A structural interpretation of the Fish Creek Slide (Lower Cretaceous), northern Alaska

AUTHO RThomas X.Homza BP Exploration(Alaska)Inc.,P.O.Box 196612,900East Benson Blvd.,Anchorage,Alaska 99519;present ad-dress:EnCana Oil &Gas (U.S.A.)Inc.,3601C Street,Suite 1334,Anchorage,Alaska 99503;tom.homza@Tom Homza is an exploration geologist for EnCana Oil &Gas (U.S.A.)Inc.He is currently working in EnCana’s Alaska Office,which he helped launch and he presently manages.He holds an M.S.degree and a Ph.D.in geology from the University of Alaska,Fairbanks and a B.S.degree in geology from the University of Vermont.Tom spent seven years at BP,where he worked in both exploration and develop-ment geology in several basins.ACKNOWLEDGEMENTSI thank BP Exploration (Alaska)Inc.,Phillips Petroleum Company (now ConocoPhillips),and Chevron U.S.A.(now ChevronTexaco)for per-mission to publish this work.Although the interpretations are my own,my understanding of the slide was enhanced by conversations with several colleagues at BP:Roy Allen,Greg Brewton,Mick Casey,Bret Chambers,Erik Hulm,Robert Humphreys,Sarah Lewis,Gary Pelka,and Jennifer Snow;Chevron:Art Bradley,Buddy Snell,Russ Warner,and Bob Wright;and Phillips:Don Griffiths,Allen Huckabay,and Frank Snyder.Paul Weimer,David House-knecht,Ken Bird,and Gil Mull reviewed early versions of this paper.I thank Steven Boyer,Russell Davies,and John Lorenz for their helpful edits.A structural interpretation of the Fish Creek Slide (Lower Cretaceous),northern AlaskaThomas X.HomzaABSTRACTModern three-dimensional seismic imaging enables a basic geo-metric description and kinematic interpretation of part of the Fish Creek Slide,a massive ( 4000km 2;1500mi 2)slope failure be-neath Alaska’s North Slope.The Fish Creek Slide is divided into a dominantly extensional organized slide zone in the west and a mostly contractional disorganized slide zone in the east.In the organized slide zone,a series of six organized slide blocks detached from the lower slope along steep escarpments during an inferred Albian earthquake focused at depth on a preexisting southeast-striking fault.Blocks were transported eastward into the basin as failure progressed catastroph-ically yet systematically by northwestward,en echelon,upslope,footwall collapse.During slide evolution,the primary flat detach-ment switched from the highly radioactive zone down stratigraphic section to the Lower Cretaceous unconformity.Although the seis-mic interpretation is hampered by a variety of slide-induced signal problems,several significant,previously unrecognized elements of the Fish Creek Slide are described:shear zones related to displace-ment variations in the slide,duplex structures in the slide blocks,postslide erosional features,and axial structural lows in slide blocks.INTRODUCTIONThe Fish Creek Slide (Weimer,1987)is a seismically defined dis-turbed zone beneath the northeastern National Petroleum Reserve–Alaska (NPRA)and the adjacent offshore area (Figure 1).Lower slope to basin deposits of the Albian Torok Formation (Figure 2)were dis-placed eastward during a catastrophic failure event to form the slide (Weimer,1987).The 450-million-bbl Alpine field underlies the easternmost part of the slide (Figure 1).The slide is important to oil explorers be-cause it directly overlies prospective Jurassic intervals and involves prospective Cretaceous rocks,both of which are being activelyAAPG Bulletin,v.88,no.3(March 2004),pp.265–278265Copyright #2004.The American Association of Petroleum Geologists.All rights reserved.Manuscript received March 19,2003;provisional acceptance June 13,2003;revised manuscript received October 20,2003;final acceptance October 27,2003.DOI:10.1306/10270303029explored.In May of 2001,Phillips Alaska Inc.(now ConcocoPhillips Alaska Inc.)and its partner Anadarko Petroleum Company announced several discoveries in the slide area (Figure 1).Six penetrations encountered three separate hydrocarbon accumulations.Detailed information about the discoveries is presently confi-dential;however,it is very likely that the Fish Creek Slide was critical to the exploration play either by providing a top seal (Fish Creek Slide detachment)for Jurassic reservoirs and/or by defining sediment dis-persal patterns for Cretaceous reservoirs between slide blocks.Constraining the geometry and kinematics of the Fish Creek Slide may help define prospect ele-ments both in the Jurassic and Cretaceous plays.The approximately 4000-km 2(1500-mi 2)Fish Creek Slide involves a variety of complex structures in a kidney-shaped region at an average present depth of approximately 6000ft (1800m).The slide has ele-ments characteristic of both an organized slide zone and a disorganized slide zone in Field’s (1981)classification of mass transport deposits.Only the former is discuss-ed because of its