seismic evaluation of differential tectonic subsidence...

ABSTRACT

Quantitative analysis of depth-converted reflectiontimes defines long-term differential motion acrossindividual structures in a central Appalachian interiorbasin known as the Rome trough. Differential motiondecreases exponentially with time. Rotation about ahinge defining the trough’s west margin reachedapproximately 37% of total displacement in about63–78 million years (m.y.). Displacement across thetrough’s faulted east margin occurred more rapidlyand reached 37% of the total in 13–51 m.y. A majorfault in the interior of the trough developed rapidlywith 37% of total displacement reached in from 16 to23 m.y. Longer term rotation across the west marginmay be due to its participation in the overall subsi-dence of the craton during the Paleozoic. The timespanned by the formation of the East-Margin andInterior faults was restricted to the Cambrian in thenorthern part of the area, but to the south, move-ment along the East-Margin fault continued throughthe Middle Ordovician.

The general effects of differential compactionand loading for a single lithology model are com-puted from the standard compaction and Airyisostasy equations. Depth-dependent compactionrequires that thicker strata over a hanging wall orsubsiding fault block undergo greater compactionthan occurs in thinner strata over the footwall orstructurally high areas. Seismic interpretation oftrough structures does not reveal the presence ofcompaction faults or of long-term differential com-paction. The observations suggest that in this inte-rior basin differential compaction of major strati-

376 AAPG Bulletin, V. 84, No. 3 (March 2000), P. 376–398.

©Copyright, 2000. The American Association of Petroleum Geologists.All rights reserved.

1Manuscript received June 16, 1998; revised manuscript received June22, 1999; final acceptance September 3, 1999.

2Department of Geology and Geography, West Virginia University,Morgantown, West Virginia 26506-6300; e-mail: [email protected]

This study was funded in part through U.S. Department of Energycontract DE-AC22-90BC14657. Comments on earlier portions of this work byRobert Milici, Robert Shumaker, Byron Kulander, Lee Avary, Kevin Biddle,and Jory Pacht were greatly appreciated. The paper also benefited fromdiscussions with Robert Shumaker, Richard Beardsley, Joe Lemon, andJovita Dominic. Appreciation is extended to Debbie Benson for herassistance in figure preparation. GTS Corporation provided some of theseismic data used in this study.

Seismic Evaluation of Differential Tectonic Subsidence,Compaction, and Loading in an Interior Basin1

Thomas H. Wilson2

graphic intervals was nearly complete prior todeposition of later sequences. Local isostatic com-pensation of differential loads across faults or faultblocks requires movement along near-vertical crust-penetrating faults and abrupt thinning by differen-tial amounts across the base of the crust. This possi-bility seems unlikely within the context of currentmodels of intracratonic extension. Analysis indi-cates that differential thickening of strata acrosstrough structures portrayed in reactivation historydiagrams defines long-term tectonic movementrather than a mixture of tectonic, compaction, andload-related motion. The analysis also suggests thatestimates of the time at which a given horizonentered the oil or gas window and estimates of thetotal depth reached by a horizon during subsidencemay also be in error if simple depth-dependentcompaction corrections are used.

INTRODUCTION

Modeling the tectonic development of sedimen-tary basins has become an integral part of explo-ration activities in the past two decades. Informationabout the timing of deformation relative to the tim-ing of hydrocarbon migration is critical to deter-mine if hydrocarbons will be trapped within agiven structure. Deformation history of individualextensional structures within a basin is the netresult of tectonic and thermomechanical processesresponsible for basin formation, combined withlithosphere loading, sediment compaction, andvariations of these factors throughout the basin(Ungerer et al., 1984).

Models of basin development have becomeincreasingly complex. Sleep (1971) noted that anexponential thermal-cooling model accurately por-trays subsidence histories of the Atlantic continen-tal margin and several basins within the interior ofthe North American craton. The model derived bySleep (1969) was initially used to explain mid-oceanridge topography as the result of thermal cooling ofisostatically compensated hot materials intrudedalong the ridge axis. McKenzie (1978) noted thatsimple thermal models require significant surface

erosion of thermally uplifted regions to account forthe elevated Moho observed beneath them. Usingthe North Sea as an example, McKenzie (1978)noted that significant erosion has not been observedand suggests that stretching may have initiallythinned the crust. Sclater and Christie (1980) pre-sented a detailed analysis in support of initial conti-nental stretching followed by thermal subsidence.

Long-term activity across fault zones and differ-entially subsiding fault blocks or margins is ob-served in seismic data. Yet, in the absence ofdetailed well-bore data required to specify porosity,density, lithology, and water depth across suchstructures, apparent and continued displacementresulting from compaction and loading (respective-ly) cannot be estimated accurately and separatedfrom tectonic subsidence. In this paper, differentialmotion across structures of the central AppalachianRome trough (Figures 1, 2) is measured from seis-mic profiles. Thickness variations computed fromdepth-converted reflection traveltimes are used todescribe relative motion across faults within thecomplex as a function of geologic time. Compactionand isostasy equations are solved to predict thepotential influence of compaction and loading forgeneralized single-lithology models. Observationsare compared to the generalized predictions. Thiscomparison is used to evaluate the degree to whichsediment compaction and isostatic responses to

differential loading influence differential motionobserved across faults and rotating fault blocks.

MAJOR BASEMENT STRUCTURE BENEATHTHE STUDY AREA

Extension during the Early Cambrian deformedthe Precambrian foreland basement in the centralAppalachian study area. The extensional basementcomplex formed by this event (Figure 2) is knownas the Rome trough (McGuire and Howell, 1963) orthe Eastern Interior aulacogen (Harris, 1978). TheRome trough is part of a larger system of grabens(Shumaker, 1986a, b; Thomas, 1991, 1993) that formedduring the opening of the Early Paleozoic Iapetusocean. Several wells drilled in the area along with lim-ited seismic data confirm the presence of the troughand a thick sequence of synrift rocks (e.g., Shumaker,1986a, b; Allen, 1988; Read, 1989; Ryder et al., 1992;Ryder, 1992; Shumaker and Wilson, 1996).

Interpretations of four seismic lines (Figure 1)define major basement structure beneath the studyarea (Figure 2). An extensive normal fault (Figure 2)known as the East-Margin fault (Shumaker, 1993)defines the east margin of the trough along itslength. With local exception, subsidence across thewestern margin of the trough occurs through rota-tion about a hinge zone (Figure 2) known as the

Wilson 377

Figure 1—Structure map of the central Appalachian foreland basin of North America. The locations of seismic linesanalyzed in this study are shown (modified from Shumaker and Wilson, 1996).

Ohio-West Virginia hinge zone (Ryder et al., 1992).In the southern part of the area along line 1 (Figures2, 3) the trough forms a simple asymmetrical graben;however, to the north the trough widens and is sub-divided by one or more interior faults. Minor inter-nal faults are ubiquitous within the trough interiorand are not represented at this scale.

