You are on page 1of 17

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 103, NO.

B10, PAGES 24,141-24,157, OCTOBER 10, 1998

Clinoform developmentby advection-diffusionof suspended


sediment:Modeling and comparisonto natural systems
Carlos Pirmez 1
Lamont-Doherty
EarthObservatoryof ColumbiaUniversity,Palisades,
New York

Lincoln F. Pratson
Instituteof Arctic andAlpine Research,Universityof Colorado,Boulder

Michael S. Steckler
Lamont-Doherty
EarthObservatoryof ColumbiaUniversity,Palisades,
New York

Abstract. Clinoformsare the buildingblocksof progradingstratigraphic sequences.


Thesesig-
moid-shaped surfacescanbe foundformingtodayonmodemdeltas.Sedimentation rateprofiles
overthe clinoformsurfaceof thesedeltasshowlow ratesof sedimentaccumulationon bothtopset
andbottomsetregions,with a maximumaccumulation rateon theupperforesetregion.We pres-
ent a modelfor the formationof clinoformsthatrelieson the interpretation
of modemclinoform
sedimentation as a result of the distribution of shear stresses at the mouth of a river. Model clino-
form surfacesare generatedusingan equationfor the conservation of suspended sediment
concentration, togetherwith a conservation of fluid equationfor simpletime-averaged flow
velocityfields.In themodel,suspended sedimentis advectedhorizontallyintoa basin,and
gravitationalsettlingof sedimentparticlesis counteracted by verticalturbulentdiffusion.In
shallowwater, shearstresses are too largeto allow deposition,and sedimentbypasses the topset
region.With increasing waterdepth,near-bedshearstresses decrease, andsedimentis allowedto
depositat the foresetregion,with graduallydecreasing ratestowarddeeperwater.This
sedimentation patternleadsto progradation of the clinoformsurfaces throughtime. The clinoform
surfacesproducedby themodelcapturethefundamental morphological characteristics
of natural
clinoforms.Theseincludethe gradualsloperolloverat thetopsetandbottomset,steeperforeset
slopeswith increased grainsize,andan increasein foresetslopethroughtimeasclinoforms
progradeintodeeperwater.Becausetheparameters controllingthemodelclinoformshavea
directrelationto physicalquantitiesthatcanbe measuredin naturalsystems, themodelis an
importantsteptowardunravelingthephysicalprocesses associated with thesedeposits.

1. Introduction mixed siliciclastic-carbonate


environments.The shapeof their
variable, basinwarddipping profile is affected by the same
The morphologyand stratigraphyof deltaicdepositsandtheir factorsthat affectthe sequences
they comprise:relativesealevel,
intemalstructurewerefirst systematicallyinvestigated
by Gilbert sediment supply, depositionalregime, and sediment type.
[1885, 1890], who identified the topset,foresetand bottomset
However, what clinoforms indicate about these factors remains
bedsof the deltaicdepositsand their clinoform-shaped surfaces
just partiallyunderstood,
andonly in qualitativeterms.
on the shoresof Pleistocene Lake Bonneville.Clinoform-shaped
The shapeof clinoform surfacesis thoughtto indicatedif-
deposits,however,extendbeyondthe Gilbert-typedelta [Bates,
ferentalepositionalenvironments.Sangreeand Windmier[1977]
1953] andrangein scalefrom deltaicdepositsa few metersthick
defined two basic clinoform shapes:sigmoid and oblique.
to continentalshelf-slopedepositsseveralhundredsof meters
Clinoformssurfacescharacterized by aggradational topsetsand
thick. The stackingof clinoform-shaped depositsmakesup the
gradualincreasein slopefrom the topsetto the foresetare termed
bulk of stratigraphicsequences[Mitchurnet al., 1977], and, as
sigmoid.Obliqueclinoformsdisplayabruptchangein slopefrom
such,clinoform surfacescan be thoughtof as the fundamental
topsetto foresetand a topsetregioncharacterized by little or no
buildingblocks of the infill of sedimentarybasins.Clinoforms
aggradation, suggesting sediment bypass and/or erosion.
are ubiquitousin the modem sedimentaryrecordof continental
Sigmoidal clinoforms are generally interpretedto represent
margins,includingcarbonateplatforms,siliciclasticshelves,and
relatively low-energydeltaic environments.Oblique clinoforms
commonlyhavesteeperforesets(up to -10 ø) and are interpreted
to form in high-energyenvironmentsof shallowcoastalwaters
[Sangreeand Windmier,1977]. Thesecharacteristics suggestthat
NowatExxonProduction
Research
Co.,Houston,
Texas. the dimensions, shape,andsedimenttype of clinoformscontaina
recordaboutthe environmentof deposition,includingthe'energy
Copyright1998by the AmericanGeophysical
Union.
of transportprocessesand grain size of the sedimentparticles.
Papernumber98JB01516. The interpretationof this recordin termsof sea level changes,
0148-0227/98/98JB-01516509.00 basinsubsidence, and sedimentsupplyhas been one of the key
24,141
24,142 PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION

elementsin the sequencestratigraphicapproachto investigating addressesthe fundamentalquestionof what factorscontrol the


sedimentary sequences [Vail et al., 1977; Vail, 1987; shapeof subaqueousclinoforms.Thosethat have been used to
Posarnentieret al., 1988; Van Wagoneret al., 1990]. examinethe shapeand characteristics of subaqueous clinoforms
While there is a generalunderstanding of the factorsthat af- follow a slope-drivendiffusion of topographyapproachrather
fect the stackingpatternsof clinoform packageswithin sedi- than a fluid-sediment interaction approach, contain the
mentary sequencesand the evolution of the sequencesthem- fundamental physicalprinciplesassociated with fluid flow but do
selves,we are still hamperedin our ability to quantifythe factors not generateclinoforms,or still are not suitablefor a detailed
that influence the shapes of clinoform surfaces and their analysisof the factorsthat controlthe characteristicsof clinoform
implicationsfor estimatingthe depositionalregime, sediment depositionalsurfaces.
type, and sea level changes.Studiesof modem clinoformsin Models that focus on the developmentof large-scaleconti-
deltaic depositionalsystems[e.g., Kuehl et al., 1986, 1989; nentalmargin sequencesoften generateclinoformsby stacking
Wrightet al., 1988, 1990;Alexanderet al., 1991;Nittrouer et al., sedimentpackagesaccordingto predefinedgeometricrules[e.g.,
1995, 1996; Nittrouer and DeMaster, 1996] revealed that ,lervey,1988; Ross, 1989; Kendall et al., 1991; Reynoldset al.,
clinoformshapeand growth patternsare determinedto a large 1991; Thorneand Swirl, 1991; Bowmanand Vail, 1993]. These
extentby the spatialand temporaldistributionof energyin the models can be useful as a tool to understand the effects of sea
water column. The hydrodynamiccharacteristics of the water level, sedimentsupply,and tectonicson the grossgeometryof
column and the resulting sedimentdispersalpatternsultimately stratigraphicsequences,but they do not addressthe physical
determinethe shapeof the sedimentation rate profile acrossthe processesthat are responsiblefor the formationof clinoform
clinoformsurface.A maximumin the sedimentation rate profile units,of whichthesesequences arecomposed.
in activelyprogradingmodernclinoformsystemsoccursnearthe Diffusion-based models of clinoform formation assume that
upperforeset,just below the clinoformbreak,or rollover point sedimenttransportis a function of topographicslope [Kenyon
[Kuehl et al., 1986; Alexanderet al., 1991]. Another importantand Turcotte, 1985; Flemings and Jordan, 1989; Jordan and
observationthat is borneout from the studiesof modem systems Flemings, 1991; Thorne, 1995]. These models result in a
clinoformgeometrythat resemblesthat in naturalsystems,but
is that the clinoformrollover is, in general,locatedat variable
distancesfrom the shorelineand in variablewater depths.Sucha implicit in these models is the interpretationthat sediment
distinction may have important consequencesfor the transportis a functionof slope-drivenprocesses, suchas creep,
characterization of depositionalfacies in ancientclinoformsas sliding,and slumping[e.g., Kenyonand Turcotte,1985]. While
well as for estimatingthe magnitudesof sealevel changesfrom theseprocessescertainly contributeto sedimentdistributionin
shifts in the position of onlap within depositionalsequences modem clinoforms (e.g., Mississippi delta [Coleman et al.,
[Greenleeand Moore, 1988]. 1983]), they are by no meansthe most significantprocessin
In this study we introducea mathematicalmodel to generate manymodernsystems(e.g., Amazondelta [Adamset al., 1986]).
clinoform surfaces. The model uses conservation and sediment In addition,slope-drivenmass-transport processes,by removing
dispersionprinciples, including the interactionbetween fluid sedimentfrom the steeperforesetslopes,may act as a retreating
flow in the water columnand changesin the basinmorphology agent,counteracting the effectsof clinoformprogradation[Ross
due to sedimentaccumulation.The model is an attemptto quan- et al., 1994; Pratson and Coakley, 1996; Pratson and Haxby,
of 1996]. Another shortcoming of slope-driven diffusion
tify the physical processesthat control the characteristics
clinoforms and generate stratigraphicsequences,taking into approaches is that only concave-upprofilesareproduced,which
account the observationson modem clinoform systems.Our results in an active clinoform break that is coincident with the
modeladdresses threekey pointsthat have beeneitherneglected shoreline[Flemingsand Jordan, 1989; Kenyon and Turcotte,
or de-emphasized in previous modeling attempts: (1) 1985].This shortcoming hasbeencircumvented by introducinga
quantificationof the factorscontrollingthe shapeof clinoform water-depthdependentdiffusion coefficientof Kaufmanet al.
surfacesand their significanceto interpretationof sedimentary [ 1991]. A different approachwas used by Thorne [1995], who
sequences;(2) the separationbetweenthe clinoform rollover defined a mathematical sediment accumulation function,
point and the shorelinethat is apparentin many observationsof following the earlier work of Thorne and Swift [1991], and
both the modem and ancientrecord;and (3) the integrationof numericallycomputeda sigmoidclinoformshapeusingcalculus
observations on modernclinoformsand the implicationsfor the of variations.The approachusedhere is similar in that sediment
interpretationof the ancient sedimentaryrecord. We begin by accumulationrates also are computedbut differs in that it is
discussingkey elementsof existing models for clinoform based on the distribution of shear stresses within the water
formation and then summarize observations on selected modem column above the basin floor.

