressive surface (MRS) anddiastemic shoreline ravinement (SR-D).

The lowermost systems tract was called thelowstand systems tract (LST) and was boundat the base by the subaerial unconformity(SU) on the shelf and basinward by the slopeonlap surface (SOS) and the facies changebelow the turbidites. The LST was bound atthe top by the “transgressive surface” (SR-D+ MRS) and it encompassed both marine strataand nonmarine strata. The middle systemstract was named the transgressive systemstract (TST) and it was bound by thetransgressive surface (SR-D + MRS) belowand the maximum flooding surface (MFS)above. The upper systems tract was calledthe highstand systems tract (HST) and it wasbound by the MFS below and the sequenceboundary (SU / SOS / facies change) above(Figure 11.1, page 52).

There are a few arguable issues associatedwith both the LST and the HST as definedand applied by Van Wagoner et al. (1988,1990). The main one is that the highlydiachronous facies change at the base of theturbidites is not a well defined boundingsurface for either a sequence or a systemstract, as discussed in Embry (2009a). Thisfacies change at the base of the turbidites isused as both the basal contact of the LST andthe upper contact of the HST. As defined,part of the bounding surface of these systemstracts does not constitute a sequencestratigraphic surface. Furthermore, the useof a diastemic shorel ine ravinement(landward portion of their “transgressive

Page 54: Embry Sequence Stratigraphy

52

surface”) as the upper boundary of the LSTon the basin flank is a problem because it isa highly diachronous surface.

Another issue involves the application ofthese three systems tracts to a ramp setting.The mutual boundary between the HSTbelow and the LST above in a ramp settingequates to the sequence boundary. Asdiscussed in Embry (2009a), Van Wagoner etal. (1988, 1990) and many others (e.g.,Burchette and Wright, 1992) placed thisboundary at a highly diachronous, facies

Figure 11.2. The boundaries of a material-based, depositional sequence (SU, SR-U, MRS) are shown in red onthis sequence model characterized by a ramp setting with a fast initial base-level rise rate. The occurrence ofthe low diachroneity maximum flooding surface (MFS) allows the sequence to be subdivided into two systemstracts – a transgressive systems tract (TST) and a regressive systems tract (RST). Note that all nonmarinestrata between the SU and the MFS are placed in the TST.

change at the base of a prograding shallow-water facies. Such a boundary is not definitiveenough for either a systems tract or asequence boundary.

Two Systems Tracts: Embry (1993) andEmbry and Johannessen (1993) offered asolution to the problem of employing highlydiachronous facies changes or a diastemicshoreline ravinement as systems tractboundaries. In ramp and shelf / slope / basinsettings, the only material-based, lowdiachroneity, sequence stratigraphic surface

that occurs within a depositional sequenceis the maximum flooding surface (MFS). Onthis basis, Embry (1993) proposed that adepositional sequence be subdivided intotwo systems tracts: a lower transgressivesystems tract that follows the definition ofVan Wagoner et al. (1988) and an upper, newlydefined, regressive systems tract (Figures11.2, 11.3).

These two systems tracts are best definedby the key sequence stratigraphic surfacesthat form their lower and upper boundaries.Thus, a TST is defined as a sequencestratigraphic unit bound by correlativesurfaces below and a maximum floodingsurface and its correlative surfaces above.The RST is just the opposite being definedas a sequence stratigraphic unit bound by amaximum flooding surface and its correlativesurfaces below and by a maximum regressivesurface and its correlative surfaces above.

The RST, as defined above, includes both theLST and HST of Van Wagoner et al. (1988) andthis is a consequence of eliminating a facieschange surface (proposed HST / LSTboundary) as a systems tract boundary(Figure 11.3). Also, the nonmarine strataassigned to an LST by Van Wagoner et al.(1988) (Figure 11.1) are much better placedin a TST (Figures 11.2, 11.3) as discussed bySuter et al. (1987) and many others.

The genetic stratigraphic sequence has MFSsas its bounding surfaces as discussed by Embry(2009a). Like the depositional sequence, itcan also be subdivided into a TST and an RSTby using the internal composite boundary ofan SU / SR-U / SOS / MRS as a boundary forboth systems tracts (Figures 11.2, 11.3).

In summary, two material-based systemstracts – transgressive systems tract andregressive systems tract – can be recognizedin both a depositional sequence and a geneticstrat igraphic sequence in a lmost a l ls i tuat ions. Such recognit ion can beaccomplished in a very objective manner.

Time -based Systems TractsTime -based Systems TractsTime -based Systems TractsTime -based Systems TractsTime -based Systems TractsAs described in Embry (2009c), two abstract,time-based surfaces are recognized in thet ime-based approach to sequencestratigraphy. These are the basal surface offorced regression (BSFR) which equates tothe time surface at the start of base-level falland the correlative conformity (CC) whichequates to the time surface at the start ofbase-level rise. The defined, time-basedsystems tracts use both material-basedsurfaces and the two abstract, time-basedsurfaces (BSFR, CC) as boundaries. Twoclassification systems of time-based systemstracts have been proposed – one defines foursystems tracts for a depositional sequence

Figure 11.1. The Van Wagoner et al. (1988) depositional sequence boundary for a shelf / slope setting consistsof a subaerial unconformity (SU) on the shelf, a slope onlap surface (SOS) on the slope, and the facies changeat the base of the submarine fan deposits in the basin (modified from a portion of Figure 2 of Van Wagoneret al., 1988). Two surfaces are recognized within such a sequence – the transgressive surface (TS) and themaximum flooding surface (MFS) and allow the delineation of three systems tracts – lowstand (LST),transgressive (TST), and highstand (HST).

Page 55: Embry Sequence Stratigraphy

53

Figure 11.3. The boundaries of a material-based, depositional sequence (SU, SR-U, SOS, MRS) are shown inred on this sequence model characterized by a shelf / slope / basin setting. The internal maximum floodingsurface allows such a sequence to be subdivided into a transgressive systems tract (TST) and a regressivesystems tract.

and the other three.

Four Systems Tracts: Hunt and Tucker(1992) proposed subd i v id ing adepositional sequence into four systemstracts (Figure 11.4) that, in ascendingorder, were named lowstand ,transgress ive , h i ghstand , and forcedregressive systems tracts. The theoreticalba s i s o f th i s four s y s tems t r ac tclassification scheme was elaborated uponby Helland-Hansen and Gjelberg (1994).The bounding surfaces for these units willd i f f e r s l i gh t l y depend ing on thephysiography of the setting (ramp or shelf/ slope / basin) and the speed of the initialbase-level rise (fast, slow). Thus, thesedifferent specific systems tracts are bestdefined by a key sequence stratigraphicsurface that is common to all models and itscorrelative surfaces for both the lower andupper boundary. The four key surfaces thatare used are used to define either the loweror upper boundary of the four systems tractsare two time-based surfaces – correlativeconformity (CC) and basal surface of forcedregression (BSFR) – and two material-basedsurfaces – maximum regressive surface (MRS)and maximum flooding surface (MFS). On thisbasis, the four systems tracts of Hunt andTucker (1992) are defined below:

Lowstand Systems Tract (LST) – A componentunit of a sequence defined by a correlativeconformity (CC) and its correlative surfacesas the lower boundary and a maximumregressive surface (MRS) and its correlativesurfaces as the upper boundary. This is a time-based systems tract.

Transgressive Systems Tract (TST) – Acomponent unit of a sequence defined by amaximum regress ive surface and itscorrelative surfaces as the lower boundaryand a maximum flooding surface (MFS) andits correlat ive surfaces as the upperboundary. This is a material-based systemstract, which was originally defined by VanWagoner et al. (1988).

Highstand Systems Tract (HST) – Acomponent unit of a sequence defined by amaximum flooding surface and its correlativesurfaces as the lower boundary and a basalsurface of forced regression (BSFR) and itscorrelative surfaces as the upper boundary.This is a time-based systems tract.

Forced Regressive (Falling Stage) SystemsTract (FRST, FSST) – A component unit of asequence defined by a basal surface of forcedregression (BSFR) and its correlative surfacesas the lower boundary and a correlativeconformity (CC) and its correlative surfacesas the upper boundary. This is a time-basedsystems tract.

Figure 11.4. The boundaries of a time-based, depositional sequence (SU, CC) based on Hunt and Tucker(1992) are shown in red on this sequence model characterized by a ramp setting with a slow initial base-levelrise rate. Such a sequence contains three sequence stratigraphic surfaces within it – maximum regressivesurface (MRS), maximum flooding surface (MFS), and basal surface of forced regression (BSFR). This allowsthe time-based sequence to be subdivided into four systems tracts: three time-based ones (highstand (HST),falling stage (FSST), and lowstand (LST)) and one material-based one (transgressive (TST)).

In summary, the Hunt and Tucker (1992)class i f icat ion scheme included threetimebased systems tracts (LST, HST, and FRST)of their definition and one material-basedsystems tract (TST), which had originally beendefined by Van Wagoner et al. (1988). Figure4 illustrates these four systems tracts for asequence model with a ramp setting and aslow initial base-level rise. Figure 11.5 (page54) is a model with ramp setting and a fastinitial rise. Note that the key, definingsurfaces for the four systems tracts remain

the same in both models but in some casesthe correlative surfaces are different. Thesame comment would apply if models with ashelf / slope / basin setting were illustrated.

In the Hunt and Tucker (1992) classificationscheme, all the defined systems tracts, withthe exception of the TST, utilize one or bothof the two time surfaces (CC, BSFR) as partof their boundaries (Figure 11.4) . Asdiscussed in Embry (2009c), these timesurfaces have no characteristic physical

Page 56: Embry Sequence Stratigraphy

54

properties and thus cannot be delineatedobjectively in most settings. The use of suchboundaries is potentially problematic. Thus,I would not recommend the LST, HST, andFRST (FSST) as defined by Hunt and Tucker(1992) for use in sequence stratigraphy.

Three Systems Tracts: The second, time-based systems tract classification scheme isthat of Posamentier and Allen (1999). Asdescribed in the previous article on time-based sequences (Embry, 2009b),Posamentier and Allen (1999) placed thecorrelative surface for the depositionalsequence boundary along the basal surfaceof forced regression (BSRF) which is thetime surface at the start of base-level fall.They divided such a depositional sequenceinto three systems tracts : lowstand,transgressive, and highstand (Figure 11.6).

The transgress ive systems tract andhighstand systems tract of this classificationfollow the definitions used by Hunt andTucker (1992) in that they have exactly thesame key surfaces defining their lower andupper boundaries. Only the lowstandsystems tract has a new and differentdefinition and it contrasts with the twoprevious definitions of an LST provided byVan Wagoner et al. (1988) and Hunt andTucker (1992).

Posamentier and Allen (1999) defined alowstand systems tract as being bound at itsbase by the basal surface of forced regressionand its correlative surfaces and at its top by

Figure 11.6. The boundaries of the Posamentier and Allen’s (1999) proposed time-based, depositionalsequence (SU, BSFR) are shown in red on this sequence model characterized by a ramp setting with a slowinitial base-level rise rate. Only two internal surfaces, MRS and MFS, are recognized and accordingly Posamentierand Allen (1999) subdivided their depositional sequence into three systems tracts – lowstand (LST), transgres-sive (TST), and highstand (HST). The TST and HST followed the original definitions of these units but the LSTwas redefined by making the BSFR as the key bounding surface at its base.

the maximum regressive surface and itscorrelative surfaces (Figure 11.6). Note thatPosamentier and Allen (1999) do not usethe correlative conformity as a systems tractboundary. However, they do suggest thattheir lowstand systems tract might be

subdivided into an “early LST” and a “lateLST” by recognition of the CC within an LST.

As was discussed in the previous section,the highstand systems tract as defined byHunt and Tucker (1992) and adopted byPosamentier and Allen (1999) has limited usein practice because it is bound in part by anabstract time surface (BSFR) with no physicalcharacteristics. The same comment appliesto Posamentier and Allen’s (1999) reviseddefinition of a lowstand systems tract thathas the abstract BSFR as the key surface forits lower boundary.

In summary, two classification schemes havebeen proposed in the time-based approachto sequence stratigraphy. Both systems tract,the TST, which is valid and of practical value.Two different, timebased definitions for alowstand systems tract, which was originallydefined as a material-based unit, have beenproposed. Both use an abstract time surfaceas part of the lower boundary (BSFR for oneand CC for the other), which limits theirpractical applications. The other specificsystems tracts proposed in these schemes(HST, FRST) are also bound in part by anabstract time surface that limits their use.

SummarySummarySummarySummarySummaryA systems tract is best def ined as acomponent unit of a sequence which is boundby sequence stratigraphic surfaces. Specificsystems tracts are defined by key surfaces

Figure 11.5. The boundaries of a time-based, depositional sequence (SU, SR-U, CC) based on Hunt and Tucker(1992) are shown in red on this sequence model characterized by a ramp setting with a fast initial base-levelrise rate. The same four systems tracts with the same key, defining surfaces as shown on Figure 4 can bedelineated. Note that the correlative surfaces of the key surfaces differ from those shown in Figure 4. Forexample the correlative surfaces of the CC in this figure are the SU and SR-U whereas in Figure 4 the onlycorrelative surface of the CC is the SU.

Page 57: Embry Sequence Stratigraphy

55

Figure 11.7. A summary of the various systems tract classification schemes that have been proposed. The sequence boundaries are in red and internal systems tractboundaries are in blue.

and their correlative surfaces for both thelower and upper boundaries of the unit. Forexample, a transgressive systems tract isdefined as a sequence stratigraphic unitbound at its base by a maximum regressivesurface and its correlative surfaces and at itstop by a maximum flooding surface and itscorrelative surfaces.

Four different classification schemes forsubdividing a depositional sequence intosystems tracts have been proposed – twomaterial-based schemes and two timebasedschemes. These four proposals aresummarized and compared in Figure 11.7. Inthe material-based approach to sequencestrat igraphy, the three systems tractclassification scheme of Van Wagoner et al.(1988) is problematic because the LST andHST use a highly diachronous facies changeas a key surface for both of these units. Theelimination of the facies change as a boundingsurface and the combination of the HST andLST into a single systems tract, which istermed a regressive systems tract (RST)(Embry, 1993), results in a more practical,two systems tract classification scheme(Figure 11.7).

In the time-based approach to sequencestratigraphy, both a four systems tractclassification scheme and a three systemstract one have been proposed. Both

classification schemes are problematic in thatmost of the defined systems tracts havelimited practical value due to the use of anabstract time surface as one or both ofbounding surfaces of the unit.

The most useful systems tracts which havebeen proposed so far are the transgressivesystems tract and the regressive systemstract. The most confusing and contentiousunit is the lowstand systems tract in that ithas been defined in three different ways.