Sonic logs from deep wells in the area (seeShumaker and Wilson, 1996) were used for velocitycontrol. A constant velocity function derived fromthe sonic logs was used to convert reflection travel-time to depth. There are a limited number of soniclogs in the region, but comparison between them ofthe average velocities for major stratigraphic inter-vals does not reveal significant velocity variation.Line 1 (Figure 3) is a reprocessed single-fold seismicline, whereas lines 2, 3, and 4 are multichannel (opti-mal 24-fold) Vibroseis™ data. The quality of the datashown in line 1 is typical of the quality of data avail-able for this study. Interpretations are based on themost coherent seismic reflections observed in thedata. Although different interpretations of theselines are possible, the major features of the interpre-tations are not expected to vary significantly.Identification of reflections on line 1 is based on syn-thetic seismic correlation (Shumaker and Wilson,1996). The strata associated with the reflectionsshown on line 1 range in age from Early Cambrian tomiddle Mississippian (Figure 4). Depth conversionof major reflections across line 1 (Figure 5) reveals

that the trough, at regional scale, is a simple asym-metrical graben. The western limb gradually subsid-ed relative to the Ohio-West Virginia hinge zone thatlies off the line to the northwest (Figure 2). The East-Margin fault forms the eastern boundary of thetrough, although minor structures are observed far-ther east in the footwall on line 1 (Figure 3). Minorstructural irregularity is observed within the trough(Figure 3). A low-relief structural high is presentacross the trough in the Silurian Rose Hill throughDevonian Huron reflections (Figures 3, 5). Depth-converted seismic traveltimes reveal a structural riseof approximately 100 m in the Devonian Onondagareflector west of the East-Margin fault into the inte-rior of the trough. Structural relief observed on con-tour maps of well log data presented by Gao andShumaker (1996) of the Onondaga Limestone inthis area also reveal approximately 100 m of struc-tural rise into the trough. This general agreementbetween seismic and well log derived structure fur-ther supports the absence of significant velocityanomalies. Uplift of the trough interior is interpret-ed to have produced the structural high. This upliftmay have occurred in the Late Devonian since theinterval between the Devonian Onondaga andMississippian Greenbrier reflections (Figures 3, 5) isthinner across the trough than it is to the east of thetrough margin; however, because the area lies alongthe distal edge of detached structures produced dur-ing the Alleghany orogeny, that disharmony could be

378 Subsidence, Compaction, and Loading

Figure 2—Contour map of acousticbasement across the Rome troughderived from lines 1 through 4.

produced by detachment and thickening of the sec-tion just east of the trough margin (Shumaker andWilson, 1996).

Line 2 (Figures 2, 6A) extends across an unfault-ed portion of the western trough margin. West mar-gin collapse is largely rotational about the well-defined Ohio-West Virginia hinge zone. Approximately1.2 km of offset is observed across the East-Marginfault. Interior faults form a double-step acrosswhich the basement drops about 0.9 km to thenorthwest. Lower and Middle Cambrian sectionsthicken significantly eastward toward the Interiorand East-Margin faults. The Interior fault observedon line 2 is interpreted to extend northward acrossline 3 (Figures 2, 6B). On line 3, the east margin ofthe trough is cut by two normal faults, which havea combined displacement of nearly 2 km. Faultingalong the Interior and East-Margin faults is accom-panied by fault-block rotation. Upper level struc-tures on the east end of line 3 are detached above adecollement in the Ordovician Martinsburg Shale.These detached structures formed during theAlleghany orogeny. Along line 4 (Figures 2, 6C), the

western margin of the trough also lacks major base-ment faults. Normal faults similar to those that sub-divide the interior of the trough to the south aremore numerous along line 4. Maximum depths ofaround 6 km (subsea) observed on line 1 to thesouth increase northward to more than 8 km alongline 4.

The basement interpretation presented here(Figure 2) is intended as a generalized view ofmajor trough structures in the study area and dif-fers from previous interpretations of the trough(e.g., Kulander and Dean, 1993). For example, theinterpretation of Kulander and Dean (1993) sug-gested that offset along the East-Margin faultdecreases steadily from about 1.5 km to less than0.5 km southwest to northeast across the area;however, lines 1–4 reveal that displacement alongthe East-Margin fault increases from approximately1 km along line 1 to 1.7 km along line 3 (Figure 6).Farther to the north along line 4, the displacementdecreases to approximately 1 km. Harris (1978),Donaldson and Shumaker (1981), and Kulanderand Dean (1993) described the west margin of the

Wilson 379

Figure 3—Line 1 (courtesy of GTS Corporation) crosses the Rome trough in the southern part of the study area (seeFigures 1, 2).

trough as a continuous normal fault through thearea. Lines 1–4, however, do not reveal significantfaults on the west flank of the trough. Line 2 revealsRyder’s Ohio-West Virginia hinge zone (Figure 2)across which the west margin rotates down into thetrough. Line 3 extends high up onto the west mar-gin of the trough but does not cross the hinge.Normal faults marking the western edge of thetrough may be concentrated along the Ohio-West

Virginia hinge zone, which is observed only on line2 (Figure 6A). Tegland (1977), Swimm (1986), andR. W. Beardsley (unpublished seismic) showedlocal faults along the west margin; however, theinterpretations presented here (Figure 6) do notsupport the presence of a continuous west marginfault. West margin relief appears to be accommo-dated largely through rotation about the Ohio-WestVirginia hinge zone.


Figure 4—Depth, traveltime, and interval velocity are presented for the Paleozoic section of the Appalachian fore-land. Major reflections observed in the seismic data of the foreland are noted with an asterisk (taken from Shumak-er and Wilson, 1996).

SEISMIC ANALYSIS

Quantitative analysis of individual structures wasdone using reactivation history diagrams. A schemat-ic representation (Figure 7) of the major structuralfeatures of the trough (Figure 5) is used to illustrateconstruction of the reactivation history diagram.Deposition of sediments across an active normalfault (e.g., southeastern end of Figure 7) producesthicker sediment accumulations over the hangingwall and thinner accumulations over the footwall.In the absence of strike-slip movement, increasedthickness within any given interval across the faultwill be related to vertical displacement across thefault during deposition of that interval. In Figure 7,for example between points C and D, the thicknessof the basal unit 1 thins from 880 m over the hang-ing wall to 250 m over the footwall. The thicknessdifference is 630 m. An identical relationship existsfor strata deposited across the area deformed byrotation on the northwest f lank of the graben(Figure 7). Unit 1 (Figure 7) thickens in thedowndip direction between points A and B. Thethickness difference across the northwest flank isalso 630 m.

The difference in interval thickness across a struc-ture provides a measure of the structural displacementthat occurred during the span of geologic time associ-ated with deposition of that interval. Syndepositionalsubsidence and continued movement of trough struc-tures actively influenced sedimentation (e.g., Figures5, 6). Thickness differences that occur over severalcontiguous time intervals provide an historical recordof structural development. Reactivation history dia-grams (Wilson et al., 1994; Dominic and Wilson, 1995;Shumaker and Wilson, 1996) constructed from thick-ness differences portray historical activity across indi-vidual structures.

Thickness differences (d) measured from Figure7 are listed in Table 1. The thickness of unit 1, forexample, decreases 630 m (250 to 880 m) from theinterior of the trough onto its f lanks. Negativethickness differences imply subsidence of thetrough relative to its margins. Positive thickness dif-ferences (e.g., those for units 4, 7, and 8 in Figure 7and Table 1) indicate that the trough interior wasuplifted during deposition of those units. Thecumulative increases in thickness reach the pres-ent-day total of –985 m (Table 1), which also coin-cides with the present-day basement relief acrossboth margins of the trough (Figure 7). To allow forexponential fitting (see discussion in following sec-tion), the cumulative difference (Table 1) is expressedrelative to a maximum vertical displacement (d0) ofthe trough interior relative to its margins. In thisexample, d0 is set equal to 1400 m (Table 1) butwill vary from line to line. The cumulative differ-ences are subtracted from the maximum value toobtain the adjusted cumulative differences (Table1). The maximum value (d0) is chosen to preventvalues from becoming negative or zero and can bethought of as the maximum possible vertical dis-placement across a given structure.