and ancient clinoform systems.We then advance the basic Syvitskiet al. [1988] advanceda new classof multiprocess
equationsthat describeour model and how the model works models that includes sediment advection by fluid flow and
gravity-driven sedimenttransport,in addition to slope-driven
througha seriesof runs that illustratethe model sensitivityto
variousparameters.Finally, we compareand contrastour model diffusion.This approachhas proven to be a usefultool to un-
resultswith naturalclinoformsystems. derstandthe various sedimentaryprocessesthat affect the ge-
ometry of sedimentarybasin infill and has been successfully
appliedto simulatespecificdepositionalsettings,in particular
2. Background f•ord basins[Syvitskiet al., 1988; Syvitskiand Daughney,1992;
Syvitskiand Alcott, 1993, 1995]. The usefulnessof thesemodels
2.1. Existing Models
to investigatethe formation of clinoform surfacesand the
Existingsedimentationmodelscanbe generallyclassifiedinto fundamentalcausesof their variableshapeis limitedbecauseof
geometric,diffusion-based,and process-basedmodels.Many of the needto specifya largenumberof inputparameters, including
the existingmodelshave advantages and disadvantageswhich a detailed descriptionof temporal fluctuationsin water and
are only briefly discussedhere. None of the existingmodels sedimentdischargethrougha model run. Our model approach
PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION 24,143

differs in that we assume a small set of flow and sediment total relief H of these clinoforms reaches in excess of 250 m for
parametersthat are thought to representthe time-averaged some of the Miocene surfaces.Presumably,there are several
conditions for the formation of a model surface. factorsinfluencingthe slope and height of clinoforms.To first
A few models have attemptedto examine sedimentation order,the clinoformheightprobablyis determinedby the space
processesthrough basic principles of fluid motion and sedi- available to accumulatesediment(basin depth) and the factors
mentation. Models such as those describedby Tezlaff and limiting vertical accumulation(such as sedimentflux, sediment
Harbaugh [1989] can be efficientlyusedto simulatea specific availability and water column energy). High-relief clinoforms
basin setting or depositionalsystem,but the large numberof generallyare associated with steeperclinoformfronts(e.g., New
parametersmakes it difficult to evaluatethe fundamentaleffects Jerseyand Lake Mead; see Table 1 and also Steckler et al.
of flow and sedimentvariabilityon the shapeof the deposits.On [1998]).
the other hand, one can attemptto model the physical flow 3. As clinoforms become steeper, the separationbetween
processes to a high degreeof detail, but thesecomputationally r.olloverand inflectionpointstend to approach zero. It is im-
intensiveand complex parameterizationsof flow and sediment portantto note, however,that the distancenevervanishes,since
transport have not yet been applied to the generation of it would requirean infinite curvatureat the rollover.Many of the
stratigraphy[e.g., Jewell et al., 1993]. Swift and Thorne[1991], profilesdisplaya very sharprollover,with smallseparationfrom
Thorne and Swift [1991] and Thorne et al. [1991] used basic the foresetinflection point (e.g., Rhine and Mississippi,Figure
physicalprinciplesof fluid motionunderthe effectsof wavesand 3). Two groupsof clinoformscan be identifiedon the basisof
currents, together with regional observationson continental slopedistributionandrelativeelevationof the inflectionpoint.In
shelves,to develop a sedimenttransportmodel basedon the the first group, which includes the Lake Mead, Rhine,
regime concept. One critical assumptionembedded in their Mississippi, and New Jersey clinoforms, the surfaces are
modelis that the equilibriumshapeof the coastalprofile follows distinctly asymmetrical.The asymmetryis characterizedby a
a power law function of distancefrom the coastline.The model rapid slopeincreasefrom the topsetto the inflectionpoint andby
has been used to successfullypredict changesin the coastal a more gentle slope decreasefrom the inflection point seaward
morphologydue to erosionand deposition[Thorneet al., 1991], (Figures2b and 3a-3d). In the secondgroup,which includesthe
althoughonly clinoformsconstrainedby an assumedequilibrium large deltas of the Ganges,Amazon, and Yellow Rivers, slope
shapearegeneratedby this model. distributionis nearly symmetric.The distinctionbetweenthe two
2.2. Natural C!inoform Systems groupsis clearly seenthroughthe normalizedelevationof the
inflectionpoint,y/H (Figure2a). For the largesubaqueous deltas,
Modem clinoform surfaces are seen on the deltas of the y/H<0.4, whereasfor the smallerdeltas,y/H is larger.
Amazon,Colorado(in Lake Mead), Ganges,Mississippi,Rhine 4. Studiesof modernsubaqueous deltason continentalshelves
(in Lake Constance)and Yellow Rivers,amongothers(Figures off major rivers indicate that modem, actively prograding
1-3). Ancient clinoform surfacesare common in the Miocene clinoforms show a maximum sedimentationrate at the upper
sedimentarysequencesof many continentalmargins[Bartek et foreset,seawardof the clinoform rollover point [e.g., Kuehl et
al., 1991; Steckleret al., 1993] (Figure la). The topographic al., !986; Alexander et al., !991] (Figures 4a and 4b).
profileswere digitizedfrom figurespublishedby severalsources Sedimentationrates are small at both the topsetand bottomset
(Table 1). These depositionalsurfaces display significant portionsof the clinoform. Similar sedimentationpatternsappear
variability of scales (Figures 1-3 and Table 1), and direct to prevailover moderncontinentalshelves,as indicatedby both
comparisonbetweenthem (and modelclinoforms)requiresthat long-termand modernaccumulationpatterns.Seismicprofiling
the vertical and horizontal dimensions be normalized. We use the
and drilling results on the eastern U.S. margin show that
vertical dimensionH as the scalingfactor for the clinoform preferentialsedimentaccumulationoccurredon the upperslope
profiles,defined as the elevationdifferencebetweenthe toe and (foreset) during the Quaternary[Hathaway et al., 1976; Poag,
top of the clinoform surfaces.The toe elevation (zero) is 1985; Poag and Mountain, 1987; Pratson et al., 1994]. The
determinedat the point where the clinoform surfacebecomes present-daysediment distribution over the continentalshelf
horizontalor conformable(i.e., with the samegradient)with the surfaceshowsa sand-mudtransitionregion on the seafloorjust
underlyingsurface.The top is definedasthe shallowest pointat below the shelf break, indicatingbypassof mudsover the sandy
the mouth of the river or the topsetsurfaceof the clinoform shelf [Stanleyet al., 1983; Nittrouer and Wright, 1994]. Modem
where horizontal. The distance axis also was shifted for ease of sedimentaccumulationratesdeterminedfrom sedimenttraps on
comparison,with the origin placed at the point where slope the U.S. eastcoastare also enhancedon the upperslope(Figure
reachesa maximum value, that is, at the clinoform face inflection 4c) [Biscayeand Anderson,1994]. These observations suggest
point. The rollover point is defined as the point of highest that continentalshelf sedimentationrate profiles behave in a
Curvature landwardof theinflectionpoint. similar fashionto thoseobservedfor subaqueous deltas,and as
Examplesof normalizedclinoformsurfacesare seenin Fig- suggestedby Thorne [1995], shelvesmay be viewed as large-
ures 2 and 3. Among the key elementsthat characterizethese scale analogsof the deltaic clinoform surfacesobservedat the
clinoformsurfacesarethe following: mouth of rivers.
1. All clinoform surfacesdisplay a gently sloping topset. We interpretthese observationsto indicatethat suspended
Depthsof the clinoformrollover pointsinvestigatedhere vary sedimentoriginatingfrom the coastalareas or river sourceis
significantly,being relatively shallow in low-energylake en- preventedfrom being depositedon the topsetregion of clino-
vironmentssuchas Lake Mead (Figure la) and the Rhine deltas, forms becauseof high near-bedshearstresses. As shearstresses
andmuchdeeperin open-shelfclinoformsandsubaqueous deltas decreaseseawardbecauseof increasedwater depth, sedimentis
suchas the New Jerseyclinoforms(Figure l a), or the Ganges allowed to deposit,and sedimentationrate increases,reachinga
andAmazonRivers(seevaluesin Table 1). maximum at the foreset. As rates of depositionincrease,less
2. The maximum slopeof clinoformforesetsvariesfrom less sedimentremainsin the water column, resultingin a decreaseof
than 0.5ø to 6.3ø in the clinoformsanalyzedhere (Table 1). The the depositionrate farther seawardover the bottomsetregion of
24,144 PIRMEZ ET AL.' CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION

(A) o

lOO

•:•'
200

300
m3
m4
..... m5
m5.2
400 ........... m5.3
.... m5.4
......... m5.6
rn6
...... ol (pre-Miocenesurface) V.E.~ 110x
500 i i I i i i i '<""
0 25 50 75 100

Distance (km)
200
(B) Lake Mead
1935
- .... 1937
1938
150
• • 1939
..... 1941

100
1• =I 1948

50

10 15 20
Distance (km)

Figure1. (a) Topographic profilesof naturalclinoformsurfacesfromMiocenesedimentary sequencesoffshore


NewJersey (surfaces andtheirrespective
names aresequence
boundaries; seeTable1).Eachsurface representsthe
paleoelevation at the time of formationof eachsequence boundary, reconstructedusingtwo-dimensional
backstripping (datafrom Steckleret al. [1998]).(b) Topographic profilesfrom Lake Mead. Each profile
corresponds to a lakebottomsurveyduringtheperiodof lakefilling afterdammingthe ColoradoRiver in 1935.
Lakelevelsat eachsurvey vary;depthof rolloverforeachprofileis indicatedin Table1. Position
of downlap of
eachprofileoccurs -12 kmfromtherivermouth,andflattening of profilebeyond12kmrepresents lakeinfill.Data
digitizedfromSmithet al. [1954](in Figure12.16of Graf[1984]).

theclinoform.
Also,it isevident
fromthestudyof sedimentationclinoforms.No singlemodelhascombinedthe elementsof water
on modemclinoformsthat the peak in sedimentaccumulation column energy distributionand sedimentdispersalinto the
oftenis detachedfromtheshoreline, particularly
in energeticcharacterization
of sedimentation
accumulation profilesandhow
shelfenvironments[Nittrouer
andWright,1994].Thelocationof they affect the shapeof clinoformsurfaces.Our modeling
theclinoformbreakisthereforeindependent
of theshoreline
and approachstartsfrom the basicobservationof the characteristics
is determined,
to a largeextent,by the distribution
of shear and processeson modernclinoforms:their shape,sediments,
stresses
in the watercolumn.Theseobservations
providethe overlying fluid flow characteristics,and sedimentationrate
foundationfor our sedimentdispersion
modelof clinoform profiles.The presentmodel is a step towardaddressing the
development. intermediate-
to long-termformationof stratigraphic
sequences
In summary,existingsedimentation
modelshave not ad- and the factorsthat controlthe shapeand sedimentdistribution
dressed
key issuesdealingwith the formation
of subaqueousontheclinoformsurfaces
thatmakeupthesediment
packages.
PIRMEZET AL.: CLINOFORMDEVELOPMENTBY ADVECTIONDIFFUSION 24,145

1.0

(A) • • .c•'• .......Rhine-


1951

0.8
"'"• __",,, -• ..<>..e..
Mead-
1948
- "% •,• • -: NewJersey(m5.4)
'x•. •[• -- Ganges
Yinfiection •x
>0.6 x • •• • •:• --- Shandong
0.6
'.......
"';':;'
.......................................................
'•'•-..........
X •.• ...... Amazon
' H=70
(ref
1)
".........,. •'•-•:••
...........
• .....Amazon,
H=70
(ref
'•
2)
",,. • :: ......
0.4
..................
•............................................................
;.•,....
X[ •.•.....
- Yinfiection<0.4 •, • • "•....
ß ....
0.2
- •,,,,•
• , -..• +• '-.,,•.