ReferencesReferencesReferencesReferencesReferencesBrown, L. and Fisher, W. 1977. Seismicstratigraphicinterpretation of depositional systems: Examplesfrom the Brazilian rift and pull-apart basins. In:Seismic stratigraphy: application to hydrocarbonexploration. C. Payton (ed.). AAPG Memoir 26, p.213-248.

Burchette, T. and Wright, V. P. 1992. Carbonateramp depositional deposits. Sedimentary Geology,v. 79, p. 3-57.

Embry, A. F., 1993, Transgressive-regressive (T-R)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embr y, A . F. 2009a. Pract ica l sequencestrat igraphy IX: the uni ts of sequence

stratigraphy; part 1, material-based sequences.Canadian Society of Petroleum Geologists. TheReservoir, v. 36, issue 2, p. 23-29.

Embr y, A . F. 2009b. Pract ica l sequencestratigraphy X: the units of sequence stratigraphy;part 2, time-based sequences. Canadian Societyof Petroleum Geologists. The Reservoir, v. 36, issue3. p. 21-24.

Embr y, A . F. 2009c . Pract ica l sequencestratigraphy VIII: the time-based surfaces ofsequence stratigraphy. Canadian Society ofPetroleum Geologists. The Reservoir, v. 36, issue 1,p. 27-33.

Embry, A. F. and Johannessen, E. P. 1993. T-Rsequence stratigraphy, facies analysis andreservoir distribution in the uppermost Triassic-Lower Jurassic succession, western Sverdrup Basin,Arctic Canada, In: Arctic Geology and PetroleumPotential. T. Vorren, E. Bergsager, O. A. Dahl-Stamnes,E. Holter, B. Johansen, E. Lie, and T.B. Lund (eds.).NPF Special Publication 2, p. 121-146.

Embry, A., Johannessen, E. P., Owen, D., andBeauchamp, B., 2007. Recommendations forsequence stratigraphic surfaces and units(abstract). Arctic Conference Days, Abstract Book.Tromso, Norway.

Helland-Hansen, W. and Gjelberg, J. 1994.Conceptual basis and variability in sequence

Page 58: Embry Sequence Stratigraphy

56

stratigraphy: a different perspective. SedimentaryGeology, v. 92, p. 1-52.

Hunt , D. and Tucker, M. 1992. Strandedparasequences and the forced regressive wedgesystems tract: deposition during base-level fall.Sedimentary Geology, v. 81, p. 1-9.

Posamentier, H. and Vail, P. 1988. Eustatic controlson clastic deposition II - sequence and systemstract models. In: Sea level changes: an integratedapproach. C. Wilgus, B. S. Hastings, C. G. Kendall,H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner,(eds.). Society of Economic Paleontologists andMineralogists, Special Publication 42, p. 125- 154.

Posamentier, H. Jervey, M. and Vail, P., 1988.Eustatic controls on clastic deposition I-conceptualframework. In: Sea level changes: an integratedapproach. C. Wilgus, B. S. Hastings, C. G. Kendall,H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner,(eds.). Society of Economic Paleontologists andMineralogists, Special Publication 42, p. 109- 124.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. Society of Economic Paleontologistsand Mineralogists, Concepts in Sedimentology andPaleontology, no. 7, 210 p.

Suter, J., Berryhill, H., and Penland, S. 1987. LateQuaternary sea level fluctuations and depositionalsequences, southwest Louisiana continental shelf.In: D. Nummedal, O. Pilkey, and J. Howard (eds.).Sea-level changes and coastal evolution. Societyof Economic Paleontologists and Mineralogists,Special Publication 41, p. 199-122.

Van Wagoner, J. C., Posamentier, H. W., Mitchum,R. M., Vail, P. R., Sarg, J. F., Loutit, T. S., and Hardenbol,J. 1988. An overview of the fundamentals ofsequence stratigraphy and key definitions. In: Sealevel changes: an integrated approach. C. Wilgus,B. S. Hastings, C. G. Kendall, H. W. Posamentier, C.A. Ross, and J. C. Van Wagoner, (eds.). Society ofEconomic Paleontologists and Mineralogists,Special Publication 42, p. 39-46.

Van Wagoner, J. C., Mitchum, R. M., Campion, K.M., and Rahmanian, V. D. 1990. Siliciclasticsequence stratigraphy in well logs, cores andoutcrops. American Association of PetroleumGeologists, Methods in Exploration, no. 7, 55 p.

Page 59: Embry Sequence Stratigraphy

57

Practical Sequence Stratigraphy XIIPractical Sequence Stratigraphy XIIPractical Sequence Stratigraphy XIIPractical Sequence Stratigraphy XIIPractical Sequence Stratigraphy XIIThe Units of Sequence Stratigraphy,Part 4: Parasequences by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionThere are three basic types of stratigraphicunits in use in sequence stratigraphy –sequence, systems tract, and parasequence.Various proposals for defining and delineatingspecific types of sequences and systems tractswere reviewed and discussed in the last threearticles in this series (Embry, 2009a, b, c). Inthis article I will discuss the last unit type –the parasequence.

Original DefinitionOriginal DefinitionOriginal DefinitionOriginal DefinitionOriginal DefinitionThe term parasequence was originallydefined by Van Wagoner et al. (1988, p. 39). Inkeeping with sequence stratigraphic practice,they defined a parasequence by means of itsbounding surfaces: “a relatively conformablesuccession of beds or bedsets bound bymarine-flooding surfaces.” To understand thedefinition of a parasequence, one needs adefinition of its defining bounding surface –a marineflooding surface.

A marine-flooding surface, which is muchmore commonly called a flooding surface (FS),was defined by Van Wagoner et al. (1988, p.39) as “a surface separating younger fromolder strata across which there is an abruptincrease in water depth.” This definition doesnot provide much insight into what aflooding surface actually is and how onewould recognize one. All stratigraphicsurfaces separate younger from older strata(Law of Superposition) leaving “an abruptincrease in water depth” as the only criteriafor recognit ion. Given this is a veryinterpret ive cr iter ia , rather than anobservable one, i t is not a suitablecharacteristic for defining a material-basedsurface. Because adequate definitions werenot provided, one is left wondering what aflooding surface actually is and, consequently,what a parasequence is.

Van Wagoner et al. (1990) provided muchmore information and insight into what theyactually meant by the terms flooding surfaceand parasequence, although the actualdefinitions remained the same as thoseoriginally offered. They provided a model forthe development of a parasequence (VanWagoner et al., 1990, Figure 4) (Figure 12.1herein) and portrayed parasequences as unitsof shallowing-upward strata separated bysurfaces which marked a transgression. As

Figure 12.2. Parasequence boundary as defined and illustrated by Van Wag-oner et al. (1988, 1990). These authors placed the boundary at the changefrom sandstone below to shale / siltstone above and called such a surface aflooding surface (FS). As shown, a flooding surface (FS) lies between a maxi-mum regressive surface (MRS) below and a maximum flooding surface (MFS)above. As defined, a flooding surface is a lithostratigraphic surface rather thana sequence stratigraphic one. Modified after a portion of Van Wagoner et al.Figure 3B.

Figure 12.1. The Van Wagoner et al. (1990) model for parasequences. Eachparasequence consists of a succession of coarsening-upward, regressive strata.The lack of any transgressive strata is a problematic flaw in this model. Modi-fied after a portion of Van Wagoner et al. (1990), Figure 4.

shown on Figure12.1, Van Wagoner etal . (1990) did notinclude the formationof any transgressivestrata in theirparasequence model.Given the Law of theConservat ion ofMatter, this lack of anytransgressive stratais nonactualistic asdiscussed by Arnott(1995).Consideration of thisaspect of the modelhelps to put incontext problemsassociated with theVan Wagoner et al.(1988, 1990) conceptof a flooding surfaceand a parasequence.These are discussedbelow.

Parasequence as aParasequence as aParasequence as aParasequence as aParasequence as aLithostratigraphicLithostratigraphicLithostratigraphicLithostratigraphicLithostratigraphicUnitUnitUnitUnitUnitAn inspect ion ofvarious diagrams inVan Wagoner et al.(1990) (e.g., Figures3b and 7 in VanWagoner et al., 1990)reveals that what VanWagoner et al. (1988,1990) meant by aflooding surface is acontact between amarine sandstonebelow and a deeper-water, marine shale /s i l tstone above(Figure 12.2). Goingby the illustrations inVan Wagoner et al.(1990), such a contact can be gradational(conformable) or scoured (diastem). Asshown on Figures 12.2 and 12.3, a floodingsurface, as conceived and applied by VanWagoner et a l . (1988, 1990), is bestcategorized as a within-trend facies contactthat is developed within a transgressivesuccession. A flooding surface (FS) lies

between two surfaces of sequencestratigraphy – a maximum regressive surface(MRS) below and a maximum flooding surface(MFS) above (Figures 12.2 and 12.3). Notably,a flooding surface does not represent achange in depositional trend. Rather, itrepresents a change in lithology (sandstone/ l imestone to shale / marl) within a

Page 60: Embry Sequence Stratigraphy

58

depositional trend (e.g., within a fining-upward succession) and consequently is bestconsidered to be a lithostratigraphic surfacerather than a sequence stratigraphic one.

With the understanding of what a floodingsurface real ly is , i t fol lows that aparasequence is a unit bound bylithostratigraphic surfaces. Furthermore, thismeans that a parasequence as defined andused by Van Wagoner et al. (1988, 1990) isreally a lithostratigraphic unit and not asequence strat igraphic one . Anothercomplication is that the term flooding surfacehas also been sometimes inappropriatelyapplied to well defined and characterized,materia l -based surfaces of sequencestratigraphy including a maximum regressivesurface, a maximum flooding surface, and ashorel ine ravinement (Embry, 2005;Catuneanu, 2006). This practice has createdadditional uncertainty and confusion as towhat a flooding surface and a parasequence

Figure 12.3. A schematic section of atransgressiveregressive succession that is not inter-rupted by any unconformities. Two sequence strati-graphic surfaces, maximum regressive surface (MRS)and maximum flooding surface (MFS) as well as alithostratigraphic surface, flooding surface (FS), canbe delineated in such a succession.

Although Van Wagoner et al. (1988, 1990) put theparasequence boundary at the flooding surface,others have used the MRS or the MFS. Such vari-ance in boundary delineation has resulted in con-siderable confusion regarding the placement of aparasequence boundary.

Figure 12.5. The upper portion of a small scale, regressive-transgressive succession is shown. A maximumregressive surface (MRS) is delineated at the base of the red weathering, massive, highly burrowed, calcareoussandstone unit and coincides with the change from a coarsening- and shallowing-upward succession (regres-sive) to a fining- and deepening-upward succession (transgressive). A flooding surface (FS) occurs within thetransgressive strata at the boundary between the burrowed sandstone and the overlying unit of shale andsiltstone. Middle Devonian Bird Fiord Formation, Bathurst Island

Figure 12.4. A schematic cross-section showing the correlation of three transgressive-regressive successions.Van Wagoner et al. (1988, 1990) recommend the use of the diachronous, flooding surface (FS), a lithostratigraphicsurface, as the boundary for a parasequence. Herein, the maximum regressive surface (MRS) which has verylow diachroneity and is a sequence stratigraphic surface, is recommended as the defining bounding surfacefor a parasequence. A unit bound by maximum flooding surfaces (MFS) has already been defined andnamed a genetic stratigraphic sequence.

really are and how they can be objectivelydelineated. Additional confusion has resultedfrom cases where an MFS coincides with thelithological change from sandstone to shale.

Redefining a Parasequence as a SequenceRedefining a Parasequence as a SequenceRedefining a Parasequence as a SequenceRedefining a Parasequence as a SequenceRedefining a Parasequence as a SequenceStratigraphic UnitStratigraphic UnitStratigraphic UnitStratigraphic UnitStratigraphic UnitThe parasequence is a widely used unit insequence stratigraphic analysis despiteuncertaint ies concerning boundaryplacement and consequent variations in use.To rectify this confusing situation, it is

necessary to define a parasequence usingbona fide sequence stratigraphic surfaces forits defining boundaries. The schematic crosssection of Figure 12.4 i l lustrates twodifferent placements for a parasequenceboundary. Van Wagoner et al. (1988, 1990)put the boundary at the diachronous, facieschange from sandstone to shale (floodingsurface) whereas, I would contend a betterplacement is at the maximum regressivesurface (MRS), which is a sequencestrat igraphic surface with very low

Page 61: Embry Sequence Stratigraphy

59

diachroneity. As shown in Figures 12.5 and12.6, the MRS occurs below an FS. Figure12.7 illustrates an outcropping parasequencewhich has MRSs for boundaries. The highlyburrowed, massive sandstone that wasdeposited during transgress ion anddeepening is placed at the base of theparasequence rather than at the top asproposed by Van Wagoner et al. (1988, 1990).Arnott (1995) also reached the sameconclusion that a parasequence boundary isbest placed at the base of such transgressivestrata.

I suggest a parasequence be defined as “asmall-scale, sequence stratigraphic unitbound by maximum regressive surfaces(MRS) and their correlative surfaces”.Because an MRS is an accepted, material-basedsurface of sequence stratigraphy (Embry,2002, 2008), this change ensures that aparasequence is a bona f ide sequencestratigraphic unit. Furthermore, such adefinition does not alter the basic meaningor utility of a parasequence and matches hownumerous practitioners have already appliedthe term (e.g., Arnott, 1995).

Parasequence vs. SequenceParasequence vs. SequenceParasequence vs. SequenceParasequence vs. SequenceParasequence vs. SequenceAs shown on Figure 12.8 (page 60), aparasequence is delineated and extended byrecognizing MRSs and correlating them.

Figure 12.6. A portion of Figure 7 of Van Wagoner et al. (1990) showing the Van Wagoner et al. (1990)parasequence boundaries (FSs in blue) versus the recommended MRSs in red. Modified after a portion of VanWagoner et al. (1990) Figure 7.

Figure 12.7. A parasequence is delineated in outcropping Bird Fiord Formation (Middle Devonian) strata onBathurst Island, Canadian Arctic Archipelago. The boundaries are placed at readily recognizable maximumregressive surfaces (MRS). Tom Oliver for scale.

Sometimes a maximum flooding surface(MFS) can replace an MRS and in this situationthe MFS would act as the parasequenceboundary because it would be a correlativesurface of the MRS. If both bounding MRSs

were everywhere replaced by MFSs, theresultant unit would be a geneticstratigraphic sequence (Embry, 2009a) ratherthan a parasequence.