The adjusted cumulative difference (Table 1) isgraphed as a function of geologic time (Figure 8).This plot is referred to as a reactivation history dia-gram (RHD) to distinguish it from a subsidence his-tory curve. RHDs are distinguished from subsidencecurves in that they portray differential movementacross individual structures rather than total subsi-dence at a given point. Differential movements rep-resented by reactivation history diagrams are notcorrected for compaction and loading effects.Compensation for compaction and loading are dis-cussed following presentation of RHDs for the majorstructures of the trough. The issue of correction is

Wilson 381

Figure 5—Line 1 (Figure 3) depth section. Depths arereferenced to sealevel.

critical to interpretation of the RHD because non-tectonic displacements are expected throughoutgeologic time as a result of differential compactionand loading of different thickness sediment deposit-ed across a fault or tilted fault block.

DIFFERENTIAL MOVEMENT INFERRED FROM RHD

RHDs for the west and east margins of the trough(Figures 9, 10, respectively) and for the Interior fault

(Figure 11) portray differential motion that has occurredacross the major structural elements defining the trough.Fowler (1990), Lerche (1990), and McKenzie (1978)model subsidence (d) as a thermal process that, in first-order approximation, varies exponentially as

(1)

In equation 1, d0 and t0 are constants. Although d0

and t0 are constants, their values may change from


Figure 6—Depth-converted reflectionsdigitized from (A)line 2, (B) line 3, and(C) line 4 located inFigure 2. Depths arereferenced to sealevel.

d d et t= −

( )0

0 1

one data set to another within a basin. The con-stant t0 represents the thermal time constant of thelithosphere.

(2)

(Fowler, 1990) where L is the thickness of the litho-sphere and κ is the thermal diffusivity. Although

tL

0

2

2=

π κ

the mechanism or mechanisms responsible fortrough formation are not known with certainty, dif-ferential motion across trough structures shown inthe RHDs follows an exponential decay processand is modeled by equation 1. In the current appli-cation, equation 1 is solved in the form

(3)d d d et t

0 00+ = −( )

Wilson 383

Figure 6—Continued.

Figure 7—Schematicrepresentation of theRome trough. Unitnumbers are shownin the center and onthe edges of the profile along withunit thicknesses inand out of thetrough. Comparisonsare made betweenpoints A and B andpoints C and D inTable 1.

In this expression, d0 is approximately equal to themaximum possible offset across the fault or rotat-ing margin along this line. As noted, the adjustedcumulative differences (Table 1) are obtained bysubtracting the cumulative difference from d0 ateach stage. In this example (Table 1), d0 is 1400 m.The variable d (Table 1) represents the change inthickness of each interval into the trough. Smallervalues of t0 are associated with more rapid differen-tial subsidence than are larger values of t0. Timeconstants derived from exponential fitting of theadjusted cumulative differences are negative, andthe negative sign is shown explicitly in equation 3,which defines an exponential decay process. Timeconstants (t0) were computed by fitting exponen-tial functions only to the subsiding portions of theRHDs (Figures 10–13). At time t = t0, the exponentin equation 3 equals –1 and the term on the rightbecomes d0/e or approximately 0.37d0. Thus, inthis application, t0 represents the time it takes for37% of the maximum possible displacement (d0)across a structure to occur.

The time scale of the observations (Figure 8)extends from the early Cambrian–middle Mis-sissippian or approximately 570–340 Ma. Growth oftrough structures is assumed to have begun in theEarly Cambrian. Time zero is taken as 570 Ma. Thetime constant (t0) calculated for each seismic pro-file across the trough’s unfaulted west marginranges from 62.5 to 77.5 m.y. (Figure 9) and impliesthat approximately 37% of the total displacementacross the margin occurred over a time interval of62.5–77.5 m.y. These large time constants reflectlong-term rotational collapse of the trough’s westflank throughout much of the Paleozoic. Time con-stants along the east margin range from 13 to 51m.y. (Figure 10). The exponential curve fit to line 1was derived from a 230 m.y. period extending from

the Early Cambrian (570 Ma) into the Mississippian(340 Ma). On line 1 the motions of the east andwest trough margins closely parallel each other.Slight inversion or tectonic uplift of the troughinterior begins to occur along the east margin after120 m.y. (or 450 Ma in the Middle–Late Ordo-vician). Inverse movement along the East-Marginfault on lines 2 and 4 is more pronounced than on


Table 1. Thickness Differences Between the Reference Points in the Schematic Representation of the Trough*

Thickness Thickness AdjustedDifference Difference Cumulative Cumulative

A to B D to C Difference Differenced d – d0 = 1400 m

(m) (m) (m) (m)

Unit 1 –630 –630 –630 770Unit 2 –300 –300 –930 470Unit 3 –300 –300 –1230 170Unit 4 50 50 –1180 220Unit 5 –20 –20 –1200 200Unit 6 –20 –20 –1220 180Unit 7 75 75 –1145 255Unit 8 160 160 –985 415Total –985 –985 – –

*See Figure 7. d = thickness difference. Adjusted cumulative differences are used to construct the reactivation history diagram.

Figure 8—Example reactivation history diagram con-structed from the schematic representation of thetrough shown in Figure 7. Adjusted cumulative differ-ences are plotted with respect to time in millions ofyears since the onset of deformation (m.y.) and in mil-lions of years before present (Ma) extending back to thebeginning of the Cambrian. Time intervals spanned byindividual geologic time periods are also marked.

line 1, whereas on line 3, interior uplift is restrictedto the Early and Middle Devonian from about 375to 350 Ma. The time constants computed for lines2–4 range from 33 to 13 m.y. and are much lessthan that computed for line 1. The coupled re-sponse of the east and west margins along line 1sets that area apart from the area to the northwhere differential motion across the east and westmargins differ considerably through time.

Time constants of 16–23 m.y. were calculated foroffsets occurring across the interior faults on lines2–4 (Figures 2, 11). Time constants were computedfor the initial 70 m.y. (Cambrian) during which nor-mal offsets occurred. Uplift of the trough interior(inversion) began early in the Ordovician on line 2.On line 3 to the north, a slight amount of interioruplift began during the Early Silurian. On line 4 inte-rior uplift occurred during the Middle–Late

Cambrian. In general, normal offsets on the Interiorfault appear to have been confined to the Cambrianperiod. Significant interior faulting is not observedon line 1 (see Figures 2, 3, 5).

Lack of longer term normal displacements onthe East-Margin and Interior faults suggests theywere decoupled from continued rotation of thewest margin following the Cambrian. These obser-vations suggest that rotational subsidence of thetrough’s west margin continued in response toIapetan shelf loading, whereas the East-Margin andInterior faults remained locked or served as zonesof weakness along which slight reverse movementsoccurred.

Continued movement on the East-Margin fault ofline 1 after the Cambrian is anomalous by comparisonto that observed on lines 2–4. Gao (1994) reportedsimilar reactivation along the east margin of the

Wilson 385

Figure 9—Reactivationhistory diagrams forthe west margin ofthe Rome trough.

trough just south of line 1. The analysis presentedhere in combination with Gao’s observations sug-gests that normal offsets along the East-Margin faultnorth of line 1 ceased following the Cambrian butcontinued through the Silurian in the area south ofline 2. These two areas may have experiencedslightly different stress histories. Differencesbetween these two areas are also observed in theform of significant differences in coal cleat trendsbetween these two areas (Kulander and Dean,1993). Crustal scale gravity models derived alonglines 1–4 (Gurshaw, 1995; Morgan, 1996; T. H.Wilson, unpublished results) suggest that the crustthickens more than 10 km from north to southbetween these two areas.

