Balanced Structural Cross-Sections of the Central Salt Range and Potwar Plateau of Pakistan: Shortening and Overthrust Deformation

AN ABSTRACT OF THE THESIS OF

Dan M. Baker for the degree of Master of Science in Geology presented on June 4, 1987. Title: Balanced Structural Cross-Sections of the Central Salt Range and Potwar Plateau of Pakistan: Shortening and Overthrust Deformation

Abstract approved: Redacted for Privacy 5obert J. Lillie

The Salt Range and Potwar Plateau are part of the currently active foreland fold-thrust belt of the Himalaya in Pakistan. Seismic reflection data have been combined with surface mapping, drillhole and paleomagnetic information to construct balanced structural cross-sections of the current frontal thrust zone (Salt Range), and to estimate the amount of horizontal contraction and timing of deformation for the central Salt Range/Potwar Plateau. The data show that a 1 km offset of the basement acted as a buttress that caused the central Salt Range/Potwar Plateau thrust sheet to ramp to the surface, exposing Mesozoic and Paleozoic strata at the range front. The basement offset may have been formed during Eocam- brian rifting; however, an origin related to Neogene crustal flexure is supported. The frontal portion of the thrust sheet was passively folded as it overrode the subthrust surface upon a thick, ductile layer of salt. The proposed ramping mechanism includes deformation prior to overriding the basement offset and suggests a smoother thrust trajectory than the ideal, stairstep profile. Balanced sections do not indicate a lateral change in shortening in the central Salt Range, as would be expected during proposed regional rotation during ramping. Just north of the Salt Range, the thrust sheet is undeformed, except for minor folding related to salt redistibution. Lack of internal deformation in the Soan syncline is due to decoupling of sediments from the basement along the thick Eocambrian salt layer. Still farther north, in the Northern Potwar Deformed Zone (NPDZ), a hinterland dipping duplex is disharmonic with mapped surface structure, suggesting an upper-level detachment. Duplexing ends southward in a "triangle zone" geometry, and an apparent imbricate fan. Beneath the northern limb of the Soan syncline, the duplex and imbricate fan are separated by an oblique ramp. Section balancing suggests about 45 km of shortening across the NPDZ. The timing of structural events is constrained by stratigraphic evidence, paleomagnetic dating of unconformities and sediment accumulation rates. Early to middle Pliocene (approx. 4.5 Ma) clastic deposition in the southern Potwar Plateau, previously interpreted in terms of compressional deformation, may instead be due to passage of the peripheral bulge (Sargodha High) through the Indian lithospheric plate causing basement normal faulting, uplift, and erosion. Although the timing of deformation in the central Potwar Plateau is poorly constrained, the Salt Range thrust sheet is estimated to have overridden the basement offset from 2.1-1.6 mya. Cross-section balancing gives a minimum of 19 to 23 km of shortening across this frontal ramp. The rate of Himalayan convergence which can be attributed to underthrusting of Indian basement beneath sediments in the Pakistan foreland is therefore 9-14 mm/yr. Deformation front migration appears to be controlled by the distribution of salt and basement buttressing in the central Salt Range/Potwar Plateau. Rapid facies progradation, measured in the Jhelum Reentrant, is thought to be controlled by thrust propagation rates and not indicative of typical foredeep migration. Restorations suggest that the lower and middle Siwalik molasse depocenter was in the southern Potwar Plateau, prior to jumping south of the Salt Range about

2 mya. Balanced Structural Cross-Sections of the Central Salt Range and Potwar Plateau of Pakistan

Shortening and Overthrust Deformation

Dan M. Baker

A THESIS

submitted to

Oregon State University

in partial fulfillment

of the requirements for the

degree of

Master of Science

Completed June 4, 1987

Commencement June, 1988 APPROVED:

Redacted for Privacy

Professor of Geoeogy in charge of major

Redacted for Privacy

Cailan of Depar ent of Geology

Redacted for Privacy

Dean of Gradua School

Date thesis is presented June 4, 1987 ACKNOWLEDGEMENTS

Thesis work in northern Pakistan was supported by NSF grants INT- 8118403, EAR-8318194, INT-8609914, and EAR-8608224. Additional assistance was provided by Amoco Production Company and Chevron Inc. I am grateful to the Oil and Gas Development Corporation and the Ministry of Petroleum and Natural Resources of Pakistan for the use of reflection and gravity data discussed in this study. Depth conversion of seismic sections performed with the AIMS seismic modeling package, donated to Oregon State University by Geoquest, International. Thanks to Bob Lillie for his time, help in writing, and enthusi- asm. The time and consideration given by other committee members,

R. Yeats, V. Kulm and D. Bibee, plus M. Yousuf was appreciated. I would also like to thank my colleagues in this study for their help and willingness to share ideas, especially M. Leathers whose hard work got things going. I am grateful to Mrs. Hamerick, a grammar school teacher, whose special attention got me interested in science. The moral support of my parents and family members helped me see past the poverty that accompanied graduate school. Special thanks to my wife, Linda, for her assistance, especially in typing, motivation to return to school, and her support throughout the process. TABLE OF CONTENTS

Page INTRODUCTION 1 Objectives 1 Previous Work 2 Tectonic Setting 6 Methods 9 Depth Sections 9 Balanced Sections 10 Restorations 16 Calculation of Shortening 17

STRATIGRAPHY AND SEDIMENTATION 19 Stratigraphy 19 Sedimentation and Depositional Models 21 Depotectonics 22 Depocenter Migration across the Central SR/PP 22 Foredeep Migration Rates 23

STRUCTURE OF THE CENTRAL SALT RANGE AND POTWAR PLATEAU 29 General Description 29 Northern Potwar Deformed Zone 32 Description 32 Analysis 41 Transition from the NPDZ to the Soan Syncline 44 Description 44 Analysis 45 Southern Potwar Plateau/Soan Trough 45 Transition from the Southern Potwar Plateau to the Salt Range 49 Description 49 Analysis 49 Salt Range 58 General Description 58 Eastward Termination of the Salt Range Thrust Sheet 61 Structures within the Central Salt Range 64 Ramping of the Salt Range Thrust Sheet 66 Mechanism 66 Removal of Molasse from Beneath the Salt Range 68

TIMING CONSTRAINTS ON THE EVOLUTION OF THE FOLD-THRUST BELT 72 Migration of Folding and Thrusting into the Foreland 72 Decollement Propagation Rate Calculation 73 Motion of the Thrust System 74 Tectonic Rotations 75 Timing of Basement Faulting 76

SHORTENING AND ITS IMPLICATIONS 82 Amount of Shortening 82 Salt Range 82 Southern Potwar Plateau/Soan Trough 82 Northern Potwar Plateau 83 Frontal Shortening Rate and Comparison to Plate Convergence Rates 83 Tectonic Rotation Versus the Distribution of Shortening 85

CONCLUSIONS 87 Future Work 89

BIBLIOGRAPHY 90

APPENDICES 101 (A) Development of the Himalayan Frontal Thrust Zone: Salt Range, Pakistan 101 (B) Restoration Results and Implications 116 (C) Factors Influencing Shortening Calculations 118 (D) Petroleum Geology 120 LIST OF FIGURES

Figure Page

1. Regional location map. 3

2. Tectonic map of northern Pakistan. 5

3. Map of seismic and well control. 8

4. Example of velocities used for seismic time to depth conversion. 11

5. Structure map of the SR/PP with paleomagnetic sample locations. 14

6. Sedimentation model, showing southward shift of molasse depocenter. 24

7. Restored positions of the Sargodha high, Salt Range, basement offset and Baun locality at 4.5 mya. 26

8. Block diagram illustrating a change in deformation sytle across an oblique ramp in the northern Potwar deformed zone. 30

9. Seismic transect across the northern Potwar deformed zone west of the oblique ramp. 33

10. Seismic transect across the northern Potwar deformed zone east of the oblique ramp. 35

11. Seismic transect across the northern limb of the Soan syncline. 37

12. Summary of deformation sytles in the northern Potwar Plateau. 40

13. Regional balanced cross-section and restorations. 42

14. Seismic transect across the southern Potwar Plateau. 46

15. Seismic profile across the frontal basement ramp. 50

16. Geologic map of the central Salt Range area. 51

17. Frontal balanced cross-sections. 52 18. Frontal ramping diagram showing preramping deformation, ramp shallowing, and footwall collapse. 56

19. Block diagram illustrating evolution of the central Salt Range. 59

20. Hangingwall/footwall cutoff separation map of the east-central Salt Range. 62

21. Restorations illustrating stratigraphic thickness changes versus the age of basement faulting. 69

22. Schematic diagram illustrating allochthonous and autochthonous origin of Upper Siwalik conglomerates preserved at the Baun locality. 79

23. Schematic crustal section demonstrating potential distribution of convergence absorbtion. 84

Al. Regional location map. 103

105 A2. Structure/geology map of the SR/PP.

A3. Schematic cross-section of the Himalayan foreland fold-thrust belt in Pakistan. 107

109 A4. Seismic profile of the frontal ramp region.

A5. Balanced cross-section and restorations demonstrating a possible mode of origin of the Salt Range. 112 BALANCED STRUCTURAL CROSS-SECTIONS OF THE CENTRAL SALT RANGE AND POTWAR PLATEAU OF PAKISTAN: SHORTENING AND OVERTHRUST DEFORMATION

INTRODUCTION

The active western Himalayan foreland fold-thrust belt presents an opportunity to study a developing stage of collision. Approximately 3000 km of seismic reflection profiles (supplied to Oregon State University by the Pakistan Oil and Gas Development Corporation andthe Ministry of Petroleum), have been interpreted in conjunctionwith surface maps (especially those of Gee, 1980), and published andunpub- lished well and gravity data. These data were used to construct balanced structural cross-sections of the central Salt Rangeand Potwar Plateau and to estimate amounts of shortening. The same data set was used by Leathers (1987) and Pennock (in prep) to construct cross-sections to the west and east of the thesis area, by Jaume(1986) to study the mechanics of thrusting and by Duroy (1986) to analyselithospheric flexure for the Himalaya of Pakistan. Appendix A (Baker et al., in review) is a brief overview of portions of this study,concentrating on the timing of frontal ramping and the calculationof a frontal shortening rate.

Objectives

The Pakistani foreland contains a wide variety of structures related to the combined effects of compressional andsalt tectonics. A major objective of this study is to describe structuralstyles and examine the relationships between different structures. In addition, recent sedimentological and stratigraphicstudies in the Salt Range/ Potwar Plateau (SR/PP) area have utilizedmagnetostratigraphy and tephrachronology to age-calibrate individual horizonsin the Siwalik molasse (e.g Johnson et al., 1986). These data constrain the timing of individual deformational events and the progressionof the thrust front

across the foreland basin. When used in combination with structural information, the evolution of the fold-thrust belt,especially the Salt 2

Range, can be inferred. The effect of the progressing deformation front and basement faulting on foredeep sedimentation is also considered. An attempt is made to solve the following problems: (1) the origin of the down-to-the north, basement offset that controls Salt Range ramping; (2) how ramping in the Salt Range compares with previously developed models of thrust ramping; (3) the role of thrust sheet ramping in the overall rotation documented in the area; (4) the amount of horizontal contraction of cover rocks; and (5) the rate of shortening in the frontal portion of the Pakistani foreland fold-thrust belt.

Previous Work

The Salt Range, with its excellent exposures of a wide range of Phanerozoic strata, has long attracted the interests of geologists. In addition to the stratigraphic significance of its outcrops, the presence of highly deformed Mesozoic and Paleozoic rocks sofar outboard of the major Himalayan deformation zone has attracted the attention of structural geologists. Stratigraphic classification of the Tertiary Siwalik succession across the SR/PP is an ongoing problem. Maps and general accounts of the stratigraphy were prepared by Wadia (1928), Cotter (1933), and Gill

(1952). More recently, attempts at an absolute chronostratigraphy through the application of paleomagnetic stratigraphy and fission track dating, have been made (e.g. Barndt et al., 1978). Detailed maps (1:50,000), recently published by Gee (1980), demonstrate the structural complexity of the Salt Range. Attempts to unravel the history of less spectacular structures in the Potwar Plateau were made by Pinfold (1918), Cotter (1933), Gill (1952), Martin (1962) and later by Voskrensenskiy (1978). A summary of some of the data and results from subsurface studies undertaken during hydrocarbon exploration can be found in Khan et al. (1986). The significance of the position of the SR/PP on the west sideof the bend in the Himalayan trend at the Hazara-Kashmir Syntaxis(HKS; fig. 1), and on the east end of the sinuous Pakistani fold beltremains a problem. The Ceozoic festoons in both India and Pakistan are thought Fig.1: Regional location map, showing position of the Salt Range/Pot- war Plateau (SR/PP) near the west end of the overall Himalayan trend. Sinuous lines through southern and central Pakistan represent the Pakistan fold-thrust belt. HKS=Hazara-Kashmir Syntaxis; ISZ=Indus Suture Zone; MBT=Main Boundary Thrust; MKT= Main Karakorum Thrust; SH=Sargodha High. Rectangular box denotes area of figure 2. 4 to be controlled by the reactivation of Precambrian basement lineaments by Swaminath and others (1964), or by the distribution of evaporites at the base of the sedimentary sequence (Seeber et al., 1981). A more general account concerning possible causes of the scalloped outline of Pakistan's mountain fronts was given by Sarwar and DeJong (1979). The Hazara region (including the SR/PP) was genetically shown to be related to the Himalaya by Wadia (1931), who noted that the rocks of the Hazara are a stratigraphic continuation of those exposed in the Himalaya. He suggested that the two belts are sections of the same mountain system, bent around a horst-like projection of the Indian shield. Crawford (1974) offered an alternative, hypothesizing that the Salt Range has been rotated 75° counterclockwise from a position in line with the main Himalayan front. His theory is in part supported by recent work which shows that thrusting in the foreland is accompanied by counterclockwise rotation of portions of the Salt Range (Opdyke et al., 1982). A conflicting result, however, was presented by Powell (1979), who showed that the Salt Range and the main Himalaya have an east-west component of convergence, suggesting that the HKS may be the collision point of two different parts of the orogen. Davis and Engelder (1985) suggest that relative movements between the Salt Range and the Himalaya can also be explained by the lubricating effects of salt within the decollement zone west of the syntaxis. Recent work by Yeats and others (1984), who documented that deformation in the eastern Salt Range is younger than 0.4 mya, and by Lillie and Yousuf (1986), who explained the emplacement of the Salt Range by ramping over a basement fault, have set up this investigation into the evolution of the central Salt Range. A gross balanced section across this area (Coward and Butler, 1985) provides a basic framework for investigation, but has the main decollement at the base of the molasse instead of within the Eocambrian evaporites, as shown on reflection profiles and includes several thrusts not rocognized in the field (Yeats and Hussain, in press). Previous estimates of shortening represented by overthrusting at the Salt Range are 35 km by Gee (1934) and 50 km by Sarwar and DeJong (1979). An estimate of total shortening from the Salt Range to the Main Mantle Thrust (fig. 2) is 470 km (Coward and Butler, 1985). 5

7Q° 74° 37°69' ° NORTHERN PAKISTAN

OC Generalized Tectonic Map Thrust (After Kazmi and Rana, 1982) LEGEND (FOR AREA SOUTH OF M B T 36° Neogene Malone Chitral. Paleocene to Eocene Monne to Continental Sedimentary Rocks Penman to Jurassic KOHISTAN

.1111PrIcantbrian M Campion Sedimentary Room ARC

12c1Preeartaxian Indian Shook'

Oregon State University 1986 35°

Fire Line

Ka*/ Rivav KoDai AJottod PESHAWAR

J!,!' or abad sr=go.2. 34° Pesnowor B. 1.

ono,

POTWAR Soon Rnn 33° PLATEAU

32°

PLAIN Lyallpur

3'° Fig. 2: Generalized tectonic map of northern Pakistan, illustrating features in, and north of the western Himalayan foreland (after Kazmi and Rana, 1982). Salt Range is the area of Precambrian to Eocene sedimentary rocks north of the Salt Range Thrust. Dotted line north of the Salt Range marks the subsurface position of the basement offset. Bold rectangles represent the areas of figures showing available seismic coverage. Right rectangle denotes area of figure 3. Left rectangle represents western SR/PP coverage (see Leathers, 1987). X-X' and Y-Y' are cross-sections of Leath- ers (1987) and Z-Z'is a crustal cross-section (see Duroy, 1986; Duroy et al., in press). HR=Hill Ranges; JR=Jhelum Reentrant; KH=Kirana Hills; MR=Mianwali Reentrant; and SuR=Surghar Range. 6

Tectonic Setting

The Himalaya of Pakistan is a product of the ongoing collision of the Indian subcontinent with Eurasia. The continent-continent collisional event began in Late Eocene, after docking of the Kohistan/Ladakh arc (fig. 2) with Asia in late Cretaceous(Windley, 1983). Seafloor reconstructions indicate that since the collision began, 1500 to 3500 km of convergence between Eurasia and India has occurred.Convergence is accommodated by giant underthrusting of one continental block beneath the other and/or diffuse deformation over a broad zone (Molnar and Tapponnier, 1975; Tapponnier et al., 1982). Some underthrusting of the Indian subcontinent beneath Eurasia is supported by reconstructions of Gondwana, and by gravity and seismic refraction studies indicating twice the normal thickness of continental crust beneath the high Hima- laya in Tibet and northern Pakistan (Finetti et al., 1981). Diffuse deformation involves crustal thickening via folding and thrusting, and lateral strain by strike-slip faulting (Molnar and Tapponnier, 1975). Continental fragments north of the Himalaya may be squeezed out of the way of the advancing Indian subcontinent alongstrike-slip faults ("escape block tectonics"). To the south, in the lesser Himalaya of northern Pakistan(Hill Ranges; fig. 2), detachments at upper crustal levels occur along a series of south-verging thrusts (Yeats and Lawrence, 1984). The southern margin of the collision zone in Pakistan (Salt Range, Kohat and Potwar Plateaus) is an active foreland fold-thrust belt formedin response to the underthrusting of cratonicIndia beneath its own Phanerozoic sedimentary cover. On the basis of microseismicity, the Salt Range, Potwar Plateau and Hill Ranges are interpreted as thin skinned, underlain by a low-angle decollement which surfaces at the Salt Range front (Seeber and Armbruster, 1979). At the Hazara-Kashmir Syntaxis, the fold-thrust belt changesin trend and broadens in Pakistan (fig. 1). In India east of the Punjab reentrant, there is strong coupling between sediments andbasement; consequently the zone of underthrusting is narrow and has ahigh angle of cross-sectional taper (Davis et al., 1983; Davis and Engelder,

1985). In Pakistan, on the west flank of the Hazara-KashmirSyntaxis, 7 low-strength evaporites of late Precambrian to Early Cambrian age constitute the zone of decollement (Crawford, 1974). As a result and in contrast to India, the zone of underthrusting extends far out over the foreland (Seeber and Armbruster, 1979), more than 100 km southof the Main Boundary Thrust (fig. 2), and it has a narrowcross-sectional taper. Further evidence for the decoupling properties of the evaporites lies in the large detachment earthquakes which occur beneath the foreland in India, but not in Pakistan (Quittmeyer and Jacob, 1979; Seeber et al., 1981). In northern Pakistan, a southward progression of major compressional deformation has been documented (Raynolds and Johnson,1985; Johnson et al., 1986), although out of sequence deformationalso occurs (Burbank and Raynolds, in press). At the southern margin of the Peshawar basin, major thrusting occurred prior to 2.8 mya(Burbank, 1982), while in the eastern Salt Range, major deformation may be younger than 0.4 mya (fig. 2; Johnson, etal., 1979). The deformation front in the central Salt Range may now be shifting to theLilla structure, a thrust-cored anticline in the JhelumRiver plain 15 km south of the Salt Range. South of the fold-thrust belt, sediments slope gentlynorthward from the Sargodha High, an area of moderate seismicity,including exposures of Precambrian rocks at theKirana Hills (fig. 2; Farah et al., 1977). It is often assumed that the Sargodha High is anold Precambrian feature analogous to the Aravalli Range ofIndia. However, similar stratigraphy in wells drilled north and southof the Sargodha High, plus sharp topographic breaks between the Kiranahills and the surrounding plains, suggest that this ridge is a relatively young feature (Yeats and Lawrence, 1984). This ridge parallels the overall northwest-southeast Himalayan trend and is perhaps an"outer swell" related to the downflexing of the underthrustingIndian plate (Seeber et al., 1981; Duroy, 1986; Duroy et al.,in press). Alternatively, the Sargodha High could represent an incipient Himalayanbasement thrust analogous to the Main Central Thrust (MCT) in India(Yeats and Law-

rence, 1984). In addition to the Himalayan orogeny,evidence of past and continuing epeirogenic movements exist in the stratigraphicsections of the 8

7.5 +3,5 7 3' 61 745

33.45.