,.....

0.0 _ , . , . . ...._
-400
-300 -200 -100 0 •H 100 200 300 400
(B)

1.0

0.8

(.9 0.4

0.2

0.0
-200 -100 0 x/H 100 200
Figure2. (a)Normalized
topographic
elevation
profiles
overseveral
deltas
(references
fromwhich
profiles
were
digitized,
andclinoform
dimensions
areinTable1).Notelowelevation
ofinflection
point(x/H=O)
forlarge
deltas
versus
otherclinoforms.
(b)Bottomslopedistribution
of clinoform
profiles.

3.1. Conservation of Sediment


3. Model Description
The steadystateequationfor the conservation
of sediment
We model clinoform formationin two dimensionsusing the
dischargeis
conservationequations
for fluid and suspended
sedimentcon-
centrationand assumingsteadystatesedimentinput. In the tYqi
=0 (1)
model a sediment-laden river enters the basin with a mean ve-
locityU andmeanconcentration
C. It is assumed
thatriverand
basin waters are well mixed and that there are no external forces, where
qi isthesediment
discharge
in thexi direction.
Sediment
suchas waves,tides, or wind. Sedimentgrainsare advected dischargein the vertical,
laterallyat thesamevelocityasthefluidanddeposit
depending
on the balancebetweenthe gravitational
settlingand uplifting qz=(w+ Ws)c-
Kz•zz (2a)
vertical eddiesof turbulence.The basicelementsof the model are
described
belowandillustratedin Figure5. Symbolsaregivenin represents
the balancebetween
the verticalsettlingflux minus
the notation section. theresuspension
flux,whereKz is the coefficient
of vertical
24,146 PIRMEZ ET AL.' CLINOFORM DEVELOPMENTBY ADVECTION DIFFUSION

(A) •.o .......


,, •x ..............
1948
_ ,.: 1946

';,..•
.. Lake
Mead
......
1947
0.8
1945
1944
•" 0.6 - "'- • "'- 1941
,.,,.,• ... 1940
.......................
1939
• 0.4 1938
1937
0.2

0.0

-200 -1O0 0 x/H 1O0 200 300 400


1.0
Rhine
(B)
...... 1961
0.8
1951
- -- 1941
............. 1921
1885

0.0
0 100

1.0

(c) 0.8
Mississippi
1947
...... 1921
.......
--::::-
........1874
, 1838
> 0.4

0.2

0.0
0 100 200
x/H

(D) o.8 '"' - New Jersey


....... m2.3
.... m3
'• 0.6 .......... 1115

:• 0.4 - m5.2
m5.3
..... m5.4
m6

0.2
0.0
. -
- v• _

-lOO o x/H lOO


Figure3. Normalizedelevationprofilesfor clinoformsurfaces surveyed
at multipleinstants
duringbasinevolution
on (a) LakeMead(datafromSmithet al. [1954]),(b) RhineRiverdelta(datadigitizedfromKenyonand Turcotte
[1985,Figure9]), (c) MississippiRiverdelta(datadigitizedfromKenyonandTurcotte[1985],Figure12]),and(c)
New JerseyMioceneclinoforms(datafrom Steckleret al. [1998]). Figures3e-3h showthe maximumclinoform
front slope(slopeat the inflectionpoint)andseawardmigrationdistanceof the inflectionpointas a functionof
time.
PIRMEZ ET AL.' CLINOFORMDEVELOPMENTBY ADVECTIONDIFFUSION 24,147

Thesolutionof (1) withthedischarge


governed
by equations
1.2
(E) LakeMead - 3O (2a) and(2b) describes the suspended
sediment
concentration
field,givenappropriateboundary conditions
andprofilesfor u,
w,andKz.In thecurrent
version
ofthemodel
we usethelayer-
averaged
conservation
equations.
The layer-averaged
equation

Table 1. Scales of Natural Clinoforms

. • -- Clinoformadvance H, rn Maximum Rollover


slope,deg depth,rn
Clinoformslope
0.0
i i 0 NewJerseya
0 12
o1(Green-2,
36Ma)b
Years (from 1937) rn6(Pink-3,24 Ma) 100 1.17 50
mS.6(Blue-2,22 Ma) 155 1.2 85
rn5.4(Sand,19.5Ma) 140 1.27 70
6
(F) Rhine mS.3(18.7 Ma) 137 1.96 68
mS.2(Ochre,18.3Ma) 130 1.37 65
rn5(Green-1,16.6Ma) 119 1.0 60
rn4(Pink-2,14.1Ma) 216 2.77 87
rn3(13.6Ma) 232 3.1 87
m2.3 (13.2 Ma) 202 2.86 138
Lake Mead c

m 2 % 1935 d
1937 24 0.09 15
1938 41 0.37 20
1939 59 0.32 12
o 1940 63 0.33 10
0 I I ,
0 20 40 60 8o 1941 82 0.46 10
1944 93 0.49 7
Years (from1885) 1945 88 0.48 7
lO
1946 88 0.67 7
1.5-(G) Mississippi 1947 84 0.74 8
8 1948 86 1.23 10

MississippiDelta e
1838 106 1 5-10
1874 106 1.4 5-10
1921 99 1.23 5-10
1947 99 1.14 5-10
x0.5 Rhine Delta •
1885 48 6.28 5
1921 69 1.04 2
1941 67 1.3 0
0.0 o 1951 67 1.4 0
0 40 80 12o 1961 67 2.31 5
Years (from1838) AmazonDelta f
Kuehl
etal. [1986]
g 70 0.16 38
- (H) NewJersey Alexanderet al. [1991] 60 1.23 40
? Nittrouer
etal.[1996]
h 50 0.51 40
ShandongDelta
Alexanderet al. [1991] 51 0.29 30
GangesDelta
/ 1,00 Kuehlet al. [1989] 100.4 0.27 25
Model Examples
Flat basin
D = 45 gm 8.5 0.1 -
D = 90 gm 8.5 0.3 -
Slopingbasin
• 0 t= 2 17 0.25 -
0 5 10 15 20 25 t = 10 30 0.9 -
t= 15 35 1.2 -
Years(myfrom.36Ma)
aNamesandagesofunconformities
afterSteckleret al. [1998].
Figure 3. (continued) bBasalsurface,
slope= 0.1ø- 0.2ø.
cDatafromSmithet al. [1954],measured
in Figure12.16of Graf
[1984].
turbulent diffusion. We assume that horizontal diffusion dPre-darn
riverprofile,
meanslope
= 0.02ø.
(Kx•/&) isnegligible
such
thatlateral
sediment
dispersal
isdue eDatameasured
fromprofilesin KenyonandTurcotte[1985].
solelyto advection: rDatameasured
ondifferent
profiles
published
byseveral
authors.
gProfile indicatedasref. 2 in Figure2b
qx = u c (2b) hProfileindicated
asref.1 in Figure2b
24,148 PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION

(A) Amazon shelf of sedimentfrom the water columnby depositionor erosion,


o
respectively.
We usethe formulationof McCaveandSwift[1976]for R(x):
50

lOO
Water depth (m) R(x)=Wsco!l-•cl
p (4)
Equation(4) resultsfrom work on cohesivesediments[Krone,
lO
_ Accumulation 1962,1993].Theterm(1- XO/Xc)
reflects
thebalance
between
floc
rate (cm aggregationas clays settle into zonesof higher concentration,
andfloc disaggregation
andresuspension by the flow turbulence.
5
Wherex0 > Xc,deposition
doesnot occur,andthesediment
is
advectedlaterally; at present,erosionis not simulatedin the
0 I I I ' I i I I model.WherexO<Xc, deposition
resultsfromthe net balance
0 40 80 120 160 betweenthe settlingflux Ws co and the upwarddirected
Distance(km) resuspension
flux determined
by theratioxo/x
c. Thevaluep
(B) Gulf of Bohai(YellowR.) represents
the probabilitythat a particlewill remainat the bottom
oncedeposited andis hereassumed to be unity.
30 Water
depth 3.2. Basin Floor Evolution
: S- profile!
Assuming
the sediment
hasa bulkdensityPBat the basin
•:• SE
-profil
:1
floor in the absenceof sedimentcompactionand tectonicsub-
70 I sidence,the changein bottomtopography
with time is described
0.8 by theone-dimensional Exnerequation

0.4

0 whichaftersubstituting
the right-handsidewith (3) and(4) be-
comes
0 100 200
Distance(km)

o
(C) Mid-AtlanticBight •'=PBLWsco
1- (6)
o.1

0.2 Therelationof co to C is determined


by firstassuming
thatthe
verticaldistributionof suspended
sedimentconcentration
follows
the open-channel Rouseprofile[Graf 1984,p. 173],
0.5

C(Z)
=Crefh-z h-zre
f
1.5
andsecond,
assuming
thata constant
factor,ro,relates
thenear-
bedandaverageconcentrations:
i