An MRS often correlates with anunconformable shoreline ravinement (SRU)on the basin flanks and thus an SR-U can actas one of the bounding surfaces of aparasequence . However, when bothbounding MRSs of a previously delineatedparasequence can be shown to be correlativewith SR-Us (Figure 12.9, page 60), then theunit must be designated as a depositionalsequence rather than a parasequence. Thus aparasequence can be seen as a “depositionalsequence in waiting” and any parasequencewhich has been recognized is potentially adepositional sequence upon subsequentwork and extended correlation.

Given the above, a case can be made fordropping the term parasequence all togetherand including a unit bounded by MRSs withinthe definition of a depositional sequence.Hopefully, this question will be decided overthe next few years.

Parasequence and ScaleParasequence and ScaleParasequence and ScaleParasequence and ScaleParasequence and ScalePart of the proposed definition of aparasequence limits the term parasequenceto small-scale units (< tens-of-metres thick)and this is dictated by current practice. Large-scale (hundreds-of-metres thick) sequencestratigraphic units bound by MRSs are bestdesignated as depositional sequences. Thereason for this is almost all larger magnitudeMRSs correlate back to unconformities (SR-Us, SUs) on the basin margin.

Page 62: Embry Sequence Stratigraphy

60

Final remarksFinal remarksFinal remarksFinal remarksFinal remarksThe term parasequence is best appliedexclusively to small-scale, transgressiveregressive units bound by MRSs and theircorrelative surfaces. Parasequences areformed either during a small-scale, base-levelfall-rise cycle (correlative SR-U not yetrecognized) or during a reduct ion insediment supply during base-level rise (no

SR-U generated). Finally, I recommend thatthat a flooding surface (a lithostratigraphicsurface between a marine sandstone /limestone below and a shaly lithology above)be allowed as a proxy for a parasequenceboundary when available data do not allowthe MRS to be reliably or easily delineated.However, I would emphasize that it isdesirable to use MRSs whenever possible.

ReferencesReferencesReferencesReferencesReferencesArnott, R. W. 1995. The parasequence definition –are transgress ive depos i ts inadequate lyaddressed? Journal of Sedimentary Research, v.65, p. 1-6.

Catuneanu, O. 2006. Principles of SequenceStratigraphy. Elsevier, New York, 375 p.

Embry, A. F. 2002. Transgressive-Regressive (T-R)Sequence Stratigraphy, In: Sequence stratigraphicmodels for exploration and production. J.Armentrout and N. Rosen (eds.). Gulf Coast Societyof Economic Paleontologists and MineralogistsConference Proceedings, Houston, p. 151-172.

Embry, A. F. 2005. Parasequences in ThirdGeneration (3G) Sequence Stratigraphy. Searchand Discovery Article #110022. http://www.searchanddiscovery.net/documents/2005/av/embry/softvnetplayer.htm.

Embry, A. F. 2008. Practical Sequence StratigraphyVI: The Material-based Surfaces of SequenceStratigraphy, Part 2: Shoreline ravinement andMaximum Regressive Surface. Canadian Societyof Petroleum Geologists. The Reservoir, v. 35, issue9, p. 32-39.

Figure 12.8. As shown on this schematic cross-sec-tion, parasequences are delineated and correlatedon the basis of maximum regressive surfaces (MRS).Correlative surfaces of an MRS often include a maxi-mum flooding surface and a shoreline ravinementand these sequence stratigraphic surfaces can alsobe used as a parasequence boundary.

Figure 12.9. When both maximum regressive surface (MRS) boundaries of a previous delineated parasequence(e.g., see Figure 8) are found to correlate with unconformable shoreline ravinements (SR-U), as illustrated onthis schematic cross-section, the unit becomes a depositional sequence. Thus a parasequence can be re-garded as a “depositional sequence in waiting.”

Embr y, A . F. 2009a. Pract ica l SequenceStrat igraphy IX: The Uni ts of SequenceStratigraphy, Part 1, Material-based Sequences.Canadian Society of Petroleum Geologists. TheReservoir, v. 36, issue 2, p. 23-29.

Embr y, A . F. 2009b. Pract ica l SequenceStrat igraphy X: The Uni ts of SequenceStratigraphy, Part 2, Time-based Sequences.Canadian Society of Petroleum Geologists. TheReservoir, v. 36, issue 3, p. 21-24.

Embr y, A . F. 2009c . Pract ica l SequenceStrat igraphy XI : The Uni ts of SequenceStratigraphy, Part 3: Systems Tracts. CanadianSociety of Petroleum Geologists. The Reservoir, v.36, issue 4, p. 24-29.

Van Wagoner, J. C., Posamentier, H. W., Mitchum,R. M., Vail, P. R., Sarg, J. F., Loutit, T. S., and Hardenbol,J. 1988. An overview of the fundamentals ofsequence stratigraphy and key definitions. In: Sealevel changes: an integrated approach. C. Wilgus,B. S. Hastings, C. G. Kendall, H. W. Posamentier, C.A. Ross, and J. C. Van Wagoner (eds.). Society ofEconomic Paleontologists and Mineralogists,Special Publication 42, p. 39-46.

Van Wagoner, J. C., Mitchum, R. M., Campion, K.M., and Rahmanian, V. D. 1990. Siliciclasticsequence stratigraphy in well logs, cores andoutcrops. American Association of PetroleumGeologists. Methods in Exploration, no. 7, 55 p.

Page 63: Embry Sequence Stratigraphy

61

Practical Sequence Stratigraphy XIIIPractical Sequence Stratigraphy XIIIPractical Sequence Stratigraphy XIIIPractical Sequence Stratigraphy XIIIPractical Sequence Stratigraphy XIIISequence Stratigraphy Hierarchy by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionThe various surfaces and units of sequencestratigraphy have been described in theprevious articles of this series. It is mostimportant that sequence stratigraphicsurfaces be assigned to a hierarchy i fnumerous sequence stratigraphic surfacesare used for regional correlation or ifindividual sequences are delineated andmapped (Embry, 1993, 1995). The main reasonfor this is that very many sequencestratigraphic surfaces of greatly varyingmagnitude occur in a given succession and,without a hierarchy, any two recognizedsequence boundaries, regardless of theirmagnitude, (e.g., two MFSs in the case ofgenetic stratigraphic sequences and anycombination of two SU, SR-U, or MRSs in thecase of a depositional sequence) could, intheory, be used to form the boundaries of asequence (Figure 13.1). This would result ina huge number of potential sequences andthe only way to escape such chaos is toestablish a hierarchy of surfaces.

It is widely recognized that there is a greatvariation in the magnitude of sequencestratigraphic surfaces and that there is a needto separate large magnitude sequences /sequence boundaries from much smaller-scale ones. This is a natural consequence ofthe recognition that sequence boundariesand the enclosed sequences are not scaledependent. Notably, two very differentmethodologies for developing such ahierarchy of sequences and sequenceboundaries have been proposed – atheoretical, model-driven method and anempirical, data-driven method.

Model-driven HierarchyModel-driven HierarchyModel-driven HierarchyModel-driven HierarchyModel-driven HierarchyThe model-driven approach has beenchampioned by Exxon scientists (e.g., Vail etal., 1977; Mitchum and Van Wagoner, 1991;Vail et al., 1991; Posamentier and Allen, 1999).Such an approach is based on the hypothesisthat sequence stratigraphic surfaces aregenerated by eustasy-driven, sinusoidal base-level changes and that such eustatic cyclesincrease in amplitude with decreasingfrequency. Thus, very large amplitude changes,driven by tectono-eustasy (changes in volumeof ocean basins), occur rarely and theresulting sequence boundaries are assignedto either a 1st or 2nd order category. Suchorders are usually referred to as low-orderboundaries although a few authors refer tosuch boundaries as high-order boundaries. I

Figure 13.1. Because a depositional sequence is de-fined as a unit bound by subaerial unconformitiesand correlative surfaces, it is essential that a hierar-chy of sequence boundaries be defined. If a hierar-chical system is not used, up to 45 sequences couldbe defined in the above schematic succession with10 recognized sequence boundaries (e.g., 1-2 1-3,1-4, 1-5, etc.). Such a chaotic and unacceptablesituation is avoidable only by separating the se-quence boundaries into different classes (orders)and arranging them into a hierarchy.

follow the practice of referring to 1st, 2ndand 3rd order boundaries as low-orderboundaries and 4th, 5th and 6th orderboundaries as high-order boundaries.

In the model-driven hierarchy, high-orderboundaries are related to climate-driven,Milankovitch cycles, which drive high

frequency, eustatic changes in the 20,000 yearto 400,000 year band. In such a model-drivenapproach, a sequence is assigned to a givenorder based on the amount of t imerepresented by the sequence – that is, theamount of time which lapsed between thedevelopment of each of its bounding surfaces.

This model-driven approach culminated in apublication by Vail et al. (1991) in which sixorders of boundaries were defined solelyon boundary frequency. The six orders andtheir characteristic boundary frequencies inthis hierarchical scheme are:

1st order – >50 MA2nd order – 3-50 MA3rd order – 0.5-3 MA4th order – 0.08-0.5 MA5th order – 0.03-0.08 MA6th order – 0.01-0.03 MA

Such a model-driven approach to establishinga hierarchy of sequences is highly prone tocircular reasoning. Because any givenstratigraphic section contains numerousdeposit ional sequence boundaries(unconformities and MRSs), any desiredfrequency of boundary occurrence can bedetermined simply by selecting only theboundaries that fit the desired result. Forexample, if fourteen sequence boundarieswere recognized within a succession spanning20 MA, there are many combinations ofboundaries that could be chosen to delineatea sequence with a boundary frequency of 10MA (Figure 13.2a, page 62). As shown inFigure 2b, Haq et al. (1988) applied such amethodology for the delineation of 2ndorder cycles on their global sequence charts.The boundaries of the second order cycles(sequences) on the charts have beensubjectively selected to fit the desired result(sequence duration of ~ 10 MA).

The fundamental f law of the above,modeldriven methodology is that you can’tdetermine the frequency of occurrence ofan entity or a phenomenon until you have aclear definition of the entity or phenomenon.It simply comes down to the premise that, ifone wants to determine the frequency of2nd order sequence boundaries, one mustbe able to empirically recognize 2nd orderboundaries in the first place. Boundaryfrequency is a conclusion that can be only bereached once the di f ferent orders ofboundaries are defined with reasonable

Page 64: Embry Sequence Stratigraphy

62

objectivity. Duration is not an observablecharacteristic of a sequence.

Data-driven HierarchyData-driven HierarchyData-driven HierarchyData-driven HierarchyData-driven HierarchyEmbry (1993, 1995) advocated for the use ofan empirical, data-driven methodology forestabl ishing a hierarchy of sequencestratigraphic boundaries and enclosed units.Such an approach is based on reasonablyobjective scientific criteria rather than on apriori assumptions and untested hypotheses,as is the case for the model-driven approachdescribed above.

In the data-driven approach, a hierarchy ofboundaries is established on the basis of theinterpreted relat ive magnitude of theboundaries . The interpreted relat ivemagnitude of a boundary would reflect themagnitude of base-level shift that generatedthe boundary in the first place. A base-levelchange of 500 m is going to result in arelatively large magnitude sequence boundarythat has different attributes than a smallermagnitude sequence boundary that wasgenerated by a base-level change of 10 m orless. In a given basin, the largest magnitudeboundaries (i.e., the sequence boundariesgenerated by the largest interpreted base-level changes) are assigned to the 1st ordercategory in the hierarchy and the smallestmagnitude boundaries recognized (i.e., thosegenerated by the smallest interpreted base-level changes) would be assigned to thehighest order established (e.g., 4, 5, or 6).

To apply such a methodology, it is necessaryto find observable, scientific criteria whichallow the characterization of the relativemagnitude of a sequence boundary. Suchcriteria would reflect the magnitude of thebase-level change which generated theboundaries. The attributes of a sequenceboundary I have found useful to estimate therelative magnitude of a sequence boundary,and indirectly the amount of base-levelchange that generated the boundary in thefirst place, are listed below. Such observablecharacteristics are placed in order of theirimportance for assessing the relat ivemagnitude of a given depositional sequenceboundary with the first one being mostimportant.

1) The degree of change of the tectonicsetting across the boundary.

2) The degree of change of thedepositional regime and sedimentcomposition across the boundary.

3) The amount of section missing belowthe unconformity at as many localitiesas possible. Localities close to thebasin edge are very helpful.

4) The estimated amount of deepening atthe maximum flooding surface abovethe sequence boundary where it is anunconformity.

5) How far the subaerial unconformityand associated shoreline faciespenetrate into the basin.

It is important to note that not all thesecharacteristics can be applied for eachboundary, but in many cases most of themcan be. In many instances, the largestmagnitude boundaries in a basin, whichwould be 1st order boundaries for that basin,mark a significant change in tectonic andsedimentary regime and are associated withlarge amounts of erosion and significantdeepening. The unconformity and shorelinefacies usually penetrate far into the basin.Such sequence boundaries are most oftenreadily apparent and correlatable and wouldbound 1st order depositional sequences.Because of the tectonic and sedimentaryregime changes, there is little doubt that suchboundaries were generated by tectonics(Figure 13.3) and denote very large, base-level changes both within the basin and inthe surrounding hinterlands.

2nd order boundaries also mark a change inthe tectonic and depositional regimes, aswell as large changes in base level. Similar to1st order boundaries, they are mainly drivenby tectonics as evidenced by the tectonicregime change across the boundary. Evidenceof tectonic movements including faulting,folding, and tilting can often be discernedbeneath 1st and 2nd order boundaries. 2ndorder boundaries differ from 1st order onesin that the amount of base-level change isdistinctly less as evidenced by less erosionand basinward penetrat ion of theunconformities. Also the magnitude oftectonic regime change is significantly less.The separation of 1st order boundaries from2nd order ones can be somewhat subjective,but in most instances the two orders can beconsistently differentiated from each otherwithin a basin.

3rd order boundaries exhibit no tectonicregime change but do have a noticeablechange in sedimentary regime across them.Once again it is likely that tectonics is themain driver of 3rd order boundaries becauseit is very difficult to explain the noticeablechange in sedimentary regime on the basisof eustasy. The amount of erosion and basinpenetration of the unconformable portionsof 3rd order boundaries, as well as thesubsequent deepening during the followingtransgression, are less than that for 1st and2nd order boundaries (Figure 13.3). Notably1st, 2nd, and 3rd order boundaries can usuallybe correlated throughout most or all of abasin and are the main ones recognized onseismic sections.