53. 06

010,A 490010, 33,5

33,X

3245

HIMALAYAN FORELAND THRUST BELT NORTHERN PAKISTAN SALT RANGE POTWAR PLATEAU

E xplorofion 119 .1 *619 and 0B A

5666c Lines .. air 4a. C.00.000 004.000C.CI . San P1.00. row Ca0n 32.15' O. 4a.01000.40 Pomo. eso.. %ow Kxxx 0 699060 99666 C. o 1006.0. Amos. 0100. 0...00. Cs. sre, llo 0080 6600*.L..Nmaavol

SAR000A

Down Mar um.soy 1966 31500' 73.30' 73.45' '4 00 7205 7e 30' r295 731;,0 73.15'

Fig. 3: Map of available seismic andwell control in the central and eastern SR/PP. See Khan et al. (1986) foradditional information and Leathers (1987) for welldata and structure contour maps. Portions of seismic lines illustratedin text (figs. 9, 10, 11, 14, and 15) and used to constructbalanced cross-sections are figure 5. denoted by solid lines. Rectangular box denotes area of 9

Salt Range, Potwar Plateau and Jhelum reentrant. The concentration of Eocambrian evaporites into distinct basins, uplift of the western Salt Range region relative to the eastern Salt Range during or after the Cambrian, and the subsidence of the western Salt Range region relative to the eastern Salt Range during the Mesozoic attest to pre- orogenic movements. Magnetic-polarity stratigraphy indicates that since the onset of the Himalayan orogeny, the Potwar Plateau was the locus of molasse sedimentation; it apparently subsided rapidly during early Siwalik deposition (approx. 15-5 mya). Since then, the regions to both the west (Trans Indus Salt Range) and east (Jhelum Reentrant) became the locus of sedimentation in the foredeep (fig. 2; Khan, 1984). Today, the Jhelum Reentrant remains a structural trough subsiding more rapidly than adjacent parts of the molasse basin (ie. Potwar Plateau; Johnson et al., 1979).

Methods

The two main tools used in the structural analysis of thrust belts are balanced cross-sections (Dahlstrom, 1969) and the interconnections or relationships among faults (Boyer and Elliot, 1982). The interconnections of faults are studied qualitatively using observations of branch and tip lines, for example, to construct branch line maps, while balanced sections require a more analytical approach. For this reason, only the techniques involved in section balancing, including restoring sections, are explained here. Section balancing, including the deter- mination of shortening, is subdivided into the following categories: (1) conversion of seismic time sections to depth sections; (2) reinterpretation of the depth sections to balanced sections, incorporating surface and drilling data; (3) restoration of the balanced sections; and (4) shortening calculations. Depth Sections: Surface constraints on the interpretation of the seismic reflection transects (fig. 3) were the geologic maps of the Salt Range (1:50,000 scale) by Gee (1980), and maps by Cotter (1933) and Gill (1952) for the Potwar Plateau. Faults and fold axes identified on Landsat imagery (approx. 1:250,000 scale) were used to supplement geologic maps in the northern Potwar Plateau. Strikes and dips, corrected 10 to apparent dips, were projected along strike into the line of section from a couple of kilometers away in the frontal sections to up to five kilometers away in the Potwar Plateau. Well information (fig. 3; see Leathers [1987] for additional well data), particularily formation thicknesses, were also projected onto sections paying particular attention to the depositional strikes of units. An additional subsurface constraint was the structure of the basement surface, as determined via gravity modeling by Duroy (1986; Duroy et al., in press). Velocities used for depth conversions (e.g. fig. 4) were obtained by incorporating: (1) the range of possible velocities for a given lithology (Telford, 1976); (2) stacking velocities from seismic lines; and (3) the velocity structure calculated by Leathers (1987) from sonic logs. The velocity structure was fine tuned on the frontal sections by modeling velocity effects using the AIMS (trademark Geoquest, Int.) program, including automatically migrating sections during depth conversion (fig. 4). The main depth section was extended to the north from the AIMS sections by hand calculations to depth using the velocity constraints listed above. Balanced Sections: Frontal section lines were chosen along seismic lines (fig. 4), which are slightly oblique to the transport direction, while the main balanced section was constructed parallel to the transport direction of the Salt Range thrust, taken as S 15° E (Yeats et al., 1984). Vertical, rather than downplunge sections were constructed, because major structures along the transects have negligible plunge. Data were assembled by drawing topographic profiles and then plotting the outcrop constraints. Next, the geometry of the sections were modified by extrapolating and interpolating via the Busk and kink methods. The Busk method assumes concentric folding whereby bed contacts are drawn between inflection points around their centers of curvature. The kink method assumes straight limbs and angular hinges, but can be used to approximate all curved shapes. Axial surfaces bisect the angles between fold limbs and in this way the kink method has been used to assure a retrodeformable section (Suppe, 1985). The sinuous bed method, which requires that beds have a constant length throughout the stratigraphic section so that beds extend into their original stratigraphic wedge when retrodeformed (Dahlstrom, 11

Fig. 4: Example of velocities used for seismic time to depth conversion (i.e. for section C-C1). (A) Digitized time section showing velocities in m/sec. (B) Cross-section after depth conversion, but prior to balancing. During cross-section balancing, velocity effects are smoothed out and details added from surface maps and drilling information. Formation abbreviations: T-QM=Tertiary to Quaternary molasse; MS=Middle Siwaliks; DP=Dhok Pathan; Na=Nagri; LS= Lower Siwaliks; Ch=Chinji; RM=Rawalpindi Group(molasse); e-E=Cambrian to Eocene (i.e. platform strata); SRF=Salt Range Formation; pe=Precambrian (crystaline) basement. 10km 20km 30km r 1 0 i 2400 TOPOGRAPHIC SURFACE,.. 2900 310 0 a z 3100 2 o 3500 3 600 U 360 0 4 500 W co 4500 3 6000 6000 A 1 . 4

0 20km 30km

1 1 TOPOGRAPHIC SURFACE SEA SEA LEVELMS (DF:No) LEVEL u) CC LLI 2 2 1w 2 0 4 4 -.1 ;2

Figure 4 13

1969), was used to check the viablity of all balanced sections (see Elliot, 1983). Sections were constructed at a scale of 1:78,740. Because a single bed length measurement does not constitute section balancing, eroded beds were restored so that consistency of bed length could be used to check the validity of interpreted structural styles. Measured bed length variations were held to under 1%, with the exception of the NPDZ portion of the main balanced section which had a 1.3% error. Due to ductile flow within the Salt Range Formation, only the overlying horizons could be checked by measuring bed lengths. Follow- ing the assumption of conservation of area in cross-sectional view (Dahlstrom, 1969), the Salt Range Formation was area balanced (e.g. Gwinn, 1964), to correspond to a restored model, on the main balanced section using a planimeter. Keuffel and Esser Co.'s compensating polar planimeter (model # 62 0000) could be used to measure areas to within 0.01 sq. in., which is the equivalent of 0.04 sq. km. on the cross-section. Area balancing of the Salt Range Formation assumed that equal amounts of salt entered and left the section duringdeformation. Salt areas originally north and south of the basementoffset were balanced separately, thereby supplying a constraint on the original thickness of the Salt Range Formation at the basement offset (see restoration methods and appendix B). A pair of reference lines, called pinlines, marking the ends of balanced sections, were chosen in areas of little or no interstratal slip. The undeformed foreland was chosen as the frontal pinline on all balanced sections. The rear pinlines were placed along the axis of the Baun syncline on the frontal sections (see sections A to E, fig.3) and near the rear of a thrust slice in the NPDZ forthe main balanced section (F-F', fig. 5). In addition, local pinlines were used within the main section, at the fold axes of the Baun and Soan synclines and just south of the Salt Range frontal thrust, to further constrain the shortening across individual structures (e.g. Woodward and Boyer,

1985). The most important criterion in constructing a retrodeformable section across a faulted terrane is matching hanging-wall and footwall cutoffs. In this way footwalls exposed at the surface reflect the 14

Fig. 5: Structure map of the SR/PP including locations of paleomagnetic sample sites (with paleomagnetic pole directions where available) and other figures illustrated in text. Solid structure symbols are more pronounced features, mapped from LANDSAT photos. Dashed symbols denote less prominent structures whose locations were derived from seismic, surface maps, and wellcontrol. Dotted pattern represents buried structures. Structure abbreviations: Ba=Balkassar Anticline; BF=Basement Fault; BS=Baun Syncline; Bu=Buttar Anticline; C=Chambal Ridge; CF=Chhumbi Fault; D=Dhulian Anticline; DR=Domeli Ridge; HR=Hill Ranges; J=Jabbar Anticline; JM=Joya Mair Anticline; JT=Jogi Tilla Anticline; K=Khaur Anti- cline; KDF=Karangal/Diljabba Fault; KF=Kalabagh Fault; KK=Kotal Kund Syncline; KKF=Kallar Kahar Fault; L=Lilla Anticline; Le=Lehri Anticline; M=Mahesian Anticline; MA=Meyal Anticline; MBT=Main Boundary Thrust; MF=Murree Thrust; 0=Oblique Ramp; P=Pabbi Hills Anticline; R=Rhotas Anticline; RF=Riwat Fault. Paleomagnetic sample locations: (1) Andar Kas; (2) Baun; (3) Bhangala; (4) Dala Nala; (5) Dina; (6) Ganda Paik; (7) Hasal Kas; (8) Jalapur; (9) Kamliel Kas; (10) Kas Dovac; (11) Khaur Composite; (12) Kotal Kund; (13) Nagri; (14) Pabbi Hills; (15) Rada Kas; (16) Rhotas; (17) Soan; (18) Tatrot. Paleomagnetic sites adapted from Keller and others (1977), Barndt and others (1978), Opdyke and others (1979), Tauxe and Opdyke (1982), Johnson and others (1982), Opdyke and others (1982), Raynolds and Johnson (1985), and Burbank and others (in press). 15

730 720

330

SALT RANGE FRONTALTH L 730 0 50 k m ..),....y/ THRUST FAULT; TEETH ANT !CLINE PALEOMAGNETIC ON UPTHROW BLOCK ti) SAMPLE SITE ,...Y NORMAL FAULT; TICK ...... 15/ SYNCLINE r...... , DIRECTION OF PALEO- MARKS ON DOWNTHROWN BLOCKS MAGNETIC POLE

Figure 5 16 shape of eroded hanging-walls, and hanging-walls seen at thesurface represent the shape of burried footwalls. However, a special problem exists when attempting to balance a thrust sheet across a ramp. Slip may not be conserved on cutoffs along thethrust fault. This variation in slip is currently a point of controversy (see Woodward and Boyer,

1985). Slip is thought to increase upwards on a thrust fault(Elliot, 1976), or thrusts lose displacement upwards (Fain, 1967; Dahlstrom, 1969; Thompson, 1981; Suppe, 1983) as a thrust sheet overrides a ramp. The selection of pinline positions north of the ramp,within the Baun syncline, followed Suppe's (1983) assumption that there is nointer- stratal slip at the trailing edge of a fault-bend fold. Therefore, shear due to ramp climb was transferred to the leadinglimb. As a result, shear steepened the hanging-wall ramp so thathanging and footwall ramp shapes differ slightly. If, alternatively, the footwall and hanging-wall ramps were made to match, shear would impartinterstratal slip on the back limb (i.e. the remainder of the thrust sheetextending to the northern Potwar Plateau; Woodward and Boyer,1985). Restorations: In order to restore a section to its undeformed state, the original state of the strata must be known or assumed. In thrust faulted terranes, the basement or strata beneath thedecollement are held fixed, and fault displacements are reversed bymoving the hanging-wall rocks along the thrust fault surface. However, in the present study area there are two major problems. First of all, the salt hori- zon has flowed, and therefore itsoriginal depositional geometry is unknown. Secondly, the lower plate has been flexed downward to the north by an unknown amount. Retrodeformation was accomplished by restoring bed thicknesses, starting at the base of the sedimentary section, relative tothe sinuous bed length distance from the frontal pin line. In areas of nonparallel folding near the Salt Range front,undulations were smoothed and the average bed thickness across theentire fold was used. The earliest restoration was performed first. For the later restoration, the basement and overlying sediments were flexedby the desired amount and additional sediments stackedabove. Because molasse deposition was contemporarywith thrusting, restoration required knowing the precise ages ofsediments and timing 17 of deformation events. Individual mapped formations in the Pakistan foreland are time transgressive, therefore formations cannot simply be stripped away. For these reasons, convenient times of restoration were chosen such that intervals of strata, representingdistinct assemblages of lithologies whose ages are constrained, could be sequentially removed. These restoration times correspond to post-Eocene/pre-Murree deposition (approx. 35 mya), and post-Middle Siwalik/pre-UpperSiwalik deposition (approx. 5 mya; see F-F' fig. 5). In consideration of the previously mentioned complexities and the times of restoration, the following assumptions were made: (1) salt originally north and southof the basement offset could be (area) balanced independently;(2) for the 35 mya restoration, the top of Eocene north of thebasement offset was horizontal; (3) south of the basementoffset, post depositional erosion of Eocene and older strata may have occurred; and(4) a topographic (i.e. depositional) slope existed towards the axis of the foreland basin during molasse deposition. Assumptions (2) and (4) aid in estimating basement dips at the respective times of restoration. In this way, the restoration process itself geometricallyconstrains the amount of flexure. Appendix B discusses the results and implications of these assumptions and the retrodeformation. Finally, in a deviation from standard restoring techniques,Indian crust along with its sedimentary cover, was pulledback through time according to its present flexed shape (e.g. Duroy, 1986; Duroy etal., in press). Because the shape of flexural bulges can vary withtime, a generalized flexure model was assumed. From south to north, this model consists of an area of flat basement corresponding to the crestof the bulge (currently the Sargodha High), steep basementdip on the flank of the bulge, and then a zone of shallow to moderatebasement dip (currently the Jhelum River plain and Salt Range) giving way to steep basement dip in the area of maximum flexure(currently the Potwar Plateau and Hill Ranges). Calculation of shortening: Shortening was calculated bysubtracting the present horizontal distance from the deformedbed lengths (Dahlstrom,

1969). These values differ slightly from those obtained bysubtracting the straight-line distance between pinline cutoffs(Suppe, 1985), or the basement length, because beds weretilted prior to thrusting. 18

Subtracting the horizontal distance was preferred because reversal of interpreted extension during the basement faulting event is not taken into consideration. The results are also directly comparable to plate motions. True shortening in the horizontal plane should also use the horizontal projection of the original bed length. However, errors introduced in this study by pre-deformational tilted beds are at most 0.1 km, or about 0.5%. Shortening values were corrected for obliquity using a cosine correction, which is appropriate for the low angles involved here (Cooper, 1983). Experimental error was minimized by using the same instruments and technique to determine both the deformed and original bed lengths. Uncertainties such as those derived from imprecise measurements, bed tapers, and most other factors (see appendix C) are covered bythe arbitrarily assigned 5% error. An exception is the balanced section across the NPDZ (see F-F', fig. 5), where only amoderate degree of confidence is assigned to its structural interpretation; estimates of shortening there, based on deformation style and not precise interpretations of individual structures, have been arbitrarily assigned a 33% error based on these uncertainties. 19

STRATIGRAPHY AND SEDIMENTATION

Stratigraphy

The basement beneath the SR/PP is probably a northeastern extension of the Indian craton. Precambrian metasediments and plutonic rocks constitute the most northernly exposure of the Indian Shield at the Kirana Hills (Sargodha High; fig. 2). Resting directly on these crystalline rocks north of the ridge, the Salt Range Formation or "saline series", consisting of mainly gypsum, salt marls and rock salt, has been correlated with the Hormuz Formation of the Persion Gulf (Zarkov, 1981). Both Precambrian and Cambrian ages have been assigned to the Salt Range Formation by various workers; therefore,its age is referred to as Eocambrian. Existence of evaporites in the Hazara Slate Series suggests that the Salt Range Formation may extend northward from its outcrops at the Salt Range at least as far north as Hazara (Latif, 1973). This is a significant hypothesis for the sub- Himalaya of Pakistan, because the evaporites form the zone of decollement. Nonmarine to shallow marine sandstones and shales of the Cam- brian Jhelum group overlie the Salt Range Formation, except in the western Salt Range area. Similarity of early Paleozoic stratigraphy throughout southwestern Asia implies the developement of a widespread Gondwanan passive continental margin following late Precambrian orogen- esis (StOcklin, 1974; Bender, 1975; Cherven, 1986). Above the first of three major unconformities, lower Permian strata overlap the Cambrian westward. The Gondwanan Tobra Formation, at the base of the Permian succession, contains distinctivePrecambrian granitic clasts from the Indian shield. Marine influence increases as clastics are replaced by calcareous rocks in the upper Permian. A Mesozoic sequence of predominantly marine clastics overlies the Permian in the western half of the Salt Range. The strike of Mesozoic units prior to deposition of overlying rocks in the Potwar Plateau suggests a west-northwest facing, Mesozoic passive margin (Yeats and Hussain, in press). The next major unconformity is at the base of the Tertiary, where marine strata of Paleocene and Eocene age overlap Mesozoic and Permian rocks eastward. Basin analysis indicates that Paleogene strata 20 were deposited in a basin, centered to the north near theKala Chitta Range (Wells, 1984). Above the last of the major unconformities are fine-grained clastics of the Murree Formation, derived from the Himalaya. The Murree Formation, along with the lower Siwalik Kamlial Formation make up the Rawalpindi Group. Succeeding the Murrees across the Potwar Plateau are massive sandstones and clays of the upper Miocene to Pliocene Siwalik molasse. Confusion abounds concerning classification and subdivision of the Siwaliks. Early workers attempted to subdivide the Siwaliks using fossil assemblages; however, due to the scarcity of fossils it became more useful to subdivide the Siwaliks into litho- stratigraphic units. This is the present system of classification. Pilgrim (1910) subdivided the Siwalik Group into the Upper, Middle and Lower Siwalik subgroups. In 1913, Pilgrim subsequently subdivided these subgroups into formations; this classification remains today, with only minor changes. The Lower Siwaliks correspond to the Kamlial and Chinji Formations, while the Middle Siwaliks are dividedinto the the Nagri and Dhok Pathan Formations. Facies changes, and therefore formation contacts, are time transgressive. Sandstones within the Soan Formation (Upper Siwalik) grade laterally into boulder conglomerates, indicating the encroachment of Himalayan thrust sheets. Reflection data show that the undeformed molasse section is as thick as 5500 m in the Soan syncline. Post Siwalik deposits are critical for dating deformation; unfortunately much of the younger strata have been removed by uplift and erosion. An exception is the Lei conglomerate preserved in the Soan syncline south of Rawalpindi, which has a basal age of about 1.9 m.y.

(Raynolds, 1980). The Lei conglomerate, like the Kalabagh conglomerate of the southwestern Salt Range, is a valley fill conglomerate with Eocene clasts. Post Siwalik conglomerates are related by similarity of source rocks and mode of formation, whiletheir ages are related to local events which controlled their deposition. At the top of the section in portions of the Potwar Plateau is the Potwar Silt. The age of this formation (see appendix A [Baker et al., in review]) is significant because it is believed to have been ponded behind the rising Salt Range (Yeats et al., 1984). Recent work has shown that the 21

Potwar Silt is a loess derived from an easterly source (Shroder, in prep).