0 5 10 20 25 c• =roC (8)
Distance(km)
Strictly,ro depends
on the flow conditions,
in particular
on the
Figure 4. Sedimentationrate measuredacrossmodem clino- exponentin (7). For simplicity, we assumea constantvalue
forms:(a) Amazondelta(basedon 21øpb data,modifiedfrom ro= 1.5, which simplifiesthe calculations
considerably
and
Kuehlet al. [1986];(b) Shandong
delta(basedon 21øpbdata, appearsto be supportedby experimentswith open-channel
modifiedfrom Alexanderet al. [1991]); and (c) total sediment
suspensionsdescribedby Garcia and Parker [1987] andParker
flux on the easternU.S. shelfand slopemeasuredwith sediment
traps(modifiedfrom Biscayeand Anderson[1994]), interpreted et al. [1986]. Equation(8) allows the changein bed elevation
asa proxy for sedimentationrate. described
by (6) to be relatedto the layer-averaged
conservation
of sedimentdischargedescribed by (3).
3.3. Conservation of Fluid
for the conservationof suspendedsedimentis (for a full
derivation,see,for instance,
Parkeret al. [ 1986]): The layer-averagedequationfor conservation
of fluid,
•UCh •YUh
tYx- R(x) (3) •Yx -0 (9)

whereUCh is the layer-averaged suspendedsedimentdischarge determinesthe lateral variation of water discharge,with h(x)
andR(x) isthenetverticalflux of sediment
at anypointalongthe representing
the waterdepthand U(x) representingthemeanflow
profile.Here we modelR(x) to represent
the removalor addition velocity.
PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION 24,149

The near bed shear-stress is rates measured in natural systems with model clinoforms
2 generatedby the model.However,the modelallowsfor the
r b = s pCo Uloo (1O) comparisonof shapesbetweennaturaland model clinoform
surfaces
andthe interpretation
of variationsin shapewith regard
whereul00istheflowvelocityata reference
level1 m abovethe
bed and s is a dimensionless
scalingfactor. The flow velocity
to changesin basin physiography, grain size, and the spatial
distributionof near-bedshearstresses.
The effectsof basinshape,
nearthe bed,u•00,is determined
fromthe meanvalueU by
assuminga logarithmicvelocityprofile, sealevelchanges,andinteraction
betweenthevelocityfield and
the evolving topographyare also examinedduring model
executionthroughsimulatedtime stepping.
tt(X,Z)
U•CD((h-z)+zo
=
/() In
z0
(11)
3.4. Model Execution

with the heightabovethe bed,(h-z), setto 1 m. For a particularmodel run a set of initial parametersis
Sedimentdepositionis controlledvia a criticalvalue for the specified,includingthe initial bathymetry,the initial mean
near-bedshearstress,tc, usingthe suspension-bedloadthreshold sediment discharge (U0, COandh0),andthegrainsettling ve-
[e.g.,Middleton, 1976] locity(or its Stokes-equivalent
graindiameter).The criticalshear
2 stressis then calculatedfrom (12a) (Figure 5). At any particular
r c = pu, c (12a)
time stepthe meanflow velocityand near-bedshearstressare
first determinedfrom the dischargeat the model entranceand
u,c = Ws (12b) using (9)-(11). The flow velocity distributionwith depth,

whereu,c is thecriticalshearvelocity.In thispaper,thesettling calculatedfrom (11) and the sedimentconcentrationprofiles


velocity WS is reported through its Stokes-equivalentgrain implied by (7), are shownin Figure 5a. The settlingand
diameterfor simplicity. resuspensionfluxes(Figure5b) arethencalculated by computing
In orderto providefor properscalingin the modelwith re- the two terms in the right-handside of (6). Their difference,
spectto naturalsystems,that is, to relatemodel time to actual scaled by PB,is thesedimentation rateprofiledh/dt(Figure5c).
time, thereis a needto introducethe scalingfactors in (10). The The sedimentation rate profile leadsto a sedimentlayer with
factors equatesto the ratiobetweenthe averagemagnitudeof the thicknessdefined by an arbitrarytime step (Figure 5d). The
instantaneous velocityfield andthe residualvelocity: temporalevolutionis simulatedby recalculating the flow field
with the newly formedbasinbathymetryresultingfrom changes
llulat in elevationdueto deposition duringthe previoustime stepand
t
any changesin sealevel height.
ludt
t

4. Results
The residualadvectionvelocity(denominator in (13)) in a spe-
cific basinis thoughtto be very small, althougha net seaward Figures6 and 7 illustratethe clinoformgrowth and progra-
dationsimulatedby the modelunderseveraldifferentconditions.
flow probablycan be assumed,giventhat a sedimentary deposit
progrades outwardfrom a river mouth.The averagemagnitudeof In the simulations,dischargeat the model entranceis kept
constant,and the initial basin floor is linear, with or without a
the velocity (numeratorin (13)) is primarily a functionof the
energylevel of a particularenvironment.The factors obviously bottom slope. The maximum height of the clinoform,or the
dependson the time spanconsidered in the integrationandon the minimumwaterdepth,is controlledby the near-bedshearstress
significanceof extreme events, that is, on the magnitude- at themodelentrance
andby thesediment
grainsizethroughXc
frequencydistributionof the velocity.The valueof s shouldbe a (equation(5)). For simplicity,waterdepthat the modelentrance
characteristicof a particularenvironmentand shouldbe greater is initially setto 10 m deeperthanthe depthof the criticalshear
for high-energyenvironments. An estimateof the magnitudeof s stress
fordeposition
fora givenu(x=0)andWs in all modelruns.
can be obtainedfor existingclinoformsystems.For instance,on Key resultsand comparisons
to naturalsystemsare described
below.
the Amazon delta a mean residual seaward flow of the order of
10'2 m/s over a tidal cycle is estimatedfrom physical
4.1. C!inoform Growth
oceanographic models using tidal and shelf current
measurements
[Geyeret al., 1996]. Time seriesof instantaneous Clinoformsaggradeand progradeundera sedimentation-rate
currentsindicate velocities of the order of 1 m/s near the bed, maximumthat evolveswith clinoformgrowth.Above an initially
suggestings ---10[dewellet al., 1993; Geyer et al., 1996]. The horizontalbasinfloor the near-bedshearstressis everywhereless
advectivesedimentand flow dischargemodeledthrough(3) and than the critical shear stress. Sedimentation rate falls off from a
(9), respectively, determinesthe rate at which sediment maximum at the model entrance as a function of the sediment
accumulates. In the model the same flow exerts a shear stress advectionand settlingvelocities(Figures6a and6b). Duringthis
nearthe bed, calculatedthrough(10) and (11), which, in tums, initial phase,clinoformgrowthis dominantlyaggradational. As
determineswhetherthe sedimentgrainsarekeptin suspension or deposition raisesthe basinfloorandwaterdepthdiminishes, flow
whetherthey are removedfrom the water column and become velocityincreases to maintaina constantdischargeat the model
part of the substrate.Because the instantaneousvelocity entrance. As a result, near-bed shear stress increasesto the
experiencedby the sedimentgrainsis, in general,higherthanthe criticalstress,and sediments
bypassthe region.A sedimentation
residualvelocity,the accumulationratescalculatedin the model rate maximum forms over the evolving foreset slope, and
will also be higher than the long-term accumulationrates clinoform growth reaches an equilibrium form, becoming
measuredin a specificsetting.Thus, at this stagewe cannot dominantlyprogradational (Figures6a and 6b). As clinoforms
perform direct comparisonsof accumulationand progradation prograde,a compositetopset surfaceis formed at a constant
24,150 PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION

(A) VelocityField& CalculatedConcentration


Profiles depthwherexb= xc.Theselatterstages
approximate
thesediment
0 Clino
model
run:
Example._Ramp8
Qtim©=9 accumulation rateprofilesandclinoformgeometriesobservedon
modemdeltas(e.g., Figure4). The simulationoveran initially
flat surfaceprovidesa simpletestof thepresentmodel,because it
shows that clinoforms can be generatedunder physical
conditions independent of preexisting
topography,
thatis, solely
asa resultof gravitationalsettlingunderan advectiveflow field.

$2oSediment 4.2. C!inoform Shape


Model clinoform surfaces exhibit the characteristic foreset-
topset rollover and foreset-bottomset
flattening observedin
veoci naturalclinoforms.
The rollover(e.g.,Figure6a, left-handside)
resultsin the modelfrom the gradualincreasein sedimentation
0 ' 1(•00 2000 rate as shearstressesdiminishover the foreset.The flattening
Distance(m) resultsfrom the gradualdecreasein sedimentationrate over the
bottomsetas the water columnis depletedof sediment.In all
(B) Sedimentation
Fluxes casesthe clinoformshapeis asymmetrical,that is, slopeincreases
1.5
morerapidlyacrossthe rolloverand decreases moreslowlyfrom
SerflingFlux theforesetto thebottomset(e.g.,Figure7a, right-hand.side).

4.3. Grain Size Effects

ResuspensionFlux
Increasinggrainsizeresultsin a steeperclinoformfrontwith a
",,t=9",, more abrupt clinoformrollover and foreset-bottomset transition.
rv' Fastersettlingassociated with largergrain sizesleadsto narrower
0.5
sedimentation rate profiles,with depositionincreasinglyfocused
on the upper foreset (compareFigures 6a and 6b). Relatively
steep foreset slopes are observed,for instance,on Miocene
clinoforms buried beneath the New Jersey continental shelf
o
o lOOO 2000 where sedimentsare predominantlysand (Figures 3d and 3h)
Distance(m) [Steckleret al., 1998]. Muddy clinoformsforming the Amazon
delta, in contrast,displayvery gentleforesetslopes,of the order
(C) Sedimentation
Rate of 0.5ø or less(Figure2).
,,_

0.5
4.4. Basin Physiography

0.4 The shapeof the receivingbasin,in particularthe initial slope,


affects the evolution of the clinoform shape. An initially flat
basinfloor leadsto the developmentof an equilibriumclinoform
surfacethat is constantthroughtime (Figures 6a and 6b). An
rv' initially sloping basin floor results in a foreset slope that
0.2
increaseswith time (Figure 6c). The latter is the consequence of
more rapid seawardexpansionof the flow field as the relief
0.1 betweenthe topsetand bottomsetincreasesbasinward.This type
of foreset steepeningis clearly observedin clinoformsof the
0 . Lake Mead delta (Figures3a and 3e), where repeatsurveyswere
o ' ooo carriedout sincethe dammingof the lake, showingprogressively
Distance(m)
steeper clinoform fronts advancing over the Colorado River
(D) Modelclinoforms valley (meangradientof 0.02ø;Table 1).
o

lO

Figure 5. (opposite)Exampleof modelexecutionat a given


time step. (a) Computedvelocity and sedimentconcentration
•'20, ••••_•/t= 10 distributions.The vertical velocity distribution follows a
logarithmic distribution(equation (11)). The concentration
distributionis calculatedusingthe Rousedistribution(equation
(7)). The Rousedistributionis assumedanddoesnot playa role
30.
in themodelexceptin thedefinition
of ro in equation
(8). (b)
Computedsettlingflux andresuspension flux (termsin brackets,
equation (6)). (c) Sedimentationrate, computedfrom the
' ' 10•30 ' 2000 differenceof the curvesin Figure 5b. (d) Sedimentlayer
Distance(m) resultingfrom the sedimentation
rate curve.
PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION 24,151