greater impact on the interpretation and space constraints (see Weimer,1987,for more on the disorganized slide zone).An east-west seismic line shows the two basic fea-tures of the organized slide zone (Figure 3):(1)north-northeast–striking (slope-parallel)elongate slide blocks and (2)interblock onlap fill (not discussed in detail,see Weimer,1987).Six individual slide blocks,as much as nearly 1km (0.6mi)thick,3km (2mi)wide,and tens of kilometers long,constitute the organized slide zone (Figures 3,4).In general,the slide blocks decrease in size and increase in structural complexity in the general transport direction:east,away from the Albian shelf edge.This trend corresponds to both the original taper of the Torok clinoforms and the generaleastwardFigure 1.Location map for the Fish Creek Slide (shaded)showing the 3-D seismic survey outline,well locations,and Al-pine oil field (hatchured).Recent exploratory wells are shown as X and as if they encountered hydrocarbons.The 3-D seismic survey does not cover that portion of the slide where the detachment lies at the highly radioactive zone level (dark shading);instead,the detachment in the survey is at the Lower Cretaceous un-conformity (light shading).The boundary between these two domains is covered only by 2-D seismic data (lines omitted)and is poorly constrained.Locations of Figures 3,5,7,and 8are shown.266A Structural Interpretation of the Fish Creek Slide (Lower Cretaceous),AlaskaFigure2.Simplified stratigraph-ic cross section of the North Slopeshowing the positions of major structural detachments and the progressive erosion of older stra-tigraphy northward,beneath the Lower Cretaceous unconformity (LCU,thick wavy line).HRZ= highly radioactive zone.The thick straight line indicates the Fish Creek Slide detachment and corresponds to the stratig-raphy that is truncated beneath the Fish Creek Slide detachment. This line is red where the de-tachment is at LCU level and black where it is at HRZ level.Homza267transition from coherent,extensional blocks of the or-ganized slide zone to internally deformed,contraction-al blocks of the downslope disorganized slide zone (e.g.,Cobbold and Szatmari,1991)(Figures 3,4).A steep ramp/escarpment characterizes the de-tachment along most of the western edge of the Fish Creek Slide and flattens abruptly to a mostly layer-parallel detachment that underlies approximately 98%of the disturbed zone.Velocity and seismic character variations throughout the slide make it difficult to pick consistently and precisely the detachment and inter-pret seismic data within and beneath the slide.The purpose of this report is to use new three-dimensional (3-D)seismic data to (1)describe the geol-ogy and geophysics of the southwestern part of the Fish Creek Slide,focusing on the detachment and slide blocks in the organized slide zone,and (2)propose an evolutionary model for the slide.SimilardescriptionsFigure 3.(a)Seismic line 188through the Fish Creek Slide showing the general eastward taper of the slide,four horizon picks,the Fish Creek platform boundary fault,the west Fish Creek well with time-based gamma-ray log and synthetic seismic trace,and the slide-induced velocity anomalies.See Figures 1and 4for location of the profile.(b)Seismic line 188flattened on the Sag River Formation reflection showing a clearer image of the slide blocks (labeled)and the Lower Cretaceous unconformity/detachment.Note the axial lows in blocks 4and 5,the apparent rollover structures on the west sides of blocks,and the polarity reversal of the detachment re-flection from beneath blocks to beneath the onlap fill.A post-slide event erosional channel into which block 5slumped is labeled ‘‘a.’’268A Structural Interpretation of the Fish Creek Slide (Lower Cretaceous),Alaskaof other parts of the slide must await further data and work.The two main conclusions are (1)that a tecton-ically triggered en echelon series of footwall collapse events,driven by dextral slip at depth,stepped north-westward sequentially and very abruptly to form the slide;and (2)that both the Aptian–Albian highly ra-dioactive zone and the Hauterivian Lower Cretaceous unconformity (Figure 2)served as the primary de-tachment in separate areas at slightly different times.The complexity of the Fish Creek Slide limits the scope of this study to a basic geometric description and kin-ematic model,and whereas the new data enable more detailed descriptions than previous two-dimensional (2-D)-based studies,many questionsremain.Figure 4.(a)Two-way time map of the top Sag River horizon pick with coherency data superimposed.The impression of the slide geometry on the Sag River reflection is a result of the velocity anomalies induced by the slide.Lines AA 0and BB 0indicate the position of shear zones in the slide blocks.BB 0directly overlies the Fish Creek platform boundary fault (see Figure 6).Note the subtle z shape of slide blocks 2,3,and 4and the structural lows that trend along the axes of blocks.