The variations in t0 provide a quantitative measureof differences in the rate of structural development

within this area of the Rome trough. Because theanalysis is made of differential motion across struc-tural elements in the basin, it is not directly compa-rable to subsidence; however, the RHDs allow oneto evaluate the degree to which individual struc-tural elements participated in the overall subsi-dence of the basin. Previous studies of subsidencein the central Appalachians are limited. Ettensohn etal. (1992a) discussed the sedimentary-stratigraphicresponse and subsidence history of the Ap-palachian basin in Kentucky. Their study areaincluded the western limits of the Rome trough.Ettensohn (1992b) interpreted the sedimentationhistory of the region within the context of litho-sphere response to orogenic loading and subse-quent relaxation. Sedimentation in the Kentuckyforeland is interpreted by Ettensohn (1992b) to


Figure 10—Reactivationhistory diagrams forthe East-Margin faultof the Rome trough.

reveal a four-stage response to individual orogenicevents, consisting of the initial cratonward migra-tion of a peripheral bulge (Quinlan and Beaumont,1984), foreland subsidence, relaxation and periph-eral bulge ref lection, and unloading. Goodman(1992) computed tectonic subsidence from twowells located in the Kentucky portion of the Rometrough. Computed tectonic subsidence revealsfour distinct regional subsidence events associat-ed with (1) initial Iapetan rifting, (2) MiddleOrdovician Blountian and Middle–Late OrdovicianTaconic orogenies and Silurian Salinic distur-bance, (3) Devonian Acadian orogeny, and (4)Alleghanian orogeny.

Donaldson (1994) made subsidence computa-tions within the Rome trough near its east margin,just north of line 1 (Figure 2). Donaldson used gen-eral lithological information from an Exxon deepwell to evaluate subsidence along a seismic profilethrough the Granny Creek oil field in West Virginia;

however, detailed well log data in the Rome trougharea is limited, so Donaldson (1994) used North Seacompaction factors (Allen, 1990; Sclater and Christie,1980) to decompact the section. Donaldson (1994)incorporated sea level variations from Hallam(1992) and paleobathymmetry based on generalfacies interrelationships to sequentially remove orbackstrip (Bond and Kominz, 1984) stratigraphicintervals represented in the seismic profile.Donaldson’s results revealed roughly uniform subsi-dence throughout the Paleozoic interrupted by abrief (∼ 5 m.y.) period of uplift within the troughduring the Late Cambrian and equally brief periodof increased subsidence during the Late OrdovicianTaconic orogeny.

Reactivation history analysis presented previous-ly (Figures 9–11) reveals additional informationabout movement along individual trough structuresand accompanying regional subsidence; however,with the exception of subsidence directly associat-

Wilson 387

Figure 11—Reactivationhistory diagrams forthe Interior fault within the Rometrough.

ed with the trough-forming rifting, differentialmovement of trough structures does not appear toaccompany individual Paleozoic orogenies. Westmargin subsidence continues into the Mississippian.The development of the East-Margin fault is quitevaried. Both the East-Margin and Interior faultsaccommodated minor uplift of the trough interiorfollowing the Cambrian.

COMPACTION

The RHDs presented are not corrected for differ-ential sediment compaction or loading amplifica-tion. The potential influences of these processeson the preceding results are examined in this sec-tion using generalized compaction models. Theeffect of differential sediment compaction is mod-eled stage-by-stage through time (Figure 12) using asingle lithology. This model study is not intended toreplicate the details of movement observed acrossa specific structure within the trough, but only toillustrate the general effect of depth-dependentcompaction assumed in several compaction models(e.g., Baldwin and Butler, 1985). Model develop-ment employs the compaction model presented bySclater and Christie (1980) in which porosity (φ)decreases with depth following the relationship

(4)

Here, φ0 is the initial porosity after deposition, φ isthe present-day porosity at depth z, and c is a lithol-ogy-dependent compaction constant. During com-paction, it is also assumed that the rock volume Vw

+ sg equals Vsg + Vw where V refers to volume andthe subscripts ω and sg to water and matrix,respectively. Use of this relationship assumes thattotal grain volume remains unchanged and thusthat significant diagenesis has not occurred(Angevine et al., 1990). Use of equation 4 allowsone to express the uncompacted thickness of agiven interval in terms of its compacted thickness.

An interval encountered in a well or seismic sec-tion between depths z1 and z2 will have originalthickness z ′ (or , where = 0 at the surface)defined as

(5)

z ′ must be determined numerically. Given intervalthickness z ′ immediately following deposition, it ispossible to compute the mean sediment density ofthe interval and thus the loading effect of the origi-nal water-saturated sediment column. The processof unloading layers in this fashion is referred to as

′ = − − −( ) + −( )− − − ′z z zc

e ec

ecz cz cz2 1

0 01 2 1φ φ

φ φ= −0e

cz

backstripping (e.g., Bond and Kominz, 1984). It is asimple matter to carry the computations in the for-ward direction and compute interval thickness (z2

– z1) after burial. To construct the forward model,take an interval of thickness z ′ at the surface andmove its base to depth z2. In this process, z1 is cal-culated rather than z ′, and rearrangement of equa-tion 5 yields

(6)

The terms on the right are all known. Substitutionyields the constant

(7)

then substitution of equation 7 into equation 6yields

(8)

and z1 can be solved iteratively.Equation 8 was used to generate a forward-compaction

model (model I) across a fault that was active onlyduring deposition of the basal sedimentary layer.Values for c and φ0 taken from Sclater and Christie(1980) for North Sea shale (5.1 ×10–4/cm and 0.63,respectively) were used. The initial stages of devel-opment in model I are shown in Figure 12. Twokilometers of throw is introduced across a base-ment fault during stage 1 deposition (e.g., deposi-tion of the basal sedimentary layer). No additionalfault offsets were introduced. During stage 2 and allsubsequent stages, the hanging wall and footwallwere dropped uniformly in 0.5 km steps per stage.Differential compaction of sediments across theunderlying fault produces normal offset of theinterface separating stage 1 and stage 2 sediments(Figure 12). The process is repeated, and continuedgrowth occurs across the fault as a result of differ-ential compaction of sediments on either side ofthe fault. A total of nine stages was computed lead-ing to the final stage shown in Figure 13A.

As the sediments of model I continued to bedeposited, differential compaction led to increasedoffset of individual horizons crossing the fault.Interval thickness across the fault decreases as burialdepth increases. The reactivation history diagram forthe basement fault in model I (Figure 13B) suggeststhat the initial vertical displacement across the faultwas only about 1.6 km. The fault appears to havebeen moved continually during the following stages(stages 2–9) until the maximum displacement of

z ac

e cz1

0 1= − −φ

ac

zc

e zc

ecz cz'= + + − ′ +− −φ φ φ02

0 02

zc

ec

zc

e zc

ecz cz cz1

0 02

0 01 2+ = + + − ′ +− − − ′φ φ φ φ


′ − ′z z2 1 ′z1

2 km was finally reached. If the process of com-paction were to continue to operate over long timeperiods as predicted by continuous porosity-depthrelationships such as that defined by equation 4,we would expect to see the general effect of differ-ential compaction represented in Figure 13A. Ifcompaction actually follows equation 4, then theRHD (Figure 13B), which is uncorrected for com-paction, misrepresents the timing of basement faultdevelopment because the entire 2 km of fault dis-placement occurred entirely during stage 1.