(A) D = 45 !•m (Ws=1.8cm/s,Uo=0.3m/s,Co=0.01


)
Model Clinoforms: Flat initial bottom Sedimentation Rate
10 , 0.5

I i Maximum
Slope
=0.06
deg

•0.2
0.1

20 --•-•:•- ••;'•..... .., - 0


0 le+04 2e+04 3e+04 le+04 2e+04 3e+04
Distance(m) Distance(m)

(B) D = 90 !•m (Ws=6.4cm/s,Uo=0.3m/s,C0=0.01)


Model Clinoforms: Flat initial bottom Sedimentation Rate
10 1.5

' Max..SI;pe
=0.2deg 1.2

0.6 I',,A A :',.A A ;', A A ;', A A ',

0.3
',/\ :',,L :'," :VV\,
2O 0 ,
_ .__

0 le+04 2e+04 3e+04 0 1 e+04 2e+04 3e+04


Distance(m) Distance(m)

(C) D = 90 i•m, S = 0.1 deg(U0=0.3m/s,C0=0.01)


Model Clinoforms: Initialslopingramp Sedimentation Rate
10 , , 1.5 , ,

Slope = 1.2 deg


1.2

4O
0.3

50 ,, . ,
0 le+04 2e+04 3e+04 0 le+04 2e+04 3e+04
Distance(m) Distance(m)

Figure 6. Model clinoformsproducedundervaryingsedimentgrainsize' (left) clinoformsurfacesproducedafter


15 time stepsand (right) corresponding
sedimentation
rate curvesat eachtime step.Input parametersfor eachrun
are specifiedaboveeachgraph.

4.5. Sea Level Changes sedimentationfocusingmore sharplyon the foreset.As a result,


the foresetslope increasesthroughtime in a fashionsimilar to
The shapesof the model clinoformsand the stackingpatterns that displayedin the slopingbasinfloor model(Figure6c). With
of individual clinoform packages are clearly affected by falling sealevel, modelclinoformsprograderapidlybasinwardas
changing sea level. Rising sea level progressivelyincreases shoalingwater depth leads to sedimentbypassingof the topset
accommodation spaceabovethe topset,and clinoformsdisplay (Figure7c). In this instance,sedimentation rate profilesbecome
simultaneousaggradationand progradation(Figure 7b). As sea broader,andthe foresetslopedecreases. In the caseof falling sea
level rises, sedimentation rate profiles narrow, with level the compositetopset surfaceceasesto be flat and dips
24,152 PIRMEZET AL.: CLINOFORMDEVELOPMENTBY ADVECTIONDIFFUSION

(A) Constant Sea Level, InitialFlat Bottom


o
Model clinoforms BottomSlope

0.2

•:•10
• . • 0.1

20 , ß ß ß ß '......, 0 -. . . . '•, ,,
0 DiStance
(m) le+04 0 Distance
(m) le+04

(B) RisingSea Level, InitialFlat Bottom


Model clinoforms BottomSlope
2
• Sealevel
change 0.2

'20 •0 •0

2O

0 Distance
(m) le+04 0 Distance
(m) 1e+04

(C) FallingSea Level, InitialFlat Bottom


Model clinoforms Bottom Slope

ot 0.2

,,[,10
A8

20

Time (y)
0 30 60
28 ,
o
1•+04 o
Distance(m) Distance
(rn) Ie+04

Figure7. Modelclinoforms producedunder(a) constant


sealevel,(b) risingsealevel,and(c) fallingsealevel.
Insertshowsthesealevelcurveappliedduringthemodelrun(1/4 cycleforrunsin Figures7b and7c).Rightside
panelsshowcorrespondingbottomslopecurvesat eachtime step.Inputparameters commonto all threerunsare
initialslopeS = 0, D = 45gm,U0= 0.1m/sandCO= 0.02.
PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION 24,153

gently basinward. During a sea level rise the aggradational shorterbottomsettail when comparedto our model.While we do
topsetsdisplaya minor gradientthat tendsto becomehorizontal not includethe effectsof waves,currents,andtidesexplicitly, the
asthe clinoformsprograde(Figure7b). For the sameratesof sea model stressdistributionused here can be interpretedas being
level changeand sedimentdischargeat the model entrance,the causedby the integratedeffectsof all thesefactors.Clearly,more
compositetopsetsurfaceatop the clinoformpackageis steeper detailedmodelingwouldbe desiredin orderto determine the ac-
during a relative sea level fall in comparisonto that produced tual stressdistributionin any particularenvironmentandhow it
duringa sealevel rise.However,erosionwhichis not accounted may affectsedimentdeposition and erosion.In our modelwe
for in this version of the model is likely to affect the profile also assumea singlegrain size and neglectthe effectsof sedi-
shapesignificantly,particularlyduringthefallinglimb of the sea ment compaction.Thesesimplificationsmay be at the sourceof
level curve. someof the featuresof naturalclinoformsthat are not captured
by the currentversionof the model, suchas the tendencyfor
5. Discussion modelclinoformpackages to form a horizontalcomposite
topset
surface. Nevertheless,our model appearsto capture several
A basic feature of the current model is the feedback relation- fundamental characteristics of natural clinoform surfaces that are
shipbetweenthe evolvingtopographyandthe hydrodynamics of not producedby previousmodeling efforts. We now address
the water column.The flow field expansionin the model results some of the benefitsof the model by comparingnormalized
in an overalldecreaseof near-bedshearstresses with increasing model and natural clinoform surfaces.
water depths.As sedimentaccumulationproceeds,so doesthe In Figure 9 we plot normalized profiles of the Rhine,
bottom configuration,which, in turn, affectsthe distributionof Mississippi,andLakeMead c!inoforms togetherwith a sampleof
shearstresses.During the first executionstep, when the basin modelclinoforms.Comparison betweenthe surfacesin Figure9
floor has a constantgradient [h(x) - Sx], flow expansion would suggestthat the Rhine delta clinoformsare formed by
determined through(9) leadsto a decrease of meanflow velocity coarser sedimentthan either the Lake Mead or Mississippi
U proportionallyto I/x, and, consequently,near-bedshearstress clinoforms.Given informationon the basin physiography,the
calculated through(10) fallsoff as -1/x2. As the basinfloor model could be used to quantitativelyassessthe differencesin
evolvesand ceasesto be linear, the relationshipbetweennear- grain size and water column energy betweenthese systems.
bed shear stressand distancebecomesmore complex and is Comparingclinoform surfaceswithin the samebasincan alsobe
solvednumericallyin the model. facilitatedthroughthe model,For instance,the overallsteepening
The model velocity field is basedon the conceptof a river of the Lake Mead clinoformsappearsto resultprimarilyfrom the
enteringan enclosedbasin,with dischargebeingkept constant. progradationof the system over the sloping Colorado River
In natural systems,say over a continentalshelf or enclosedlake valley ratherthan from a changein sedimentgrain size. On the
basin, near-bedshearstressesgeneratedby stirringwaves and other hand, the abrupt steepeningof the clinoformson the New
tides will also diminishwith increasingwater depth [Nittrouer Jerseymargin seeInat the middle Miocene (from the time of
and Wright, 1994; Dyer, 1989]. The net transportof sediment surfacem5; Figuresla and3h) couldbe a resulteitherof change
due to waves over a period of time can be negligible,but the in the gradientof the underlyingsurfaceor of an increase in the
important effect of wave motion stirring is that once in sedimentgrain size (and associated highershearstressneededto
suspension, sedimentswill be advectedby evenvery smalllateral carry the coarser material) or still a combination of both.
currents[Grant and Madsen,1979]. In our model,the expansion Doublingthe grain size (e.g., from 45 to 90 mm) representsan
of the flow field and consequent decreaseof shearstresses with increasein settling velocity by a factor of about 3.5. In the
increasingwater depth is intendedto simulatethe long-term model, given the sameclinoformheight(-9 m; Figure 6), this
average effects of waves, tides, and wind-driven currents.For results in an increaseof the maximum equilibrium foreset
comparisonwith our model the depthdependence of the wave- gradientby aboutthe samefactor(0.06ø to 0.22ø;Figures6a and
stirringeffectson the near-bedshearstresscan be estimatedas 6b). The slope increasefor the middle Miocene New Jersey
follows.The horizontalorbitalvelocitydueto wavesnearthe sea clinoformsis from about 1ø to 3ø (betweensurfacesm5 and m4;
bed is [e.g.,Allen, 1985]: Table 1 and Figure 3). While the increasein sedimentflux
2•0• indicatedby the rapid middleMioceneprogradation
(Figure3h)
Uor
b=,4sinh(k
h) (14)would suggestthat an increasein sedimentgrain size is a
plausiblesourcefor the observedsteepening,the model results
with the angularfrequency0) definedby the dispersionrelation would suggestthat the steepeningcould alsobe accommodated
0)2= g k tanh(kz), wherek isthewavenumber
andA isthewave by the increasein gradientof the underlyingbasinfloor 8urface
amplitude.The shearstresscausedby wave orbitalvelocitiescan (Figuresla and 6c).
be computedusingthe samesquarelaw usedhere (equation The shapesof clinoformsoverthe largedeltasof theAmazon,
(10)). This wave-induced shearstressdecayswith depthat a rate Shandong,and GangesRivers is characteristicallydifferentfrom
thatis fasterthanin thepresentmodelanddecreases to negligible the other clinoformsexamined,with a broad, gently sloping
values,that is, insufficientto suspendpracticallyany sediment topsetand an inflectionpoint occurringat low elevationwith
grains, below about 30-40 m for a wavelengthof 50 m and respectto the overall profile (Figure 2). The shape of these
amplitudeof 1 m (Figure 8). Wave stirringis one componentof clinoformsystems
is reproduced
onlyin generalized
formin our
the total shearstressoccurringin naturalsystems,
with the effects model simulations. The simplified hydrodynamicsover the
of tides and wind-driven currentsfurther increasingthe stress model basin floor, and the sedimentationby simple advection
level and affectingthe way stressvarieswith depth. The model and gravitationalsettlingof a single grain size, cannotcapture
distributionof shear stressas illustratedin Figure 8 has an the complexityof thesehighly energeticclinoformSystems. The
importanteffecton the shapeof the clinoformsurfacesgenerated lateral advective flux simulated in our model is an extreme
by the model. For instance,clinoformsurfacesmodeledwith the simplification of the actual processesthat transportsediment
wave-stirring sty:essdistribution have a sharper rollover and across continental shelves [Nittrouer and Wright, 1994]. In
24,154 PIRMEZET AL.: CLINOFORMDEVELOPMENT
BY ADVECTIONDIFFUSION