(b)Time slice 1608ms showing the transition from the organized slide zone to the disorganized slide zone (approximate).The erosional low between blocks 4and 5is labeled ‘‘a.’’The serrated block edges near shear zones A and B are labeled ‘‘b,’’and areas of contraction are labeled ‘‘c.’’Note the thrust-belt-like map pattern of the duplex of Figure 8.(c)Simple jigsaw puzzle-type reconstruction of slide blocks.The blocks are treated as rigid units (i.e.,no internal deformation has been retrodeformed)and are reconstructed to a fixed escarpment edge to yield a best-fit geometry.This exercise yields an average extension for the Fish Creek Slide of approximately 65%and a clockwise block rotation of approximately 10j .Homza269Database/MethodsMore than30industry and government wells pene-trate the Cretaceous section within10mi(16km)of the Fish Creek Slide,and many of these penetrated the central part of the Fish Creek Slide(Figure1)(see Schindler,1988,for a history of the government pro-gram).Logs from all available wells help establish the regional geologic framework,and the five wells in the slide were used to constrain its internal elements.No additional lithologic descriptions or core analyses were carried out specifically for this study,but previous analyses were used and noted in the text.Some2000 line mi of multivintage2-D seismic data,including sev-eral hundred line miles of public data,were used to define elements of the Fish Creek Slide.The primary database is a high-quality(processed to a nominal fold of60),proprietary,530-mi2(1372-km2),3-D seismic survey collected by Northern Geophysical of America and Western Geophysical Company in1998and1999, respectively.About270mi2(700km2)of this survey covers the west-central part of the slide,accounting for about18%of the total slide area(Figure1).The3-D seismic data were interpreted on near-offset and coherency data(see Brown,1986for an in-troduction to coherency)using Landmark Graphics’SeisWorks package.Paradigm’s VoxelGeo product was used for volume-flattening/3-D visualization,and Land-mark Graphics’StratWorks software was used in well-log analysis.Previous WorkWeimer(1987)provided the first detailed description of the Fish Creek Slide.Although he was limited to 2-D seismic data,his primary conclusions are corrob-orated in the present study.Weimer(1987)interpreted the disturbed zone as a tectonically triggered submarine slide and accurately described its general geometry.Pei-kert(1985)interpreted the disturbed zone as a series of contourite mounds,an interpretation that was sound-ly refuted by Weimer(1987).The feature is widely referenced as a submarine slide(e.g.,Molenaar,1981; Kerr,1985;Bird,1987;Kirschner et al.,1992). GEOLOGIC SETTINGBoth subregional stratigraphy and structure influenced the geometry and kinematics of the Fish Creek Slide. Only pertinent aspects of the regional setting are discussed;a synoptic North Slope stratigraphic column is provided(Figure2)in lieu of a regional stratigraphic discussion(see Bird,1987;Hubbard et al.,1987;Moore et al.,1994,for regional geology).StratigraphyIn the study area,a Hauterivian rift-related unconform-ity truncates a series of Ellesmerian and Beaufortian sequence sedimentary rocks at a moderate angle(Fig-ures2,5).Only Beaufortian rocks are truncated in the 3-D area.During this erosion,the pre-Lower Creta-ceous unconformity rocks were primarily south-dipping (all coordinates in present terms);thus,the Lower Cretaceous unconformity erodes the older section pro-gressively northward.Above the Lower Cretaceous un-conformity,rift-related/transgressive sands of the Hau-terivian C-sand member of the Kuparuk Formation were deposited and subsequently transgressed by the Kalu-bik Formation and the highly radioactive zone of Bar-remian through Albian age(Pebble shale/lower Hue shale/gamma-ray zone(GRZ)equivalents in U.S.Ge-ological Survey terminology,e.g.,Bird,1987).The east-ward-younging highly radioactive zone represents the distal toesets of the thick(>5000ft;1.5km)east-prograding foreland basin sequence of the Aptian through Cenomanian Torok Formation(distal facies) and Nanushuk Group(proximal).No significant thick-ness of post-Cretaceous rocks is present in the study area.Of this stratigraphy,only the C-sand member of the Kuparuk Formation,the Kalubik Formation,the highly radioactive zone,and the slope-basin deposits of the Torok Formation are displaced as part of the Fish Creek ing the public data discussed by Weimer (1987),the timing of the Fish Creek Slide is estimated to be middle to late Albian.StructureTwo long-lived structural elements dominate the area: the Fish Creek platform(Kirschner and Rycerski,1988), a