Although considerable compaction of the sedi-ments in the trough must have occurred, syndeposi-tional compaction of earlier unconformity boundedsequences does not appear to cause the systematicdecreases in normal displacements of younger strataacross the East-Margin or Interior faults as predicted

by the depth-dependent compaction process.Model I (Figure 13A) also suggests that compactioneffects will be greater if the thickness difference inthe basal sequence deposited across the fault isgreater. The anticipated effect is not observed inseismic data across the trough. For example, alongline 1 thickening of the Rome Formation across theEast-Margin fault is about 0.76 km. Total thicknessof the Rome over the hanging wall is approximate-ly 1.16 km and that over the footwall is 0.4 km.Additional normal displacement of 0.35 kmoccurred during the Middle Cambrian, followed byeven greater normal displacement of 0.37 km dur-ing the Late Cambrian–Middle Ordovician (Rut-ledge through Trenton events of Figure 5).Thickening of the Rome across the East-Marginfault on line 2 increases to 1.2 km (0.44 km greater

Wilson 389

Figure 12—The influence of sediment compactionacross a fault developed duringdeposition of stage 1sediments (A) is followed throughsuccessive stages ofsedimentation (B–D).

than on line 1) and on line 3 to approximately 1.3km; however, thickening over the hanging wall inthe overlying Middle and Upper Cambrian sectionin these areas is only about 0.4 km, which is muchless than the 0.76 km of thickening observed online 1. If the thickening of shallow intervals on line1 is due primarily to compaction, then thickeningalong lines 2 and 3 should have been greater thanthe thickening observed on line 1. The compactioneffect should increase in proportion to increases inthe thickness of the strata across the fault.

More pronounced differences between the com-paction model (Figure 13A) and actual seismicresponse are illustrated in the RHDs for the East-Margin fault on lines 2 and 4 (Figure 10). TheseRHDs are characterized by a period of rapid growth(initial 70 m.y.) somewhat similar to that observedin the RHD derived from the model compaction

response (Figure 13B); however, instead of the con-tinued thickening over the hanging wall predictedby the model (Figure 13B), thinning associatedwith gradual uplift of the hanging wall occurs. Aclose-up view of the seismic data across the fault online 2 (Figure 14), for example, reveals subtle thin-ning to the west over the hanging wall in the RoseRun to Trenton and Trenton to Lockport intervals.In the forward compaction model (Figure 13) post-tectonic compaction-related offset is about 20% ofthe total. In general, RHDs for lines 1–4 (Figures9–11, respectively) are not similar to the RHD com-puted from the compaction model (Figure 13B).The observations suggest that the majority of sedi-ment compaction occurred prior to deposition ofthe overlying sequences, thus diminishing or alto-gether eliminating later differential compactionacross the deeper fault.

These observations provide a further example ofhow use of exponential or other continuous poros-ity-depth relationships in subsidence calculationsfor this area will lead to erroneous or irreconcilableportrayals of subsidence history. Given the present-day offset across the East-Margin fault on line 2(Figures 6B, 14) of approximately 1.2 km, itsdecompacted thickness, for example, immediatelyfollowing deposition should have been 1.96 km. Ifthat were so, then 0.76 km of normal offset is miss-ing from the East-Margin fault. The required 1.96km of normal offset would have to have been fol-lowed gradually in time by 0.76 km of reversal toleave the present-day 1.2 km of offset. This reversalwould have to have occurred in such a way as toleave the overlying strata unfaulted.

Although unlikely, it is possible for the basin toevolve in such a way that these compaction-relatedoffsets do not appear. The solution is illustrated inthe sequence of stages shown in Figure 15. In thismodel (model II), the initial basement offset is 2.89km (Figure 15A). As subsequent layers of sedimentblanket the area, the hanging wall is forced to rise by anamount equal to the additional compaction of thebasal package of sediments. Thus sedimentsdeposited early in the history of model II (Figure15B) cause the basal sequence to thin from 2.89 to2.31 km. If the hanging wall did not rise, this addedblanket of sediment would have produced anapparent normal offset of 0.58 km in the top of thebasal sequence (Figure 15B). Thus as each blanketof sediment is deposited, reverse movement mustoccur along the fault by exactly the amount re-quired to balance compaction of the basal intervalresulting from the increased weight of overlyingsediments. In the final present-day configuration(Figure 15C), there is 2 km of vertical displacementacross the fault, and although there is no displace-ment in the overlying horizons, the fault movedvertically upward 0.89 km during deposition of the


Figure 13—Forward compaction model I. (A) Final stagein the model after nine stages of sedimentation. Thereactivation history diagram (B) plotted in cumulativedifference form has not been corrected for compaction.

overlying strata. The reactivation history diagramfor this model (Figure 15D) erroneously suggeststhat the present-day offset of 2 km occurred earlyin the development of the basin and that no addi-tional displacements occurred. This scenario(Figure 15) requires that the hanging wall subsidemore slowly than the footwall so that the verticaloffset across the fault decreases over time at a ratethat matches the rate of compaction of the sedi-ments covering the hanging wall. In the followingsection, I consider the possibility that this faultreversal could occur through the differential iso-static adjustment.

LOADING

Total subsidence is a combination of tectonicand isostatic adjustments to loading. Tectonic sub-sidence during basin formation is associated pri-marily with thermal cooling or stretching and thin-ning of the crust and lithosphere along with

thermal cooling. The isostatic component of totalsubsidence represents the additional subsidence ofthe lithosphere or crust resulting from the addedweight of sediment deposited in the basin. The tec-tonic component of subsidence is computed usingequation 9 (e.g., Sclater and Christie, 1980;Angevine et al., 1990)

(9)

In this equation, Zi is the water-loaded basementdepth after removal of the ith layer; Si is the com-paction-corrected sediment load; ρm, ρs, and ρw arethe densities of the mantle, compacted sediment,and water, respectively; Wdi is water depth; and∆SLi is sea level rise or fall relative to present-day sealevel. The tectonic component equals the water-loaded basement depth (Z) after sediment has beenstripped away. To determine how the tectonic andloading components of total subsidence must vary

Z S Wd SLi im s

m wi i

m

m w

= −−

+ −−

ρ ρρ ρ

ρρ ρ

∆

Wilson 391

Figure 14—Close-up view of seismic data across the East-Margin fault along line 2.

during the development of model II, layers in themodel (Figure 15) were removed (top to bottom)and the unloaded basement depth was computed.We make the simplifying assumption that thedepression created by faulting is completely filledby sediment and that Wd and ∆SL are zero. Underthese assumptions, equation 9 reduces to

(10)

Sediment density (ρs) is computed using equation11 (after Sclater and Christie, 1980)

(11)

where is the mean porosity of the ith layer, ρw isthe water density, ρsg is sediment grain density inthe ith layer, and is the thickness of layer i.Sediment grain density (ρsg) of 2.72 gm/cm3 is used.The average porosity of a layer is determined asfollows (see Sclater and Christie, 1980)

(12)

where φ0 is the uncompacted sediment porosity, cis the lithology-dependent compaction constant,and and are the depths to the top and bot-tom, respectively, of the sequence at an earlierstage.

Equations 10–12 were used to backstrip layersfrom the model shown in Figure 15C. Sedimentgrain density (ρsg) of 2.72 gm/cm3 and mantle densi-ty (ρm) of 3.33 gm/cm3 were used for all layers. Thetime required to deposit each layer shown in thepresent-day configuration (Figure 15C) is assumed toincrease exponentially toward the surface. Total andtectonic subsidence are plotted for the hanging wall(Figure 16A) and footwall (Figure 16B). The isostaticresponses of the hanging wall and footwall paralleleach other (Figure 16C). The differential isostaticadjustment remains constant (Figure 16D) with theexception of the initial episode of fault develop-ment. The results of the modeling indicate that dif-ferential subsidence across the fault and the forcesdriving reversal of the hanging wall must be tectonicin origin. The offset across the fault decreases fromthe maximum offset of 2.89 km during the initialstage of formation to the present-day offset of 2 km.This relative uplift is required to compensate forcompaction of the overlying sediments with depth.