2O The resultson clinoformgeometryunder varyingsea level


Uorb = A 2t•ro/sinh(•h) (Figure7) are consistent
with the clinoformstratalpatterns
in-
•o2 = g•:tanh(r•h) ferred by the conceptualsequencestratigraphicmodel
•c=2rd•,wavenumber [Posamentier et al., 1988]. In that model the geometryof
re=2rdT,angularfrequency clinoformsurfacesdependsprimarilyon the rate at whichnew
•,--wavelength,T=period spaceis producedfor sediments to accumulate
andontheinflux
of sediments.As relative sea level rises, during the lowstand
• •:(Uorb),
(wave
orbital
velocity) systemstract,the additionof spaceat thetop of the clinoforms
• •,=100 would result in a progressivechangefrom progradingto
II 8 aggradingpackages.
Conversely,
duringa relativesealevelfall,
•=..
•• ';(U),
Uh
=(constant
discharge) waterdepthremainsconstant
or decreases abovethe foreset,
resultingin sedimentbypassand clinoform progradation.
Whetherdifferences
in the geometryof clinoforms,suchas the
systematicallydifferenttopsetslope betweenlowstandand
highstandclinoform packages(Figures7b and7c, respectively),
1•0 20 30 40 50 60 70 80 90 arepreservedin thegeologicrecordremainsto betestedagainst
Depth, h (m) naturalsystems.In general,the topsetsurfacesof highstand
clinoformswill be subjectto erosionassealevelfalls.However,
Figure 8. Near-bedshearstressdistributionas a functionof these surfacescould be at least in part preservedwhere the
depthas calculatedin our modelversusthat computedfrom a changein waterdepthis relativelysmall,preventing
marked
wave model. For our model, h(x)- 5+Sx, with S-0.004. The erosionof the topsets.
wave-induced orbitalvelocityUor
b is calculated
usingequation Where clinoformsprogradeacrossa dippingbasinfloor, the
(14),withA = 1. U(x=0)issetequalto Uorb(X
= 0; 3,= 100m) for foresetslopesteepens
with time (Figure6c). This occursbecause,
thiscomparison.Thenear-bedshearstress wascomputed in both
caseswiththe quadraticstresslaw (equation(10)). with a constantsedimentinput and advectionvelocity,sediment
is allowedto fill the availablespaceup to the waterdepthwhere
tb equalstc. The modelat presentdoesnot limit the foresetslope
from attainingan unrealistichigh gradient.Steepeningforeset
particular,hyperpycnalflows, eitherdirectlyfrom high sediment slopes of natural clinoforms would at some point fail.
concentration rivers [e.g., Wright et al., 1990] or after Presumably,mass wastingof the steepforesetswould act to
concentrationof fluvially derived sedimentsover the topsetas retreatthe clinoformfront, with the foresetreturningto smaller,
fluid-mudlayers[Kinekeet al., 1996], havebeenshownto play more stableslopes.Alternatively,the actionof stormwavesand
an importantrole in the growthof thesedeltas.In the caseof the currents could erode material from and/or induce mass-failures at
Amazon,seawardadvectionof fluid mudsfrom topsetto foreset the clinoform front. A possibleexampleis seenon the Rhine
appearsto occur at an approximatelyyearly cycle, with water delta clinoforms, which show a rapid decreasein gradient
columnenergysufficientlystrongto preventincorporation of the followed by a progressive increasethroughtime (Figure 3f).
fluid mudsinto the topset.The actualmechanisms of transportof Material removed from the clinoform front by mass wasting
these fluid muds are not yet well understoodand cannot be woulddepositoverthebottomset region,raisingthe seafloorand
simply characterizedby either mass-movementor diffusive thusdecreasing the relief betweentopsetand bottomsetregions
processes[Nittrouer et al., 1996]. The advectionof fluid muds of the clinoform.In the modelthis would resultin spreadingthe
appearsto involve stirring and transportby waves and tidal sedimentation rate profile over a broaderregion. The model
currents,as well as transportdue to excessdensityof the fluid resultssuggestthat masswastingis likely to becomemore
mudsdownslope[Kinekeet al., 1996]. significantrelativelylatein theevolutionof a clinoformpackage,
Our model presentlysimulatesa fluvial sourcethat remains as the systemadvances into deeperwaterand foresetslopesrise
fixed as clinoformsprogradebasinward.The depthof the topset to somecriticallimit. The Lake Mead deltais a goodexampleof
surfaceand clinoformrollover is dependenton the magnitudeof a systemthat showsclinoformsteepening duringprogradation
the shearstressin the water columnvis-h-visthe grain settling over an inclinedsurface(the pre-damColoradoRiver valley).
velocity. The model clinoforms,as in most natural clinoform Apparently,the Lake Mead delta sediments have largelybeen
surfaces,displaya clinoformrolloverat a waterdepthdepending ableto sustainclinoformfrontgradientsup to 1.2ø after 11 years
primarily on the energy conditionsprevailing in the water of deltagrowth(Figure3e). The earlyMioceneclinoforms off
column. In both modern and ancient clinoforms the shoreline and New Jerseyshow a rapid steepeningof the middle Miocene
"offiap break,"or subaqueous clinoformfront [Vail et al., 1991], clinoformfronts(Figure3h), with clinoformheightand foreset
can vary from being coincidentto >100 km apart(e.g., Figures anglebeing directlyrelated(Table 1; Steckleret al. [1998]).
1-3; Molenaar and Baird, [1992]). However, sequence Patterns of increasing slopethroughtime are detected internally
stratigraphicconceptshave not yet been fully modified to within some of the early Miocene sequences, and significant
accountfor the possiblevariations in distancebetween these steepeningalso is seen in the late Miocene to Pleistocene
features,nor for the causesfor thesevariations.The offiap break sequences off New Jersey[Steckleret al., 1993; G. Mountain,
and the shorelinepositionsare critical, for they are used to personalcommunication, 1996]. Rosset al. [1994] observeda
interpretthe extentof subaerialexposureof the shelfduringsea similartendencyof clinoformsurfacesto steepenwith time as
level falls and to predictfaciesdistributions.Natural clinoform they progradedinto deeperwater.
surfaces investigatedhere commonly show that the topset The model sensitivitytestsperformedthus far suggestthat
surfacesdip at a small angle. In energeticshelf environments, quantitative constraints on sedimentgrainsize,relativesealevel,
suchas in the Amazon and Ganges,the clinoformbreak can be sedimentsupply, and depositionalenvironmentmight be
widely separated from the coastline. obtained from clinoform geometry, particularly in instances
PIRMEZ ET AL.' CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION 24,155

1.0
- - Flat (D=45 pm)
ß ß ß Flat(D=90pm)
Sloping(t=2)
0.8
Sloping(t=15)
a a New Jersey (m5.4)
++ Mississippi-1947
• 0.6
o 0 • • • • Rhine-1951
t•O • • ......
LU 0.4

fl % o
0.2

• • '''....
Oo O
0.0 , i

-100
. i

0
,
100 200 300 400
•H

Figure 9. Comparisonof normalized model clinoformsand selectednatural clinoforms.Model clinoforms


correspond
to thoseshownin Figures6a-6c.Naturalclinoforms
aretakenfromthoseshownin Figure3.

where one or more of thesefactorsare known (e.g., Figure 9). rollover betweentopsetand foresetsegmentsof a clinoform,as
Geometricmodels suggestthat stratigraphicsequencesare a well as between the foreset and bottomset,resultingfrom the
complexand possiblynon uniquecombinationof thesefactors lateral variation in near-bed shear stressesaffecting sediment
plus subsidence[Burtonet al., 1987]. However, in assuminga accumulation
rate; (2) the separation
betweenthe point of
predefined clinoform shape, these models forfeit formative sedimentinput,or river mouth/coastline,andthe rolloverpoint is
constraintsprovidedby clinoform geometries.Diffusion-based dueto high shearstressespreventingdepositionon the topset;(3)
models suffer similar limitations. While it is a common the equilibriumforesetslopeincreases with increasinggrain size
observationthat steeperclinoformfronts in siliciclasticsystems input to the system;(4) foreset slope increasesprogressively
tend to be associatedwith higher-energyenvironmentsand through time where clinoformsadvanceover a sloping basin
coarsergrainsizes[Sangreeand Windmier,1977], therehasbeen floor (equilibriumforesetslopein thesecasesprobablyis limited
no attempt to quantify this importantproperty of clinoform by a combinationof processes suchas slopestability,sediment
morphologyin terms of grain size or facies.Diffusion models supply and accumulationrate, and erosion causedby storm
displaysteepfrontswhen the diffusioncoefficientis small,that events);and (5) temporalchangesin clinoformmorphologywith
is, in situationswhere sedimenttransportis interpretedto be varying space available for sedimentaccommodation,that is,
inefficient.In the presentmodel the clinoformfront slopewill relativesealevel changes.
vary with time if the clinoformadvancesinto deeperwater or if Perhapsmostsignificantly,the modelclinoformshapesresult
sealevelheightchanges. The equilibriumslopeachieved under from a balancebetween sedimentsettling velocity, horizontal
aninitiallyflatbasin,however,indicates
thatthemaximum slope advection, andthe turbulentenergythaipreventsdepositionand
depends ona combination of parameters,
in particular
thesettling keeps sedimentin suspension.The linkage between the flow
velocity (grain size) and the near-bedshear stress,the latter velocityfield and basinphysiography resultsin a feedbackloop
dependingon the flow velocityfield. If informationon sediment between sedimentationand the water column energy that is
propertiesis available,sayfrom samples or well logs,it maybe intrinsicto progradingclinoformsystems.
possibleto use the model in orderto place constraints on the
prevalentwatercolumnenergyconditions
duringdeposition.
Notation
6. Conclusions
b subscript
to indicatea near-bedquantity,b << H
We presenta sedimentdispersalmodel in which settling C meansedimentconcentration in waterlayer.
c suspendedsedimentconcentration (by volume).
sediment grainsareadvectedlaterallyor deposited
according
to
cb suspended
sedimentconcentration
nearthebed.
the shearstresslevelnearthe bed.The modelcouplessedimen- Co Dragcoefficient
(setto0.005).
tation with a fluid flow model that simulates a decrease in shear h layerthicknessor totalwaterdepth.
stresswith increasedwater depth, thoughtto representthe H heightof clinoform.
variationof shearstresses
in a basindueto the long-termaver- K von Karman'sconstant(setto 0.4).
agedeffectsof waves,tides, and wind-drivencurrents.While the K/ coefficient
ofturbulent
diffusion.
3, wavelength.
fluid flow modelusedhere is extremelysimple,the resulting k wavenumber,
model clinoformsreproducesome of the most fundamental p probabilityof sedimentparticle"sticking"to bed(=1).
characteristics
of naturalclinoformsystems
andprovidethebasis qi sediment
discharge
vector.
R rate of sediment accumulation.
for somepredictionsthat can be testedin the field and with
p fluid density.
further modeling. Among the characteristics of clinoforms PB bulk-density
ofdeposited
layer
(setto1800kg/m3).
reproducedhere are (1) the gradualchangein slope at the ro ratioof near-bed
to layer-average
sediment
concentration.
24,156 PIRMEZ ET AL.: CLINOFORM DEVELOPMENT BY ADVECTION DIFFUSION