Sedimentation and Depositional Models

As a result of continent-continent collision, a structural trough, called the Indo-Gangetic foredeep was formed on the Indian plate. Coarse-grained alluvial-fan sediments were deposited along the basin margin (proximal), while finer grained fluvial sediments were deposited downstream (medial and distal). In response to continued northward movement of the Indian crust, the foredeep migrated relatively southward across the foreland. In the Indian foreland, Acharyya and Ray (1982) documented that the southward progression of the thrust front is accompanied by migration of the Siwalik depositional axis. The alluvial plain there can be subdivided into a northern domain where rivers flow southfrom the mountains across broad distributary fans, a central area where south-flowing rivers merge into the Ganges and flow parallel to the Himalayan front, and a southern domain where smaller rivers flow northward from the Indian shield. Ideally, sediments deposited in these three domains are juxtaposed by the southward migrating molasse basin, such that the molasse sequence is underlain by shield derived alluvial sands and topped by Himalayan fan gravels. In Pakistan, where the Siwalik molasse outcrop is broader and has been age-calibrated more extensively by magnetostratigraphy and tephrachronology, a more detailed tracking of facies migration has been documented. Distal, white sand sandstones deposited by a major trunk stream with east-northeasterly paleocurrents (Raynolds, 1981) are replaced upsection by brown sandstones with a nearby Himalayan prove- nance (medial) and south flowing paleocurrents. The lithologies of the white and brown sandstones resemble those which are carried by the modern Indus and Jhelum rivers respectively (Burbank and Raynolds, in press). In the Jhelum reentrant, the observed coarsening-upwards of the Himalayan molasse, due to outward displacement of the orogenic front, is capped by the Soan boulder conglomerate. By measuring the southward younging of the first appearance of this 3 to 1.5 m.y. 22 boulder bed, Raynolds and Johnson (1985) estimate that the rate of southward progradation of lateral facies has been up to 30 mm/yr. An exception to the ideal foredeep migration model, in which facies contacts would be time-transgressive to the south, is the Nagri-Dhok Pathan contact (approx. 7.9 to 8.6 m.y.) which is northward time-transgressive in the central Potwar Plateau (Behrensmeyer and Tauxe, 1982; see Leathers, 1987). Numerous magnetostratigraphic sections have been measured across the SR/PP and within the Jhelum reentrant (located on fig. 5). By correlating these sections to each other and to a magnetic reversal time scale (e.g. Cox, 1969), sedimentation rates for different time intervals can be derived. At most locations plots of accumulated sediment thickness versus time show low initial sedimentation rates in a distal foredeep, followed by maximum sedimentaccumulation rates in the foredeep depocenter, and finally an attenuation of sedimentation rates due to the encroachment of deformation. Sedimentation ceases as the area is deformed and positive relief attained. The resulting sigmoidal plots suggest that the depocenter has migrated southward at over 20 mm/yr (Raynolds and Johnson, 1985). When a thrust of regional extent ramps toward the foredeep basin, a new intermontane basin may form on the back of thethrust sheet. Behind the thrust, pre-existing fluvial systems may be ponded. Burbank (1982) suggests that the Peshawar, Campbellpore and Kashmir basins were formed in this way (fig. 2). These intermontane basins are probably ephemeral features whose sediments will most likely be removed by later stages of deformation (Burbank, 1983). In the Potwar Plateau, the Potwar silt may represent an intermontane basin deposit ponded by the rising Salt Range. The Potwar silt is currently being dissected and within a few tens of thousands of years will probably be completely eroded (R. Yeats and R. Lillie, writ. comm., 1984).

Depotectonics

Depocenter Migration across the Central SR/PP: Depositional models based on Upper Siwalik sediments in the Jhelum reentrant (e.g. Raynolds and Johnson, 1985) may not apply to older molasse deposition across the 23 central SR/PP. For example, the southward shift of isochron thickness maxima observed in the Jhelum Reentrant is not found across the Potwar Plateau (see Leathers, 1987); in fact, local northward progradation has been documented (Behrensmeyer and Tauxe, 1982). In addition, the Lower and Middle Siwaliks (Chinji through Dhok Pathan), thin gradually across the Potwar Plateau, and then appear to thin abruptly farther towards the south across the Salt Range. The Toot #2 well in the western Potwar Plateau penetrated 2952 m, while 2510 m of Lower and Middle Siwaliks were drilled by the Kot Sarang and Karsal wells in the central Potwar Plateau. Only 1425 m of molasse was penetrated by the Lilla well (see fig. 3 for well locations). The total thickness of the Chinji through Dhok Pathan molasse is estimated to be 3100 m in the Soan syncline and 2700 m in the Baun syncline (fig. 6). South of the central Salt Range the entire molasse section is less than 2 km; however, most and possibly all of this molasse is interpreted to be young (i.e. Upper and Post Siwaliks). It is proposed that depocenter position was relatively stationary during the deposition of Lower and Middle Siwaliks from about 15-5 mya (Johnson et al., 1985; Johnson et al., 1982). Basin subsidence and therefore the locus of sedimentation may have been fixed by downdropping along the basementnormal fault

(fig. 6). Since this time, the depocenter jumped south of the Salt Range. The rise of the Salt Range may have caused a reorganization of drainage patterns (e.g. Burbank and Raynolds, in press), resulting in the shift in depocenter positions. ForedeeR Migration Rates: Foredeep position is controlled by the location of the lithospheric load; similarly, foredeep migration is influenced by movement of the load. Because the Himalaya functions as the main lithospheric load, Molnar and Deng's (1984) slip rate of 18 mm/yr for underthrusting the Himalaya in India is an important indicator of foredeep migration; it approximates the northward movementof Indian crust relative to the Himalaya. Direct evidence of the foredeep migration rate was obtained by Lyon-Caen and Molnar (1985), who calibrated the southward younging of basal molasse sediments overlying the Indian shield at 10-15 mm/yr in India. The above rates are, however, significantly less than the facies and depocenter migration rates determined by Raynolds and Johnson 24

Fig. 6: Schematic diagrams illustrating the general southward shift of molasse deposition. (A) Near the end of distal Rawalpindi molasse deposition across the Potwar Plateau (approx. 15 mya). Assuming steady state basin shape, there is a problem with restoring old molasse in the foredeep using derived migration rates (discussed in text). Molasse above jagged line eroded as area passed across peripheral bulge (implying sediment bypass as bulge emerged). Alternatively, the peripheral bulge did not exist or the foredeep basin was much wider. If so, the basement would have continued its north dip and Rawalpindi molasse lapped against the edge of the basin. (B) At the end of Middle Siwalik molasse (Dhok Pathan) deposition (approx. 5 mya). Lower and Middle Siwalik molasse deposition was concentrated in the Potwar Plateau attaining its maximum preserved thickness in the Soan syncline. The depocenter may have been confined in this region by basin downdropping along the basement normal fault. Lower and Middle Siwaliks thin abruptly south of the Baun syncline across the basement offset. The bulk of the sediments was derived from the deformation front located somewhere to the north; however, evidence suggests sporadic erosion and transport from the crest of the peripheral bulge. (C) Present situation. Upper and post Siwalik molasse deposition was concentrated south of the Salt Range and to the east in the Jhelum Reentrant. The bulk of the sediments south of the Salt Range are young, because this region has only recently subsided into the foredeep. The present depocenter for sediments derived from the Salt Range and elsewhere is located in the Jhelum River Plain. HINTERLAND SOURCE UNPRESERVED PROXIMAL POT WAR PLATEAU MOLASSE PRESENT TOP OF RAWALPINDI MOLASSE DISTAL MOLASSE DEPOSITION

EMERGI G PERIPHERAL: BULGE .

BAUN SYN.

. A UPPER-POST SIWALIKS CAMBRIAN TO EOCENE ° o :AND QAL LOWER- MIDDLE TERTIARY EOCAMBRIAN SALT RANGE SIWALIKS MOLLASSE FORMATION

RAWALPINDI GROUP PRECAMBRIAN BASEMENT Figure 6 26

Fig. 7: Restored positions of the Sargodha High, Salt Range Frontal Thrust, basement fault, and Baun locality at 4.5 mya, holding Indian crust fixed. The Salt Range thrust sheet has been pushed back 25 km, N15°W. The Sargodha High is assumed to be a peripheral bulge through which Indian crust has migrated northward, controlled by plate convergence directions (see Minster and Jordan [1978] for pole of rotation). The restoration direction of the bulge is not controlled by plate convergence directions, but rather by the shift of the lithospheric load (Himalaya) versus the Indian crust. In other words, the relative slip direction between the Indian crust and the western and central Himalaya in northern Pakistan, India, Nepal and Tibet controls the direction of bulge migration. The slip direction and therefore bulge migration is assumed to have been perpendicular to the northwest-southeast trend of the Himalaya and Sargodha High. The minimum rate of movement between the bulge and Indian crust is constrained by the derived frontal shortening rate, while the maximum rate was taken as Molnar and Deng's (1984) estimated slip rate beneath the Hima- laya (18 mm/yr). The illustration depicts collision of the Sar- godha High and the western Salt Range, resulting in thrust ramping (Leathers, 1987). -Note that the basement fault restores to the northern flank of the peripheral bulge. Dark areas along the Sargodha High represent exposures of Precambrian rocks forming the Kirana Hills. RESTORED POSITION OF SALT RANGE FRONTAL THRUST RESTORED POSITION OF BAUN SITE 25 SALT RANGE BASEMENT BAUN SITE k FRONTAL THRUST OFFSET POSSIBLE RESTORED POSITIONS OF PERIPHERAL BULGE AXIS

18mm/yr 28

(1985), who suggest that there is a rough equilibrium between plate convergence rates and facies/depocenter migration. Instead, the high progradation rates were influenced by increased source area uplift rates and thrust sheet advance (Yeats, in prep). The encountering of a salt-lubricated decollement causing a rapid migration of the deformation front may be responsible for the higher rates obtained by Raynolds and Johnson (1985). There is also a problem with restoring Lower Siwalik and Murree molasse in the narrow foredeep using derived rates of foredeep and peripheral bulge migration (e.g. fig. 7). Even at the lowest rate of 10 mm/yr, assuming a steady state foredeep basin shape, these sediments restore south of the peripheral bulge (Sargodha high; fig. 2). Obvi- ously there is a problem with the age of Himalayan molasse in the present, narrow Pakistani foredeep. There are several explanations for this paradox: (1) the foredeep basin was broader in the past because either the Sargodha high is a remnant structure and moves with the Indian plate, or the foredeep is narrowing due to bulge migration towards the thrust front; (2) foreland basin deposition was not re- stricted by a peripheral bulge because sediments bypassed the peripheral bulge or the peripheral bulge did not exist (fig. 6); and (3) the relative rate of movement between Indian crust and the peripheral bulge was slower than presently indicated, as would have occurredif a larger amount of convergence between the Indian plate and Eurasia were absorbed north of the collision zone (see Molnar et al., 1987) in the past. 29

STRUCTURE OF THE CENTRAL SALT RANGE AND POTWAR PLATEAU

General Description

The Salt Range and Potwar Plateau are partof a 200 km wide salient protruding southward between theMianwali and Jhelum Reentrants

(fig. 2). Approaching the area from the south, the Salt Range escarp- ment appears to rise abruptly from thePunjab plains. To the north, the Potwar Plateau is bordered by theKala Chitta and Margala Ranges, which are collectively referred to as the Hill Ranges. In the Hill Ranges, intensely deformed, Precambrianslates plus Mesozoic and Tert- iary strata are exposed at the surface along aseries of south directed thrusts. This fault zone is believed to be a westwardextension of the MBT of India (figs.1 and 2; Yeats and Lawrence, 1984). The Salt Range is an east-northeasttrending highland, bent northward at both ends. The thrust front at the Salt Range isstrongly emergent (terminology of Morley,1986). To the west, it is offset from the weakly emergent thrust front at theSurghar Range by the right-lateral Kalabagh tear fault (fig. 2; McDougall,1985). To the east the frontal thrust becomes blind and the Salt Rangedies out into the Chambal Ridge and Rohtas anticline (Yeats etal., 1984). The Pabbi Hills anticline (fig. 5) appears to be the eastwardextension of the deformation front across the Jhelum River. The Potwar Plateau is an elevated areaof little relief, sometimes referred to as the Potwar trough (e.g.Voskresenskiy, 1979) or Potwar depression (e.g. Khan et al., 1986), due toits thick sedimentary fill. The Potwar Plateau can be divided into theNPDZ, which merges along strike with the Kohat Plateau to the west, and asouthern area of little deformation, called the Soan trough becauseof its alignment

with the Soan River. The SR/PP is dominated by the Salt Rangethrust sheet. It extends from a branch line beneath the northernlimb of the Soan syncline (fig. 8) to the Salt Range front, terminated tothe west by the Kala- bagh fault and to the east by theKarangal/Diljabba fault (KDF) and a diffuse zone ofdeformation in the eastern Potwar Plateau. Overall, the shape of the Salt Range thrustsheet, presence of oblique folds 30

Fig. 8: Block diagram illustrating change in deformation style across oblique ramp. (A) Duplexing east of the oblique ramp and imbrication west of it are consistent with surface geology. Surface geology adapted from Wadia (1928) and Cotter (1933). Central portion of map interpreted because geologic contacts and faults, on bordering published maps, do not match. (B) Cross-section (G-G' in A) of oblique ramp marking the westward termination of NPDZ frontal duplexing. Note that the roof thrust dips eastwards. A transect from the syncline (see Pariwali well in [A]) to G' would show more typical oblique ramping or a culmination wall (e.g. Butler, 1982). Portion of a branch line map (e.g. Hossack, 1983) shown at surface. Branch lines denoted by broken lines. Right side (with semicircles) = trailing branch line of the Salt Range thrust sheet, left side (with semicircles) = lateral/trailing branch line of NPDZ frontal horse, and left side (with teeth) = leading branch line of NPDZ frontal horse. Branch lines picked at base of platform sequence. Note trailing edge of Salt Range thrust sheet in (A) and (B). Rawalpindi Group is divided into the Kamlial (K1) and Murree (Mu) Formations. Other abbreviations are the same as in figure 4. 31

WSW ENE

.4;11:11:19 :7;:Vf,(:ViNs,=12=^0xFilAt;"4st, 44Ass4^:tr ,.. .01 lb l"! ASMANNNO4r

V.E..1:1 10 km Figure8(cont.) 32 near its lateral terminations and concentrationof layer parallel shortening (via salt growth) near the range front resemble Fischer and Coward's (1983) model of a surge zone with compressional flow.

Northern Potwar Deformed Zone

Description: South of the Hill Ranges, folds and thrusts expose Eocene and younger strata in the NPDZ. Surface structure consists of a folded zone with anticlines, isoclinally overturned to thesouth, giving way to thrusting farther south. In the southwest, map patterns of Murree and Siwalik strata depict large anticlines and synclines cut by thrust faults (see Cotter, 1933). To the north and east, only Murrees are exposed, making folds and thrusts difficult to map; however, somelarge folds are visible on Landsat photos. Along the southern margin of the NPDZ (i.e. steeply upturned north limb of the Soansyncline), published maps show four major, south-verging thrustfaults that merge eastward into one (Gill, 1952; Voskresenskiy, 1978). The existence of other faults is suggested in the detailed mapping of the Dhulian, Khaur,and Meyal-Kharpa anticlines (fig. 5) by Martin (1962). The complex seismic expression of overthrust strata in the NPDZis indicative of the intense folding and thrusting (figs. 9 and 10). Detailed interpretations are not possible; instead, observations are made from which structural style can be inferred. Interpretations of lateral changes in structural style are hindered by the lack of strike lines and the wide spacing (5-10 km) of available seismic transects (fig. 3). On most lines the easily discernible basement reflections of the Soan syncline can be traced northward beneath the NPDZ(fig. 11). Farther north, basement reflections become more difficult toidentify

(fig. 10). The discontinuous, subhorizontal reflections of the basement merge with platform sequence reflections, as the transparent zone corresponding to Eocambrian strata thins (see Lillie et al., 1987; their fig. 4). The platform sequence is deformed into concave downward arcs and "s" shaped features,interpreted to represent thrusted anticlines and horses (packages of rock bounded by thrust faults;figs. 9 and 10; Boyer and Elliot, 1983). The disharmonic nature of these 33

Fig. 9: Uninterpreted (A) and interpreted (B), unmigrated seismic line. Note disharmonic folding where syncline (beneath Mianwali well) overlies anticline. The lower (Rawalpindi) molasse appears to be tectonically thickened beneath the syncline (shown schema- tically by arrows beneath syncline). No attempt was made to interpret the connections between mapped surface thrusts and subsurface thrusting. Mianwali well penetrated 2116 m of Siwal- iks, 2733 m of Rawalpindi and 166 m of platform strata. Arrows at the top of reflection profiles mark the positions of mapped surface thrusts. See figure 3 for location. TM=Tertiary molasse; SM=Siwalik molasse, other abbreviations are the same as in figure 4. rn - 35

Fig. 10: Uninterpreted (A) and interpreted (B), unmigrated seismic transect showing reflections within the "trianglezone", near the southern terminus of the NPDZ. Reflections may represent stacked horses of platform strata although antiforms are tighter and smaller than they appear on this unmigrated section. The decollement is probably just above the 4 second line. Dhurnal wells penetrated an average of 2170 m of Siwaliks, 1560 m of Murrees, and 1120 m of platform strata before penetrating the Salt Range Formation. R=roof thrust (upper-level detachment) other abbreviations and symbols are the same as in figure 9. See figure 3 for location. ch SECONDS n.) . mm:0+5+-I vIz:t,N (A -I'tic IfitlAlike...-' 4-5,v---,":3%.::44q-...:";'': .,,J - - - 0 z is? rCD orm c m Im .s 3V! 11j tn 37

Fig. 11: Uninterpreted (A) and interpreted (B), unmigrated seismic line across the northern limb of the Soan syncline. Note thickened lower molasse section interpreted to be the result of back thrusting along an upperlevel bedding plane detachment or roof thrust ("R"). Reflections become chaotic on the left (north) end of the section, towards the center of the "triangle zone". For this reason, the northward extent (i.e. footwall cutoff) of platform strata is unknown. Platform strata may also be present between the upper two thrusts. Abbreviations are the same as in figure 9. See figure 3 for location. 38

0 5km

: 4A.N.

Figure 11 39 reflections, compared to the subhorizontal ones below, support the interpretation of an intervening decollement. Overall, the seismic signature becomes less discernable to the north, in accordance with the increased intensity of deformation in that direction (left side fig.

10). An oblique ramp marks a lateral change in deformation style

(fig. 12). The change in structural style is shown at the surface by coalescing thrust faults and in the subsurface by different seismic signatures (compare figs. 9 and 10). Farther south, the change in deformation style is marked at the surface by the curved axis of the Khaur anticline (fig. 5), where the oblique ramp represents the western edge of horses involved in duplexing beneath the northern limbof the Soan syncline (fig. 8). West of the oblique ramp, several lines show thestacking of reflectors into an apparent imbricate fan. Surface maps indicate that thrusts cut progressively younger rocks to the south(see Cotter, 1933), suggesting an imbricate fan involving the entire allochthonous section. On one of the few migrated lines, a series of five to seven "s" shaped horses form a pattern resembling a hinterland-dippingduplex (see Boyer and Elliot, 1983). In an attempt to interpret the interconnections of faults, two observations are made. First, faults interpreted at depth do not always appear to connect with thrusts mapped at the surface. Second, there are more thrusts at depth than mapped at the surface (figs. 9 and 10). The Mianwali well (figs. 3 and 9) penetrated almost twice the normal thickness of the Rawalpindi Group, suggesting that a thrust cuts at a low angle throughthis formation. A bedding-plane detachment may therefore be present within the Rawalpindi Group, possibly near its base. Furthermore, evidence of disharmonic folding between the reflective platform sequence and near-surface reflectionsis seen on several seismic lines

(fig. 9). East of the oblique ramp, seismic observations arelimited, due to a general decrease in dataquality. Two important patterns of reflections are noted near the southern margin of the deformed zone. One is the triangular pattern formed by the south-dipping, northlimb of the Soan syncline (fig. 11) in combination with the northdipping reflec- 7 2° 30' HILL RANGES 44/RREE FAULT (A4 10km 33°30' HINTERLAND

/DIPPING N DUPLEX 33°30'

IMBRICATE FAN SOUTHWARD WITH "TRIANGLE ZONE" DECREASING JP 0°. .c.= I.E. TERMINATION OF ONTHRUSTS 44: PASSIVE ROOF DUPLEX WITH ERODED OR ANDREACTIVATED :6?FORELAND SLOPING DUPLEX I eel DUPLEX (?) o OR ANTIFORMAL STACK GEOMETRY SHARP (FAULT- PROPAGATION FOLDS ?) 0:4 Of SC)" NORTH ER NLIMB R. Fig. 12: Interpreted deformation styles in the NPDZ and beneath the northern limb of the Soan syncline. See text for discussion and figure 5 for location. 41

tions from either tilted strata or thrust faults. Between the opposed reflections is a transparent zone (far left, fig. 11). This pattern has been referred to as a "triangle zone" (Lillie et al., 1987); however, the genetic connotations of this term are ambiguous and in need of clarification. The "triangle zone" is similar to that described by Price (1981) and Jones (1982) for the foothills of the Canadian Rockies, consisting of foreland-dipping reflections representing a tilted layered sequence above an upper-level detachment, and hinterland-dipping reflections from blind thrust slices beneath the upper- level detachment. A problem with this interpretation is that it requires the presence of a south-dipping, upper-level detachment surface. No south dipping thrust faults have been mapped in this area. However, because the upper-level detachment could follow bedding planes, it may be difficult to detect from surface geology (Jones,

1982). Seismic profiles (fig. 11) show that the lower part of the Murrees thickens dramatically beneath the northern limb of the Soan syncline, as compared to thicknesses penetrated by wells to both the north and south. This thickening may be indirect evidence for the upper-level detachment surface. Seismic sections also reveal overlapping of concave-downward, platform reflections near the southern margin of the NDPZ (fig. 10). Drilling, which resulted in the discovery of the Dhurnal field (fig. 3) confirms that these arcuate reflections represent horses of platform rocks deformed into anticlines (figs. 8 and 13). Unfortunately, seismic data gaps and transparency of the "triangle zone" prevent a complete interpretation of the structural pattern. This pattern is unlike the arrangement of horses in the northern part of the NDPZ, or the stacking of thrust sheets at its southern margin, west of the lateral ramp. Analysis: Figure 12 depicts the preferred interpretation of deformation style across the NPDZ. Thrust offsets of platform strata, observed on reflection profiles in the northern half of the NPDZ, probably do not die out into the folds described at the surface in the NPDZ. This would require a rapid loss of displacement towards the surface, implying large scale ductile shortening in the hanging-wall rocks (Vann et al., 1986). An upper-level detachment surface, above the hinter- 42

SOUTHERN POT WAR PLATEAU "M BT" N P D Z SOAN SYN.