Sequence boundaries which exhibit nochange in tectonic or depositional regime,are associated with l itt le erosion andsubsequent drowning, and the unconformityand shoreline facies do not extend past thebasin margin, would be high-order, low-

Figure 13.2a. A schematic diagram illustrating the faulty logic of using boundary frequency for the establish-ment of a sequence hierarchy. In a succession which spans 20 Ma and contains 14 sequence boundaries,many combinations of boundaries (only three shown) can be used to create second-order sequences with a10 MA year frequency.

Figure 13.2b. A diagram from Haq et al. (1988) that illustrates the use of the concept of boundary frequencyto define a series of second-order sequences (LZB-2, etc.). The boundaries of the second-order sequenceshave been subjectively selected to fit the desired result (sequences of 10 Ma duration).

Page 65: Embry Sequence Stratigraphy

63

magnitude boundaries (e.g., 4th, 5th, and 6thorder) (F igure 13.3) . Parasequenceboundaries would constitute the highestorder, lowest magnitude boundaries in thehierarchy and they reflect little to no base-level change. Correlation of these high-orderboundaries is usually limited to local areaswith widespread correlation being possibleonly if control points are close and numerous(e.g., WCSB).

For sequence boundaries generated during“Greenhouse” conditions (i.e., no continentalglaciers), there tends to be a consistency foreach of the five criteria to point to the sameresult in regards to assignment of an order toboundaries. In these cases, the magnitude ofthe boundary correlates closely with thebasinward extent the unconformity, with theamount of section eroded and with the amountof subsequent deepening. Those largemagnitude boundaries in which theunconformity extends far into the basin andfor which significant erosion and subsequentdrowning are present almost always have asignificant change in depositional regime, if notalso tectonic regime.

Problems with assignment sometimes occurfor sequence boundaries formed during“Icehouse” conditions when continentalglaciers were intermittently present. Duringsuch times, relatively large base-level changes(up to ~ 120 m) due to climate-driven,eustatic sea level changes were oftenaccompanied by essentially no change indepositional and tectonic regimes. In general,changes in these latter two criteria mostoften ref lect major base- level-changeepisodes and should be used as the finalarbiters for recogniz ing the greatestmagnitude, low-order boundaries. Thus aboundary with a substantial amount of changeof depositional regime and /or tectonicregime would be ranked higher (lowerorder) than one with no change in theseregimes, even if it seemed that the one withno regime change had similar properties onthe basis of the last three criteria.

It must be emphasized that for each basinthe interpreter must establish his or herown hierarchy based on the listed criteria.Thus there is no characteristic, generic first-order sequence boundary that can be defined.First-order boundaries in a given study arethose that are interpreted to have the largestmagnitude in the basin. Thus, a first-orderboundary recognized in one basin may besomewhat different from a first-orderboundary recognized in another. Once ahierarchy has been established for a basin,that is, each recognized order has beenassigned a specific set of characteristics, theassignment of a given boundary to a givenorder can involve some subjectivity but in

most cases can be done with reasonableconsistency and objectivity.

This methodology emphasizes theestablishment of a hierarchy based on theinterpreted relat ive magnitude of thedepositional sequence boundaries. Thus, ifone wants to establish a hierarchy forsequences rather than boundaries, thevarious sequence boundaries must be rankedfirst. The order of a sequence is equal to theorder of its lowest magnitude (highestorder) boundary. Thus a sequence with afourthorder boundary at the base and a first-order boundary on top is a fourth-ordersequence.

This brings us back to our original problemof trying to avoid a chaotic delineation ofsequences in a succession with multiple

sequence boundaries of varying magnitude.With the establishment of a hierarchy ofsequence boundaries as described above,one simple rule of hierarchies now allowsus to recognize a sensible and orderlysuccession of sequences. This rule states thata sequence cannot contain within it asequence boundary that has an equal orgreater magnitude (equal or lower order)than that of its lowest magnitude (highestorder) boundary. For example, a second-order sequence cannot contain a second- orfirst-order boundary. It can contain manyhigher order (3rd-6th) boundaries. This is ofmost importance and is the only way that anorderly delineation of sequences can beproduced (Figure 13.4, page 64).

Figure 13.5 (page 69) illustrates an outcropof Lower to Upper Triassic strata on the

Figure 13.3. A schematic depiction of five orders of sequence boundaries which are determined by observ-able criteria, such as degree of tectonic regime change and penetration of the unconformity into the basin.The criteria needed to build such a hierarchy must reflect the amount of base-level change which resultedin the formation of the sequence stratigraphic surfaces.

Page 66: Embry Sequence Stratigraphy

64

eastern flank of the Sverdrup Basin of ArcticCanada along the north side of Greely Fiord,Ellesmere Island. Three large-magnitude, 2ndorder depositional sequence boundaries arepresent and consist of prominent,unconformable shoreline ravinements. Theyseparate sequences which have differenttectonic and depositional regimes and theyrecord major falls of base level followed bylarge rises. In various areas of the basin,tectonic tilting is present beneath theseunconformities (Embry, 1991, 1997). A third-order boundary is delineated in the LowerTriassic 2nd order sequence and it separatesred-weathering, fluvial strata of the Smithian3rd order sequence from the grey-weathering, shallow marine strata of theSpathian 3rd order sequence. Fourth- andfifth-order sequence boundaries (MRSs) can

be delineated in the 3rd order Spathiansequence (Figure 13.5).

Figure 13.6 illustrates an outcrop of theMiddle Triassic 2nd order sequence in theSverdrup Basin at the head of Otto Fiord onnorth-central Ellesmere Island. The sequenceis bound by two 2nd order boundaries, amaximum regressive surface at the base andan unconformable shoreline ravinement atthe top. Large base-level changes arerepresented by these boundaries and theyare readily correlated over the entire basin.On the basin margins they are associatedwith tectonic tilting. Substantial changes insubsidence rate (> 5X) also occurred acrossboth these large magnitude boundaries. A3rd order sequence boundary (MRS) occurswithin the 2nd order sequence and

subdivides it into two 3rd order sequences.Both 3rd order sequences contain 4th ordersequence boundaries which at this localityare maximum regressive surfaces.

F igure 13.7 (page 66) i l lustrates thecorrelation of various orders of sequenceboundaries in the Middle Triassic 2nd ordersequence of the Sverdrup Basin in thesubsurface of the Lougheed Island area.Notably this area is about 650 km southwestof the outcrop section of Figure 13.5 andthe same boundaries are present. Two 2ndorder unconformable shoreline ravinementsbound the Middle Triassic sequence and a 3rdorder unconformable shoreline ravinementoccurs within it. A number of 4th orderboundaries (MRSs) occur within each 3rdorder sequence.

SummarySummarySummarySummarySummaryIt is necessary to assign the recognizedsequence boundaries and other associatedsequence surfaces of a basin to a hierarchyso as to allow the delineation of variousorders of sequences. Such a hierarchy is bestgenerated by the use of observable criteriathat relate to the magnitude of base-levelchange that resulted in the generation of thesequence surfaces. The largest magnitudesequence boundaries within a basin areassigned to the first order and sometimesup to six orders of boundaries can bedetermined.

ReferencesReferencesReferencesReferencesReferencesEmbry, A. F. 1991. Mesozoic history of the ArcticIslands. In: Innuitian Orogen and Arctic Platform:Canada and Greenland. H.P. Trettin (ed.).Geological Survey of Canada, Geology of Canadano. 3 (also GSA, The Geology of North America, v.E), p. 369-433.

Embry, A. 1993. Transgressive-regressive (T-R)sequence analysis of the Jurassic succession ofthe Sverdrup Basin, Canadian Arctic Archipelago.Canadian Journal of Earth Sciences, v. 30, p. 301-320.

Embry, A. 1995. Sequence boundaries andsequence hierarchies: problems and proposals.In: Sequence stratigraphy on the northwestEuropean margin. R. Steel, et al. (eds.). NPF SpecialPublication 5, p. 1-11.

Embry, A. F. 1997. Global sequence boundaries ofthe Triassic and their recognition in the WesternCanada Sedimentary Basin. Bulletin CanadianPetroleum Geology, v. 45, p. 415-433.

Haq, B., Hardenbol, J., and Vail, P. 1988. Mesozoicand Cenozoic chronostratigraphy and cycles ofsea level change. In: Sea level changes: anintegrated approach. C. Wilgus, B. S. Hastings, C.G. Kendall, H. W. Posamentier, C. A. Ross, and J. C.

Figure 13.4. Principles of determining the order of a sequence based on hierarchical rules. A sequence cannotcontain a sequence boundary with the same or larger magnitude (same or lower order) as its smallestmagnitude (highest order) boundary. Thus a 2nd order sequence cannot contain any 1st or 2nd orderboundaries. In this way the chaos illustrated in Figure 13.1 can be avoided.

Page 67: Embry Sequence Stratigraphy

65

Figure 13.5. An outcrop of Triassic siliciclastic strata on Greely Fiord Ellesmere Island. Three 2nd order depositional sequence boundaries are present. These areunconformable shoreline ravinements across which there are major changes in tectonic and depositional regimes. Substantial erosion occurred beneath eachboundary. A 3rd order boundary occurs within the Lower Triassic 2nd order sequence and 4th and 5th order boundaries (MRSs) are indicated.

Figure 13.6. An outcrop of the Middle Triassic 2nd order sequence at the head of Otto Fiord, Ellesmere Island. The sequence is bound by a 2nd order boundary, amaximum regressive surface at the base, and an unconformable shoreline ravinement at the top. Once again significant changes in tectonic and depositional regimesoccur across these boundaries. A prominent 3rd order boundary (MRS) occurs with this 2nd order sequence and subdivides it into two 3rd order sequences. Smallermagnitude 4th order boundaries are delineated in each 3rd order sequence. Unlike the 2nd and 3rd order boundaries, no change in depositional regime occurs acrossthese 4th order boundaries.

Page 68: Embry Sequence Stratigraphy

66

Figure 13.7. The 2nd order unconformable shoreline ravinements that bound the Middle Triassic 2nd ordersequence are correlated between two wells in the Lougheed Island area of the western Sverdrup Basin usinggamma ray logs. This is the same sequence illustrated in Figure 13.6 but this locality is 650 km to thesouthwest. The 3rd order boundary within the sequence is a significant unconformable shoreline ravinementin contrast to the maximum regressive surface which formed this same 3rd order sequence boundary onFigure 13.6. Tilt related truncation is apparent on this boundary indicating that tectonics was the primarydriver of boundary formation. More subtle, lower magnitude, 4th order boundaries occur within both 3rdorder sequences.

Van Wagoner (eds .) . Society of EconomicPaleontologists and Mineralogists, SpecialPublication 42, p. 40-45.

Mitchum, R. and Van Wagoner, J. 1991. Highfrequency sequences and their stacking patterns:sequence strat igraphic evidence for highfrequency eustatic cycles. Sedimentary Geology,v. 70, p. 131-160.

Posamentier, H. and Allen, G. 1999. Siliciclasticsequence strat igraphy – concepts andapplications. Society of Economic Paleontologistsand Mineralogists Concepts in Sedimentology andPaleontology, # 7, 210 p.

Vail, P. et al. 1977. Seismic stratigraphy and globalchanges in sea level. In: Seismic stratigraphy:applications to hydrocarbon exploration. C .Payton (ed.). American Association of PetroleumGeologists Memoir 26, p. 49-212.

Vail, P. et al. 1991. The stratigraphic signatures oftectonics, eustasy and sedimentology – anoverview. In: Cycles and events in stratigraphy, G.Einsele, et al. (eds.), Springer-Verlag, New York, p.611-159.

Page 69: Embry Sequence Stratigraphy

67

Practical Sequence Stratigraphy XIVPractical Sequence Stratigraphy XIVPractical Sequence Stratigraphy XIVPractical Sequence Stratigraphy XIVPractical Sequence Stratigraphy XIVCorrelation by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionIn previous articles in this series, I havedescribed the various types of sequencestrat igraphic surfaces that have beenrecognized, as well as the different types ofsequence stratigraphic units that have beendefined on the basis of those surfaces.However, it must be emphasized that theprimary contribut ion of sequencestratigraphy to petroleum geology is that itprovides an excellent methodology forcorrelating strata and this topic is addressedherein.

Stratigraphic correlation is accomplished bymatching distinct stratigraphic surfaces orhorizons recognized in a stratigraphicsuccession at one locality to their equivalentcounterparts in a succession at anotherlocal i ty. This a l lows the extension ofrecognized stratigraphic units and surfacesinto new geographic areas and potentially toareas around the world.

One of the main goals of correlation is toestablish an approximate time-stratigraphiccorrelation framework so as to allow faciesrelat ionships to be determined andpredictions of facies occurrences to be made.Interpretations of depositional history and

paleogeographic evolution also depend uponsuch a framework built by the correlation ofstratigraphic surfaces that have a lowdiachroneity or are time barriers. Lowdiachroneity surfaces are often delineatedin biostratigraphy, magnetostratigraphy, andchemostratigraphy but such methods areoften not available for subsurface studies.Furthermore they can be very costly and timeconsuming.

Sequence stratigraphy is very useful forconstruct ing an approximate t ime-strat igraphic framework because, aspreviously described, a number of thesurfaces of sequence stratigraphy are eithertime barriers or have low diachroneity. Mostimportantly, sequence stratigraphy is readilyapplicable to subsurface studies and can bedone with seismic, well log, and / or coredatabases. In this article, the use of sequencestratigraphy for correlation is discussed anda number of examples of correlations usingsequence stratigraphy with well logs areprovided.

Sequence Stratigraphic Surfaces UsefulSequence Stratigraphic Surfaces UsefulSequence Stratigraphic Surfaces UsefulSequence Stratigraphic Surfaces UsefulSequence Stratigraphic Surfaces Usefulfor Correlationfor Correlationfor Correlationfor Correlationfor CorrelationAs discussed in previous articles, sequencestrat igraphic surfaces are those that

represent breaks in the stratigraphic recordor changes in depositional trend. Six,material-based surfaces have been definedand their relat ionships to t ime werediscussed in Embry 2008a, b, and c. Themateria l-based surfaces of sequencestratigraphy that are either time barriers orhave low diachroneity, and are thus usefulfor establishing a correlation framework,are:

• Subaerial unconformity (SU) (timebarrier),

• Unconformable shoreline ravinement(SR-U) (time barrier),

• Slope onlap surface (SOS) (timebarrier),

• Maximum regressive surface (MRS)(low diachroneity), and

• Maximum flooding surface (MFS) (lowdiachroneity).

The material-based surfaces of sequencestrat igraphy that are not useful forconstruct ing an approximate t imecorrelation framework are those that arevery diachronous. These are the regressivesurface of marine erosion (RSME) and thediastemic portion of a shoreline ravinement(SR-D) (Embry, 2008a, b). However, it is usefulto correlate such surfaces as part of thedelineation of facies distributions within thecorrelation framework.