φ φ= −′ − ′

− ′ − ′

0

2 1

1 2

c

e e

z z

cz cz

ρφ ρ φ ρ

s

i w i sg

iiS

z=+ −( )

∑ ′

1

Z Si im s

m w

= −−

ρ ρρ ρ

To put these results into perspective, the isosta-sy equation is solved for the extreme cases repre-sented in models I and II. Isostatic balance requiresthinning of the crust by an amount hm in responseto the addition of a sediment load of thickness hs,where

(13)

The density of the crust (ρ c) is taken as 3gm/cm3. The thinned crust is replaced by uppermantle with density, ρm, which is taken as 3.33gm/cm3.

Thinning was computed for sediment thickness-es corresponding to the initial hanging wall offsetsof 2 and 2.89 km and final depths of 5 km for thehanging wall and 7 km for the footwall. The origi-nal thickness of the unloaded crust was assumed tobe 45 km. The results (Figure 17) indicate that ifthe initial offsets of 2 km (model I) and 2.89 km(model II) are isostatically balanced, then subsedi-ment crust must thin from 45 to 37.26 km and34.78 km, respectively. After thinning, the base ofthe crust rests at 39.25 km in model I (Figure 17A)and 37.77 km in model II (Figure 17B). In the finalconfiguration (Figure 17C), fault offset is 2 km andthe crust has thinned to 25.9 km beneath the hang-ing wall and 29.9 km beneath the footwall. Thisplaces the base of the crust at 31 and 37 kmbeneath the surface on the footwall and hangingwall sides of the fault, respectively.

Isostatic compensation across individual faultsrequires that the fault extend almost verticallythrough the crust and that the lithosphere has noelastic strength. Recent deep-crustal seismic data hasimproved our understanding of extension tectonics.Current models of intracontinental extension suggestthat the crust is divided into shallow brittle and deep-er ductile regions (Strahler, 1998). Allmendinger et al.(1987) presented COCORP data across the Basin andRange Province in the western United States support-ing the view that deformation of the lower crust isductile. Listric normal faults that sole out into thedeeper ductile regions of the crust (Miller et al., 1983;Echtler et al., 1994; Gibbs, 1984; Kligfield et al., 1984)or that sole-out into a low-angle crust-penetratingshear zone (Wernicke, 1981) are believed to accommo-date extension in the upper brittle region of the crust.Individual faults in the shallow crust do not have crust-penetrating roots and are not isostatically compensat-ed. Rather the extended region is isostatically compen-sated by regional scale crustal thinning as is shown bydeep seismic profiling, for example, across the UpperRhine graben (Echtler et al., 1994). Within this context,it is unlikely that the base of the crust will adjust

h hm sc s

m c

=−( )−( )

ρ ρρ ρ


φi

′z1 ′z2

′zi

φ

isostatically to individual offsets along trough struc-ture as represented in Figure 17. The thickness differ-ences across these structures occur over distancestoo small to produce Airy isostatic compensation.

BURIAL CONDITIONS

The presence of oil in a given reservoir intervalpartly depends on whether potential source rocksgenerated hydrocarbons, and this will depend, inpart, on accurate assessment of source rock burialdepth and the length of time spent at a certain depth.As an example, subsidence history curves correctedand uncorrected for compaction are compared in

Figure 18. Sediment compaction must certainly occur,but the actual compaction process may follow a curvethat stair-steps between the continuous compactionand no-compaction (uncorrected) predictions. Thereis an uncertainty in our knowledge of when the hori-zon at 4 km, for example, actually reached that depth.In the example (Figure 18) it could have reached thatdepth anytime between 57 and 95 m.y. after subsi-dence began. For a potential source rock just belowthe oil window (at 2 km in this example), there is anuncertainty of about 50 m.y. in the estimated timethat the potential source rock actually passed throughthat depth. If rapid unroofing began around 275 m.y.,then the continuous compaction prediction wouldindicate that the horizon spent very little time at that

Wilson 393

Figure 15—Forwardcompaction model II.Fault with 2.89 km ofoffset develops in theinitial stage (A) and isfollowed by depositionof successive layers(e.g., B) until thepresent-day configuration (C) isreached. Reactivationhistory diagram (D)plotted in cumulativedifference form hasnot been correctedfor compaction.

depth and is unlikely to have generated oil. Theamount of time spent inside the oil window increasesas the compaction rate is reduced.

Estimates of maximum burial depth are alsoaffected by this potential source of error. The analy-sis of appetite fission track data presented by Roden(1992, 1999, personal communication) andDonaldson (1994) indicates that rocks currently atthe surface in the West Virginia area of the Rometrough were once buried at depths of approximately3 km. Rocks at various present-day depths through-out this area would have been buried at depths of 3km and greater during the early Mesozoic. Estimatesof the maximum depth reached by any given hori-zon depend on whether the estimates are correctedor uncorrected for compaction effects. Considertwo horizons, one at present-day depth of 1.5 kmand the other at 7 km (Figure 19A). If rocks current-ly at the surface are taken to a new depth of 3 km,

what total depths will these two horizons actuallyreach? If compaction continues to operate on thealready deposited rocks, then the new depths ofthese two horizons will be 3.94 and 9 km, respec-tively (Figure 19B). If, however, these strata do notundergo further compaction, then their new depthswill be 4.5 and 10 km, respectively (Figure 19C).With continuous compaction, the shallow intervalnever descends into the gas window, and sourcerocks at this depth are potentially oil producing.Conversely, if further compaction is insignificant,then source rocks descend into the gas window.

CONCLUSIONS

Block rotation and vertical displacements throughtime across normal faults in the trough follow anexponential decay process. The exponential time


Figure 16—Total andtectonic subsidenceof the (A) hangingwall and (B) footwall,(C) the load-inducedor isostatic componentof subsidence, and (D) the differentialtotal, differential tectonic, and differential isostaticresponses of model II.

constant (t0) provides a measure of the rate ofdeformation associated with individual struc-tures. The time constant (t0), derived by exponen-tial fitting, represents the time taken for 37% ofthe maximum possible offset to occur across agiven structure. Differential subsidence acrossthe western margin of the trough is characterizedby time constants that range from 63 to 78 m.y.Time constants for the East-Margin fault increasefrom 13 to 51 m.y. northeast to southwest alongthe fault. Development of the Interior fault is con-sistently of short duration with time constantsranging between 16 and 23 m.y. Time constantsfor the west margin are in all cases larger thanthose for the East-Margin and Interior faults andindicate that rotational collapse of the west mar-gin continued over a longer time period. TheInterior fault developed quickly during the 70m.y. time span of the Cambrian period. Develop-ment of the trough’s east margin is also a relative-ly short-term event confined primarily to theCambrian period, with the exception of line 1where normal displacements across the East-Margin fault continued through the Ordovician.Longer term subsidence of the trough is character-istic of the southern West Virginia segment of theRome trough (Gao, 1994; Gao et al., 2000). Theboundary between the southern and northern

West Virginia segments of the trough occursbetween lines 1 and 2 (Figure 2).