20 km (SOAN TROUGH, SALT RANGE, LILLA) + 45km (NPOZ)65km TOTAL SHORTENING

200 km 150km 100km 50km 0 km

Fig. 13: Main (regional) balanced cross-sections and restorations. (A) Post- Eocene/pre-molasse restoration (approx. 35-25 mya) or salt-wedge model. Assumes flat top of Eocene. Note platform strata above jagged line are interpreted to be eroded during subsequent passage of crust across the peripheral bulge and associated basement faulting. (B) Post Middle Siwalik/pre Upper Siwalik (approx. 4-5 mya) restoration. This restoration depicts the foreland basin after basement faulting and prior to compressional deformation. Deformation in the NPDZ and salt flow are not pictured, but may have already been initiated. Basement dip determined by assuming a 0.50 surface slope towards the depocenter (Baun syncline?) and restoring the intervening strata. Formation contacts and faults are dashed above the upward limit of preserved strata in (A) and (B). (C) Balanced cross-section depicting present-day structure. Dashed lines above topographic surface represent eroded sediments. Eroded sediments and footwall strata are not restored in the NPDZ and south of the Salt Range, respectively. Dashed lines beneath topographic surface in the northernmost Potwar Plateau indicates the questionable nature of the interpretation. Anticline, just south of the NPDZ, interpreted in a gap in seismic coveage (fig. 3). See figure A5 for a detailed interpretation of the Salt Range portion of this cross-section. See figure 5 for location. Formation symbols are the same as in figure 6. 43 land-dipping duplex is therefore suggested by the contrast in deformation style seen at the surface and observed, via seismic, at depth. Cross-section balancing suggests that duplexing represents over 50% shortening (fig. 13). Molasse above the upper-level detachment must also be shortened by 50%. The deformation style of strata above the upper-level detachment is unknown; however, these strata are often eroded as they are elevated by underlying duplex thrusting(Banks and Warburton, 1986). Regardless of the upper-level deformation style, discrepancies between surface and subsurface structural style suggest that strata above and below the upper-level detachment behave as different lithotectonic units (e.g. Mitra, 1986). In the southern half of the NPDZ, west of the oblique ramp,local disharmonic folding (fig. 9) is explained by the stacking of hanging- wall synclines above anticlines of the underlying thrust sheetduring imbrication, consistent with an imbricate fan interpretation (fig. 8). Alternatively, this pattern is consistent with the preservation of synclines above passive roof thrusts, suggesting continued duplexing (e.g. Banks and Warburton, 1986). East of the ramp, the precise structure within the "triangle zone" is difficult tointerpret, because the geometry of the structure disperses seismic energy. If the upper- level detachment is a backthrust, and overlying strata haveremained relatively stationary, then this type of structure is the frontal end of a passive roof duplex (Banks and Warburton, 1986; Vann et al.,

1986). In figure 13, the geometry of the "triangle zone" isconstrain- ed because thrust slices, exposed at the surface, are folded by underlying thrusts (see also fig. 8). Another constraint is that the roof thrust slopes toward the foreland over a duplex height exceeding an individual ramp height. In addition, the exposure of Eocene strata suggests local triple stacking of horses(fig. 13). Therefore, the geometry within the "triangle zone" consists ofeither partially overlapping ramp anticlines indicating a foreland-sloping duplex (Mitra, 1986) or completely overlapping ramp anticlines (stacked horses), suggesting an antiformal stack (Boyer and Elliot, 1982). A hinterland-dipping duplex with equally spaced faults is transformed to these geometries by increased displacements on younger thrusts(see central portion of NPDZ, fig. 13). In other words, instead of 44 propagating a new thrust, movement continues on the previous thrust as would occur if the ease of sliding increases (Mitra and Boyer, 1986). Therefore, a southward facies change toward weaker rocks (more salt) in the decollement zone, may be responsible for the foreland-sloping duplex or antiformal stack geometry.

Transition from the NPDZ to the Soan Syncline

Description: Along the southern boundary of the deformed zone is the sharply upturned, northern limb of the Soan syncline. As seen on seismic secttions (fig. 11) and supported by drilling, the near-surface structure of this transition zone consists of a foreland dipping monocline of molasse strata (fig. 13). This type of structure is common in mountain fronts around the world (Vann et al., 1986). Underlying the south-dipping molasse, platform rocks either continue subhorizontally (as in fig. 11), or are parallel to the structure of the molasse. Where platform rocks are folded along with the molasse, the transition from thrusting to nondeformation is marked by folding. There is a gradual decrease in deformation intensity as anticlines show reduced structural relief and the frontal thrust has significantly less throw than thrusts to the north. In this case, the frontal monocline (north limb of the Soan syncline) is the south limb of an anticline cored by thickened salt (e.g. Khaur anticline, fig.

5). This pattern is only observed west of the lateral ramp. Diverging platform and molasse reflections are observed on all available lines east of the lateral ramp (fig. 11). Gaps in available seismic coverage and ambiguous seismic interpretations do not exclude the presence of this style of deformation west of the oblique ramp. The continuity of platform reflections beneath folded molasse may be explained by an intervening, upper-level detachment surface. The presence of an upper-level detachment isconsistent with the previous interpretation of a "triangle zone", as discussed above. Beneath platform strata, the transparent Eocambrian layer thins eastward from about 0.9 to 0.3 seconds (i.e. about 2.0 to 0.7 km) over a distance of 22 km; the molasse basin appears tothicken as the salt thins eastwards. Evaporite thinning corresponds to the area above and 45 east of the interpreted position of the oblique ramp(fig. 6). Due to lack of strike lines and poorly known velocities, it is not known if salt thinning is associated with basement faulting. Analysis: The interpreted oblique ramp does not appear to function as a tear fault along which the northern Potwar frontal thrust west of the ramp is displaced to the south; the branchline of the upper-level detachment to the east is along strike with the Khaur anticline to the west (fig.6). West of the oblique ramp, anticlines may be fault-propagation folds (Suppe, 1985) in which folding precedes faulting and thrust-shortening is replaced by folding upsection. East of the oblique ramp, duplex thrusting appears to die out beneath a major back-thrust or upper-level detachment. Strata above the upper-level detachment are passively elevated by the southward driven wedge of deformed rocks to form the forward dipping monocline.

Southern Potwar Plateau/Soan Trough

The Soan trough is a molasse basin containing a large thickness of Lower and Middle Siwaliks occupying the southern Potwar Plateau. The general shape of the Soan trough is that of a broad, asymetric synclinorium with a steep north limb and a gentle south limb bordering the Salt Range uplift (fig. 13). In the area of study corresponding to the central Potwar Plateau, the Soan trough is divided into the Soan and the Baun synclines by an intervening anticlinal trend (Joya Mair and Balkassar structures; fig. 14). To the west, where uninterrupted by major anticlines, the Soan syncline encompasses the entire south half of the Potwar Plateau and is synonymous with the Soan trough. In the central Potwar Plateau, post Eocene sedimentary fill within the Soan trough consists of a southward thinning wedge of Rawapindi molasse (approx. 1.5 km in the Soan syncline and 0.8 km in the Baun syncline), a large thickness of Lower and MiddleSiwaliks (over 3 km in the Soan syncline), plus smaller amounts of Upper Siwalik and post Siwalik sediments, including the Potwar Silt. Reflection data and surface mapping indicate that the Soan trough is relatively undeformed. Structure within the trough consist of broad folding with wavelengths of approximately 20-30 km (see Lillie et al., 46

N BALKASSAR JOYA MAIR S A. ANTICLINE ANTICLINE

...-S

.rv .. " I n l - wl

-w.

:, .. L n Iry ...M+" .....wC^..M1w' =1T.wr.,1w.I!lM' -- ,y,wr

r . - w. . `."yr.^w,u:w.w. II ._Wr . L... ~.11T,i . . ...w.nrd r « .b l I 5 km B. r r w ,.;1 424 ti'iw nfi"sll SM ..-- ,- ..1,- r! SM..a

W Ir w RM -E -E -E . SRF S R F yam s::.'':!a_ r I... IItJNNI wL' I i ....,. 111 L ,W,C nw'IU ..J PV .nu":....'h_..w.u.'w' -.

Fig. 14: Uninterpreted (A) and interpreted (B), migrated seismic line across the Balkassar /Joya Mair anticline complex in thesouth central Potwar Plateau. This transect represents the only discernable deformation in the south central Potwar Plateau. The small syncline bisecting the broad anticline was possibly formed by subsurface salt dissolution (e.g. Lohmann, 1972). Basement and overlying reflectors are disharmonic, implying that supracrustal rocks are detached from the basement. The decollement (dotted line) is thought to be a diffuse zone involving salt flow, rather than a discrete surface. Locally steep basement dip is perhaps related to a basement fault interpreted north of the syncline. Formation abbreviations are the same as in figure 9. See figure 3 for location. After Lillie and others (1987). 47

1987; their fig. 4). Folds in the southern Potwar Plateau consist of a gently-folded, competent layer(platform strata and molasse), overlying a plastic layer of Eocambrian evaporites (fig. 14). Salt flows from beneath synclines into the cores of adjacent anticlines, analogous to folding beneath the Appalachian Plateau(Wiltshko and Chapple,

1984). Salt can have both an active and passive role in the formation of folds. An active role implies that deformation results from the buoyancy of salt (Martinez, 1974). Salt beneath the broad Soan trough may have been squeezed southward towards theSalt Range due to increased isostatic pressure exerted by the molasse basin fill(Gee, 1983). Salt may also have behaved ductily due to compressive stresses suggesting a passive role (see Martinez, 1974). Broad folding in the south-central Potwar Plateau may represent the accommodation of slip at the rear of the thrust sheet(during basement buttressing) by internal deformation. This process often occurs in association with buried thrusts fronts, as would have been present prior to frontal ramping. The Soan trough constitutes the trailing portion of the Salt Range thrust sheet, which has been translated southwards with very little internal deformation. This lack of deformation is indicative of thrust surfaces that have propagated rapidly and steadily towards the surface. Lack of intraplate and trailing edge shortening represents an atypical style of development associatedwith strongly emergent thrust fronts (e.g. Pine Mountain thrust; Morley, 1986) and is related to the decoupling properties of salt in the decollement zone. The Soan trough appears to be analogous to the Molasse Basin in the foreland of the Alps, which is a 50 kmwide zone of undeformed clastic deposits thathave been translated 23 km over Triassic evaporites (Mugnier andVialon, 1986). Deformation style throughout much of the western portion of the southern Potwar Plateau is similar to the central area(Leathers, 1987), except for the zone of strike-slip faulting in the extreme west (J. McDougall, in prep). However, the deformation style is quite different in the eastern Potwar Plateau, where folds and thrusts are aligned en echelon, in a SSW-NNE direction (fig. 5). Thrusts in the cores of anticlines and exposed at thesurface verge northward as well as southward, while folding is characterizedby broad, flat synclines 48 separated by narrower, more intensely deformed anticlines where decollement splays have cut intermittently towards the surface (Johnson et al., 1986). The seismic expression is remarkably similar to that observed across the Parry Islands fold belt in arctic Canada, which is also underlain by evaporites (Fox, 1983; Davis and Engelder, 1985). Eastern Potwar Plateau folds appear to be fault-propagation folds which are indicative of a slow rate of strain accumulation under low stresses. Thickened salt in the cores of eastern Potwar Plateau anticlines is thought to represent layer parallel thickening, during sticking of the propagating decollement surface (e.g. Fischer and Coward, 1983). Therefore, the progressive build-up of strain along the eastern edge of the thrust sheet is consistent with a slower rate of thrust-tip propagation (Williams and Chapman, 1983) and may be explained by a stick-slip style of motion (e.g. Morley, 1986). Thrusts within the cores of eastern Potwar Plateau folds may form because fold shortening is limited due to unavailability of salt for flow. In other words, it is easier to thrust than keep moving increasingly viscous flow zone material (due to facies change?). If there is an eastward facies change to a less salt-rich composition, then the broader wave- length folding of the central Potwar is explained by a greater visco- sity contrast between the Eocambrian layer and overlying competent strata (see Wiltschko and Chapple, 1984) Flat topography, in combination with shallow basement dips beneath the southern Potwar Plateau (avg. 2.70), indicates that the overthrust wedge has a narrow cross-sectional taper. Narrow cross-sectional taper and lack of internal deformation in an overthrust that extends far out over the foreland are consistent with recent work by Davis andEngelder (1985) on the mechanics of thrusting above a salt layer. For this reason, the cause of contrasting deformation style and thrust-tip propagation rates in the eastern and central Potwar Plateau is thought to be a consequence of salt thinning eastwards resulting in an increase in basal friction (Davis and Engelder, 1985). Alternatively, the lessening of basement dip eastward (i.e. from ) 2.5° to (1°) requires internal deformation to maintain critical taper (see Davis et al., 1983) and may explain the change in deformation style (Jaume, 1986; Jaume and Lillie, in press). A third possibility is that basement 49

ramps are responsible for the increase in internal deformation in the eastern Potwar Plateau; an undulatory basement structure is interpreted from gravity anomalies (Leathers, 1987), and reflection profiles (E. Pennock, in prep.) across the eastern SR/PP.

Transition from the Southern Potwar Plateau to the Salt Range

Description: The northern flank of the Salt Range is an eroded monocline; the south limb of the Baun syncline/Soan trough synclinorium. Reflection profiles show that north-dipping sediments exposed at the surface overlie a fault with the basement offset as much as one kilometer, down-to-the north (fig. 15). Throw on the basement fault diminishes to both the east and west, gradually dying out into basement monoclines (figs. 5 and 16) The basement offset correlates with a local steepening of the Bouguer gravity gradient (Farah et al., 1977), and apparently acted as a buttress which controlled ramping and emplacement of the Salt Range thrust sheet (Lillie and Yousuf, 1986). Therefore, the monocline at the north flank of the Salt Range is the surface expression of the footwall ramp. Similar ramps that are controlled by basement offsets have been recognized beneath the forelands of other collisional mountain belts (Lillie and Yousuf, 1986). Seismic profiles and the detailed surface maps of Gee (1980) show that the form of the northern flank of the Salt Range varies along strike. Within the central area of study, the monocline can be divided into three zones corresponding to different shapes. In the west-central SR/PP, the molasse ramps gently to the surface at dips of 15-30° (fig. 17a and b). Farther to the west, surface dips are flatter, probably due to ramping across a shallowly-dipping, basement monocline (Leath- ers, 1987). The central regime begins east of the high-angle Kallar Kahar fault (fig. 5), where the monocline is offset by a pair of back-thrusts having a total throw of about 1.5 km (fig. 17c and d). Farther east, the back-thrusts gradually die out as the monocline steepens to 30° to 650. Analysis: The north flank of the Salt Range was the location of compressive flow and diapirism which are interpreted to have played roles in pre-ramping deformation (fig. 18), as well as the development Mb.

1,,,,...-.7 ....s... 1...1 ,:, t.Az i et...... ,-, ,,..7

I ,,4,1%.,..-Itg,1-,.!, ,..-.1.-.z1 'i.,;.r.-J-:'1, '4,f,:44; t444,14:414.0: RA 0,41 I- -ii---,:t:.-f!...!- ti; le.:4.1.44 Al4 h.-."--. r'lfr''' e-'1":"1%/lyli .4 4.'' :;;4';,e,71-ysloc. li:=1,4..le..i....4.4;11. oN.:04, z.4:11-,,...;::A-r,-,1,;:,AYt. '2.-i .:011:1:;e7,-sti ..$:-*;','4*.e ;;;-14'' c- is t,e..1: ,,,10'.??.., 1:,;14-.01? 4.:-..Q-7 44,-,-z---v- vps ill, ..T4.44. 0.166.' - 1-----vAL,-,-,-.1,-.9.: ...t.,1: es.,441.1.%-t-A11-. .i..:44...t .411. mi-..14--:..:,. .i0.2-44,3ge. :: :r i' c7;1 z:I.11qill''11 :411=='NF'4,'Pi*:;e1f.:14z44 ." ----: ....:,. :4XIV 4ki - --'-'4.41V. 'i, '..1q . I, 1'4%it . :I. -V"..0....!,1.e., - r,4,0; i.....ti.,4.4,44 4t1 i:ti,tt-.S; 11/I V t I " . 0 .41 4/11;411.;.. 7../ls 1/ _.4 =-1;:j"... 11-- - --.:: i',.n,IY :i1.1 cutt I V_ ,ilk -^ SO l'.014l'..-1.'i i.r ,-;* '-v. z 4;#,.--,.--- /' *.i -;4N Et il 144"....,'"';,,11.:47,.;t4111' C.T...:XNA- 1-..A'. 'till 4,. '..Ail'1.4'.wit Ile/AWL" --41iFS1V-?' '41:.i: kl'i**1;441:-'1.rifr'it:: IF '1+.:' dp ;c.v., -4r i:c-'. -1=-:u:41.0 ..4'ir' tem );14^,71.; iv,;Nimie . . . .1.ZWto, Airs 'f'il '...... 1'4''74 i 41.444'A, 1,-:.-444:;' t'ir' ii):0. 11.11'1.400; 1)13041glili g-41:11 'le e"j_.-. `4*-1:.,,.. ' ir.1);77_1' z-4-41'41.44i 't; --illt".11',1,4 I 0, ,_i.P-1.,,,,;:: rorpowt .,..:,.... :,...-,:',;,;. t._1.....?, 7.!:.1.+1 %.,;,;:rzITI',e.:Ze-ii:7;,,;.:11.;; -.. ...11.'..:,- --'1'--:1' ' , L'''',L --f.----12". 1;.....t1T-:-...,..;Iffiti..,...,',.'....**1 4 IA -.4- -,. -...... - .,k, v.,..1:31,. !,:`..:!4I' -4....-..... ;pi:14344 ,..... -,,jr. .4.., ...... ,...,f;, -...., . to:, 1.41;:i .1a, 51 /ri)11I .1414111. ,1111,1i4P V14'i,s?,;.';`,"zs,',g4sIlifil:kil.l'ititi'044-: 4411,1?11: 4i.i'S:40.0: 411114' !IA 1'1141 -.--',,;.3.; 4:4; A ,;1g:-V,1,1;;;Ats.14.:, 044.00 kki4v,g4411, ?);.:1. Pr.., 9.1ic..-t li 0 111,. '.4- 3 ,,,,,,r.; ,,,--;44-..:- , 1`,..q. : .,.- v 4 . INglitotp. $.. Oareom (414 'ATI':2 -11rWi -.11( 1 '11°,ilq,Z? .A;''.+i:,:c Iiilt,A., ',;A,p,'$,. ;T:44 Mt It T. ixootv $41.:.....i.44. 41.41e, .16 ',,,cajj;.7i,t,(441;'N. f,'fr,',41.114 0.?Irort, '0.".".474lkii; ilitImpiq ,;',--,.. 1..,1;`{4.0:-.r fit4,47.? 4.4,f:1- .4144JA 141,U,QVI. 4A::,,, ,41,..,21.4741.1-,-.11.4:yo.,..VA:;11,11;, 4s$9 .14 11.44 \ 440401,14.14.44,1i:$ii ',.--; #.4 KR ;g1:004.111 v+a-'. ' =IS A:14:11% tw,a it,ii a2 ...... ttta. ,..t.,.41v14:1g A41'11441/. 4...,-....,...44 ,,114444.0.4141 1;41 411,14)VAT /11414$0:011141.:11;1}; is:7,1/N 4.111,...... 11,,r_,104.1?,,y,.iy ,.1.11; ?,,k,,,14,161/40 a ''/<`3 - h 011P+Tik:11".# -''.)2. ':,,,z,vie It, li_,Z, ...21' I :.4..ks,, 11.3.33.1; .1, 11 ,t.,t- .. 1 1.._10,..,-.... )0