s scaling
factor,ratiobetween
Urm
sandUresidual Greenlee,S.M., and T.C. Moore, Recognitionand interpretation of de-
S bottomslope. positional sequencesand calculationof sea-level changesfrom
x0 near-bed shear
stress. stratigraphicdata- offshoreNew Jerseyand AlabamaTertiary, in
Xc critical
shearstress
fordeposition. Sea-LevelChanges:An IntegratedApproach,editedby C.K. Wilgus
t time. et al., pp. 329-353, Soc.for Sediment.Geol.,Tulsa,Okla., 1988.
U meanhorizontalfluid velocity in layer. Hathaway,J.C. et al., Preliminarysummaryof the 1976 Atlantic Margin
u horizontalcomponentof fluid velocityvector. CoringProjectof the U.S. GeologicalSurvey,U.S. Geol. Surv.Open
u, shear
velocity
[u,= •l(xb/ p)]. File Rep., 76-844,217 pp., 1976.
u,c criticalshear
velocity
fordeposition. Jervey,M.T., Quantitativegeologicalmodelingof siliciclasticrock se-
u•00horizontalcomponentoffluidvelocity
1mabove
bed. quencesand their seismicexpression,in Sea-LevelChanges:An In-
Ws grainsettlingvelocity. tegratedApproach,editedby C.K. Wilguset al, pp. 47-69, Soc.for
co wave angularfrequency,2x/period. Sediment.Geol., Tulsa, Okla., 1988.
x horizontal distance from sediment source. Jewell, P.J., R.F. Stallard, and G.L. Mellor, Numerical studiesof bottom
z depthbelow sealevel. shear stress and sediment distribution on the Amazon continental
zo Nikuradse
roughness
parameter
(1/30thofgraindiameter). shelf,J. Sediment.Petrol., 63, 734-745, 1993.
Jordan,T.E., and P.B. Flemings,Large-scalestratigraphicarchitecture,
eustaticvariation, and unsteadytectonism:A theoreticalevaluation,
Acknowledgments. The authorsthank the supportof the National J. af Geaphys.Res.,96, 6681-6699, 1991.
Science Foundation through grant NSF-EAR-92-05623. Coauthors Kaufman, P., J.P. Grotzinger,and D.S. McCormick, Depth-dependent
L.F.P. and M.S.S. also acknowledgesupportfrom the Office of Naval diffusion algorithm for simulation of sedimentationin shallow
ResearchthroughgrantsN00014-97-1-0016 andN00014-95-1-0076, re- marine depositionalsystems,in SedimentaryModeling: Computer
spectively.Constructivediscussionswith Bill Ryan and Gary Karner Simulationsand Methodsfor ImprovedParameterDefinition,edited
were usefulin improvingour understandingof clinoformdevelopment. by E.K. Franseenet al., pp. 490-508, KansasGeol. Surv. Bull.,
Chris Paola providedvery constructivecommentsand criticismsthat Lawrence, 1991.
helpedimprovethispaper.Christopher Kendall,JohnCarey,Don Swift, Kendall, C.G.S.C., P. Moore, J. Stroble, R. Cannon, M. Perlmutter, J.
the AssociateEditor Herbert Wang, and an anonymousreviewerare Bezdek, and G. Biswas,Simulationof the sedimentaryfill of basins,
thankedfor theirthoughtfulcomments andconstructive
suggestionsthat in SedimentaryModeling: ComputerSimulationsand Methodsfor
helpedimprovethe manuscript.This is L-DEO contribution5797. ImprovedParameterDefinition,editedby E.K. Franseenet al., pp. 9-
30, KansasGeol. Surv. Bull., Lawrence, 1991.
Kenyon,P.M., and D.L. Turcotte,Morphologyof a deltaby bulk sedi-
menttransport,Geol. Sac.Am. Bull., 96, 1457-1465, 1985.
References Kineke,G.C., R.W. Steinberg,J.H. Trowbridge,andW.R. Geyer,Fluid-
mudprocesses on the Amazoncontinentalshelf,Cont.ShelfRes.,16,
Adams, C.E., Jr., J.T. Wells, and J.M. Coleman, Transversebedformson 667-696, 1996.
the Amazoncontinentalshelf,Cont.ShelfRes.,6, 175-187, 1986. Krone,R.B., Flume Studiesof the Transportof Sedimentin Estuarial
Alexander, C.R., D.J. DeMaster, and C.A. Nittrouer, Sedimentaccumu- ShoalingProcesses,110 pp., Univ. Hydraul.Eng. Lab., Berkeley,
lationin a modemepicontinental-shelf
setting:The Yellow Sea,Mar. Calif., 1962.
Geol., 98, 51-72, 1991. Krone, R.B., Sedimentationrevisited,in Nearshoreand Estuarine Cohe-
Allen, J.R.L., Principlesof PhysicalSedimentology,
272 pp., Allen and siveSedimentTransport, Coastaland EstuarineStud.,vol. 42, edited
Unwin, Winchester,Mass., 1985. by A.J. Mehta, pp. 108-125,AGU, Washington,D.C., 1993.
Bartek, L.R., P.R. Vail, J.B. Anderson,P.A. Emmet, and S. Wu, Effect Kuehl, S.A., C.A. Nittrouer, and D.J. DeMaster, Nature of sedimentac-
of Cenozoicice sheetfluctuationsin Antarcticaon the stratigraphic cumulationon the Amazon ContinentalShelf, Cont. Shelf Res., 6,
signatureof the Neogene,J. Geophys.Res.,96, 6753-6778, 1991. 209-225, 1986.
Bates, C.C., Rational theory of delta formation,Am. Assoc.Pet. Geol. Kuehl, S.A., T.M. Hariu, and W.S. Moore, Shelf sedimentationoff the
Bull., 37, 2119-2162, 1953. Ganges-Brahmaputra river system:Evidencefor sedimentby-passing
Biscaye, p.E., and R.F. Anderson,Fluxes of particulatematter on the to the BengalFan, Geology,17, 1132-1135,1989.
slopeof the southemMid-Atlantic Bight: SEEP-II, Deep Sea Res.,
McCave, I.N., and S.A. Swirl, A physicalmodelfor the rate of deposi-
Part II, 41,459-509, 1994.
tion of fine-grainedsediments in the deepsea,Geol.Sac.Am. Bull.,
Bowman, S.A., and P.R. Vail, Carbonatesedimentationprocessesin 87, 541-546, 1976.
PHIL, paper presentedat Am. Assoc.Pet. Geol. Annu. Cony., New Middleton, G.V., Hydraulic interpretationof sandsize distributions,J.
Orleans, Louisiana, 1993.
Geol., 84, 405-426, 1976.
Burton,R., C.G.S.C. Kendall,and I. Lerche,Out of our depth:On the Mitchum, R.M., Jr., P.R. Vail, and S. Thompson III, Seismic
impossibilityof fathomingeustaticsea level from the stratigraphic stratigraphyand global changesin sea level, 2, The depositional
record,Earth Sci. Rev., 24, 237-277, 1987. sequenceas a basic unit for stratigraphicanalysis, in Seismic
Coleman,J.M., D.B. Prior, and J.F. Lindsay,Deltaic influenceson shelf Stratigraphy-Applicationsto HydrocarbonExploration,edited by
edge instabilityprocesses, in The Shelfbreak:Critical Interface on C.E. Payton,pp. 165-184,Am. Assoc.Pet.Geol., Tulsa,Okla., 1977.
ContinentalMargins, edited by D.J. Stanleyand G.T. Moore, pp. Molenaar,C.M., and J.K. Baird, Regionalstratigraphiccrosssectionsof
121-137, Soc. for Sediment.Geol., Tulsa, Okla., 1983. Upper Cretaceousrocks acrossthe San Juan Basin, northwestern
Dyer, K.R., Sedimentprocessesin estuaries:Future researchrequire- New Mexico and southwestem Colorado,U.S. Geol. Surv.OpenFile
ments,J. Geophys.Res.,94, 14,327-14,339,1989. Rep.,92-0257, 3 sheets,1992.
Flemings,P.B., and T.E. Jordan,A syntheticstratigraphicmodel of Nittrouer,C.A., and D.J. DeMaster,The Amazonshelfsetting:tropical,
forelandbasindevelopment,J. Geophys.Res.,94, 3851-3866, 1989. energeticand influencedby a large river, Cont. ShelfRes., 16, 553-
Garcia,M., and G. Parker,On the numericalpredictionof turbiditycur- 573, 1996.
rents,in Third InternationalSymposiumon River Sedimentation,ed- Nittrouer,C.A., and L.D. Wright, Transportof particlesacrossconti-
ited by S.T. Wang, H.W. Shen,andL.Z. Ding, pp. 1556-1565,Univ. nentalshelves,Rev. Geophys.,32, 85-113, 1994.
of Miss. Press,St. Louis, 1987. Nittrouer,C.A., S.A. Kuehl, R.W. Sternberg,A.G. FigueiredoJr., and
Geyer, W.R., R.C. Beardsley, S.J. Lentz, J. Candela, R. Limeburner, L.E.C. Faria, An introductionto the geologicalsignificanceof sedi-
W.E. Johns,B.M. Castro,and I.D. Soares,Physicaloceanography of ment transportand accumulationon the Amazon continentalshelf,
the Amazonshelf,Cont.ShelfRes.,16 (5/6), 575-616, 1996. Mar. Geol., 125, 177-192, 1995.
Gilbert, G.K., The topographicfeaturesof lake shores,U.S. Geol. Surv. Nittrouer,C.A., S.A. Kuehl, A.G. FigueiredoJr., M.A. Allison, C.K.
Annu.Rep.,5, 104-108, 1885. Sommerfield, J.M. Rine, L.E.C. Faria, and O.M. Silveira, The
Gilbert, G.K., Lake Bonneville,U.S. Geol. Surv.Monogr., 1, 438 pp., geologicalrecordpreservedby Amazon shelf sedimentation, Cont.
1890. ShelfRes., 16, 817-841, 1996.
Graf, W.H., Hydraulics of SedimentTransport,Water Resour.Publ., Parker,G., Y. Fukushima,andH.M. Pantin,Self-accelerating turbidity
Fort Collins, Colo., 1984. currents,J. Fluid Mech., 171, 145-181, 1986.
Grant, W.D., and O.S. Madsen, Combined wave and current interaction Poag,C.W., Depositionalhistoryandstratigraphic
referencesectionfor
with a roughbottom,J. Geophys.Res.,84, 1797-1808,1979. the centralBaltimoreCanyontrough,in GeologicEvolutionof the
PIRMEZET AL.:CLINOFORM
DEVELOPMENT
BY ADVECTIONDIFFUSION 24,157