Understanding the cause for variation in the his-torical development of different trough structuresrequires further study. Long-term rotation of thetrough’s west margin may reflect, in part, its partic-ipation in the overall subsidence of the Appa-lachian foreland. Differential subsidence across thewest margin of the trough continued throughoutmost of the Paleozoic, whereas development of theEast-Margin and Interior faults were relatively short-lived events restricted, in general, to the Cambrianperiod; however, uplift of the trough interior oftenoccurred along the East-Margin and Interior faultsfollowing the Cambrian. This uplift may indicatethat the East-Margin and Interior faults were passiveparticipants in foreland subsidence of theAppalachian basin.

The import of the observations presented in thisstudy to hydrocarbon exploration is related to theirdefinition of the time periods over which develop-ing structures are likely to have had an effect onpotential reservoir intervals. Beardsley and Cable(1983) noted that recognition of the concept ofreactivation of specific structures is critical to thedevelopment of frontier exploration workinghypotheses. In the northern part of the study area,the East-Margin and Interior faults developed rapid-ly. Their impact on sedimentation was greatest dur-ing the Cambrian. Potential reservoirs may have

Wilson 395

Figure 17—The loading effect of sediments for (A)model I, (B) model II, and (C) the final stage.

Figure 18—Hypothetical subsidence curves are shownfor deep and shallow horizons. Curves are shown withand without the compaction correction. A horizondeposited at 95 m.y. is followed along corrected anduncorrected subsidence paths. Hypothetical locations ofoil and gas windows are shown.

been created in these deeper intervals. Minoruplift occurring across the East-Margin andInterior faults later in the history of the basin mayhave created secondary fracture porosity in shal-lower intervals. The Interior and East-Marginfaults represent potential deep exploration targetswithin the northern West Virginia area. In south-western West Virginia the East-Margin fault wasactive over a longer period of time, and late-stageuplift and inversion are more pronounced (Gao,1994; Gao et al., 2000); hence, the influence ofthis structure on reservoir formation in the south-ern West Virginia area extends through much ofthe Paleozoic (Gao, 1994; Gao et al., 2000; Yang,1998). Long-term collapse across the west marginhad a significant inf luence on sedimentation(Ryder et al., 1992; Ryder, 1992) and the creationof secondary fracture porosity along its length.The early formation of these structures placesthem in potential trapping positions early in thehistory of the basin. Continued movement acrossthese structures through time influences currentreservoir conditions in sequentially deposited

depositional systems overlying these early struc-tures (Beardsley and Cable, 1983).

Seismic interpretation and compaction modelingindicate that reactivation history diagrams accurate-ly describe the history of vertical movementsobserved across individual structures within theRome trough. The final configuration of layers inmodel II (Figure 15C) suggests that a fault with 2 km of vertical offset occurred early in the deposi-tional history of the model. Model II (Figure 15C) issimilar in appearance to the seismic responseobserved across line 2 (Figure 14). It seems unlike-ly that the East-Margin fault along line 2 evolvedthrough the delicate match of tectonic displace-ment to compaction rate that was required to elim-inate long-term compaction-related offsets acrossthe fault in model II. Interpretation and modelingsuggest that compaction of major stratigraphicintervals must terminate prior to deposition of sub-sequent intervals. Local isostatic adjustments of thelithosphere to differential loading across faults orrotating margins also appear unlikely to occur with-in the context of current models of intracratonicextension.

Observations and modeling suggest that reactiva-tion history diagrams, when uncompensated forcompaction and loading, accurately portray differ-ential tectonic subsidence across Rome troughstructures. The results suggest that the standardapproach to backstripping may lead to erroneousresults. Corrections for compaction and loadingeffects may lead to errors in the magnitude of esti-mated tectonic strain and portrayal of basin history.The application of these corrections must be con-sidered basin by basin.

REFERENCES CITEDAllen, J. P., 1988, Stratigraphic analysis and depositional structural

influence on Cambrian units of the Rome trough in WestVirginia: Master’s thesis, West Virginia University,Morgantown, West Virginia, 123 p.

Allen, P. A., 1990, Basin analysis: principles and applications:Brookline Village, Massachusetts, Publishers’ BusinessServices, 586 p.

Allmendinger, R. W., T. A. Hauge, E. C. Hauser, C. J. Potter, and J. Oliver, 1987, Tectonic heredity and the layered lower crust inthe Basin and Range Province, western United States, in M. P.Coward, J. F. Dewey, and P. L. Hancock, eds., Continentalextension tectonics: Geological Society Special Publication no. 28,p. 223–246.

Angevine, C., P. Heller, and C. Paola, 1990, Quantitativesedimentary basin modeling: AAPG Continuing EducationCourse Note Series 32, 133 p.

Baldwin, B., and C. O. Butler, 1985, Compaction curves: AAPGBulletin, v. 69, no. 4, p. 622–626.

Beardsley, R. W., and M. S. Cable, 1983, Overview of theevolution of the Appalachian basin: Northeastern Geology, v. 5, no. 3/4, p. 137–145.

Bond, G. C., and M. A. Kominz, 1984, Construction of tectonicsubsidence curves for the early Paleozoic miogeocline,southern Canadian Rocky Mountains: implications for


Figure, 19—Present-day depths of two horizons areshown above position A in relation to hypothetical posi-tions of the oil and gas windows. The addition of 3 kmof sediment and correction for compaction moves thesetwo horizons to new depths shown above position B.Addition of 3 km of sediment with no adjustment forcompaction moves the two horizons to depths shown atposition C.

subsidence mechanisms, age of breakup, and crustal thinning:Geological Society of America Bulletin, v. 95, p. 155–173.

Dominic, J. and T. Wilson, 1995, Regional seismic evaluation ofbasement control on Paleozoic sedimentation in Ohio andWest Virginia: AAPG Bulletin, v. 79, no. 9, p. 1411–1412.

Donaldson, A., and R. Shumaker, 1981, Late Paleozoic molasse ofcentral Appalachians: Geological Society of Canada SpecialPaper no. 23, p. 99–124.

Donaldson, K., 1994, An integrated approach to reservoirmodeling: Master’s thesis, West Virginia University,Morgantown, West Virginia, 259 p.

Echtler, H. P., E. Luschen, and G. Mayer, 1994, Lower crustthinning in the Rhine graben: implications for recent rifting:Tectonics, v. 13, no. 2, p. 342–353.

Ettensohn, F. R., S. F. Barnet, D. R. Chestnut, C. F. Eble, P. T.Goodmann, S. F. Greb, R. T. Lierman, C. E. Mason, J. C. Pashin,and G. A. Schumacher, 1992a, Changing interpretations ofKentucky geology—layer-cake, facies, flexure, and eustasy:State of Ohio Miscellaneous Report no. 5, 184 p.

Ettensohn, F. R., 1992b, Some basic flexural models, in F. R.Ettensohn, ed., Changing interpretations of Kentuckygeology—layer-cake, facies, flexure, and eustasy: State of OhioMiscellaneous Report no. 5, p. 9–12.

Fowler, C. M. R., 1990, The solid earth: an introduction to globalgeophysics: New York, Cambridge University Press, 472 p.

Gao, D., 1994, Subsurface geometry and growth history of theWarfield structure in south-central West Virginia, centralAppalachian basin: Master’s thesis, West Virginia University,Morgantown, West Virginia, 243 p.

Gao, D., and R. C. Shumaker, 1996, Subsurface geology of theWarfield structures in southwestern West Virginia:implications for tectonic deformation and hydrocarbonexploration in the central Appalachian basin: AAPG Bulletin, v. 80, p. 1242–1261.

Gao, D., R. C. Shumaker, and T. H. Wilson, 2000, Along-axissegmentation and growth history of the Rome trough in thecentral Appalachian basin: AAPG Bulletin, v. 84, p. 75–99.