I I I I ' . II H it:'I, ; I II I 11 t., A J :Ifilit'; ') 1 4;1? 111/i '..:4 1 j ijii 1 jd '4Ititli !:7%;-i:' '-;404 IAN: liA 1q11,1,di 44,-L9;."1::.4,1;:;:::")41c`:5:;" 34 ..):(t-:._:.;411,:....; 1.?0, 44t1,,,..,10,1:4, ',igiR4 .A0441; ,at 1i. )4-4.01.fiti 1": 4:,..' 1;-. . 4 .4 1Y, "1":; ,-,,. 3. .1.4 if. 5. ,7_,,,,,,=-1 $9._ "73/4 . 1 . .gr.4.4, ,i'.I 7 t':.:.it$ ii, r;4:00"4151fuir'tr._-:---..-.-- 1.: : ...,-,--,-,.;e.,-,,;74.-",-...---,1_11-.C.' -49 FA'. 41..;4P.4:4.4. . 4414.iet '=4 1::.111;i7:1- ,E-..:''''-.7': 04 ;0' 1.4J4110F°44.44,5 46Y- J1tF1 114111414fiii:fitAii 4 ,,l, 5117, .-j-i -1;1:14 itjj:14X 4, )4.4!1/4 . 4 .4e. '23:144A:t 1 01 el. s 441V..14kai;;44'vlr'.,,...4.4_,* Al:;;1-404 t4e.44,10"; , *.17,7444 ;CAST.. falliIIP grilt.A 01 MO 144:41.1.#444 totomi<-1.- - 4Ittr,: r -,067::leit

, ..:411-44..'roe10,-,.71-P.,...... II/,-:iiiii."..""f":;41.11:11:1kr.:;:i.715441:12;vtr-th:

....A* ...,4j 7.-grenfffel. ' , ...... - -,444, Ja" . ...I:of :11-1,,, 474: :1, .7.;04.4 42::::;2 411..": a7t..t ,;.,i14;f:II:j.1:;:1111itliii;;;41,11i14-r;41:10 1;;;Ili06:Ili 11;111,4;:i:"!1),:,:11.1.,; .. .44',.' V Y44 . 4444411 4`i44f ' I'd 9.' .4 -4, 4 g0041,44 41, 1 4 . , f? ( 4,14 I. In-114' 4.,.11:CIrlitS-1.77/N41'4441)1 i-4 114,4 4 ::411111111P1111, 111$1114; SIn714i 0414; :14142;e110Y4itirerAig711.:itt itl,pl.,..' 4 ),I111.t. 1.A;;I:Sli,,,,,.-11,,,;,, 41.t4ga? 4 ,4 teigl 14 11)11,1012`'WI-Iva Al 1.1.r.oWit.-.';,-2, III it .1:1012 .=1% fl. 1.:11Trt't-'iP. 0Pv, u r .. ,1 4.''''.! 14.- -4144(ja'41,tie. 4,,. v -1 --Ist.vp..41. 3-s...... ', '''''1 .4.-,,r.d, 1'. 14-44Ii:1?ir.:4,kviimon'.4i}:lk"411'411:37411.1'.

I .

III 5

I II " Ill

- - 5 . II : . I II' . I - . " a. : . . . " . III .. .' -I "

. I I .-I 51

Fig. 16: Geologic map of central Salt Range (after Cotter,1933; Gee, 1980), showing lines of cross-section balancing (figs. 13 and 17). Note the westward increase in separation betweenfootwall cutoff (i.e. basement offset, portrayed by dotted line) and preserved portion of hanging-wall ramp (rangefront); see text for explanation. Amounts of shortening determined from section balancing: A-A'=23.2 km; B-B'=21.2 km; C-C1=21.8 km; D-D'=21.9 km; E-E'=23.2 km; and F-F1=19.7 km. Shortening values are less in the reentrant between B and D, where erosion of the rangefront may have been greater (see fig. appendix C). The lowest amount of shortening (i.e. F-F') is interpreted to be in conjunctionwith the erosion of imbricate thrust slices from the rangefront (see dashed thrust faults). The southernmost dashed thrust depicts a possible restored position of the hanging-wall ramp. The actual position of the buried range front thrust is thought to lie between surface exposures of allochthonous strata and the restored position of the hangingwall ramp. Formation abbreviations: Tp=Paleocene; and P=Permian; other abbreviations are the same as in figure 4. See figure 5 for location. 7 2°30' 72°45' 73°00' 32°55' 32°55'

QAL I

* 0% 0 040 U S 0 0 co

DP, Na, Ch

RM 3 2°45' =Tp -E -C-- P

SRF

32 °33' Figure 16 0 10km 20km 30km

1 NNW 1 1 SSE _- TOPO. Ch 2km SURFACE -1- _-? RM JA-- LE IWO SL MS SRF T -QM --2km SRF -C E SRF . MIM.0 p£ --4km

Fig. 17: Frontal balanced cross-sections. Restored sediments dashed above topographic surface. Middle Siwaliks are not restored, due in part to uncer- tainty in thickness change towards restored position of basement fault. Note eastward (i.e. from A to E) steepening of hangingwall ramp across basement offset and increase in internal deformation near the toe of the thrust (especially south of KDF on E-E'). Dashed thrust splay on (E) denotes an alternative interpretation to high angle, strike-slip fault mapped by Yeats and others (1987). Surface formation contacts, strikes, dips (not illustrated) and faults are all from 1:50,000 maps of Gee (1980). See figure 5 for fault and figure 4 for formation abbreviations and figure 16 for transect locations. 0 10km 20km 30km NNW SSE 2km TOPO /SURFACE SL

0 10 km 20km 30km NNW KKF SSE

Ch 2km ft- --- _77 firsi TOPO. SURFACE SL

2km

4km

Figure 17 (cont.) 30km D.

2km TOPO - SURFACE SL

2km

4km

0 10km 20km 30km

2km

-2km

-4km

Figure 17 (cont.) 56

Fig. 18: Ramping diagram demonstrating a possible mechanism of incorporation of footwall strata as the ramp shallows.Horses/slices of molasse (exaggerated in size to demonstrate themechanism) are plucked from the footwall and eroded near the toe of the thrust sheet. Note (jagged) erosional surface through time. See figure A5 for the initial and final configurations of strata acrossthe ramp. (A) Preramping deformation. Basement buttressing causes buckling of competent strata above an elongate salt pillow or dome formed in response to southward directed salt flow. Reactivation of the normal fault plane, in strata overlying the basement, resulting in reverse fault separation allows folding without footwall deformation prior to ramping. (B) Incorporation and erosion of the first molasse slice (south dipping stipples) and position of next molasse slice (north dipping stipples).Dead zone (DZ) grows as the ramp shallows. (C) Continued erosion and footwall incorporation at the toe of the thrust sheet. Smaller slices of Siwaliks and fanglomerates have been identified at the range front (Yeats et al., 1984; Gee, 1980). S

-",;!,;(2- . _1,, (,:ja.:-10. 'tit :-\ I I / Th /-- / 01 _ 1.. f.itOtt"-; ri 1\;: I I ,, ,/-s. . 11!:Z;-i I1 Z.11'.../%-, "" ""s - 1-/. `T';,1>'-',"-,1-1,-,s'es,r,s1-',` - \'- "

Figure 18 Ln 58 of the present monocline shapes. Pre-ramping deformation helps to explain the emplacement of thick salt beneath the Salt Range and avoid an unreasonably sharp initial interlimb ramp angle during subsequent

ramping. The basement fault, acting as a buttress to both the southward salt flow (Gee, 1983) and the southward propagating decollement, may have localized the growth of an elongate salt pillow or dome as the overlying competent strata buckled against it (fig. 18). Regional horizontal compressive forces cause evaporites to become mobile, resulting in diapirism (fig. 18a; Martinez, 1974). As salt growth continued, Cambrian and younger strata on the north side of the fault would be lifted into contact with less competent molasse. Compressive stresses could then push the thrust sheet across the buttress (fig.

18b). Once ramping was initiated, differential overburden would cause salt to flow from the region of higher to lower load stresses (Jenyon, 1985); i.e. southward from the downthrown fault block to the base of the overthrust. As the central portion of the thrust sheet between high-angle faults pushed forward, the thick accumulation of salt would be accommodated beneath the anticlinal structure south of the ramp and by the steepening of dips north of the basement offset (fig. 19).Reverse faults probably didn't form until normal stresses built up as the thrust lengthened (fig. 19b; Wiltschko, 1979). Reverse faulting is a mechanism by which thrust sheets migrate through sharp, lower hinges during ramping (Serra, 1977 mode 3; Moore, 1977), and apparently occurred in conjuction with ramp steepening during diapiric salt growth. Therefore, different footwall ramp shapes formed as a result of salt accommodation and segmentation by strike-slip faults.

Salt Range

General Description: The Salt Range is a hanging-wall, ramp anticline emplaced across a low-angle step in the sub-thrust surface. The general shape of the Salt Range is similar to Suppe's (1983) model of a fault-bend fold, but without the frontal limb. Also, Suppe's precise geometric relationships are not strictly applicable because of the plastic behavior of salt at the base of the thrust sheet (fig. 19). 59

Fig. 19: Schematic diagram (to scale) illustratingthe evolution of the central Salt Range. Roman numerals represent progressivestages of ramping. I.=early, II.=intermediate, and III.=latestage of ramping. Letters correspond to: (A) east-central; (B) central; and (C)west-central slices of Salt Range. Double barbed arrows indicate the interpreteddirection of salt flow. Dashed line is the surface trace of theoriginal normal fault plane. The footwall ramp shallowsas ramping progresses. A reverse fault develops to accommodate movementover the highest portion of the basement butress in (B). In (C) footwall clipping allowsshallow-ramping without reverse faulting or a steep monoclinal flexure (as in A),and results in a far-traveled horse. There is an increased amount of internaldeformation from west (C) to east (A). Note that a large thickness of molassehas to be unroofed before platform strata issubjected to erosion. 60

Alternatively, the Salt Range resembles a thrust nappe(e.g. Ramsay et al., 1983; Charlesworth and Kilby, 1981); again, thefrontal limb is missing or thrust over during ramping. More accurately, the Salt Range's anticlinal shape is the result of buckle ordiapiric folding prior to ramping and bending moment folding duringramping. To the west, Salt Range structure becomes morecomplex, especially near the western terminus of the range,where deformation is by a combination of thrusting and right lateral, strike-slipfaulting (J. McDougall, in prep). The overall ramp anticline shape of the Salt Range does appear to continue, bending to the northwhere its genesis may be due to oblique or lateralramping. The presence of more thrust faults within the western Salt Range (see Gee, 1980) suggestsincreased imbrication. Beneath the central Salt Range, reflectionprofiles reveal an autochthon of highly-reflective, probably Cambrian, andperhaps younger, platform strata(fig. 15). Autochthonous molasse may also exist beneath the Salt Range; this possibility isdifficult to inves- tigate because the seismic expression above thehighly-reflective sequence is chaotic to transparent. The amount of molasse at depth depends on the orientation of the thrust surface,which is difficult to discern because indisputable thrust surface reflections areabsent beneath the Salt Range. Locally, the Dhariala well penetrated over two kilometers of Salt Range Formation (fig. 3; Gee, 1983; Khan etal.,

1986). This thickness of salt is probably anomalous because thewell was drilled in the zone of steepestramping and upon a truncated anticline. Thick salt suggests that the detachment remains at depth and did not cut abruptly to the surface just southof the basement offset. The decollement is not thought to have completelyflattened upon autochthonous platform strata,because a secondary mechanism would be required to cause the thrust to ramp to the surface. Accordingly, a moderately-dipping thrust is interpreted, resulting in anorthward- thickening wedge of salt and southward-thickening wedgeof molasse (figs. 13 and 17). Although overthrusting with littleinternal deformation is common in the west-central Salt Range, internal deformationincreases to both the east and west. To the east, folding is prevalent between the 61

Kallar Kahar strike-slip fault and KDF (fig. 5). South of the KDF, there is intense deformation at the toe of the Salt Rangethrust (fig. 17e), where nonparallel, locally overturned folds and reverse faults are present (fig. 17e). Although the frontal thrust is buried by fan gravels and alluvium throughout most of the central Salt Range (fig. 16), range front thrusting is exposed in the eastern and westernmost Salt Range (Yeats et al., 1984). Folds and reverse faults within fan gravels, and thrust faults between fan gravels, Siwaliks and older strata are observed. The eroded fault scarp at the range front may actually be the preserved hanging-wall ramp. Its scalloped outline is due to both curved reverse faults and the concentration of erosion andalluvial fan deposition at reentrants (Yeats et al., 1984). South of Salt Range thrusting, sediments slope gently northwards from the Sargodha High. Compressional deformation is present 15 km south of the central Salt Range at the Lilla anticline, which appears to be cored by a blind thrust within Eocambrian evaporites(fig. 13; Lillie et al., 1987). This suggests that the deformation front may be shifting southward from the Salt Range. Alternatively, the Lilla structure may represent weak crumpling of footwall strataformed during rapid emplacement of the Salt Range thrust sheet (e.g. Morley, 1986). Eastward Termination of the Salt Range Thrust Sheet: The character of the Salt Range changes southeast of the KDF (fig. 5). All available reflection profiles crossing this fault and the area to the southeast show thrusting and duplication of the platform sequence (see fig. 20). Both forward and hindward verging thrusts are mapped, dividing the eastern Salt Range into several fault blocks. Seismic lines constraining the position of autochthonous platform strata, indicates that shortening is no longer concentrated on the range-front thrust(fig.

20). Complications are indicated by the mapped changes in vergence of the KDF (fig. 5; Gee, 1980), which has been described as"flipping like a propeller blade" (Voskresenskiy,1979). Farther east, where the range bends to the north, thrustingis replaced by active folding. The KDF cannot be the eastward continuation of the Salt Range frontal thrust, since it would require an abruptreduction in shortening along strike (fig. 20). Rather, it may be a splay off of the 62

Fig. 20: Hanging-wall/footwall cutoff separation map of the east-central Salt Range and Karangal/Diljabba (KDF) zone. Patterns denote individual thrust fault surfaces between cutoffs. Patterned areas represent duplication of platform strata, and therefore show the minimum amount of horizontal shortening. Hanging-wall and footwall cutoffs and duplication constrained by seismic lines; cutoffs dashed where extrapolated beyond seismic control (fig. 3). An eastward decrease in overthrust distance (and therefore shortening) on the frontal thrust is demonstrated. However, duplication related to frontal thrusting is unknown in easternmost portions of map (see large question mark). There does not appear to be an eastward decrease in overall shortening. The KDF is the surface expression of thrusting that accounts for a large amount of shortening. The change in polarity of the KDF is explained by overlapping thrust sheets; eastern thrust sheets are interpreted to overlap western ones. Mapped positions of cutoffs may not coincide with polarity flips because mapped cutoffs mark the base of platform strata while younger strata are normally exposed at the surface (see fig. A2). Northernmost cutoffs from Pennock (in prep.). Minor thrust fault surface (less than 3 km overlap) in north-central portion of map not shown (see E. Pennock [in prep.] for detailed interpretation). See figure 5 for location and abbreviations. 73°15'

6.04r0 THRUST MAPPED AT SURFACE (GEE,I980) HANGWALL CUTOFF BASE PLATFORM STRATA ___,...../FOOTWALL CUTOFF BASE PLATFORM STRATA 1-1SALT RANGE FRONTAL THRUST \\SE FORWARD VERGING THRUST BACKTHRUST 1111111 NW FORWARD VERGING THRUST

3 2°45'

73° 73°15' Figure 20 64 frontal thrust dominated by strike-slip offsets(see Yeats et al., in prep.) near the range front (fig. 17e). Figure 20 demonstrates an eastward increase in the amount of shortening along the KDF zone. Overlapping thrust sheets explain polarity changes along theKDF, suggesting that it is the composite trace of several faults(fig. 20). A back thrust (fig. 5, where teeth are on the southeastside of "fault A") is seen to cut an earlier formed thrust to produce a"pop up" structure, or "triangle zone" (as defined byButler, 1982), indicative of a slow rate of strain accumulation. By this interpretation the transition from the central to eastern Salt Range may be analogous to the transfer zone between en echelon thrustsdescribed by Dahlstrom (1977), although complicated by additional structures. Because both thrusting, in combination withfolding, and strike-slip faulting are interpreted, the KDF appears to be a zone of transpression. Structures within the Central Salt Range: The Salt Rangeconsists of strong folding, faulting and uplift in a narrow zone. In the central Salt Range most structures can be classified into fivecategories: (1) minor folds with axes approximately parallel to the range front; (2) high angle faults; (3) minor thrusts within the allochthon; (4) diapiric salt structures; and (5) strike-slip fault zones. The east-west orientation of minor folds (fig.16) suggest that they were formed by the same compressive stresses thatemplaced the Salt Range thrust sheet. Most of this folding appears to be concentric, although nonparallel folding occurs locally, near the range front (fig. 17e). The intensity of folding increases southwards, towards the toe of the thrust, as well as toward the easternSalt Range

(fig. 17). High-angle faults within the Salt Range (fig. 16) are mapped as "normal, or probably normal" by Gee (1980), although many appear tobe reverse faults formed by compressive stresses(fig. 17). Bending- moment faulting is probably responsiblefor many of the high angle

faults. As a thrust sheet ramps up-section, its uppersurface travels through a compressive regime at the foot of the ramp,and then an extensional regime where the thrust sheet flattens; theopposite is true for its bottom surface (Wiltshko,1979). Thus, brittle platform rocks may experience normal faulting both at the footof the ramp where 65

they are below the neutral surface, and again at the crest of the Salt Range because platform rocks are above the neutral surface after overlying molasse is eroded (fig. 19). Normal faults may also be due to extension above salt diapirs. Other than large scale thrusting at the range front and along the KDF zone, low-angle reverse faults within the Salt Range allochthon are rare. These small scale thrusts (less than 5 km long), concentrated near the range front (fig. 16), may be horses clipped from the footwall, but most represent imbrication within hanging-wall strata. Sur- face maps show thrusts and higher-angle reverse faults associated with salt diapirism, both in gorges at the range front and along strike-slip faults. Salt bursts up into planes of weakness where geostatic pressure within the salt is greater than hydrostatic pressure in overlying sediments (Ala, 1974). Inspection of geologic maps along the range front (fig. 16) reveal transverse (i.e. north-south trending) folding, consisting of sharp-crested anticlines in valleys, and flat synclines making up topographic highs. This inverted topography is due to gravitative flow processes (salt diapirism) related to unloading as valleys are eroded by streams (Gardezi and Ashraf, 1974). Three fault zones, which lie within the central Salt Range and oblique to the frontal thrust have been mapped as active strike-slip faults by Yeats and others (in prep). Landsat photos show about 3 km of right-lateral offset along the Kallar Kahar Fault and 9 km of left- lateral offset along the KDF.However, the offsets of individual formation contacts are less on geologic maps (Gee, 1980), possibly a result of fault drag. The three faults appear to function as tear faults, segmenting the Salt Range; they separate different deformation styles and cut structures not found on the opposite side (fig. 19). Once the Salt Range thrust sheet was segmented, structures on either side could develop independently (e.g. Dubey, 1980). Faults interpreted to have strike-slip motions are mapped as thrusts and high-angle faults by Gee (1980). Within these fault zones, high-angle faults bound horst blocks which locally bring salt to the surface (Yeats and Lawrence, 1984). At places, strike-slip faulting is associated with folding and reverse faulting, which collectively form flower structures (fig. 17c). As previously noted, the KDF zone is dominated by 66 thrusting farther north (figs. 5 and 20); therefore, it may be partof a transpressive zone. If so, then the central Salt Range is pushing past its flanks and is a small scale example of"escape block tectonics" as suggested by R. S. Yeats (personal comm., 1985).