As discussed in Embry (2009), two, time-based surfaces have also been defined as partof sequence strat igraphy a lthough areasonable argument can be made that suchsurfaces are much better assigned tochronostratigraphy rather than sequencestratigraphy. These time-based surfaces arethe basal surface of forced regression (BSFR),which equals the time surface at the start ofregional base-level fall, and the correlativeconformity (CC), which represents the timesurface at the start of regional base-levelrise. Like all time-based surfaces, thesesurfaces have no def in ing physicalcharacterist ics and thus their use forcorrelation is very limited. This assessmentis supported by the lack of any publicationsthat have used such surfaces for correlationof well log sections.

Correlating Shallow Marine StrataCorrelating Shallow Marine StrataCorrelating Shallow Marine StrataCorrelating Shallow Marine StrataCorrelating Shallow Marine StrataThe sequence stratigraphic model forsiliciclastics in a ramp setting (Embry, 2008d)is illustrated in Figure 14.1 and shows the

Figure 14.1. Sequence stratigraphic model for a siliciclastic ramp setting (Embry, 2008d). Note that the SR-U, MRS, and MFS all occur in shallow marine strata and these surfaces are excellent for correlation in suchstrata. Towards the basin margin, nonmarine strata become intercalated with the shallow marine strata andan SU and SR-D can also be delineated and correlated.

Page 70: Embry Sequence Stratigraphy

68

three surfaces of sequence stratigraphy thatare useful for the correlation of shallowmarine strata. These are the unconformableshorel ine ravinement, the maximumregressive surface, and the maximum floodingsurface. As shown on the model (Figure 14.1,page 67), the maximum flooding surface isoften very widespread and it is usually theeasiest surface to recognize and correlate.As discussed in Embry (2008c), the MFSrepresents the change from a finingupwardtrend to a coarsening-upward one, and ongamma logs is best placed at the highestgamma horizon unless higher resolution data(e.g., core) support a different placement.

Between every two MFSs in shallow marinestrata, there will be either an MRS or anSRU. The maximum regressive surface marksthe change from a coarsening-upward trendto a fining-upward one and on gamma logs itis best placed at the lowest gamma horizon(Embry, 2008b) unless, once again, moredetailed data indicate a different placement.As seen on Figure 14.1 (page 67), the MRScorrelates laterally to an unconformableshorel ine ravinement (SR-U) and, incombination, these two surfaces allow thedelineation of a widespread correlativehorizon. On gamma logs, an SR-U is oftenmarked by an abrupt contact, overlain by afining-upward (increasing gamma) succession(Figure 8 in Embry 2008b). Confirmation ofthe existence of an SR-U requires thedemonstration of truncated strata below it.

Figure 14.2 is a three-well, gamma-raycrosssection of the lower Charlie LakeFormation (Upper Triassic) in northwesternAlberta and the data were kindly supplied bymy colleague, Jim Dixon. The Charlie LakeFormation consists mainly of interbedded

Figure 14.2. Stratigraphic cross-section of lower portion of Charlie Lake Formation, northwest, Alberta. Thedatum is the base of the Charlie Lake, a lithostratigraphic surface. Maximum regressive surfaces, maximumflooding surfaces, and one unconformable shoreline ravinement have been correlated. The SR-U truncatesstrata eastwards and very minor onlap occurs above it. Data courtesy of J. Dixon.

thinning of the section between the SR-Uand the overlying MFS.

Figure 14.3 illustrates a stratigraphic cross-section of Upper Triassic, shallow marinestrata on the southern flank of the SverdrupBasin in the Melville Island area of ArcticCanada. In this case, the wells are muchfarther apart than the previous example andmore section is present (300 m versus 50m). Because of this, only large-magnitudesurfaces have been correlated, although thereare opportunities for correlating smaller-scale surfaces. The datum is a prominentunconformable shoreline ravinement (SR-U)near the top of the succession and it passesbasinward into a readily recognizable MRSthat separates two distinctly differentdepositional regimes (2nd order boundary).Because this surface was essentia l lyhorizontal when it was formed (shorefaceerosion at sea level), it makes a very gooddatum.

Once again, MRSs and MFSs are correlatedon the basis of gamma ray signature andsample descript ions. Unconformableshoreline ravinements are delineated wheretruncation can be demonstrated. Some minordepositional thickening for individual unitsis visible downdip, but notably, most changesin thickness are due to the effects ofmarginward truncat ion beneath theunconformities. This indicates that theunconformities were generated by tectonicmovements rather than by eustasy. This topicwill be more fully explored in the next article.

Figure 14.4 also illustrates shallow marinestrata in a ramp setting (Lower Jurassic,western Sverdrup Basin) and in this case,there is a large distance between the controlpoints and the line of section is close to thedirection of depositional dip. In sectionsparallel to depositional strike or those whichextend only a short distance down dip (e.g.,Figures 14.2, 14.3), depositional dip has littleto no effect on stratal geometry of the largermagnitude surfaces. However, in this case,depositional dip is a significant factor in thegeometry of the correlated surfaces.

The succession portrayed in Figure 14.4 isconformable with the only unconformitypresent being an SR-U at the base of thesuccession. A prominent maximum floodingsurface (3rd order) is used as a datum forlack of a better one. It must be kept in mindthat this MFS datum was not a horizontalhorizon at the time of formation but slopedbasinward, approximating the sea floor dip.Using an originally sloping surface as ahorizontal datum will distort original stratalgeometries somewhat.

The larger-magnitude MRSs and MFSs are

shale, siltstone, and sandstone with lesscommon limestone and anhydrite . Thedepositional setting was a shallow, restrictedseaway and individual units can be correlatedover large areas (J. Dixon, pers. comm., 2007).

The datum for this cross-section is alithostratigraphic one, the base of the CharlieLake Formation, which is marked by a unit ofanhydrite overlying a sandstone unit at thetop of the Hal fway Formation. Alithostratigraphic contact is usually not thebest choice for a datum because of thepotential high diachroneity of such a surfacebut it is certainly objective. In this case, thiscontact appears to have low diachroneity asdemonstrated by its near parallelism with aneasily correlatable MRS about 10 m above. Ihave correlated this cross-section mainlywith MRSs and MFSs delineated on the basisof gamma ray signature.

In most cases, where control is very closeand a cross-section is not long, the MFSs andMRSs will parallel each other becausedifferences in subsidence rates tend to bevery small over short distances. The presenceof an unconformity is suspected when twodifferent sets of parallel MRSs and MFSs arepresent and are at any angle to each other.On this basis , I have interpreted theoccurrence of an unconformable shorelineravinement (SR-U) beneath a sharp-based,fining-upward, limestone unit (informallycalled the “A marker”) (Figure 14.2). Thisinterpretation is supported by the truncationof an MRS and the progressive eastwardthinning of the section between the “Amarker” and the first correlatable MRS abovethe datum. All the correlated surfaces abovethe SR-U nearly parallel it and minor,eastward onlap is expressed as a slight

Page 71: Embry Sequence Stratigraphy

69

Figure 14.3. Stratigraphic cross-section of Upper Triassic (Carnian) strata on the southwest flank of theSverdrup Basin, Melville Island, Arctic Canada. Only large-magnitude sequence stratigraphic surfaces havebeen correlated and a prominent SR-U that caps the Carnian succession is used as the datum. An SR-Uforms the base of the succession and two SR-Us occur within. MRSs and MFSs are truncated by the SR-Usand the lowermost Carnian sequence is absent in the well on the basin edge (Hecla-C-32).

Figure 14.4. Stratigraphic cross-section of Lower Jurassic strata in the southwestern Sverdrup Basin. Thedatum is a prominent MRS near the top of the Lower Jurassic (late Toarcian). The cross-section is dip orientedand extends for over 100 km. Large-magnitude MRSs and MFSs have been correlated and they all dipbasinward, approximating the original sea floor dip. Sandstone units found in the eastern two wells “shale-out”basinward beneath the MRSs.

correlated on Figure 14.4 and any smaller-scale correlation is precluded by the largedistances between control points. Thecorrelatable surfaces approximate thedipping sea f loor at the time of theirformation and thus diverge from the datumbecause of the greater water depths to thewest. The sandstones that underlie the MRSsin the east change facies to shale and siltstonebasinward as water depth increased. The firstMFS below the datum is well characterizedon the sonic log by a very slow travel time(high clay content). This surface can be readilyrecognized throughout the basin and marksthe height of a major transgression in theearly Toarcian (a global event).

In summary, maximum flooding surfaces,maximum regress ive surfaces andunconformable shorel ine ravinementsurfaces are ideal surfaces for correlation inshallow marine strata. Various orders of thesesurfaces are usual ly present and theloworder, high-magnitude surfaces are theeasiest to correlate . High-order, low-magnitude surfaces can be correlated ifreasonably close control is available.

Correlating Interbedded Nonmarine andCorrelating Interbedded Nonmarine andCorrelating Interbedded Nonmarine andCorrelating Interbedded Nonmarine andCorrelating Interbedded Nonmarine andShallow Marine StrataShallow Marine StrataShallow Marine StrataShallow Marine StrataShallow Marine StrataAs shown of Figure 14.1 (page 67), whennonmarine strata are intercalated withshallow marine strata on the basin margins,the potential for the recognition of subaerialunconformities (SU) and diastemic shorelineravinements (SR-D) exists. The reason forthis is that the occurrence of nonmarinestrata either directly above (SU) or directlybelow (SR-D) is one of the def in ingcharacteristics of these surfaces. If nonmarinestrata are not present in a succession, thenSUs and SR-Ds cannot be delineated.

Figure 14.5 (page 70) is a cross-section of aninterval of Lower Cretaceous, interbeddedmarine and non-marine strata of the lowerportion of the Isachsen Formation onEglinton Island, Arctic Canada (SverdrupBasin). Prominent subaerial unconformitiesare delineated beneath very coarse-grained,fluvial channel deposits and the basal one isused as the datum. This SU overlies offshoremarine strata and is a major 1st orderboundary in the basin. The SU’s allow thelower Isachsen Formation to be subdividedinto two depositional sequences.

MFSs can be delineated and correlated in themarine interval of each sequence and thesesurfaces subdivide each sequence into alower transgressive systems tract (TST) andan upper regressive systems tract (RST). Thecontact between nonmarine strata below andmarine strata above occurs within the TST ofeach sequence and is a diastemic shorelineravinement. The highly diachronous nature

Page 72: Embry Sequence Stratigraphy

70

Final ly, i t is wor th noting that thestratigraphic interval between the SR and theoverlying MFS is thicker and sandier whereit overlies the incised valley strata wheresupply was greater.

Correlation in Fluvial StrataCorrelation in Fluvial StrataCorrelation in Fluvial StrataCorrelation in Fluvial StrataCorrelation in Fluvial StrataCorrelation with sequence stratigraphy insuccessions of fluvial strata that have nomarine intercalations can be difficult. Theonly sequence stratigraphic surface that iscommon is the subaerial unconformity thatoccurs at the base of channel deposits or atthe top of paleosols. It is difficult to correlatesuch subaeria l unconformit ies withconfidence and it is often even harder toestablish a hierarchy of surfaces. It is alsoimportant to distinguish between subaerialunconformities that are regional truncationsurfaces and subaerial diastems (channelscours) that are the product of rivermigration during rising base level.

A subaerial unconformity at the base of anincised valley represents a regional base-level fall and is likely a large-scale sequenceboundary. It must correlate with a soilhorizon in the interfluve areas although it isusually very difficult to establish such acorrelation without excellent control. Agood example of such work is McCarthy andPlint (1998) who correlated subaerialunconformities in a well exposed successionof channel deposits and overbank strata withsoil horizons. Their work demonstrates theneed for very close control for suchcorrelations.

It is important to try to recognize andcorrelate the large-scale subaeria lunconformities that may be present. Theseare sometimes associated with a significantchange in grain composition and / or clastsize. Other stratigraphic data such aschemostratigraphic and magnetostratigraphicdata can be integrated to help identify SUs.Zaitlin et al. (2002) and Ratcliffe et al. (2004)provide a solid example of identifyingregional SUs in a fluvial succession throughthe use of changes in mineralogical andchemical composition.

Maximum flooding surfaces can sometimesbe tentatively determined in fluvial strata andmay be represented by a horizon thatexhibits a marine influence (e.g., brackishwater facies). In absence of any indication ofa marine influence, an MFS in fluvial stratacan be delineated with reasonable objectivitywithin the interval with the highest ratio ofoverbank fines to channel sandstone (i.e.,the change from fining-upward to coarsening-upward). This interval also tends to beassociated with the thickest and mostcommon coal seams. Overall, widespreadMFSs do not appear to be common in fluvial

Figure 14.5. Stratigraphic cross-section of Lower Cretaceous strata (lowerIsachsen Fm) from the southwestern flank of the Sverdrup Basin, EglintonIsland, Arctic Canada. The succession consists of intercalated nonmarine andshallow marine strata. A 1st order SU at the base of the succession is used as thedatum. The delineation of two other SUs allows two depositional sequences to bedefined. The correlation of an MFS within the marine strata of each sequenceallows each sequence to be subdivided into a transgressive systems tract (TST)and a regressive systems tract (RST). A diastemic shoreline ravinement (SR-D)occurs within each TST at the boundary between the nonmarine strata andoverlying marine strata. Because of the highly diachronous nature of the SR-Ds(climbs stratigraphically), such surfaces are not used as part of the time corre-lation framework or as a system tract boundary.

of such a SR-D is well illustrated by the upperone which climbs stratigraphically upward(i.e., becomes younger) towards the moremarginal well on the left (Figure 14.5). Notethat the correlated MFSs essentially parallelthe datum and each other (i.e., very lowdiachroneity) and no depositional dip isdiscernable for the MFSs on this section.

Figure 14.6 illustrates the correlatablesurfaces associated with an incised valleydeposit that can be considered as a plum ofnonmarine strata in a pudding of shallowmarine deposits. The cross-section consistsof four, reasonably close wells in southernSaskatchewan and includes the transitionfrom the nonmarine Mannville Group at thebase to the deep shelf, marine shales of theLower Colorado Group at the top (logs andfacies interpretation from O. Catuneanu).These strata are mainly shallow marinesandstone, siltstone, and shale but an intervalof fluvial sandstone occurs in two wells. A

prominent MRS hasbeen chosen as thedatum and it isoverlain by an easilypicked MFS. Thesurface at the top ofthe nonmarineMannville strata is adiastemic shorelineravinement (contactof marine strataoverlying nonmarinestrata).