Gibbs, A. D., 1984, Structural evolution of extensional basinmargins: Journal of the Geological Society of London, v. 141,p. 609–620.

Goodman, P. T., 1992, A Paleozoic subsidence history of theautochthonous Appalachian basin in Kentucky, in F. R.Ettensohn, ed., Changing interpretations of Kentuckygeology—layer-cake, facies, flexure, and eustasy: State of OhioMiscellaneous Report no. 5, p.12–19.

Gurshaw, E., 1995, Gravity model studies across the Rome troughof west-central West Virginia: Master’s Option 2 ProblemReport, West Virginia University, Department of Geology andGeography, 48 p.

Hallam, A., 1992, Phanerozoic sea-level changes: New York,Columbia University Press, 158 p.

Harris, L. D., 1978, The eastern interior aulacogen and its relationto Devonian shale gas production: Second Eastern Gas ShalesSymposium, U.S. Department of Energy, Morgantown EnergyTechnology Center, DOE/METC/SP-78/6, v. 2, p. 55–72.

Kligfield, R., J. Crespin, S. Naruk, and G. H. Davis, 1984,Displacement and strain pattern of extensional orogens:Tectonics, v. 3, p. 357–609.

Kulander, B., and S. Dean, 1993, Coal-cleat domains and domainboundaries in the Allegheny plateau of West Virginia: AAPGBulletin, v. 77, no. 8, p. 1374–1388.

Lerche, I., 1990, Basin analysis: quantitative methods, v. 1: SanDiego, Academic Press, 562 p.

McGuire, W. H., and P. Howell, 1963, Oil and gas possibilities ofthe Cambrian and Lower Ordovician in Kentucky: Lexington,Kentucky, Spindletop Research Center (for Kentucky Dept. ofCommerce), 216 p.

McKenzie, D., 1978, Some remarks on the development ofsedimentary basins: Earth and Planetary Science Letters, v. 40,p. 25–32.

Miller, E. L., P. B. Gans, and J. Garing, 1983, The Snake Range

decollement: an exhumed mid-Tertiary ductile-brittletransition: Tectonics, v. 2, p. 239–263.

Morgan, J. F., 1996, Central West Virginia Allegheny plateaugravity and magnetic transect: Master’s Option 2 ProblemReport, West Virginia University, Department of Geology andGeography, 58 p.

Quinlan, G. M., and C. Beaumont, 1984, Appalachian thrusting,lithosphere flexure, and the Paleozoic stratigraphy of theeastern interior of North America: Canadian Journal of EarthScience, v. 21, p. 973–996.

Read, J. F., 1989, Evolution of Cambrian–Ordovician passivemargin, U.S. Appalachians, in R. D. Hatcher, W. A. Thomas,and G. W. Viele, eds., The Appalachian-Ouachita Orogen inthe United States: Geological Society of America, The Geologyof North America, v. F2, p 42–57.

Roden, M. K., 1992, Apatite fission-track thermochronology of theAppalachian foreland basin from the Virginia Piedmont toeastern Ohio: Geological Society of America Annual MeetingAbstracts with Programs.

Ryder, R. T., 1992, Stratigraphic framework of Cambrian andOrdovician rocks in central Appalachian basin from MedinaCounty, Ohio, through southwestern and south-centralPennsylvania to Hampshire County, West Virginia: U.S.Geological Survey Bulletin 1839-K, p. K1–K32.

Ryder, R. T., A. G. Harris, and J. E. Repetski, 1992, Stratigraphicframework of Cambrian and Ordovician rocks in centralAppalachian basin from Morrow County, Ohio, to PendletonCounty, West Virginia: U.S. Geological Survey Bulletin 1839-G,p. G1–G25.

Sclater, J. G., and P. A. F. Christie, 1980, Continental stretching:an explanation of the post-mid-Cretaceous subsidence of theNorth Sea basin: Journal of Geophysical Research, v. 85, no.87, p. 3711–3739.

Shumaker, R. C., 1986a, Structural development of Paleozoiccontinental basins of eastern North America, in M. J. Aldrich,Jr., and A. W. Laughlin, eds., Proceedings of the 6thInternational Conference on Basement Tectonics: Salt LakeCity, Basement Tectonics Association, p. 82–95.

Shumaker, R. C., 1986b, The effect of basement structure on thesedimentation and detached structural trends within theAppalachian basin, in R. C. McDowell and L. Glova, eds., TheLowry volume, studies in Appalachian geology, memoir 3:Virginia Polytechnic Institute and State University, Departmentof Geology, Blacksburg, Virginia, p. 67–81.

Shumaker, R. C., 1993, Geology of the Warfield anticline and itssignificance in regional structure: Program and Abstracts, 24thAnnual Appalachian Petroleum Geology Symposium, I. C.White Memorial Fund, Publication no. 5, p. 85–106.

Shumaker, R. C., and T. H. Wilson, 1996, Basement structure ofthe Appalachian foreland in West Virginia: its style and affecton sedimentation, in B. A. van der Pluijm and P. A.Catacosinos, eds., Basement and basins of eastern NorthAmerica: Geological Society of America Special Paper 308, p. 139–155.

Sleep, N. H., 1969, Sensitivity of heat flow and gravity to themechanism of sea-floor spreading: Journal of GeophysicalResearch, v. 72, p. 542.

Sleep, N. H., 1971, Thermal effects of the formation of Atlanticcontinental margins by continental break up: GeophysicalJournal of the Royal Astronomical Society, v. 24, p. 325–350.

Strahler, A. N., 1998, Plate tectonics: Cambridge, Massachusetts,GeoBooks Publishing, 54 p.

Swimm, D. E., 1986, Seismic detection of a fractured rock facies:the Cottageville Devonian shale gas field: Master’s thesis, WestVirginia University, Morgantown, West Virginia, 212 p.

Tegland, E. G., 1977, Update concerning the three dimensionalseismic investigations in the Cottageville field: Proceedings ofthe First Eastern Gas Shales Symposium, p. 23–36.

Thomas, W. A., 1991, The Appalachian-Ouachita rifted margin ofsoutheastern North America: Geological Society of AmericaBulletin, v. 103, p. 415–431.

Wilson 397

Thomas, W. A., 1993, Low-angle detachment geometry of the latePrecambrian–Cambrian Appalachian-Ouachita rifted margin ofsoutheastern North America: Geology, v. 21, p. 921–924.

Ungerer, P., F. Bessis, P. Y. Chenet, B. Durand, E. Nogaret, A. Chiarelli, J. L. Oudin, and J. F. Perrin, 1984, Geological andgeochemical models in oil exploration: principles andpractical examples, in G. Demaison and R. J. Murris, eds.,Petroleum geochemistry and basin evaluation: AAPG Memoir35, p. 53–77.

Wernicke, B., 1981, Low-angle normal faults in the Basin andRange province: nappe tectonics in an extending orogen:

Nature, v. 291, p. 645–648.Wilson, T. H., R. C. Shumaker, and L. Zheng, 1994, Sequential

development of structural heterogeneity and its relationship tooil production: Granny Creek, West Virginia, in A. Schultz, andE. Rader, eds., Studies in eastern energy and the environment:Virginia Division of Mineral Resources Publication 132, p. 20–25.

Yang, C., 1998, Basin analysis of the carboniferous strata incentral and southern West Virginia using sequencestratigraphic principles: Ph.D. dissertation, West VirginiaUniversity, Morgantown, West Virginia, 348 p.


seismic evaluation of differential tectonic subsidence...

Documents