Ramping of the Salt Range Thrust Sheet

Mechanism: Ramping and emplacement of the central Salt Range was controlled by a down-to-the north basement offset acting as abuttress (Lillie and Yousuf, 1986). For a discussion of other factors responsible for the emplacement of the entire Salt Range, see Baker etal. (in review, see appendix A). In the central Salt Range, a pre-existing fracture through competent strata, if present, would haveinfluenced the ramping style. The presence or absence of this pre-existing fracture depends on the timing of basement faulting(discussed below). The proposed mode of ramping assumes post Eocene basementfaulting (fig. 13) and therefore a plane of weakness through platform strata. Since the work of Rich (1934), hanging-wall folding has been thought to be the result of strata moving over steps in thethrust surface. More recently, it has been suggested that propagating thrusts terminate laterally into and are preceeded by a"ductile bead" of deformation (Elliot, 1976; Coward and Kim,1981). For example, the Pabbi Hills anticline (fig. 5) may be an eastwardextension of the Salt Range frontal thrust. Fischer and Coward (1983) reported field evidence, in the Heilam thruat sheet of northwestern Scotland, indicating layer parallel shortening resulting in asymmetric foldingprior to thrusting. Therefore, an alternative mechanism of hanging-wall ramp anticline formation is possible in which thrusts cut throughpreviously folded strata, in this case formed by preramping salt growth,with smooth trajectories similar to thrust nappe emplacement(Cooper and Trayner, 1986). However, a lack of observable footwall deformationof platform rocks (fig. 15) and the presence of apre-existing ramp within thrusted strata suggest a ramp-flat geometry(Cooper and Trayner,

1986). Noting Wiltschko and Eastman's (1983) work, it islikely that a normal fault plane formed during basement faultingwould be reactivated in strata overlying the basement during compression anddiapirism 67 resulting in a reverse fault separation (fig.l8a). If so, the southern flank of the central Salt Range is not only an eroded thrustfault scarp (Yeats et al., 1984), but also representsthe preserved portion of the downthrown normal fault block. The shift from displacement along the reactivatednormal fault plane is an example of ramp shallowing. Serra (1977) reported field evidence for ramp shallowing with increasing displacement. Steeper ramps are abandoned in favor of shallower ones asweak rocks in the lower ramp region of the upper plate thicken (see Serra, 1977,modes 1 and 2) as a result of salt flow (figs. 18 and19). A shallow ramp explains the absence of the frontal, forward dipping limb. Because the Salt Range is young, it is unlikely that the absence ofthis limb is due entirely to erosion. A frontal limb forms as a thrust flattens after cutting up-section (Suppe, 1983). Therefore, a plausible explanation is that the thrust doesn't flatten, but instead slopesgently to the surface from the basement offset through relativelyincompetent molasse (figs. 13 and 17). Similar, shallow ramps are seen in the foothills of the Canadian Rocky Mountains where thrust nappes cut upsection through molasse (Charlesworch and Kilby, 1981). Hanging-wall and footwall short cuts would modify the shape of the ramp so that competent strata could move acrossthe footwall with less bending resistance (Knipe, 1985) producing a smoother and shallower thrust trajectory (e.g. fig. 18c). Footwall clipping is an example of a footwall shortcut (see fig.17a). Wiltschko and Eastman (1983) demonstrate that hanging-wall strata become immobilized or cut off against basement faults during thrusting, forming a "deadzone"

(fig. 18). Shallow ramping facilitated development of the great overthrust length found in the central Salt Range. The resulting thrust sheet profile is wedge shaped (fig. 17); wedges have a greater stablelength than rectangular shaped fault blocks (Murrel,1981). At low dips there is less work against gravity and bending resistance tocontinued thrust sheet movement (Wiltshko, 1979). Work against gravity is further reduced by the buoyancy of salt which helped tolift overlying strata across the basement offset. The largest overthrusts are formed where the basal rocks are the weakest, reducing drag(Murrel, 1981), and 68 where the toe is continually eroded, increasing the deviatoric stress (Rubey and Hubbert, 1959; Rayleigh and Griggs, 1963). Therefore, the presence of salt and unroofing of molasse probably enhancedthrusting in the Salt Range. Faulting of the basement in Eocambrian time would have created a stratigraphic situation resulting in a different ramping mechanism. Less southward salt flow would be required to emplace thick salt beneath the Salt Range (fig. 21a). The basement offset could still have localized salt growth due to convergence of flow over the basement fault, because the salt would have thinned southward across the fault (see Suppe, 1985). If pre-ramping deformation did occur, then the structural style should be similar to fault-propagation folds of the eastern SR/PP, implying footwall deformation. Because there would be no surface of weakness for reactivation through competent strata,the ramp would form by brittle fracture. In this case, a relatively low- dipping ramp should result so as to reduce bending moment strain and work (Wiltschko, 1979). Because a steep ramp is observed through platform strata and no footwall deformation is seen (fig.15), this ramping mechanism is not supported. Removal of Molasse From Beneath the Salt Range: The interpreted shallow footwall ramp through the molasse is due to two processes. First, the thrust sheet climbs up-section as it overrides its own debris(fig.

19). The Salt Range thrust sheet is known to override fanglomerates along the frontal thrust (Yeats et al., 1984). Although overriding of fanglomerates is an important factor, it is not suggested that the allochthon flattened onto a paleoland surface; restorations indicate that Siwaliks may have been present on the upthrown basement fault block prior to thrusting (see appendix B; fig. 13). In addition, molasse may have been removed from beneath or incorporated into the Salt Range thrust sheet (fig. 18). Horses of Siwaliks plucked from the footwall ramp during thrusting would be carried along by the overlying thrust. Mountain fronts advance as footwall rocks are incorporated (Elliot, 1976). Second order duplexing may be the method of molasse incorporation; footwalls of major ramp anticlines are one of the most common structural positions of duplexes (Mitra, 1986). Most of the evidence of duplexing would have been removed by erosion during 69

Fig. 21: Restorations illustrating stratigraphic thickness changes versus the age of basement faulting. Formation patterns are the same as in figure 6 except Lower to Middle Siwaliks are subdivided (upper Middle Siwaliks are cross-hatched). (A) Post Dhok Pathan restoration depicting Eocambrian basement normal faulting. Salt thickening (tabular salt model) by 1 km north of basement offset supplies approximate area balance of salt. Basement must remain approximately unflexed for forward topographic slope. Eocambrian basement faulting does not explain local southward thinning of platform strata. (B) Post Rawalpindi restoration showing early molasse and platform strata thinning across the basement offset. However, the Rawalpindi Group thins regionally to the south; the estimated pinch-out is near the basement fault. Early to middle Miocene basement faulting does explain the southward thinning of platform strata. However, platform strata thinning can not account for a kilometer of stratigraphic thickness change by itself. (C) Post Dhok Pathan restorations showing the thinning of (1) Lower to Middle (about 13-8 mya) and (2) upper Middle (about 8-5 mya; vertical-ruled) Siwalik molasse across the normal fault. These two restorations correspond to basement faulting related to Neogene flexure. Thinning of the upper Middle Siwalik molasse across the basement offset (model C2) is most consistent with restored stratigraphic thicknesses (fig. 13) based on estimated positions of the basement offset versus the peripheral bulge (e.g. fig. 7). However, it does not explain southward thinning of platform strata unless platform strata was eroded along the crest of the bulge prior to molasse progradation. An increased basement dip on this restoration would result in a thinner Lower to Middle Siwalik molasse section on the north side of the basement offset (not shown). Southward thickness changes across the basement normal fault could then be accomplished by a combination of platform strata thinning (as in model B) and Siwalik molasse thinning (models Cl and C2). Upper Siwalik and platform strata thinning is consistent with the deposition of Upper Siwalik conglomerate beds, containing recycled Permian Tobra Formation clasts, interpreted as derived from the up-thrown normal fault block (see fig. 22b). This latter version suggests basement faulting at 4.5 mya (discussed below). Iuum 71

ramping (fig. 18). Some horses of Siwaliks are preserved at the Salt Range front. Overturned lenses of Siwaliks, identified beneath allochthonous Salt Range Formation in the eastern Salt Range (Yeats et al., 1984), are evidence of the incorporation of footwall strata by imbrication or duplexing. Footwall rock failure would occur in conjunction with ramp shallowing (fig. 18), and is associated with fast displacement rates (Knipe, 1985). 72

TIMING CONSTRAINTS ON EVOLUTION OF THE FOLD-THRUST BELT

Efforts to calibrate the development of the Himalaya are hampered by the removal of sediments critical to dating structural movements. In the foreland, a large effort has been put forth to date stratigraphic horizons; tephrachronology and magnetostratigraphy have been success- fully used to age calibrate units. This has opened the door to various methods of constraining the timing of structural events (see Raynolds, 1980; Burbank and Raynolds, in press).

Migration of Folding and Thrusting into the Foreland

Although there is a general progression of deformation from north to south across the modern foreland thrust belt in Pakistan(Johnson et al., 1986), detailed studies reveal deviations from this classic hinterland-to-foreland pattern (Burbank and Raynolds, in press). The earliest documented deformation is pre-Paleocene displacement on faults within the Attock-Cherat Range resulting in a low-angle unconformity across the SR/PP (Yeats and Hussain, in press). A hiatus has been interpreted to indicate that the Peshawar basin area was active during the late Miocene and early Pliocene (Burbank and Tahrirkeli, 1985). More recently, southeastward migration of the deformation front has been documented along the eastern SR/PP (Johnson et al., 1979; Burbank et al., 1986; Johnson et al., 1986). From northwest to southeast, evidence for the southeastward younging of deformation consists of (see fig.5 for locations): (1) uplift associated with overthrusting along the Riwat fault at the Kas Dovac site between 3.4 and 2.7 mya (Raynolds, 1980; Johnson et al., 1982; Johnson et al., 1986); (2) Soan syncline control of deposition at about 3 mya; (3) surface expression of the Buttar fold after 5.5 mya; (4) surface relief of the Jabbar anticline more recently than 3 mya (Raynolds, 1980); (5) unroofing of Eocene strata at 2.5 mya along the Domeli Ridge (Johnson et al., 1982, Johnson et al., 1986); (6) deposition of angular Siwalik clasts indicating deformation of the en-echelon Lehri and Mehesian anticlines at approximately 2.3 mya (Raynolds, 1980); (7) post 2.4 myainitiation of deformation and post 0.7 mya surface expression of the Chambal Ridge 73

(Johnson et al., 1979); and (8) initiation of deformation along the Pabbi Hills anticline at about 1.2 mya, with surface expression after 0.4 mya (Keller et al., 1977). Out-of-sequence deformation is exemplified along the north limb of the Soan syncline,5 km south of Rawalpin- di (fig. 5), where tilted strata deposited 2.1 mya are overlain by flat lying strata deposited about 1.9 Ma (Raynolds, 1980). These data pre- cisely bracket an uplift event and, when compared to the ages of deformation farther south, demonstrate a backstepping of strain release. Yeats and others (1984) document deformation after 0.4 mya in the eastern Salt Range. They suggest that thrusting along the Salt Range front is active at a recurrence interval of thousands of years. An active Salt Range frontal thrust is supported by micro-earthquake studies which show that the entire Salt Range is active at low magni- tude levels (Seeber and Jacob, 1977). The timing of initiation of central Salt Range ramping is discussed in Baker et al. (in review; see appendix A). In the central Potwar Plateau, the Dhok Pathan Formation, dated as 5.1 m.y. at its upper boundary (Johnson et al., 1982), is folded and truncated along the northern limb of the Soan syncline. Therefore, NPDZ deformation is atleast as young as 5.1 m.y. There is no evidence of active thrusting in the NPDZ. However, more interior portions of the foreland fold-thrust belt may remain active, or are reactivated, from time to time. This is exemplified by accelerated uplift of the Attock Range after 0.6 mya and continuing tectonic deformation within the Peshawar basin (Burbank and Tahirkheli, 1985).

Decollement Propagation Rate Calculation

The rate of deformation front migration or decollement tip propagation is clouded by out-of-sequence deformation. This rate can be estimated along the eastern margin of the SR/PP. Truncation of the stratigraphic records at Kas Dovac and Pabbi Hills (fig. 5), indicating the initiation of surface expression along the Riwat fault andPabbi Hills anticline, respectively, represent the most widely spaced equivalent events. The events occur about 3 m.y. apart, in locations separated by 90 km, indicating a deformation front migration rate of30 74 mm/yr. This rate is the same as the facies progradation rate obtained by Raynolds and Johnson (1985) for the same area. Due to sparce timing data and varying rates of decollement propagation, an average migration rate cannot be obtained by similar calculations in the central study area. Instead, Williams and Chapman's (1983) slip/propagation rate method is used. By dividing the derived frontal shortening rate, which is equivalent to the slip rate, of 11 mm/yr (Baker et al., in review; see appendix A), by the percentage of shortening, corresponding to the area over which the slip rate was determined, then multiplying by 100, the propagation rate is obtained. Cross-section balancing indicates approximately 20% shortening (e.g. fig. 13) between the trailing branch line of the Salt Range thrust sheet in the central Potwar Plateau (fig. 8) and the present deformation front at the Lilla structure; therefore, the resulting propagation rate is 55 mm/yr. It appears that the decollement propagates into undeformed foreland sediments at a rate temporarily exceeding plate convergence rates, analogous to fault-plane slip rates exceeding plate movements during earthquakes. This represents a fast rate of deformation front migration and is explained by decoupling along the salt- lubricated decollement. By extrapolating this propagation rate back in time from the present deformation front, continuing deformation in the central portion of the NPDZ, until just before ramping at about 2 mya, is implied. Care must be taken in interpreting these results because the propagation rate is very sensitive to the derived shortening rate. An important implication of the faster propagation rate is that deformation would be occurring in the eastern Potwar Plateau, before termination of deformation in the central portion of the NPDZ.

Motion of the Thrust System

The southward migration of the deformation front across the Pakistan foreland, in the central SR/PP portion of the fold-thrust belt, appears to have occurred at a varying rate. Initial deformation south of the Hill Ranges is assumed to have taken place in the NPDZ. Intense deformation there is indicative of a relatively slow rate of decollement propagation. At some point, the decollement began to 75 develop in the salt layer and propagated rapidly across the Soan trough, before encountering the basement buttress. This added resistance caused the sole thrust to stick and temporarily ceaseforward propagation. Internal deformation of the thrust sheet, including interpreted buckling against the basement buttress (fig. 18), took place at this time. As the Salt Range thrust sheet overrode the basement offset, slip was again transferred to the frontal thrust, which ramped to the surface. In contrast, deformation style and timing data (described above) suggests a more steady rate of deformation-front migration in the eastern SR/PP. During basement buttressing, slip at the rear of the thrust system may have been accommodated by a continuation ofthrusting or out-of- sequence deformation in the NPDZ. While the deformation front surface was continuing to step out into the foreland to the east,deformation continued in the central portion of the NPDZ because strain was transferred northward from the buttress. Assuming equal amounts of shortening along strike south of the MBT, this interpretation suggests that eastern Potwar Plateau shortening is not compensated byoverthrusting in the central Salt Range, but rather by NPDZ deformation. This hypothesis is supported by the eastward narrowing of the NPDZ (fig. 5), suggesting an eastward decrease in shortening there. Modeling by Mandl and Shippam (1981) suggest that an increased push or buttressingwill cause fracturing, resulting in thrust ramping near the rearof a thrust

sheet. In the central Potwar Plateau this would be manifested by the incorporation of additional horses or imbricate slices at the southern margin of the NPDZ.

Tectonic Rotations

Early paleomagnetic measurements on Paleozoic rocks exposed in the Salt Range were used to both constrain the overall rotation ofGreater India during collision and infer tectonic rotation of the Salt Range (see Klootwijk, 1979). Both Paleozoic and Siwalik paleomagnetic measurements suggest more than 6.5° counterclockwiserotation of the entire Indian plate during the past 10 m.y. (Molnar and Tapponnier,

1975). From these same data over 200 counterclockwise rotation of the 76

Salt Range relative to the Indian shield has been suggested(Crawford, 1974; Klootwijk, 1979). Opdyke and others (1982) divided new paleomagnetic observations into two groups corresponding to declinations ranging from 3500-360° and 3200-330°, respectively (see fig.5 for locations). The first group can be accounted for by rotations of the Indianplate. Sediments from the second group have an average rotation of about 35° counterclockwise, well in excess of Indian plate rotation. By sampling sediments ranging from 12 to 2 m.y. in age, Opdyke and others (1982) found evidence suggesting that rotations were not contemporaneous and concluded that the SR/PP is segmented into regions with different times of rotations. Burbank and others (in press) sampled along a transect crossing the Soan syncline south of Rawalpindi (Soan and Kas Dovac sites; fig.

5). Sampled sediments range in age from 9 to less than 0.7 n.y. Burbank interpreted a temporal variation in rotation rates and divided the progessive rotation into three phases. During the first phase, which covers an interval from 9-2.5 mya, 10-20° rotation took place, representing Indian plate rotation plus regional rotation away from the

HKS. This is followed by little or no rotation during deformation of the Soan syncline between 2.1 and 1.9 mya. The remaining 20-25° of rotation took place rapidly after the deposition of the Lei conglomerate (approx. 1.8 mya), and before deposition of the Potwar Silt. Burbank claims that late-stage rotation is not due to local deformation but rather a regional event above the decollement surface in response to drag along the Jhelum Reentrant. Shortening values, as determined by cross-section balancing, across the central Salt Range (fig. 18) do not support a regional counterclockwise rotation of the Salt Range thrust sheet during ramping, (discussed below).