The presence of theisolated pod ofn o n m a r i n es a n d s t o n ecompl icates anotherwise standardcorrelation of MRSsand MRSs. As u b a e r i a lunconformity (SU)must be placed atbase of thenonmarine strata toexplain their isolatedoccurrence . Adiastemic shorelineravinement (SR-D)once again occurs atthe contact betweenthe fluvial strata andthe overlying marinestrata . Because as h o r e l i n eravinement is usuallynot an isolatedsurface, i t isreasonable, i f notmandatory, to extendthe SR that occurs

on top of the fluvial strata into the adjacentmarine strata. The SR in the marine stratawould be a significant unconformity (SR-U)(SU eroded) as opposed to being a minordiastem as it is when it overlies the fluvialstrata. Its placement in the marine strata isguided by the constraints that it should be atapproximately the same stratigraphic levelas the SR-D (SR is close to a horizontalsurface) and it should occur at the base of afining-upward succession.

As shown on Figure 14.6, the SR-U in themarine strata might have otherwise beeninterpreted as an MRS if the control pointswith the fluvial strata and the accompanyingSR-D were not available. Conversely, if anSR-U is interpreted to occur in a section ofshallow marine strata (e.g., Figures 14.2,14.3, and 14.4), then the occurrence ofincised valley, nonmarine deposits, whichstratigraphically hang down from the SR, is apotential exploration target for that area.

Page 73: Embry Sequence Stratigraphy

71

strata and this is supported by the generallack of regionally correlatable seismicreflectors in such strata.

In summary, SUs are the main sequencestrat igraphic surfaces avai lable forcorrelation in fluvial strata. However, it isusual ly very dif f icult to del ineate andcorrelate such surfaces and close controland data from other stratigraphic disciplinesare often required.

Correlating Deep Marine SiliciclasticsCorrelating Deep Marine SiliciclasticsCorrelating Deep Marine SiliciclasticsCorrelating Deep Marine SiliciclasticsCorrelating Deep Marine SiliciclasticsSequence analysis in deep-water siliciclasticsalso presents substantial challenges. Thesequence stratigraphic surfaces that can beexpected in this environment are the slopeonlap surface (SOS), maximum regressivesurface (MRS), and the maximum floodingsurface (MFS). In an interval of stackedsubmarine fan deposits, an MRS can be drawnon top of the units of coarsening-upward,turbiditic fan deposits (e.g., Johannessen andSteel, 2005; Hodgson et al., 2006). Such ahorizon may occur near the top or wellwithin the package of turbidites.

In the same succession, the MFS can be drawnat the horizon of the finest sediment, usuallywithin a shale unit that separates thickintervals of turbidites (Sixsmith et al., 2004).Identification of an SOS in siliciclastics isdifficult and is best done on seismic sectionswhere the stratal geometry inside the thick,shale-dominant slope succession can bedetermined. Some workers have interpretedthe base of the f irst turbidite as acorrelatable sequence stratigraphic surface(e.g., Posamentier et al., 1988; Van Wagoneret al., 1990). However, in many cases such asurface is simply a scoured, within-trendfacies contact (a diastem within a coarseningupward, regressive succession) and is not asurface of sequence stratigraphy. In general,the base of a turbidite package is verygradational and is diachronous both downdip and laterally (Hodgson et al., 2006). Insome instances, where turbidites onlap theslope, the base of the turbidite succession,where it onlaps, coincides with an SOS.

Correlating Carbonate StrataCorrelating Carbonate StrataCorrelating Carbonate StrataCorrelating Carbonate StrataCorrelating Carbonate StrataSequence analysis for carbonate strata is inmany respects very similar as that forsiliciclastic strata although some differencesdo occur. These differences are due mainlyto di f ferences in how carbonatesedimentation responds to base-level changeas compared to siliciclastic sedimentation.For example, during times of falling baselevel, rates of siliciclastic sedimentation inmarine areas are often enhanced due toincreased delivery of sediment to a basin.However, carbonate sedimentation in a shelf/ slope / basin setting, often significantlydecreases with base-level fall because much

Figure 14.6. Stratigraphic cross-section of Lower Cretaceous strata from southern Saskatchewan. The nonma-rine strata of the Mannville Group are overlain by a succession of shallow marine strata that overalldeepenupward to offshore shale and siltstone of the Lower Colorado Group. A prominent MRS is used as thedatum. Fluvial strata occur in two wells and an SU is delineated below these strata with an SR-D above. Theshoreline ravinement is correlated into the adjacent marine strata as an SR-U. The SU is cut off by the SR-U,forming the flanks of an incised valley. Data courtesy of O. Catuneanu.

of the carbonate shelf (the carbonatesediment factory) is exposed. However, thesame types of surfaces of sequencestratigraphy that are recognized in siliciclasticstrata occur in carbonate strata. They canhave a few different attributes in carbonatesthan they do in siliciclastics. Notably the SOSis often very well expressed and can bereadily delineated when present.

For shallow marine carbonates, these includea maximum regressive surface, maximumflooding surface, shoreline ravinement, andregressive surface of marine erosion.Subaerial unconformities form during timesof base-level fall but most become modifiedby subsequent marine erosion and are thusunconformable shoreline ravinements.

The determination of MRSs and MFSsdepends on facies analys is and thedetermination of sediment supply trends inthe carbonate strata. These are often not asclearly expressed on mechanical logs as theyare in siliciclastic rocks. They can often bemore easily delineated on logs when fine-grained clastic sediment is part of thedepositional system (see Wendte and Uyeno,2005). For reefs and carbonate banks, oneor more SOSs are almost always present onthe slopes.

Figure 14.7 (page 72) illustrates a sequencecorrelat ion for a carbonate-dominantsuccession that contains carbonate rampdeposits that lie both below (Nisku Fm) andabove (Wolf Lake, Blue Ridge mbrs) aninterval of reef (Zeta Lake Mbr) and off-reefstrata (Cynthia Mbr). These data weresupplied by Jack Wendte and are from theWest Pembina area of Alberta (see Wendteet al., 1995).

MRSs and MFSs are readily correlated in thecarbonate ramp deposits below the reef(Figure 7). A slope onlap surface has beendelineated on the flank of the reef and ismarked by a high gamma (starved interval) inthe off-reef well. The SOS joins with an SR-U on top of the reef and both these surfacesformed when base level fell such that mostof the reef was exposed (SR-U) and the slopewas starved of sediment (SOS). During thelater stage of base-level fall, argillaceoussiliciclastic sediment prograded into the areaand onlapped the SOS (Figure 14.7, page72).

With base-level rise and transgression,carbonate sediment production greatlyincreased and siliciclastic sediment inputceased. Carbonate ramps then built out overthe siliciclastics which had filled in the deep,inter-reef areas. MRSs and MFSs are readilycorrelated in these post-reef ramp strata.This example shows how the delineation andcorrelation of sequences stratigraphicsurfaces helps to elucidate the depositionalhistory of a succession.

SummarySummarySummarySummarySummarySequence stratigraphy provides an excellentmethodology for construct ing anapproximate time-stratigraphic frameworkthrough the delineation and correlation ofsequence stratigraphic surfaces which havelow diachroneity (maximum regressivesurface, maximum flooding surface) or aretime barriers (subaerial unconformity,unconformable shoreline ravinement, slopeonlap surface). Such a framework is essentialfor predicting facies development away fromcontrol points and for reconstructingdepositional history and paleogeographicevolution.

Page 74: Embry Sequence Stratigraphy

72

Each general depositional environment hasat least one type of sequence stratigraphicunit that is useful for correlat ion. Insuccessions with intercalated nonmarine andshallow marine, siliciclastic strata, foursurface types are often present (SU, SR-U,MRS, MFS). In deep marine, siliciclasticsettings, correlatable surfaces include MRS,MFS, and SOS, with the SOS often being hardto delineate.

In carbonate strata, subaerial unconformities(SU) are very rare and unconformity surfaceson the basin f lanks are almost alwaysunconformable shoreline ravinements (SRU).S lope onlap surfaces are common incarbonate platform / slope / basin and reefsettings and are usually readily delineatedand correlated.

Correlation from basin edge to basin centreis best accomplished with maximum floodingsurfaces. A combined maximum regressivesurface and unconformable shorel ineravinement is also useful for such regionalcorrelations.

ReferencesReferencesReferencesReferencesReferencesEmbr y, A . F. 2008a. Pract ica l Sequence

Geologists. The Reservoir, v. 36, issue 1, p. 27-33.

Hodgson, D., Flint, S., Hodgetts, D., Drinkwater, N.,Johannessen, E., and Luthi, S. 2006. StratigraphicEvolution of Fine-Grained Submarine Fan Systems,Tanqua Depocenter, Karoo Basin, South Africa.Journal of Sedimentary Research, v. 76, p. 20-40.

Johannessen, E. J. and Steel, R. J. 2005. Shelfmarginclinoforms and prediction of deep water sands.Basin Research, v. 17, p. 521-550.

McCarthy, P. J. and Plint, A. G. 1998. Recognition ofinterfluve sequence boundaries: integratingpaleopedology and sequence stratigraphy.Geology, v. 26, p. 387-390.

Posamentier, H., Jervey, M., and Vail, P. 1988.Eustatic controls on clastic deposition I-conceptualframework, In: Sea level changes: an integratedapproach. C. Wilgus, B. S. Hastings, C. G. Kendall,H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner(eds.). Society of Economic Paleontologists andMineralogists, Special Publication 42, p. 109-124.

Ratcliffe, K., Wright, A., Hallsworth, C., Morton, A.,Zaitlin, B., Potocki, D., and Wray, D. 2004. Anexample of alternative correlation techniques ina low accommodation setting, nonmarinehydrocarbon setting: the (Lower Cretaceous)Mannville Basal Quartz succession of southernAlberta. AAPG Bulletin, v. 88, p. 1419-1432.

Sixsmith, P., Flint, S., Wickens, H., and Johnson, S.2004. Anatomy and stratigraphic developmentof a basin floor turbidite system in the LainsburgFormation, Main Karoo Basin, South Africa JournalSedimentary Research, v. 74, p. 239-254.

Van Wagoner, J. C., Mitchum, R. M., Campion, K. M.,and Rahmanian, V. D. 1990. Siliciclastic sequencestratigraphy in well logs, cores and outcrops: AAPGMethods in Exploration, no. 7, 55 p.

Wendte, J., Bosman, M., Stoakes, F., and Bernstein,L. 1995. Genetic and stratigraphic significance ofthe Upper Devonian Frasnian Z marker, west-central Alberta. Bulletin Canadian PetroleumGeology, v. 43, p. 393-406.

Wendte , J . and Uyeno, T. 2005. Sequencestratigraphy and evolution of Middle to UpperDevonian Beaverhill Lake strata, south-centralAlberta. Bulletin Canadian Petroleum Geology, v.53, p. 250-354.

Zaitlin, B. A., Warren, M. J., Potocki, D., Rosenthal, L.,and Boyd, R. 2002. Depositional styles in a lowaccommodation foreland basin setting: anexample f rom the Basal Quartz (LowerCretaceous), south-central Alber ta. BulletinCanadian Petroleum Geology, v. 50, p. 31-72.

Figure 14.7. Stratigraphic cross-section of Upper Devonian carbonate strata in the west Pembina area ofwestern Alberta. Carbonate ramp strata, in which MRSs and MFSs are readily correlated, occur below andabove a reef / off-reef interval. An unconformable shoreline ravinement (SR-U) caps the reefal strata andcorrelates with a prominent slope onlap surface (SOS) on the reef flank. Siliciclastic shale and siltstone,deposited during base-level fall, onlap the SOS and are overlain by prograding ramp carbonates depositedduring the subsequent base-level rise. Data courtesy of J. Wendte.

Stratigraphy IV: The Material-based Surfaces ofSequence Stratigraphy, Part 1: SubaerialUnconformity and Regressive Surface of MarineErosion. Canadian Society of Petroleum Geologists.The Reservoir, v. 35, issue 8, p. 37-41.

Embr y, A . F. 2008b. Pract ica l SequenceStratigraphy V: The Material-based Surfaces ofSequence Stratigraphy, Part 2: Shorel ineravinement and Maximum Regressive Surface.Canadian Society of Petroleum Geologists. TheReservoir, v. 35, issue 9, p. 32-39.

Embr y, A . F. 2008c . Pract ica l SequenceStratigraphy VI: The Material-based Surfaces ofSequence Stratigraphy, Part 3: Maximum FloodingSurface and Slope Onlap Surface. CanadianSociety of Petroleum Geologists. The Reservoir, v.35, issue 10, p. 36-41.

Embr y, A . F. 2008d. Pract ica l SequenceStratigraphy VII: The base-level change model formaterial-based sequence stratigraphic surfaces.Canadian Society of Petroleum Geologists. TheReservoir, v. 35, issue 11, p. 31-37.

Embry, A. F. 2009. Practical Sequence StratigraphyVIII: The Time-based Surfaces of SequenceStratigraphy. Canadian Society of Petroleum

Page 75: Embry Sequence Stratigraphy

73

Practical Sequence Stratigraphy XVPractical Sequence Stratigraphy XVPractical Sequence Stratigraphy XVPractical Sequence Stratigraphy XVPractical Sequence Stratigraphy XVTectonics vs. Eustasy andApplications to Petroleum Exploration by Ashton Embry

IntroductionIntroductionIntroductionIntroductionIntroductionThe delineation and correlation of sequencestratigraphic surfaces allows one to build anapproximate time stratigraphic framework,which is essential for determining faciesrelationships. This is perhaps the primary useof sequence stratigraphy and it was describedin the last article of this series (Embry, 2009).Once the sequence stratigraphic frameworkhas been establ ished and the faciesrelationships resolved, the depositionalhistory of the succession can be interpretedin terms of base-level changes because thesequence strat igraphic surfaces weregenerated by changes in base level asdiscussed in Embry (2008).

For example, the recognition and correlationof a subaerial unconformity allows one tointerpret that a base-level fall occurred overthe entire extent of the unconformity. If anunconformable shoreline ravinement ismapped, it leads to the interpretation thatthe area underwent base-level fall followedby a rapid base-level rise. Thus a sequencestratigraphic correlation framework not onlyal lows the facies relat ionships to beestablished but it also provides a means ofinterpreting depositional history in termsof base- level movements. Wheninterpretations of base-level changes aremade, it is also worthwhile to try todetermine which external factor wasresponsible for the recognized base-levelchanges. This will enhance the understandingof the depositional history of the successionand will improve predictions of faciesdistributions and potential stratigraphictraps.