gently southeast-plunging,granitic,basement high, and the adjacent Nechelik trough,a broad basement low that is separated from the Fish Creek platform in this area by(1)a discrete,southeast-striking,down-to-the-northeast normal fault and(2)a north-striking, down-to-the-east splay of that fault(Figures3,5,6,7). Although these features are defined best by lower Ellesmerian isopachs and gravity and magnetics data, they continued to influence the geology of the area throughout Beaufortian and Brookian deposition,as270A Structural Interpretation of the Fish Creek Slide(Lower Cretaceous),Alaskaevidenced by stratigraphic and structural patterns at various levels.EFFECTS OF THE SLIDE ON THE SEISMIC SIGNALSeismic interpretation of the Fish Creek Slide is chal-lenging because the slide dramatically affects the seis-mic signal.Only pertinent effects are discussed:veloc-ity anomalies,polarity switching,and diffraction tails.In general,the highly radioactive zone is present in the slide blocks and absent between the blocks.Weimer (1987)correctly noted that where the highly radioactive zone is present,a seismic traveltime delay,or push-down,is observed relative to where the highly radioactive zone is absent (Figure 3a).This is caused by the relative velocities of the highly radioactive zone and Torok toesets (seismically slow)and the on-lap fill (anomalously fast).One way to decrease the push-down effect is to use modern software to flatten the seismic data on a subslide reflector (e.g.,Figure 3b).In a flattened section,polarity reversal becomes ap-parent as the detachment changes polarity from a peak(black)beneath the blocks to a trough (white)beneath the fill (Figure 3b).Further,many of the blocks are tall,narrow,and taper upward to form steeply dipping zones of varying acoustic properties.This causes mi-gration problems resulting in diffraction tails or non-geologic dipping seismic events.Diffraction tails are difficult to distinguish from onlap fill patterns and from apparent downlapping reflectors associated with roll-over structures (e.g.,Figure 3a).GEOMETRY OF THE ORGANIZED SLIDE ZONE The DetachmentThe fault separating the transported material in the slide from the undeformed material beneath the slide takes two basic forms:a steep ramp or escarpment mark-ing the edge of the slide and a flat detachment or de-collement that underlies most of the slide.The EscarpmentThe escarpment is clearly visible in the 3-D survey (Fig-ures 3,4).There,its map trace describes aleft-steppingFigure posite seismic line showing an oblique view of the interpreted ramp transition from the highly radioactive zone (HRZ)to the Lower Cre-taceous unconformity (LCU)detachment for the Fish Creek Slide.Note the angular trunca-tion beneath the detachment where the detachment lies at the LCU.The two seismic panels on the left are from 2-D lines.See Figures 1and 4for location.Homza271Figure6.Isochron from the basement to the top of the Lisburne Group showing the fault(white polygon)separating the relatively high Fish Creek platform(FCP)from the low Nechelik trough.Inset map at right fits into the box at the upper left and shows coherency data from the highly radioactive zone(HRZ)pick.Note the correspondence between the trends of the en echelon fault pattern and the Fish Creek platform boundary fault.The strain ellipse,resolved from the en echelon pattern and associated structures,indicates northwest-oriented dextral shear.Overlaid on the map are the Fish Creek Slide escarpment at the highly radioactive zone level and the north-oriented splay off the basement fault that has a similar trend to the Fish Creek Slide escarpment(see Figure7).Note that the Fish Creekplatform boundary fault directly underlies the escarpment for more than1mi(1.6km)in the central part of the survey.Figures1,4for location)show-ing the correspondence betweenthe deep north-trending splay ofthe Fish Creek platform bound-ary fault and the Fish Creek Slide(FCS)escarpment.The trends ofthese faults are roughly parallelwherever the splay is observed(Figure6).Slide blocks1and2are labeled.Note the rolloverstructures to the west of eachblock and that the rollover ad-jacent to block2is deformed bycompression,suggesting that itwas present before block1slid,and that the translation of block1caused this redeformation.272A Structural Interpretation of the Fish Creek Slide(Lower Cretaceous),Alaskaen echelon pattern of short(<2mi;3km)north-northwest–striking segments and southeast transfers or relay faults(e.g.,Peacock and Sanderson,1994).The deeper,clearly defined fault that separates the Fish Creek platform and the Nechelik trough lies directly beneath,and parallel to,a well-imaged transfer fault in the Fish Creek Slide escarpment(Figure6).Further,in some places where the escarpment does not directly overlie the deeper