Timing of Basement Faulting

In order to control ramping, the basement offset must have been present prior to the encroachment of the thrust front. Possible origins of the basement fault can be found in Baker et al. (in review; see appendix A). The lack of sufficient stratigraphic thickness 77 changes and the presence of a substantial crustal thickness(Farah et al., 1977), suggesting that the crust has not been thinned, seems to rule out Mesozoic rifting as the primary cause of the present basement offset. Similarily, basement faulting during the late Cretaceous through Paleocene is unlikely, because contemporaneous strata are not sufficiently thick to change thickness by a kilometer across the fault, and preserved strata do not thicken towards the restored position of the down-thrown block. The basement fault could have formed entirely during the Eocambrian or the Neogene, as evidenced by stratigraphic thickness changes constrained by seismic and exploration wells. An Eocambrian event is supported by the thick allochthonous salt of the Salt Range restoring to the down-thrown side of the basement fault (fig. 21 a). Eocambrian faulting may have controlled the distribution of salt. Fault control of salt deposition is also supported by the change in deformation style in the eastern Potwar Plateau, attributed to the thinning of the salt, coinciding with the dying out of the normal fault. However, the presence of cratonic crust and lack of seismic reflectors characteristic of synrift graben fill (Lillie and Yousuf, 1986) do not support an Eocambrian event. Also, the preferred ramping mechanism is inconsistent with Eocambrian basement faulting, as discussed above. The basement offset could also have formed entirely during the Neogene, because the molasse thins by over a kilometer from the Baun syncline to the Jhelum plain (Baker et al., in review; see Appendix

A). A variation of this theme would be late Paleogene to early Neogene faulting represented by thinning of platform strata and Rawalpindi molasse (fig. 21 b). However, Rawalpindi thinning is thought to represent a natural stratigraphic pinch-out and platform stratathinning is better explained by later erosion related to southward migration of the peripheral bulge (Baker et al., in review; see appendix A). Neogene faulting is thought to be related to lithospheric flexure as thrust sheets are emplaced from the north. As the bulge passes through the crust, the upper crust is placed in an extensionalregime in which normal faulting occurs (Isaacs et al., 1968). For this reason, and since seismic expression suggests anormal rather than a reverse fault (fig. 15), reverse faultingrelated to the present 78 compressional regime is considered unlikely. Flexural modeling by Duroy (1986; Duroy et al., in press) supports the interpretation of the Sargodha high as a flexural bulge with an associated extensional regime in the upper crust. Additional support of flexural extension of the Indian continental crust comes from observations of two large normal fault earthquakes which have occurred beneath the foreland of India (Molnar et al., 1976; Ni and Barazangi, 1984). A seismic profile across the Sargodha high, south of the Mianwali Reentrant, supports its interpretation as a flexural bulge. Prominent platform and basement reflections are warped into a broad, anticlinal shape, indicating a post Permian and probably a post Eocene age of development. Permian strata were penetrated beneath molasse by the Kundian well (see Leathers, 1987). The northern limb is steeper, consistent with the modeled shape of the flexural bulge (Duroy, 1986; Duroy et al., in press). In addition, molasse reflections lap on to the north limb as crust is flexed down into the foredeep.The sequence of prominent platform reflections is truncated along the southern limb and thinner on the northern limb, which is explained by erosion near the crest of the bulge. Finally, horst and graben structures are interpreted, from the reflection profile, along the crest of the Sargodha high, suggesting normal faulting. An Upper Siwalik conglomerate bed, deposited about 4.5 mya at the Baun locality (fig. 16), contains clasts of Eocene carbonates and recycled Precambrian clasts from the Permian Tobra Formation (Johnson et al., 1986; Burbank and Raynolds, in press). These clasts are important stratigraphic evidence constraining the timing of either ramping or basement faulting. If the source area of Upper Siwalik clasts is the Salt Range, then conglomerate deposition, in combination with diminished sedimentation rates and differential rotation, may indicate incipient Salt Range thrusting by 4.5 mya (Johnson et al., 1986; Burbank and Raynolds, in press). Figure 22a shows that the amount of uplift necessary to expose Permian strata as suggested by Burbank and Raynolds (in press) would lift competent hanging-wall strata (i.e. Cambrian to Eocene rocks) at the ramp above competent footwall strata. In lieu of this, regional transmission of stresses from the basement buttress resulting in the Riwat fault and MBT 79

Fig. 22: Schematic diagrams illustrating 2 possible modes of exposing source beds of Eocene carbonate and Permian TobraFormation clasts found in Upper Siwalik conglomerates at Baun. The Baun site has been "pulled back" 25 km northward, from the basement fault, to account for later shortening on these 4.5 mya restorations. (A) Allochthonous source beds uplifted by early activation of a segment of the Salt Range thrust without overriding the older basement fault (Burbank and Raynolds, in press). (B) Autochthon- ous source beds exposed to erosion viauplift along a contemporaneous basement fault. To expose beds immediately adjacent to the fault requires an average basement dip of 5.5° or 8° (measured from the down-thrown and upthrown basement blocks respectively); these dips are anomalously steep compared to present basement structure (see Duroy, 1986; Duroy et al., in press). Therefore, as a slight perturbation of (B), autochthonous source beds mayhave been exposed farther to the south along the crest of the peripheral bulge at 4.5 mya, in line with more refined restorations. An autochthonous source is supported by drilling at Lilla, where source strata have apparently been removed byerosion (fig. 13). 80 movements (Burbank et al., 1986) makes little sense. In addition, exposure of alLochthonous Salt Range source bedswould require the unroofing of approximately 2 km of molasse prior to 4.5 mya, while no evidence of recycled Siwalik and Rawalpindi clasts have been reported at Baun. Alternatively, the erosion of clasts deposited at the Baun locality are readily explained by passage of the flexural bulge. By this interpretation, Eocene and Permian clasts are interpreted as derived from the south, from the up-thrown basement fault block (fig.

22b). The Shell Lilla well (fig. 3) penetrated Cambrian rocks directly beneath Siwaliks; erosion of intervening strata, including Eocene carbonates and the lower Permian Tobra Formation, is therefore demonstrated south of the basement offset. Baun conglomerate clasts are interpreted as derived from an uplifted autochthonous source, directly adjacent to the basement fault (fig. 22b) or located farther to the south along the crest of the peripheral bulge (fig. 6b). Differential rotations recorded at Baun are not explained by passage of the peripheral bulge unless strike-slip movements accompanied dip-slip faulting on basement faults, resulting in shearing of overlying rocks. Strike-slip fault plane solutions of earthquakes along the Sargodha High (Seeber et al., 1981) may be evidence of this mechanism. Alternatively, rotation data or the interpretation of the age of unrotated Upper Siwaliks could be in error. Magnetic polarity stratigraphy depends on an unambiguous correlation between the magnetic polarity time scale and the locally measured succession of reversals. The correlation of the mostly reversed Upper Siwaliks at Baun is admittedly ambiguous (Opdyke et al., 1979). For example the observed succession could just as easily be assigned to the Gauss rather than the Gilbert Chron (see Burbank and Raynolds, in press; their fig. 4), indicating that Upper Siwaliks at Baun were deposited between 2.3 and

3.4 mya. Another possibility is that rotation occurred during early initiation of thrusting in the western Salt Range (see Leathers, 1987), which conflicts with Johnson and others (1986) interpreted time of ramping. If the Baun conglomerates are the last remnants of stratigraphic thickening across the normal fault (fig. 22b), then basement faulting 81 was occurring at 4.5 mya. Otherwise earlier faulting is suggested because it would explain a kilometer of Lower to Middle Siwalik molasse thickening between the Baun syncline and the Jhelum plain (see fig.

21c). Using the derived rate of frontal shortening and the rate of underthrusting beneath the Himalaya (see Molnar and Deng, 1984) as an indication of the relative motions between Indian crust and the flexural bulge, the past positions of the basement fault versus the Sargodha high have been estimated (fig. 7). These reconstructions suggest that the faulted basement was in the extensional regime at the crest of the bulge, or along its sharply downflexed northern flank, between 4 and 12 mya. Another possibility is that an early formed basement fault has been reactivated. A faulted upper crust could have been bevelled by erosion prior to deposition of the Eocambrian and younger section and then the fault reactivated during Neogene flexure. The great throw on this basement fault as compared to other potential basement faults identified on seismic profiles is thus explained by reactivation of a previous plane of weakness in the upper crust. The scenario of restorations constrained by salt retrodeformation, paleobasement dips and topographic slopes is most consistent with Neogene basement faulting (see appendix B).However, the controversy over whether the basement normal fault formed in Eocambrian orTertiary time may not be easily resolved because: (a) Neogene thrusting altered the original distribution of evaporites (Gee, 1983); i.e. salt thickening observed today across the normal fault could be due to salt flow during compression and (b) erosion from the top of the thrust sheet as it ramped to form the Salt Range has removed potential evidence of abrupt molasse thickening and conglomerate deposition on the downthrown side of the normal fault. 82

SHORTENING AND ITS IMPLICATIONS

Amount of Shortening

Salt Range: Six balanced sections were constructed to determine the minimum amount of frontal shortening across the basement ramp in the central Salt Range and evaluate potential variations in shortening along strike. Frontal shortening values, corrected to N 15° W, range from 19.6 to 23.3 km (fig. 16). The greatest amounts of shortening correspond to the sections farthest east and west. Therefore, the balanced sections show little or no regional variation in shortening. Because shortening doesn't vary substantially along strike and the measured amounts of shortening represent minimum values, the larger values are probably closer to the actual amount of shortening. Factors which potentially effect the accuracy of shortening values are discussed in appendix C. The main factors thought to influence the derived shortening values are erosional truncation and failure to account for burial of the frontal thrust. Apparent shortening is less across reentrants where erosion and deposition are greater; inverted topography within these reentrants (see above; Gardezi and Ashraf, 1974) suggests that present day valleys are areas of unroofing due to salt diapirism. Therefore, erosion may explain differences in measured shortening (fig. 16). Southern Potwar Plateau/Soan Trough: The Soan trough is the relatively undeformed, trailing edge of the Salt Range thrust sheet (fig. 13). Only 0.4 km of shortening is measured across broad folds between the axis of the Soan and Baun synclines, a distance of about 50 km. Short- ening along the north limb of the Soan syncline is included with northern Potwar shortening, while deformation of the southlimb of the Baun syncline is included with frontal shortening. Because of the decoupling properties of the Eocambrian salt, Cambrian and younger strata in the intervening area were translated about 23 km over basement with almost no internal deformation. The 0.4 km of internal shortening contrasts sharply with the area along strike in the eastern Potwar Plateau where Johnson and others (1986) estimate "20% (i.e. about 20 km) shortening along the breadth of the detachment". 83

Northern Potwar Plateau: The calculation of shortening in the NPDZ was hampered by poor surface control due to alluvium cover and lack of detailed maps, lack of sufficiently deep drilling, and chaotic seismic expression. The regional balanced cross-section suggests 45+15 km of horizontal shortening of competent platform strata (fig. 13).

Frontal Shortening Rate and Comparison to Plate Convergence Res

Ramping of the central portion of the Salt Range thrust sheet represents a discrete event from which timing and the amount of contraction of sedimentary cover rocks has been determined. Using the amount of shortening from balanced sections and the estimated time of ramping (Baker et al., in review; see appendix A), the minimum rate of shortening between the Indian basement and the Phanerozoic sedimentary cover is between 9 and 14 mm/yr. Several factors have not been considered in this rate calculation. Shortening within the trailing portion of the thrust sheet and across the Lilla structure are ignored but add up to less than 0.5 km. In addition, the reversal of an estimated 0.7 km of interpreted extension during basement faulting (fig. 13), is also not considered. However, these factors are offset by potential shortening prior to the thrust sheet overriding the ramp (i.e. interpreted preramping deformation; fig. 18). For these reasons, and considering timing evidence, the earlier end of the range of thrust ramp initiation is preferred (i.e. about 2.1 mya). Because cross-section balancing results in a minimum shortening value and most sources of error would increase rather than decrease the measured shortening (see appendix C), true shortening in the central Salt Range is assumed to be at least 23 km. Therefore a frontal shortening rate of 11+2 km/m.y. is thought to be representative of the amount of convergence absorbed south of the Hill Ranges(MBT) in Pakistan (fig. 23). This rate is comparable to Lyon-Caen and Molnar's (1985) estimate of 10-15 mm/yr of convergence being absorbed by underthrusting of the central Himalaya. The calculated frontal shortening rate represents about 25% of the 42 mm/yr convergence between the Indian and Asian plates at longitude 730 east (Minster and Jordan, N TOTAL = 42 mm/yr I8-38mm/yr STABLE 1-- 1-4-11mm/yr-4.- ASIAN PAMIR Y HIMALAYA I- f- CRATON TIEN SHAN PB rUPLIFTAND HR1' PP SR SH PLAIN STRAIN ("ESCAPE BLOCK TECTONICS") 'INTERNAL DEFORMATION AND UNDERTHRUSTING NORTH OF THE -7-27 mm/Yr SLIPri`rimm/yrZER 0 SLIP HIMALAYA 424 mm/yr SLIP =18 mm/yt FRONTAL SHORTENING-4Imm/yr SLIP=18 -38mm/yr 11' MOHO 100 km

Fig. 23: Schematic crustal section demonstrating potential distribution of convergence absorbtion. Minimum constaint on convergence absorbed north of the Himalaya is 4 mm/yr (Molnar, 1987), while maximum is 38 mm/yr (Lyon- Caen and Molnar, 1985). The other constraint is the derived rate of frontal shortening, 11 mm/yr (Baker et al, 1987; see appendix A). The balance of convergence (i.e. from a total of 42 mm/yr [Minster and Jordan, 1978]) is accommodated via internal deformation of the Himalaya. Slip along the decollement is equal to the sum of convergence absorbed south of the respective locality. Lyon-Caen and Molnar's 18 mm/yr rate of underthrusting is arbitrarily placed beneath the Main Mantle Thrust (MMT). Crustal structure modified from Seeber and Armbruster (1979). HR=Hill Ranges; PB=Peshawar Basin; PP=Potwar Plateau; and SR-Salt Range. See figure 1for other abbreviations. 85

1978). Therefore, most convergence is absorbed north of thefrontal thrust zone by contraction within the collision zoneand escape-block tectonics (fig. 23).

Tectonic Rotation Versus the Distribution of Shortening

Burbank and others (in press) mechanism of late-stagerotation implies rigid counterclockwise rotation of the Salt Range thrustsheet about an axis near the eastern edge of the saltbasin. However, this hypothesis is inconsistent with pre 4.5 mya rotation at Baun(Opdyke et al., 1979; Burbank and Raynolds, in press; Johnson etal., 1986). To test Burbank and others' hypothesis, the amountof shortening was calculated for 20° to 30° counterclockwise rotation using various moment arms extending to positions in theJhelum reentrant; 29 to 94 km shortening was required across the central Salt Range. The high value is considerably more than measured by balanced cross-sections. The lowest value, which is similar to estimates (fig. 16), wasobtained for a pivot near the eastern edge of theSalt Range. A pivot farther west within the eastern Salt Range is consistent with shorteningvalues. However, a westward increase in shortening would be expectedif ramping were part of a regional rotation, as suggested by timing data. No westward increase was found in the central Salt Range(fig. 16). Also, the large amount of platform strata duplication and thepotential for additional shortening in the eastern Salt Range(fig. 20), suggests that shortening there is close to or in excess of that measuredin the central Salt Range. Erosion should not have had a great effect on shortening values because ramping occurred recently (fig. 19). An increase in shortening in the western Salt Range to over 30 km,estimated from a cross-section (Leathers, 1987) and supportedby the westward increase in the width of the premolasse outcropbelt, may be consistent with gross rotation models. However, the amount of shortening across the western Salt Range is difficult to determine;the location of footwall cutoffs, interpreted from reflectionprofiles, is indiscernible. Timing data can further constrain potential regionalrotation. Deformation at the Domeli and Lehri/Mehesian structures(fig. 5) is 86 interpreted to have preceeded ramping (Johnson et al., 1986). The remaining amount of shortening, across the frontal thrust in the eastern Salt Range (fig. 20), appears to be significantlyless than in the central Salt Range. Add to this the higher estimates of shortening in the western Salt Range, and a regional counterclockwise rotation during ramping is implied. The timing of deformation along the KDF versus frontal ramping and proposed regional rotation is important because it represents a large amount of shortening. If ramping was part of a regional rotation, then strike-slip faults segmenting the Salt Range (fig. 5) may separate areas with different amounts of shortening. Opdyke and others' (1982) interpretation of rotation is supported. That is, rotation above the decollement is not rigid but segmented into regions with different times of rotation. In this case, measured rotations would be caused by local movements. Drag along the edge of the salt basin can be used to explain rotation of eastern SR/PP structures into their present allignment. These structures would be rotated individually, in response to similar stresses, resulting in their parallel orientations (fig. 5). In the central Salt Range, different orientations of the Salt Range front and the basement offset, indicate a westward increase in hanging-wall and footwall ramp separation and are therefore evidence of rotation duringramping. Instead of being the result of a regional rotation, the westward increase in overthrust distance is locally compensated by an eastward increase in internal deformation, mainly concentrated at the toe of the thrust (fig. 17), suggesting a local rotation of the hanging-wall ramp during thrusting. 87

CONCLUSIONS

Insight has been obtained as to the current structure and mode of evolution of the central Salt Range and Potwar Plateau of Pakistan. A pre-existing, down-to-the-north basement offset controlled ramping of the encroaching thrust front. Palinspastic restorations, seismic observations, and ramping style suggest that the basement offset is a normal fault formed during Neogene flexure of the crust. Alternative- ly, the basement offset may have been formed during Eocambrian rifting and therefore controlled the original distribution of evaporites. During later thrusting, the buttressing effect of the basement offset caused a local buildup of salt. Lack of footwall deformation and the presence of a steep ramp through platform strata support aramping mechanism in which the normal fault through competent platform strata was reactivated in a reverse sense duringinitial stages of ramping. Salt growth eventually lifted strata south of the offset into a position where compressive stresses could push the thrust sheet along a shallow ramp toward the surface. The thrust trajectory appears to have been smoother than would occur during typical fault-bend folding. However, bending moment stresses across the basement offset were still responsible for passively folding competent strata into the present anticlinal form of the Salt Range. The thrust sheet north of the ramp in the southern Potwar Plateau is only mildy deformed due to the decoupling properties of salt. Farther to the north, the northern limb of the Soan syncline dips steeply southward above an upper-level detachment. Strata above the upper-level detachment and NPDZ thrust slices form a "triangle zone" geometry. This style of deformation changes westward, across an oblique ramp, to an apparent imbricate fan. The upper-level detachment is interpreted to continue northward where it separates different lithotectonic units in the NPDZ. Beneath the upper-level detachment, competent platform strata are deformed into a hinterland-dipping

duplex. The timing of deformation is poorly constrained throughout the central Soan trough and NPDZ. However, changing sediment accumulation rates used in conjunction with stratigraphic evidence suggest thatthe 88 basement ramp to the south was overridden about 2 mya. Evidence of early to middle Pliocene deformation (i.e. Upper Siwalik conglomerate clasts preserved in the Baun syncline), thought to indicate early hanging-wall ramp uplift, is alternatively explained by basement normal faulting, uplift, and erosion related to migration of the peripheral bulge through the Indian lithospheric plate. Balanced cross-sections suggest that a minimum of 19 to 23 km of shortening is present across the frontal ramp, indicating a shortening rate of 9-14 mm/yr.There- fore, frontal thrusting in the Pakistan foreland accounts for a significant but not the majority of the total plate convergence. Rapid decollement propagation along the salt lubricated decollement within the central Potwar Plateau was temporarily slowed by basement buttressing. Differing thrust propagation rates between the central and eastern Potwar Plateau suggest that deformation within the central NPDZ was concurrent with shortening in the eastern Potwar Plateau. Regional rotation of the thrust sheet may then have occurred during subsequent frontal ramping as indicated by an apparent westward increase in overthrust distance. However, central Salt Range sections do not indicate a westward increase in shortening, as would be expected as a result of proposed regional rotation of the Salt Rangethrust

sheet. Instead, the westward increasing overthrust distance in the central Salt Range appears to be locally compensated by an eastward increase in internal deformation. Previous foredeep and facies progradation rates appear to be over- estimated due to the influence of recent rapid deformation front migra-

tion. Foredeep migration is instead controlled by the rate of underthrusting beneath the Himalaya, of which frontal shortening represents the minimum rate, not total plate convergence. Palinspastic restorations suggest that Lower and Middle Siwalik molasse deposition was concentrated in the central Potwar Plateau, possibly by basin downdropping along the basement normal fault. As the deformation front shifted to the Salt Range area, the depocenter jumped to the south along the Jhelum River. 89

Future Work

Although many of the problems in the Pakastani foreland have been addressed (e.g. Gee, 1983), future work is needed to fully understand the structural development of the SR/PP. Additional timing information would be helpful. For example, an indisputable age of the sediments at Baun may help to further constrain the time of Salt Range thrust sheet ramping and normal fault formation. The age of sediments at the Lilla structure would determine its relationship to Salt Range thrust ramping. Similarily, the maximum age of the Potwar silt is needed to date the initial uplift of the Salt Range. Additional rotation data would help to resolve the controversy concerning rigid regional (Burbank, in press) or segmented rotation (Opdyke et al., 1982). Better surface maps, seismic data and additional well control are needed to further resolve the structural style in the NPDZ. A branch line map is essential to understanding the interconnections of faults there. More accurate shortening data are needed in the NPDZ, and in the eastern SR/PP to determine the regional variation of shortening and its bearing on regional rotation. To this end, additional subsurface data are needed along the northern flank of the western Salt Range to locate footwall cutoffs and resolve the emerging controversy over the amount of overthrusting preserved there. In light of changing basement dips as a result of lithospheric flexure, a dynamic model of the mechanics of foreland thrusting is needed to study past movements of the thrust system. Similarily, a revised model of foreland deposition should be developed that takes changing basin shape and the effect of the Sargodha High into consideration. Finally, the rate of deformation front migration plus the rate and percentage of shortening should be studied in the Indian portion of the Himalayan foreland fold-thrust belt where a salt lubricated decollement is lacking. A comparison of results to the present study could then be used to evaluate the decoupling effects of salt and the origin of the HKS. 90

BIBLIOGRAPHY

Acharyya, S. K., and K. K. Ray, 1982, Hydrocarbon possibilities of concealed Mesozoic Paleogene sediments below Himalayan nappes reappraisal, Amer. Assoc. Petrol. Geol. Bull, v. 66, p. 57-70.