Drivers of Base-level changeDrivers of Base-level changeDrivers of Base-level changeDrivers of Base-level changeDrivers of Base-level changeThree, external (a l logenic) factors –tectonics, sea level (eustasy), and climate –have the potential to drive changes in baselevel as was first described by Barrell (1917)(see a lso Embry, 2008, in press) . Animportant question regarding the base levelhistory of a given sequence is “Which of thethree variables was the main driver of thebase-level transit cycle recorded by thatsequence?” Climate change tends to resultin local, minor base-level changes and is nota viable driver for any sequence of regionalextent and / or large magnitude.Consequently this factor is not consideredfurther herein.

Both tectonic activity and eustasy arepotentially viable drivers for sequencedevelopment at any scale. It is alwaysreasonable to ask i f the subaeria lunconformities and / or unconformableshoreline ravinements which bound a givensequence on the basin flanks were theproduct of tectonic uplift followed bycollapse or were generated by eustatic fallfollowed by rise.

The debate of whether tectonics or eustasyis the main driver of the base-level changesrecorded by sequence stratigraphic surfaceshas been going on since the surfaces wererecognized in the 19th century. The debategot quite heated in the 1930s when theorig in of Pennsylvanian cyclothems(synonymous with depositional sequences)was considered (Weller, 1930; Wanless andShepard, 1936). The interpretation that thesesmall-scale, depositional sequences weregenerated by eustasy driven by the waxingand waning of Gondwana glaciers is nowwidely accepted (e.g., Heckel, 1986).

Sloss et al. (1949) defined the term sequencefor very large-magnitude units with boundingunconformities that stretched over most ofthe North American continent. Sloss (1963)clearly demonstrated that suchunconformities were tectonic in origin. In1977, when Exxon scientists published theirrevolutionary papers on seismic / sequencestratigraphy (Vail et al., 1977), eustasy wasappealed to as the main factor for generatingal l sequences, large and smal l . Thisinterpretation was based mainly on theobservations that the same age sequencesoccurred on different continental margins.

Some researchers now simply assume thateustasy is responsible for all sequenceboundaries and have published sea levelcurves for parts or all of the Phanerozoic onthe basis of this assumption and on scatteredobservations around the world (Haq et al.,1987; Hardenbol et al., 1998; Miller et al.,2005; Haq and Schutter, 2008). The validityof such curves is highly questionable giventhe great uncertainty of the underlyingassumption, not to mention the limitedobservations. Below, both eustatic andtectonic mechanisms for sequence boundarygeneration are discussed. Also, suggestionsare offered for how one can distinguish a

tectonically generated sequence boundaryfrom a eustatically driven one.

EustasyEustasyEustasyEustasyEustasyThere can no doubt that in some caseseustasy is the main factor in sequencegeneration. Given that sequence-boundingunconformities are generated over relativelyshort intervals of time regardless of theirmagnitude, the only reasonable phenomenonfor creat ing a sequence-boundingunconformity by eustasy is through changesin sea level caused by changes in terrestrialice volumes. Rates of tectono-eustaticchange (changing volume of the ocean basins)are far too slow to generate a sequenceboundary.

Ice-volume-related, eustatic changes are welldocumented and are due mainly to climatecycles driven by changes in orbita lparameters, the so-called Milankovitch cycles(Hays et al., 1976). There are three main typesof Milankovitch cycles and each has acharacteristic periodicity – precession of theequinoxes (~ 20 kyr), axis tilt or obliquity (~40 kyr), and orbit eccentricity (100 kyr and400 kyr). It would appear that such climate-driven cycles have operated on Earth at leastfrom Proterozoic onward (Grotzinger,1986). Amplitudes of sea level changesassociated with these cycles have varied fromover 100 metres when extensive glacierswere present in both hemispheres (e.g.,Pleistocene) to perhaps 10 metres or lesswhen only mountain glaciers were present(e.g., Devonian). Other climate-relatedfactors, such as temperature-related watervolume change and varying, land-based, waterstorage also contributed to sea level changesbut were minor compared to changing icevolumes.

Figure 15.1 (page 74) i l lustrates theinterpreted sea-level changes over the pasthalf-million years based on oxygen isotopedata. A base-level transit cycle during thistime was about 100 kyr long (eccentricity)and had an amplitude of about 120 metres.Note that a base-level cycle during this timeis dominated by a long interval of overallbase-level fall which is broken by a few, veryshort intervals of minor rise. The maininterval of base-level rise for each cyclecomprises only about 20% of the cycle timeand is characterized by relatively high rates

Page 76: Embry Sequence Stratigraphy

74

of rise (four times faster than fall rates). Thequest ion becomes, “What obser vablefeatures would be expected to characterizesequence-bounding unconformit iesgenerated by eustatically driven base leveltransit cycles?” First of all, in a downdipdirect ion, the angle between theunconformity and the truncated beds wouldbe very low, being slightly greater that thedip of the sea floor (< 1°). On strike, therewould be no angularity. Secondly, therewould be no change in sedimentary ortectonic regime across such an unconformityalthough changes in sediment compositionmight occur, given the possibility of newdrainage systems being established.

Furthermore, given the high frequency of theeustat ic cycles , sequences would berelatively thin in shelfal areas and numerous,very similar-looking sequences would bestacked upon each other. Finally, one wouldexpect to find the same unconformities onal l the basin f lanks. In theory, suchunconformit ies would potentia l ly becorrelatable worldwide but, as Miall (1991)has elegantly demonstrated, the lack ofprecision of dating techniques prevents thereliable correlation of high-frequency,eustasy-driven sequence boundaries fromone basin to another.

It must be noted that a few authors (e.g., Milleret al., 2003) have postulated that rare intervalsof substantial glaciation may have occurredduring Greenhouse times (e.g., mid-Permian -Early Paleogene). Such infrequent glacialintervals would be responsible for theoccurrence of sporadic unconformities whichrecord base-level falls of up to 60 m. This isan intriguing hypothesis that needs to beproperly tested. In these cases, suchunconformities would exhibit the first twocriteria mentioned above but closely spacedunconformities would not be expected.

There are numerous examples of sequenceswhich exhibit the above characteristics inthe literature and their eustatic origin iswidely accepted. For the most part, they arehigh frequency sequences found insuccessions deposited during “Icehouse”conditions of the Carboniferous – EarlyPermian and Late Paleogene – Recent.Parasequences and highfrequency, low-magnitude sequences, which characterizesuccessions deposited during “Greenhouse”intervals also may well be the product ofeustasy-driven base- level change asevidenced by their boundary characteristics.However, as demonstrated by Catuneanu etal. (1997), not all high-frequency, low-magnitude sequences are of eustatic origin.Finally, some low-frequency, large-magnitudesequence boundaries in Greenhousesuccessions may also be of eustatic origin(Miller et al., 2003) but this interpretation isstill very much open to debate.

TTTTTeeeeectctctctctoniconiconiconiconicsssssTectonics also provides a viable mechanismfor the generation of sequences. However,unlike eustasy, we don’t have a reliable,actualistic curve shape for a tectonicallydriven, base-level cycle. I suggest thattectonic activity at various scales would besimilar to faulting (i.e., fractal relationship)with short intervals of intense activityseparated by long intervals of quiescence.Figure 15.2 illustrates a tectonically drivencurve based on this model of tectonism. Thecurve consists predominantly of relativelylong intervals of base-level rise (80+% ofthe time) which characterize the times ofrelative quiescence. It is punctuated byrelatively short intervals of tectonic upliftfol lowed by tectonic col lapse whichrepresent the times of greatly increasedtectonic activity. Such a model is empiricallysupported by observat ions on thestratigraphic geometries of low-order,

Figure 15.1. Eustatic sea level curve for the past 500 kyr based on oxygen isotopes. The eustatically driven,base-level cycles are about 100 kyr-long and have long intervals of base-level fall and short intervals of rise.

Figure 15.2. A proposed tectonically driven, base-level curve that is characterized by long intervals ofslow rise (tectonic quiescence) punctuated by hortintervals of rapid base-level fall followed by rapidrise (tectonic uplift and collapse).

Mesozoic sequences of the Sverdrup Basin(Embry, 1990) as well as by the work of(Gawthorpe et al., 1994 and 2003) on fault-driven, base-level changes in rift basins.

Large, low-frequency, tectonically driven,base-level changes would generate widelyspaced, large-magnitude ( low-order)sequence boundaries overlain by a thininterval of transgressive strata depositedduring the collapse phase and overlain bythick intervals of prograding strata depositedduring slow base-level rise related tothermally driven subsidence. However, itmust be noted, tectonic activity can occuron a variety of scales and thus it is possiblefor high-frequency, tectonical ly drivensequence boundaries to be developed intectonically active settings such as forelandbasins (e.g., Catuneanu et al., 1997; Plint,2000) and rift basins (e.g., Gawthorpe et al.,1994).

Once again, the over-riding quest ionbecomes “What are the characteristics oftectonically driven sequence boundaries thatwould a l low them to be rel iablydistinguished from eustasy-driven ones?”Perhaps the most reliable indicator forrecogniz ing a tectonical ly generatedunconformity is the presence of substantialangularity between the unconformity andunderlying sequence stratigraphic surfaces.Figure 15.3 illustrates an outcrop exampleof such an angular unconformity which wasundoubtedly generated by tectonic uplift asopposed to sea level fall. Such angularitybeneath an unconformity can be

Page 77: Embry Sequence Stratigraphy

75

demonstrated in subsurface successions withseismic and closely spaced well data (e.g.,Embry, 1997, Figure 6; Dixon, 2009, Figure31). In general, anytime an angularity of a fewdegrees or more can be determined beneathan unconformity (SU, SR-U), especially overan area of little to no differential subsidence,there can be little doubt as to the tectonicorigin of the unconformity (Figure 15.4).

Other characteristics of a depositionalsequence boundary that indicate it wasgenerated by tectonics are:

• There are major changes indepositional regime across theboundary.

• There are major changes in sedimentcomposition and direction of sourceareas across the boundary.

• There are significant changes intectonic regime and subsidence ratesacross the boundary.

Figure 15.5 (page 76) illustrates Lower toUpper Triassic strata on the flank of theSverdrup Basin of Arctic Canada. A largemagnitude (2nd-order) unconformityseparates the Lower and Middle Triassicstrata (Embry, 1988 and 1991) and significantchanges in both depositional and tectonicregime occur across this boundary. TheLower Triassic succession consists mainly ofbraided stream strata and was deposited in ahigh subsidence regime (170 mm/myr). Theoverlying Middle Triassic strata consist ofoffshore marine shale and siltstone and weredeposited under low subsidence conditions(10 m/myr). The subsidence rate decreasedby more than 90% across the unconformity,a clear indication of a tectonic origin for theunconformity.

Figure 15.6 (page 76) illustrates the sequenceboundary between Middle Triassic stratabelow and Upper Triassic strata above. Onceagain this is a large magnitude boundary (2ndorder) and is interpreted to be tectonic inorigin in part due to the dramatic shift indepositional regime across the boundary. TheMiddle Triassic strata consist of siliciclasticsandstone, siltstone, and shale whereas theoverlying Upper Triassic strata consist mainlyof shelf carbonates. Notably there is also asignificant shift in source area across thisboundary (Embry, 1988) as well as a notablechange in subsidence rate.

Another unconformity of interpretedtectonic origin is illustrated in Figure 15.7(page 77). It separates Norian (early LateTriassic) strata from Rhaetian (late LateTriassic) strata and there is an abrupt changein sediment composition between theNorian sandstones (quartz, chert, rockfragments) and Rhaetian sandstones (highly

Figure 15.4. The “tilt-test” for an unconformity (Embry, 1997, p.418). Such a stratigraphic geometry test is bestdone in areas of similar subsidence rate so as to eliminate geometric effects of differential subsidence. Apossible eustatically driven unconformity is illustrated in Figure 4A. The sequence stratigraphic surfaces bothabove and below the unconformity parallel both each other and the unconformity. In this case, base-level fallwas equal at all localities which would fit a eustatic origin, although a tectonic one cannot be ruled out. InFigure 15.4B, the stratigraphic surfaces below the unconformity parallel each other (equal subsidence rate)but are at a substantial angle to the unconformity (differential uplift). The stratigraphic surfaces above theunconformity are parallel to the unconformity and to each other (equal subsidence rate restored). Suchstratal geometries can only be generated by tectonics and thus the demonstration of such geometriesprovides excellent evidence for the occurrence of a tectonically generated unconformity (e.g., Figure 15.3).

Figure 15.3. A subaerial unconformity (wavy red line) in the Isachsen Formation on western Axel HeibergIsland, Arctic Archipelago. Underlying strata (lower Isachsen Formation, Deer Bay Formation) are truncatedat a high angle leaving no doubt as to the tectonic origin of the unconformity.

Page 78: Embry Sequence Stratigraphy

76

quartzose), Furthermore, up to 500 m ofNorian strata are truncated beneath theunconformity in some areas, removing anydoubt as to the tectonic origin of theunconformity.

Tectonically generated depositional sequenceboundaries have been commonly describedin the l i terature beginning with thecontinent-wide ones of Sloss (1963, 1988).In general, it appears that most large-magnitude boundaries, which are oftenassigned to a 1st-, 2nd-, or 3rd-order levelin a hierarchy, are tectonic in origin. In almostall cases, they exhibit two or more of thecriteria listed for tectonically generatedunconformities. As discussed above, smallermagnitude boundaries (4th-, 5th-, and 6th-order) are often of eustatic origin but insome cases are tectonic.

One point of contention has been that somelarge-magnitude unconformities that have theabove-described signature of tectonicboundaries have been interpreted as beingeustatic in origin because they are recognizedin basins on dif ferent continents. Forexample, the sequence boundary whichapproximates the Middle / Late Triassicboundary (Figure15. 6, page 75) has beenrecognized in a number of basins around theworld and consequently was interpreted tobe the product of eustasy (Biddle, 1984).However, there is little doubt that this majorsequence boundary is primarily the productof tectonic movements (Embry, 1997). Asdiscussed by Sloss (1991, 1992) and Embry(1990, 1997, 2006), a reasonable case can bemade for the generation of similar age,tectonically generated sequence boundariesin basins throughout the world by appealingto plate tectonic mechanisms (see alsoCollins and Bon, 1996). The bottom line isthe boundary characteristic of occurrencein different basins throughout the world isnot a valid criterion for differentiatingeustatically driven sequence boundaries fromtectonically generated ones.

SummaryEither eustasy or tectonics can be the mainforcing function for sequence boundarydevelopment. Each of these external factorshas a characteristic base-level curve shape(Figure 15.8) with tectonics being dominatedby slow rise and punctuated by shortintervals of rapid fall followed by rapid rise.A eustatic curve is dominated by long, slowfalls and short intervals of fast rise. Asillustrated on Figure 15.8, the start of base-level rise will nearly coincide with starttransgression for both driving factors andthus a eustatically generated sequenceboundary will often superficially resemble atectonically generated one.