United StatesAtlantic Margin, edited by C.W. Poag, pp. 217-264, Syvitski, J.P.M., J.N. Smith, E.A. Calabrese,and B.P. Boudreau,Basin
Van Nostrand Reinhold, New York, 1985. sedimentation andthe growthof progradingdeltas,J. Geophys.Res.,
Poag,C.W., and G.S. Mountain, Late Cretaceousand Cenozoicevolu- 93, 6895-6908, 1988.
tion of the New Jerseycontinentalslopeand upperrise:An integra- Tezlaff, D.M., and J.W. Harbaugh,SimulatingClastic Sedimentation,
tion of boreholedata with seismicreflectionprofiles, in Initial Re- 202 pp., Van NostrandReinhold,New York, 1989.
ports of the Deep SeaDrilling Project,editedby C.W. PoagandA.B. Thorne,J.A., On the scaleindependent
shapeof prograding stratigraphic
Watts,pp. 673-724, U.S. Gov. Print.Off., Washington,D.C., 1987. units:Applicationsto sequencestratigraphy,
in Fractalsin Petroleum
Posamentier,H.W., M.T. Jervey, and P.R. Vail, Eustaticcontrolson Geologyand Earth Processes,edited by C.C. Barton and P.R. La
clasticdeposition,I, Conceptualframework,in Sea-LevelChanges: Pointe,pp. 97-112, Plenum,New York, 1995.
An IntegratedApproach,edited by C.K. Wilgus et al., pp. 125-154, Thorne,J.A., andD.J.P.Swirl, Sedimentation on continentalmargins,II,
Soc. for Sediment.Geol., Tulsa, Okla., 1988. Applicationof the regime concept,in Shelf Sandsand Sandstone
Pratson,L.F., and B.J. Coakley, A model for the headwarderosionof Bodies:Geometry,Faciesand Distribution,editedby D.J.P.Swirl et
submarinecanyonsinducedby downslope-eroding sedimentflows, al., pp. 33-58, BlackwellSci., Cambridge,Mass.,1991.
Geol. Soc.Am. Bull., 108,225-234, 1996. Thorne, J.A., E. Grace, D.J.P. Swirl, and A. Niedoroda, Sedimentation
Pratson,L.F., and W. Haxby, What is the slopeof the U.S. continental on continentalmargins,III, The depositionalfabric: An analytical
slope?,Geology,24,3-6, 1996. approach to stratification
andfaciesidentification,in ShelfSandsand
Pratson,L.F., W.B.F. Ryan, G.S. Mountain, and D.C. Twichell, Subma- SandstoneBodies: Geometry,Facies and Distribution,edited by
rine canyon initiation by downslopeeroding sedimentflows: Evi- D.J.P. Swirl et al., pp. 59-87, Blackwell Sci., Cambridge,Mass.,
dence in Late Cenozoic strataon the New Jerseycontinentalslope, 1991.
Geol. Soc. Am. Bull., 106, 395-412, 1994. Vail, P.R., Seismicstratigraphyinterpretation
usingsequencestratigra-
Reynolds, D.J., M.S. Steckler, and B.J. Coakley, The role of the phy, 1, Seismicstratigraphyinterpretationprocedure,in Atlas of
sedimentload in sequencestratigraphy:the influence of flexural SeismicStratigraphy,editedby A.W. Bally, pp. 1-10, Am. Assoc.
isostasyandcompaction, J. Geophys.Res.,96, 6931-6949, 1991. Pet. Geol., Tulsa, Okla., 1987.
Ross, W.C., Modeling base-leveldynamicsas a control on basin-fill Vail, P.R., J. Mitchum,R.M., R.G. Todd, J.M. Widmier,S. Thompson
geometriesand facies distribution:A conceptualframework, in III, J.B. Sangree,J.N. Bubb,andW.G. Hatlelid,Seismicstratigraphy
QuantitativeDynamic Stratigraphy,edited by T.A. Cross,pp. 387- and global changes of sea level, in Seismic Stratigraphy:
400, Prentice-Hall,EnglewoodCliffs, N.J., 1989. Applicationsto HydrocarbonExploration,editedby C.E. Clayton,
Ross,W.C., B.A. Halliwell, J.A. May, D.E. Watts, and J.P.M. Syvitski, pp. 49-212, Am. Assoc.Pet. Geol., Tulsa, Okla., 1977.
Slopereadjustment: A new model for the developmentof submarine Vail, P.R., F. Audemard,S.A. Bowman,G. Einsele,and G. Perez-Cruz,
fansandaprons,Geology,22, 511-514, 1994. The stratigraphicsignaturesof tectonics,eustacyand sedimentation,
Sangree,J.B., and J.M. Windmier, Seismic stratigraphyand global in CyclesandEventsin Stratigraphy, editedby G. Einseleet al., pp.
changesin sealevel, 9, Seismicinterpretationof depositional facies, 617-659, Springer-Verlag, New York, 1991.
in SeismicStratigraphy:Applicationsto HydrocarbonExploration, Van Wagoner, J.C., R.M. Mitchum, K.M. Campion, and V.D.
editedby C.E. Payton,pp. 165-184,Am. Assoc.Pet. Geol., Tulsa, Rahmanian,Siliciclasticsequencestratigraphyin well logs, cores,
Okla., 1977. and outcrops:Concepts for high-resolutioncorrelationof timeand
Smith, W.O., et al., Comprehensive surveyof Lake Mead, U.S. Geol. facies, 55 pp., Am. Assoc.Pet. Geol., Tulsa,Okla., 1990.
Surv.Prof Pap., 295, 1954. Wright,L.D., W.J. Wiseman,B.D. Bornhold,D.B. Prior,J.N. Suhayda,
Stanley,D.J., S.K. Addy, andE.W. Behrens,The mudline:Variabilityof G.H. Keller,Z.-S. Yang, andY.B. Fan,Marinedispersalanddeposi-
its positionrelativeto shelfbreak,in TheShelfBreak:Critical Inter- tion of Yellow River siltsby gravity-drivenunderflows,Nature, 332,
face on ContinentalMargins,editedby D.J. StanleyandG.T. Moore, 629-632, 1988.
pp. 279-298, Soc.for Sediment.Geol.,Tulsa,Okla., 1983. Wright, L.D., J. Wiseman, W.J., Z.-S. Yang, B.D. Bornhold, G.H.
Steckler, M.S., D.J. Reynolds, B.J. Coakley, B.A. Swirl, and R.D. Keller, D.B. Prior, and J.N. Suhayda,Processes
of marinedispersal
Jarrard,Modeling passivemarginsequencestratigraphy,in Sequence and depositionof suspendedsilts off the modem mouth of the
Stratigraphyand FaciesAssociations, editedby H.W. Posamentier et Huanghe(Yellow River), Cont.ShelfRes.,10, 1-40, 1990.
al., pp. 19-41, Int. Assoc.Sediment.,1993.
Steckler, M.S., G.S. Mountain, K.G. Miller and N. Christie-Blick,
Reconstruction
of Tertiary progradationand clinoformdevelopment
on theNew Jerseymarginby 2-D backstripping, Mar. Geol., in press,
1998. C. Pirmez, Exxon ProductionResearchCo., P.O. Box 2189, Houston,
Swirl, D.J.P., and J.A. Thorne,Sedimentation on continentalmargins,I, TX 77252-2189. (email:Carlos.Pirmez•EXXON.sprint.com)
A generalmodel for shelf sedimentation,in Shelf Sandsand Sand- M.S. Steckler, Lamont-DohertyEarth Observatoryof Columbia
stoneBodies: Geometry,Facies and Distribution,edited by D.J.P. University, 61 Route 9W, Palisades, NY 10964. (email:
Swirl et al., pp. 3-31, BlackwellSci., Cambridge,Mass., 1991. steckler•ldeo.columbia.edu)
Syvitski,J.P.M., and J.M. Alcott, GRAIN2: Predictionsof particlesize L.F. Pratson,Instituteof Arctic and Alpine Research,Universityof
seawardof river mouths,Comput.Geosci.,19,399-446, 1993. Colorado, 1560 30th Street, Boulder, CO 80303. (email:
Syvitski,J.P.M., andJ.M. Alcott, RIVER3: Simulationof river discharge Lincøln'Pratsøn•cøløradø'edu)
andsedimenttransport,Comput.Geosci.,21, 89-151, 1995.
Syvitski,J.P.M., and S. Daughney,DELTA2: Delta progradationand ba- (ReceivedJune26, 1997;revisedMarch31, 1998;
sin filling, Comput.Geosci.,18,839-897, 1992. acceptedApril 30, 1998.)

You might also like