fault,it is possible to map a lower angle fault that links the two.The northwest portion of the deeper fault records Albian reactivation in the form of left-stepping,north-northwest–striking,en ech-elon transverse faults(i.e.,a southeast-striking,dextral, brittle,shear zone)(Figure6inset and unpublished data).The north-striking northernmost part of the es-carpment in the3-D survey directly overlies a down-to-the-east normal fault in the basement(Figure7)that splays off the southeast-striking part of the Fish Creek platform boundary fault.Together,these geometries strongly suggest kinematic linkage between the Fish Creek Slide and preexisting faults.Clearly,an additional primary influence on the po-sition of the Fish Creek Slide was the unstable geometry of the Torok slope,which trended generally north-south and dipped about5j eastward.The updip tip line(i.e., height)of the escarpment is mapped at a relatively low level of confidence because of the presence of com-plex and equivocal geometries associated with footwall collapse features and dipping reflections in the Torok (e.g.,Figure3).However,the oldest,continuous,post-slide,slope reflection is the same throughout the south-ern part of the slide(in2-D and3-D),thereby suggesting synchronous slip along the escarpment and catastrophic failure.The geometry of the escarpment outside the3-D survey is poorly constrained,but it appears en echelon with a decrease in dip along the southern edge of the slide.The Flat DetachmentFor oil and gas exploration,it is important to accu-rately pick the flat detachment on the seismic data because the placement of the detachment strongly influences the structural model.This,however,is dif-ficult in the Fish Creek Slide.Three primary tech-niques guided the pick for the Fish Creek Slide.First, well logs,synthetic seismic traces,and core descrip-tions from wells outside the slide were tied to reflec-tors that extend into the slide at various levels relative to the detachment,thereby constraining the detach-ment level.Second,hanging-wall cutoffs associated with rollovers were assumed to downlap directly onto the detachment surface.This is considered a robust assumption,and it provides solid constraints in dif-ferent parts of the slide.Third,well logs from within the slide were used to pick the detachment.Detach-ment picks are problematic insofar as there is no diag-nostic log signature for the detachment(e.g.,Figure 3a),the wells penetrate different elements of the slide, there may be loss or duplication of section in these wells,and the detachment may not reside at the same level in all places.Based on limited data,Weimer(1987)interpreted the detachment to lie near the base of the highly ra-dioactive zone throughout the slide.In the south and the east,outside the3-D survey,downlap onto the highly radioactive zone reflector is apparent,thereby supporting Weimer’s(1987)interpretation.However, this is not the case in the3-D area where the highly radioactive zone is well above,and rolls over onto,the detachment,which,in turn,clearly ties to the Lower Cretaceous unconformity(Figures3,5).Further,core from the N.Kalikpik well(Figure1),which penetrated block1and has a normal highly radioactive zone sec-tion,shows a shear/rubble zone with slickensides at the stratigraphic position of the Lower Cretaceous un-conformity.To the east of the3-D data,a broad area of no internal deformation characterizes the slide,and the detachment is difficult to pick.Farther east,the detach-ment is observed to step up to the highly radioactive zone,where it remains until the eastern line of zero displacement is reached.Thus,based on well data,seis-mic ties,and geometries,the detachment is interpreted to lie at the Lower Cretaceous unconformity level throughout the3-D area and near the base of the highly radioactive zone to the south and east of the3-D area (Figures1,5).Steps in the detachment from the highly radioactive zone to the Lower Cretaceous unconform-ity are detectable on some2-D lines to the south(Fig-ure5)and to the east of the3-D survey,but poor imaging precludes their detailed description.A Lower Cretaceous unconformity level detachment makes mechanical sense insofar as the Lower Cretaceous un-conformity represents a sharp boundary,where during the Cretaceous,weaker,less consolidated material di-rectly overlaid strong rock,making the unconformity a preferred slip surface.The dip on the detachment at the time of the slide is difficult to ascertain.A detachment-to-Nanushuk top-set isochron map(not shown)suggests that although the Torok slope deposits dipped eastward,the detach-ment actually dipped slightly( 1j)southwestward dur-ing Nanushuk deposition.A complex interplay amongHomza273foreland flexure,Nechelik trough subsidence,differen-tial compaction,and a complicated velocity field prob-ably