However, a eustatically generated sequence

Figure 15.6. Middle Triassic to Upper Triassic (Carnian) strata at Blind Fiord, Ellesmere Island, Arctic Archi-pelago. A 2nd-order sequence boundary (solid red line), which is a maximum regressive surface at this locality,marks the Middle / Upper Triassic boundary. A major shift in depositional regime from siliciclastics tocarbonates occurs across the boundary. A notable change in source area and in subsidence rates also occursacross this boundary. All these features point to a tectonic origin for the sequence boundary.

boundary has a number of di f ferentcharacterist ics in comparison to atectonically driven one. It is most importantto determine the degree of angularitybeneath a basin-flank unconformity in both

dip and strike directions and to determinethe amount of change, if any, in sedimentaryregime, tectonic regime and sedimentsource area across each boundary. With suchdata, a reasonable and reliable interpretation

Figure 15.5. Lower to Upper Triassic rocks on the flank of the Sverdrup Basin, Canyon Fiord, Ellesmere Island,Arctic Archipelago. The unconformity (unconformable shoreline ravinement, wavy red line) between theLower and Middle Triassic successions marks a major change in both subsidence rate and sedimentaryregime. These characteristics point to a tectonic origin for the unconformity.

Page 79: Embry Sequence Stratigraphy

77

Figure 15.7. A subaerial unconformity separates Norian strata from Rhaetian strata (Late Triassic) on westernRaanes Peninsula, Ellesmere Island, Arctic Archipelago. This 1st-order sequence boundary coincides with asignificant change in sandstone composition. High-angle truncation is also associated with the boundary onthe basin flanks and it is interpreted to be tectonic in origin.

Figure 15.8. A comparison of a tectonically driven baselevel curve (dominated by rise with a short, fast fall)with a eustatically driven one (dominated by fall with a short, fast rise). In both cases the start of base-level risecoincides with the start of transgression because of initial high rates of rise. This results in superficial similaritiesbetween sequences generated by tectonics compared with those generated by eustasy. However, the twosequence types can be differentiated on the basis of a variety of specific characteristics (see text).

of the origin of a given sequence boundarycan be made. Such an interpretation can beuseful for predicting facies development,stratigraphic geometries, and potential traps.

Sequence StratigraphySequence StratigraphySequence StratigraphySequence StratigraphySequence Stratigraphyand Petroleum Explorationand Petroleum Explorationand Petroleum Explorationand Petroleum Explorationand Petroleum ExplorationAn important task in petroleum explorationis the construct ion of strat igraphiccrosssections on which correlations aremade and facies relationships determined.The success of a play involv ing thedelineation of a stratigraphic trap dependson the reliability of the correlations and thesubsequent fac ies analys is within theframework. As has been demonstrated,sequence strat igraphy involves therecognition and correlation of a variety ofstratigraphic surfaces that are used to form anapproximate time (chronostratigraphic)framework. These surfaces include subaerialunconformity, unconformable shorelineravinement, maximum regressive surface, slopeonlap surface, and maximum flooding surface(Embry, 2009). A very detailed framework canbe constructed with these surfaces, especiallywhen small-scale surfaces are correlated.

A sequence stratigraphic framework isessential to guide facies analysis and one ofthe key objectives of such analysis is toidentify porous facies that may act aspetroleum reservoirs. For siliciclastics, suchfacies are usually sand bodies of nonmarine,shoreline, shallow shelf, and deep marineorigin. Each systems tract can be seen as anapproximate time stratigraphic unit thatcontains a variety of facies from nonmarineto deep marine. These wil l occur in a

identified, it can be concluded that incisedval leys that preserve the subaeria lunconformity and a section of mainlytransgressive, non-marine strata may occurin the area. Incised valleys can contain avariety of porous facies and be completelysurrounded by impermeable strata such asoffshore shales. Once the facies within anincised valley are documented at one locality,predictions can be made regarding facieschanges within the valley succession bothlandward and seaward of that locality. Ofcourse, major base-level falls that resultedin an exposed shelf edge allow a predictionof the occurrence of sand-prone slopechannel fil ls and submarine fans in theadjacent deepmarine basin area.

Sequence analysis also helps to predict howand where porous strata pinch out laterally.Within a regressive systems tract (RST), ashorel ine sandstone unit sometimesdisappears landward due to truncation bythe sequence bounding unconformity andpinches out basinward due to facies changeto impermeable offshore shale and siltstone.Impermeable shelf strata of the overlyingtransgressive systems tract (TST) of the nextsequence can seal such strata. Thus a fairwaythat has a high potential to contain porous,shoreline sandstone within a given RST canbe delineated with the available control.Seismic data can be used to reveal specificprospects along the fairway.

In other cases a shoreline sandstone willpinchout landward due to facies change toimpermeable coastal plain facies that can also

predictable lateral and vertical order withinthe systems tract. For example, if a subaerialunconformity is identified and it rests onoffshore marine shale, one can predict that apotentially porous, shoreface sandstone unitl ies basinward of that local i ty at thebasinward termination of the unconformity.

Another example would be when anunconformable shoreface ravinement is

Page 80: Embry Sequence Stratigraphy

78

provide a top seal. Similar fairways of porous,nearshore sandstone can sometimes bedelineated for TSTs and, in this case, thesandstone will often pinchout landward dueto onlap onto a shoreline ravinement. Suchsandstone is usually well sealed by overlyingshale and siltstone that were deposited astransgression progressed.

There can be no doubt that the properinterpretation of depositional facies is criticalfor successful explorat ion. The samesentiment applies to the surfaces of sequencestratigraphy and an incorrect interpretationof a given surface can lead to misdirectedexploration. Often only mechanical logs areavailable for a sequence interpretation andin this situation an unconformable shorelineravinement can be easily be mistaken for amaximum regressive surface and viceversa.On a gamma log, both surfaces are drawn atthe change from a shal low marine,coarsening-upward succession (RST) to ashallow marine, fining-upward one (TST). Ifthe underly ing coarsening-upwardsuccession terminates in shaly, mid-shelfsandstone, the explorat ionist wouldnaturally want to locate potentially porous,shoreface sandstone in that RST. If the surfaceencountered in the control point is amaximum regress ive surface, then ashoreface sandstone unit would occurlandward of the control well. However, ifthe surface is an unconformable shorelineravinement then the shoreface sandstoneunit would occur basinward of the controlwell, in exactly the opposite direction as wasdictated by the MRS interpretation. Asillustrated by this example, the correctinterpretation of sequence stratigraphicsurfaces is critical for exploration success.

Concluding RemarksConcluding RemarksConcluding RemarksConcluding RemarksConcluding RemarksThis article wraps up the Practical SequenceStratigraphy series, which has covered themain topics of sequence strat igraphyincluding historical development, thesurfaces of sequence stratigraphy, the linkagebetween base level and sequencestratigraphic surfaces, the units of sequencestratigraphy, and more general topics ofsequence hierarchies, correlation, andsequence boundary origin.

Sequence stratigraphic analysis is a coremethodology in petroleum exploration. If itis applied in an objective, pragmatic mannerwith the use of a material-based surfacesand units, it can greatly enhance petroleumexploration and exploitation. The methodinvolves:

• Identification of sequence stratigraphicsurfaces in a succession,

• Correlation of the surfaces over thestudy area,

• Determination of the faciesdistribution within the sequencestratigraphic framework,

• Interpretation of the depositionalhistory of the succession in terms oftectonic and/or eustatic base-levelchanges,

• Construction of facies maps at boththe approximate time of maximumregressive and the approximate timeof maximum transgression for eachsequence.

With the adoption of this methodology,sequence stratigraphy becomes a valuableaddition to the explorationist’s tool kit.

ReferencesReferencesReferencesReferencesReferencesBarrell, J. 1917. Rhythms and the measurementsof geologic time. GSA Bulletin, v. 28, p. 745-904.

Biddle, K. 1984. Triassic sea level change and theLadinian- Carnian stage boundary Nature, v.308, p. 631-633.

Catuneanu, O., Sweet, A., and Miall, A. 1997.Reciprocal architecture of Bearpaw T-R sequences,uppermost Cretaceous , Western CanadaSedimentary Basin. Bulletin Canadian PetroleumGeology, v. 45, p. 75-94.

Collins, J. F. and Bon, J. 1996. Mantle origin ofglobal sealevel fluctuations and geomagneticreversals: evidence from non-linear dynamics. In:C. Caughey et al. (eds.), International Symposiumon Sequence Stratigraphy in SE Asia. IndonesianPetroleum Society, p. 91-128.

Dixon, J. 2009. Triassic stratigraphy in thesubsurface of the plains area of Dawson Creek(93P) and Charlie Lake (94A) map areas ,northeast British Columbia. Geological Survey ofCanada, Bulletin 595, 78 p.

Embry, A. F. 1988. Triassic sea-level changes:evidence from the Canadian Arctic Archipelago.In: C. Wilgus, B. Hastings, C. Kendall, H. Posamentier,C. Ross, and J. Van Wagoner (eds.), Sea-levelchanges – an integrated approach. SEPM SpecialPublication 42, p. 249-259.

Embry, A. F. 1990. A tectonic origin for third-orderdepositional sequences in extensional basinsimplications for basin modeling. In: T. Cross (ed.),Quantitative Dynamic Stratigraphy. Prentice Hall,p. 491-502.

Embry, A. F. 1991. Mesozoic history of the ArcticIslands. In: H. Trettin, (ed.), Innuitian Orogen andArct ic P latform: Canada and Greenland:Geological Survey of Canada, Geology of CanadaNo. 3 (also GSA, The Geology of North America, v.E), p. 369-433.

Embry, A.F. 1997. Global sequence boundaries ofthe Triassic and their recognition in the WesternCanada Sedimentary Basin. Bulletin Canadian

Petroleum Geology, v. 45, p. 415- 433.

Embry, A. F. 2006. Episodic Global Tectonics:Sequence Stratigraphy Meets Plate Tectonics. GEOExpro, v. 3, p. 26- 30.

Embry, A. F. 2008. Practical Sequence StratigraphyVII: The base-level change model for material-based sequence stratigraphic surfaces. CanadianSociety of Petroleum Geologists, Reservoir, v. 35,issue 11, p. 31-37.

Embry, A. F. 2009. Practical Sequence StratigraphyXIV: Correlation. Canadian Society of PetroleumGeologists Reservoir, v. 36, issue 7, p 14-19.

Embry, A. F. In press. Correlating SiliciclasticSuccessions with Sequence Stratigraphy. In: K.Ratcliffe and B. Zaitlin, (eds.), Application ofModern Stratigraphic Techniques: Theory andCase Histories, SEPM Special Publication 94.

Gawthorpe, R., Fraser, A., and Collier, R. 1994.Sequence stratigraphy in active extensionalbasins: implications for the interpretation ofancient basin fills. Marine and Petroleum Geology,v. 11, p. 642-658.

Gawthorpe, R., Hardy, S., and Ritchie, B. 2003.Numerical modeling of depositional sequences inhalf-graben rift basins. Sedimentology, v. 50, p.169-185.

Grotzinger, J. 1986. Upward shallowing platformcycles: a response to 2.2 billion years of low-amplitude, high-frequency (Milankovitch band)sea level oscillations. Paleoceanography, v. 1, p.403-416.

Hardenbol, J., Thierry, J., Farley, M., Jacquin, T., DeGraciansky, P. and Vail, P. 1998. Mesozoic andCenozoic Sequence Chronostrat igraphicFramework of European Basins. In: P.C . deGraciansky, J. Hardenbol, T. Jacquin, and P. Vail(eds.) , Mesozoic and Cenozoic sequencestratigraphy of European basins, SEPM SpecialPublication 60, p. 3-14.

Haq, B., Hardenbol, J., Vail, P. 1987. Chronology offluctuating sea levels since the Triassic (250 millionyears ago to present). Science, v. 235, p. 1156-167.

Haq, B and Schutter, S. 2008. A Chronology ofPaleozoic Sea- Level Changes. Science, v. 322. p.64-68.

Hays, J., Imbrie J., and Shackleton, N. 1976.Variations in the Earth’s Orbit: Pacemaker of theIce Ages. Science, v. 194. p. 1121-1132.

Heckel, P. 1986. Sea-level curve for Pennsylvanianeustat ic mar ine transgress ive-regress ivedepositional cycles along midcontinent outcropbelt, North America. Geology, v. 14, p. 330-334.

Miall. A. 1991. Stratigraphic sequences and theirchronostratigraphic correlation. Journal of

Page 81: Embry Sequence Stratigraphy

79

Sedimentary Research, v. 61, p. 497-505.

Miller K., et al. 2003. Late Cretaceous chronologyof large, rapid, sea-level changes: Glacioeustasyduring the greenhouse world. Geology. v. 31, p.585-588.

Miller, K., et al. 2005. The Phanerozoic Record ofGlobal Sea- Level Change. Science, v. 310, p. 1293-1298.

Plint, A. 2000. Sequence stratigraphy andpaleogeography of a Cenomanian deltaic complex:the Dunvegan and lower Kaskapau formationsin subsurface and outcrop, Alberta and BritishColumbia, Canada. Bulletin Canadian PetroleumGeology, v. 48, p. 43-79.

Sloss, L. 1963. Sequences in the cratonic interiorof North America. GSA Bulletin, v. 74, p. 93-113.

Sloss L. 1988. Tectonic evolution of the craton inPhanerozoic time. In: L. Sloss (ed.), SedimentaryCover – North American Craton: U.S. The Geologyof North America, D2, Geological Society ofAmerica, p. 25-51.

Sloss L. 1991. The tectonic factor in sea levelchange: a countervailing view. Journal GeophysicalResearch, v. 96, p. 6609-6617.

Sloss L. 1992. Tectonic episodes of cratons:conflicting North American concepts. Terra Nova,v. 4, p. 320-328.

Sloss, L., Krumbein, W., and Dapples, E. 1949.Integrated facies analysis. In: C. Longwell, (ed.).Sedimentary facies in geologic history. GeologicalSociety America, Memoir 39, p. 91-124.

Vail, P. et al. 1977. Seismic stratigraphy and globalchanges in sea level. In: C. Payton, (ed.), Seismicstratigraphy: applications to hydrocarbonexploration, AAPG Memoir 26, p. 49-212.

Wanless, H. and Shepard, F. 1936. Sea level andclimatic changes related to late Paleozoic cycles.GSA Bulletin, v. 47, p. 1177-1206.

Weller, J. M. 1930. Cyclic sedimentation of thePennsylvanian period and its significance. Journalof Geology, v. 38, p. 97- 135.