contributes to this paradox and precludes a detailed analysis in this study.It is possible,however,that the slide moved updip along a gently southwest-dipping detachment.Geometric/Kinematic Interpretation of theOrganized Slide ZoneIn Field’s(1981)nomenclature,slide blocks are dis-crete,internally coherent,detached units.Collectively, multiple slide blocks are termed an organized slide zone.Slide blocks are subdivided into slump blocks (rotated slide blocks)and glide blocks(pure transla-tion,no rotation).Although Fish Creek Slide blocks are nearly pure glide blocks,parts of each block record internal deformation and/or rotation;hence,the gen-eral term slide block is used.Moderate-amplitude,horizontal-to west-dipping reflections characterize the six slide blocks in the or-ganized slide zone(Figures3,4,5,7).Each block is structurally dismembered to some degree.The pieces of each block have sharp boundaries,enabling a simple puzzle-like reconstruction both within individual blocks and among separate blocks(Figure4).Together,the six blocks define a left-stepping,en echelon fan geometry (Figure4).In the3-D area,block1is farthest northwest, thickest,and most parallel to the escarpment,whereas block6is farthest southeast,thinnest,and most dis-cordant.Blocks are less structurally complex at their south ends.Horizontal separation is greater between the escarpment and adjacent blocks than it is between blocks.Shear Zones and DuplexesBlocks2–4define a very subtle z-shaped sigmoid ap-pearance across the survey(Figure4).Three subtle yet pervasive southeast-striking linear shear zones intersect the southern bend in the sigmoid,and two of these(A and B)correspond to transfer segments on the escarp-ment.Lineation B further corresponds to the Fish Creek platform boundary fault.The lineations intersect the arcuate slide blocks at points of roughly equal curvature and of transverse failure.Figure4b shows jagged serra-tions on the western edge of several slide blocks,where those blocks are proximal to the lineaments.These serrations are interpreted as bending-induced,block-parallel extensional features associated with southeast-oriented dextral shear.An antiformal stack-type duplex structure(see Boy-er and Elliot,1982)of ductily deformed and thrust-duplicated Kalubik Creek and highly radioactive zone in block5is interpreted near the southeast limit of lineation B(and of3-D data)(Figures3,4).Here,I speculate that although the system’s mechanics pre-vented this part of block5from sliding en masse far-ther east(i.e.,it was effectively stuck),continued east-directed stress compressed and internally deformed the block.Block5is then interpreted to have served as a relatively stationary structural indentor,with respect to the relatively free face to the north,and shear oc-curred around its north side,along the shear zone B.The northern sigmoid bend is associated with the position of another indentor(Figures4,8).Here,where block3nears the northern limit of3-D data,the block is an east-vergent duplex with a floor thrust at the Lower Cretaceous unconformity and a roof thrust near the top of the highly radioactive zone reflection(Fig-ure8).The highly radioactive zone event is duplicated several times as horsts with fault-bend fold geometries (e.g.,Boyer and Elliot,1982;Suppe,1983).Flattening the entire volume of data and panning through lines and time slices on a seismic workstation spectacularly re-veals this duplex geometry.The part of block3south of the duplex slid farther east,unobstructed by this northern duplex(indentor),thereby defining a sinistral northern bend of the gentle z-shaped sigmoid.Taken together,these shear zones and duplexes suggest that the sigmoidal shape is related to variations in eastward displacement(greatest in the center of the survey)that are governed by the position of both stationary,in-dentor blocks and by the Fish Creek platform edge. Rollover(Hanging-Wall)AnticlinesSeveral slide blocks have well-imaged rollover anti-clines on their landward,west sides(Figures3,4,7), whereas no rollovers are observed on the east sides of blocks in the3-D survey.This asymmetry suggests a westward stepwise progression of failure by footwall collapse,wherein the present east side of a block was once the primary escarpment.If,in contrast,the pro-gression of failure was basinward,and one large block detached first along the present escarpment,then one would expect a more symmetric rollover distribution (i.e.,horst and graben architecture)upon subsequent extension because the single block would have a rela-tively constant thickness,and directional gravity sliding would not be preferred.Theoretically,true rollover anticlines are more pro-nounced where the hanging-wall block was detached274A Structural Interpretation of the Fish Creek Slide(Lower Cretaceous),Alaska。