Ala, M. A., 1974, Salt diapirism in southern Iran: Am. Assoc. Petrol. Geol. Bull., v. 58, p. 1758-1770.

Baker, R. A., H. M. Gehman, W. R. James, and D. A. White, 1984, Geolog- ic field number and size assessments of oil and gas plays, Am. Assoc. Petrol. Geol. Bull., v. 68, p. 426-432.

Baker, D. M., R. J. Lillie, R. S. Yeats, G. D. Johnson, M. Yousuf and A. S. H. Zamin, in review, Development of the Himalayan frontal thrust zone: Salt Range, Pakistan, Geology.

Banks, C. J., and J. Warburton, 1986, 'Passive-roof' duplex geometry in the frontal structures of the Kirthar and Sulaiman mountain belts, Pakistan, Jour. Struc. Geol., v. 8, p. 229-238.

Barndt, J., N. M. Johnson, G. D. Johnson, N. D. Opdyke, E. H. Lindsay, D. Pilbeam, and R. A. K. Tahirkheli, 1978, The magnetic polarity stratigraphy and age of the Siwalik Group near Dhok Pathan vil- lage, Potwar Plateau, Pakistan, Earth Planet. Sci. Lett., v. 41, p. 355-364.

Behrensmeyer, A. K., and L. Tauxe, 1982, Isochronous fluvial systems in Miocene deposits of northern Pakistan, Sedimentology, v. 29, p. 331-352.

Bender, F., 1975, Geology of the Arabian Peninsula: Jordon, U. S. Geol. Survey Prof. Paper 560-1, 36 pp.

Boyer, S. E., and D. Elliott, 1982, Thrust systems, Amer. Assoc. Petrol. Geol. Bull., v. 66, p. 1196-1230.

Burbank, D. W., 1982, The chronologic and stratigraphic evolution of the Kashmir and Peshawar intermontane basins, northwestern Himalaya, Ph. D. thesis, Dartmouth College, Hanover, New Hamp- shire, 291 pp.

Burbank, D. W., 1983, The chronology of intermontane-basin development in the northwestern Himalaya and the evolution of the Northwest Syntaxis, Earth Planet. Sci. Lett., v. 64, p. 77-92.

Burbank, D. W., and R. G. Raynolds, in prep., Stratigraphic keys to the timing of deformation: an example from the northwestern Himalayan foredeep.

Burbank, D. W., R. G. Raynolds, and G. D. Johnson, 1986, Late Cenozoic tectonics and sedimentation in the northwestern Himalayan fore- 91

deep: II. Eastern limb of the northwest syntaxis and regional synthesis, in P. A. Allen, and Homewood, P., eds., Foreland basins, Spec. Publs. Int. Assoc. Sediment., v. 8, p. 293-306.

Burbank, D. W., R. G. Raynolds, and D. P. Whistler, in press, Detailed chronologies of tectonic rotations: new insights from magnetostratigraphic studies in Pakistan and along the Garlock fault.

Burbank, D. W., and R. A. K. Tahirkheli, 1985, The magnetostratigraphy, fission-track dating, and stratigraphic evolution of the Peshawar intermontane basin, northern Pakistan, Geol. Soc. Amer. Bull., v. 96, p. 539-552.

Butler, R. H. W., 1982, The terminology of structures in thrust belts, Jour. Struc. Geol., v. 4, p. 239-245.

Butler, R. W. H., M. P. Coward, G. M. Harwood, and R. J. Knipe, in press, Salt, its control on thrust geometry, structural style and gravitational collapse along the Himalayan mountain front in the Salt Range of northern Pakistan.

Charlesworth, H. A. K., and W. E. Kilby, 1981, Thrust nappes in the Rocky Mountain foothills near Mountain Park, Alberta, in K. R. McClay, and N. J. Price, eds., Thrust and Nappe Tectonics, Geol. Soc. London Spec. Pub. 9, p. 475-482.

Chevren, V. B., 1986, Tethys-marginal sedimentary basin in western Iran, Geol. Soc. America Bull., v. 97, p. 516-522.

Cooper, M. A, 1983, The calculation of bulk strain in oblique and inclined balanced cross-sections, Jour. of Struc. Geol., v. 5, p. 161-165.

Cooper, M. A., and P. M. Trayner, 1986, Thrust-surface geometry: implications for thrust belt evolution and section balancing techniques, Jour. Struc. Geol., v. 8, p. 305-312.

Cotter, G. P., 1933, The geology of the part of the Attock District west of longitude 72° 45', Geol Soc. India Mem., v. 55, p. 63-156.

Coward, M. P., 1982, Surge zones in the Moine thrust zone of NW Scotland, Jour. Struc. Geol., v. 4, p. 247-256.

Coward, M. P., and R. W. H. Butler, 1985, Thrust tectonics and the deep structure of the Pakistan Himalaya, Geology, v. 13, p. 417-420.

Coward, M. P., and J. H. Kim, 1981, Strain within thrust sheets, in K. R. McClay, and N. J. Price, eds., Thrust and nappe tectonics, Geol. Soc. London Spec. Pub. 9, p. 275-292.

Cox, A., 1969, Geomagnetic reversals, Science, v. 163, p. 237-244.

Crawford, A. R., 1974, The Salt Range, the Kashmir syntaxis and the 92

Pamir arc, Earth Planet. Sci. Lett., v. 22, p. 371-379.

Dahlstrom, C. D. A., 1969, Balanced cross-sections, Can. Jour. Earth Sci., v. 6, p. 743-757.

Dahlstrom, C. D. A., 1977, Structural Geology in the eastern margin of the Canadian Rocky Mountains, in E. L. Heisey, E. R. Norwood, P. H. Wach, and L. A. Hale, eds., Wyoming Geol. Assoc. 29th Ann. Field Conf. Guidebook, p. 407-439.

Davis, D. M., and T. Engelder, 1985, The role of salt in fold-and- thrust belts, Tectonophysics, v. 119, p. 67-88.

Davis, D. M., J. Suppe, and F. A. Dahlen, 1983, Mechanics of fold-and- thrust belts and accretionary wedges, Jour. Geoph. Res., v. 88, p. 1153-1172.

Dubey, A. K., 1980, Model experiments showing simultaneous development of folds and transcurrent faults, Tectonophysics, v. 65, p. 69-84.

Duroy, Y., 1986, Subsurface densities and lithospheric flexure of the Himalayan foreland in Pakistan from gravity and seismic reflection constraints, M. S. thesis, Oregon State University, Corvallis, Oregon, 74 pp.

Duroy, Y., A. Farah, L. L. Malinconico, and R. J. Lillie, 1987, Subsurface densities and lithopheric flexure of the Himalayan foreland in Pakistan, in L. L. Malinconico, and R. J. Lillie, eds., Tectonics and geophysics of the western Himalaya, Geol. Soc. Amer., Spec. Paper.

Eisbacher, G. H., M. A. Carrigy, and R. B. Campbell, 1974, Paleodrain- age pattern and late orogenic basins of the Canadian Cordillera, Soc. Econ. Paleon. and Min., Spec. Pub. v. 22, p. 143-166.

Elliott, D., 1976, The energy balance and deformation mechanisms of thrust sheets Royal Soc. London Philos. Trans., ser. A, v. 283, p. 289-312.

Elliott, D., 1983, The construction of balanced cross-sections, Jour. Struc. Geol., v. 5, p. 101.

Faill, R. T., 1973, Kink band folding, Valley and Ridge Province, Pennsylvania: Geol. Soc. America Bull., v. 4, p. 1289-1314.

Farah, A., M. A. Mirza, M. A. Ahmad and M. H. Butt, 1977, Gravity field of the buried shield in the Punjab plain, Pakistan, Geol. Soc. Amer. Bull., v. 88, p. 1147-1155.

Finetti, I., F. Giorgetti and G. Poretti, 1979, The Pakistani segment of the DSS profile, Nanga Parbat-Kabul (1974-1975), Bollettino di Geofisica Teorica ed Applicata, v. 21, p. 159-171. 93

Fischer, M. W., and M. P. Coward, 1982, Strains and folds within thrust sheets: an analysis of the Heilam Sheet, Tectonophysics, v. 88, p. 291-312.

Fox, F. G., 1983, Structure sections across Parry Islands fold belt, and Vesey Hamilton Salt Wall, Arctic Archipelago, Canada, in Seismic Expression of Structural Styles, Amer. Assoc. Petr. Geol., Stud. Geol. 15, Vol. 3, A. W. Bally ed., pp. 3.4.1-54 3.4.1-72.

Gardezi, A. H., and M. Asraf, 1974, Gravitative structures of the Kattha Masral region of the central Salt Range, Pakistan, The geological bulletin of the Punjab University, n. 11, p. 75-80.

Gee, E. R., 1934, The Saline Series of northwest India, Curr. Sci. 2, 460 p.

Gee, E. R., 1980, Salt Range series geological maps, 1:50,000, 6 sheets, Directorate of Overseas Surveys, United Kingdom, for Govt. of Pakistan and Pakistan Geol. Surv.

Gee, E. R., 1983, Tectonic problems of the Sub-Himalayan region of Pakistan, Kashmir Jour. Geol., v.1, p. 11-18.

Gill, W. D., 1952, The stratigraphy of the Siwalik Series in the northern Potwar, Punjab, Pakistan, Geol. Soc. London Quart. Jour., v. 107, p. 395-421.

Gwinn, V. E., 1964, Thin-skinned tectonics in the Plateau and northwestern Valley and Ridge provinces of the central Appalachians, Geol. Soc. Amer. Bull., v. 75, p. 863-899.

Hossack, J. R., 1983, A cross-section through the Scandinavian Cale- donides constructed with the aid of branch line maps, Jour. Struc. Geol., v. 5, p. 103-111.

Issacs, B., J. Oliver, and L. Sykes, 1968, Seismology and the new global tectonics, Jour. Geophys. Res., v. 73, p. 5855-5899.

Jaume, S., 1986, Thrust mechanics in the Himalayan foreland of Pakistan: quantitative and qualitative assessment of an area underlain by evaporites, M. S. thesis, Oregon State University, Corvallis, Oregon, 60 pp.

Jaume, S. C., and Lillie, R. J., in press, Active tectonics of the Salt Range-Potwar Plateau, Pakistan: mechanics of a fold-and-thrust belt underlain by evaporites, Tectonics.

Jenyon, M. K., 1985, Fault-associated salt flow and mass movements, Geol. Soc. London, v.142, p. 547-553.

Johnson, G. D., N. M. Johnson, N. D. Opdyke, and R. A. K. Tahirkheli, 1979, Magnetic reversal stratigraphy and sedimentary tectonic history of the upper Siwalik Group, eastern Salt Range and 94

southwestern Kashmir, in Geodynamics of Pakistan, A. Farah and K. A. DeJong, eds., Geol. Surv. Pakistan, p. 149-165.

Johnson, G. D., R. G. H. Raynolds and D. W. Burbank, 1986, Late Ceno- zoic tectonics and sedimentation in the northwestern Himalayan foredeep: I. Thrust ramping and associated deformation in the Potwar region, in P. A. Allen, and P. Homewood, eds., Foreland basins, Spec. Pubis. Int. Assoc. Sediment., v. 8, p. 273-291.

Johnson, N. M., Opdyke, N. D., Johnson, G. D., Lindsay, E. H., and Tahirkeli, R. A. K., 1982, Magnetic polarity stratigraphy and ages of Siwalik Group rocks of the Potwar Plateau, Pakistan, Palaeo- geog. Plaeoclim. Palaeoecol., v. 37, p. 17-42.

Jones, P. B., 1982, Oil and gas beneath east-dipping underthrust faults in the Alberta foothills, Rocky mountain Assoc. Geol. Guidebook, Denver, Colorado, p. 61-74.

Kazmi, A. H., and R. A. Rana, 1982, Tectonic map of Pakistan, Geol. Surv. of Pakistan, scale 1:2,000,000, 1st. edition.

Keller, H. M., R. A. K. Tahirkheli, M. A. Mirza, G. D. Johnson, N. M. Johnson, and N. M. Opdyke, 1977, Magnetic polarity stratigraphy of the Upper Siwalik deposits, Pabbi Hills, Pakistan, Earth Planet. Sci. Lett., v. 36, p. 187-201.

Khan, M. J., 1984, Brief results of the paleomagnetic studies of the Siwalik Group of the Trans-Indus Salt Range, Pakistan, Geol. Bull. Univ. Peshawar, v.17, p. 176-178.

Khan, M. A., R. Ahmed, H. A. Raza, and A. Kemal, 1986, Geology of Petroleum in Kohat-Potwar Depression, Pakistan, Amer. Assoc. Pet. Geol. Bull., v. 70, p. 396-414.

Klootwijk, C. T., 1979, A summary of palaeomagnetic data from extra- peninsular IndoPakistan and south central Asia: implications for collision tectonics, in P. S. Saklani, ed., Structural geology of the Himalaya: Today and Tomorrow Printers and Pub., New Delhi, p. 307-360.

Knipe, R. J., 1985, Footwall geometry and the rheology of thrust sheets: Jour. Struc. Geol., v. 7, p. 1-10.

Latif, M. A., 1973, Partial extension of the evaporite facies of the Salt Range to Hazara, Pakistan: Nature, v. 244, p. 124-125.

Leathers, M., 1987, Balanced structural cross section of the western Salt Range and Potwar Plateau: deformation near the strike-slip terminus of an overthrust sheet, M. S. thesis, Oregon State Univ., Corvallis, Oregon.

Lillie, R. J. and M. Yousuf, 1986, Modern analogs for some midcrustal reflections observed beneath collisional mountain belts, in 95

Reflection Seismology: The Continental Crust, Amer. Geoph. Un., Geodynamics Series Vol. 14, M. Barazangi and L. Brown, eds., p. 55-65.

Lillie, R. J., G. D. Johnson, M. Yousuf, and R. S. Yeats, 1987, Structural development within the Himalayan foreland fold-and- thrust belt of Pakistan, in Basins of Eastern Canada and Worldwide Analogs, Canada. Assoc. Petrol. Geol., Memoir 12, C. Beaumont and A. J. Tankard, eds., in press.

Lohmann, H. H., 1972, Salt dissolution in subsurface of British North Sea as interpreted from seismograms: Am. Assoc. Petrol. Geol. Bull., v. 56, p. 472-479.

Lyon-Caen, H., and P. Molnar, 1985, Gravity anomalies, flexure of the Indian plate, and the structure, support and evolution of the Himalaya and Ganga basin, Tectonics, v. 4, p. 513-538.

Mandl, G., and G. K. Shippan, 1981, Mechanical model of thrust sheet gliding and imbrication, inK. R. McClay, and N. J. Price, eds., Thrust and nappe tectonics: Geol. Soc. London Spec. Pub. 9, p. 79-98.

Martin, N. R., 1962, Tectonic style in the Potwar, West Pakistan: Punjab Univ. Geol. Bull., v. 2,p. 39-50.

Martinez, J. D., 1974, Tectonic behavior of Evaporites, in Fourth International Symposium on salt: Northern Ohio Geol. Soc., p. 155-168.

McDougall, J., in prep., Kalabagh strike-slip fault zone, western Salt Range, Pakistan, PhD thesis, Oregon State University, Corvallis, in prep.

McDougall, J. W., 1985, Strike-slip faulting in a foreland fold-thrust belt, western Salt Range, central Pakistan, Geol. Soc. Amer., Abs. with Prog., p. 658.

Minster, J. B., and T. H. Jordan, 1978, Present day plate motions, Jour. Geoph. Res., v. 83, p. 5331-5354.

Mitra, S., 1986, Duplex structures and imbricate thrust systems: geometry, structural position, and hydrocarbon potential, Am. Assoc. Petrol. Geol. Bull., v. 70, p. 1087-1112.

Mitra, S., and S. Boyer, 1986, Energy balance and deformation mechanisms of duplexes, Jour. Struc. Geol., v. 8, p. 291-304.

Molnar, P., B. C. Burchfiel, L. K'uangyi, and Z. Ziyan, 1987, Geomor- phic evidence for active faulting in the Altyn Tagh and northern Tibet and qualitative estimates of its contribution to the convergence of India and Eurasia, Geology, v. 15, p. 249-253. 96

Molnar, P., W. P. Chen, T. J. Fitch, P. Tapponier, W. E. K. Warsi and F. T. Wu, 1976, Structure and tectonics of the Himalaya: A brief summary of relevant geophysical observations, Colloques Inter- national du Centre National de la Recherch Scientific, no. 268, p. 269-294.

Molnar, P., and Deng Qidong, 1984, Large earthquakes and the average rate of deformation in Asia, Jour. Geoph. Res., v. 89, p. 6203- 6228.

Molnar, P., and P. Tapponnier, 1975, Late Cenozoic tectonics of Asia: effects of a continental collision, Science, v. 189, p. 419-426.

Morley, C. K., 1986, Classification of thrust fronts, Am. Assoc. Petrol. Geol. Bull., v. 70, p. 12-25.

Morse, J. 1977, Deformation in ramp regions of overthrust faults in E. L. Heisey, D. E. Lawson, E. R. Norwood, P. H. Wach, and L. A. Hale, eds., Rocky Mountain thrust belt, geology and resources: Wyoming Geol. Assoc. Guidebook 29, p. 457-470.

Mugnier, J. L., and P. Vialon, 1986, Deformation and displacement of the Jura cover on its basement: Jour. Struc. Geol., v. 8, p. 373-388.

Murrell, S. A. F., 1981, The rock mechanics of thrust and nappe formation, in K. R. McClay, and N. J. Price, eds., Thrust and nappe tectonics, Geol. Soc. London Spec. Pub. 9,p. 99-109.

Ni, J., and M. Barazangi, 1984, Seismotectonics of the Himalayan collision zone: Geometry of the underthrusting Indian plate beneath the Himalaya, Jour. Geoph. Res., v. 89, p. 1147-1163.

Opdyke, N. D., N. M. Johnson, G. D. Johnson, E. H. Lindsay, and R. A. K. Tahirkheli, 1982, The paleomagnetism of the middle Siwalik formations of northern Pakistan, Paleogeo., Paleoclim., and Paleoecol., v. 37, p. 43-62.

Opdyke, N. D., E. Lindsay, G. D. Johnson, N. Johnson, R. A. K. Tahirk- heli, and M. A. Mirza, 1979, Magnetic polarity stratigraphy and vertebrate paleontology of the Upper Siwalik subgroup of northern Pakistan: Palaeogeog. Palaeoclim. Palaeoecol., v. 27, p. 1-34.

Pennock, N., in prep., Balanced structural cross-section of the eastern Potwar Plateau, Pakistan: Shortening and deformation at the edge of a salt basin, M. S. thesis, Oregon State University, Corvallis.

Pilgrim, G. E., 1910, Preliminary notes on a revised classification of the Tertiary freshwater deposits of India, Rec. Geol. Surv. India, v. 40, p. 185-205.

Pilgrim, G. E., 1913, The correlation of the Siwaliks with mammal horizons of Europe, Rec. Geol. Soc. India, v. 43, p. 264-326. 97

Pinfold, E. S., 1918, Notes on structure and stratigraphy in the North-west Punjab: Rec. Geol. Survey India, v. 49, p. 137-160.

Powell, C. M., 1979, A speculative tectonic history of Pakistan and surroundings: Some constraints from the Indian Ocean, in A. Farah and K. A. DeJong, eds., Geodynamics of Pakistan, Geological Survey of Pakistan, p. 5-24.

Price, R.A., 1981, The Cordilleran foreland thrust and fold belt in the southern Canadian Rocky Mountains, in K. R. McClay and N. J. Price, eds., Thrust and Nappe Tectonics, Geol. Soc. London, Black- well Scientific Pubs., p. 427-448.

Quittmeyer, R. C., and K. H. Jacob, 1979, Historical and modern seismicity of Pakistan, Afghanistan, northwestern India and southeast- ern Iran, Bull. Seismol. Soc. Amer., v. 69, p. 773-823.

Ramberg, H., 1974, Experimental and theoretical study of salt dome evolution, in A.H. Coogan, ed., Fourth Symposium on salt, Northern Ohio Geol. Soc., Cleveland, OH, v.1, p. 261-270.

Ramsay, J. G., 1969, The measurement of strain displacement in orogenic belts, in P. E. Kent, G. E. Satterthswaite, and A. M. Spencer, eds., Time and Place in Orogeny, Geol. Soc. London Spec. Pub. 3, p. 43-79.

Ramsay, J. G., M. Casey, and R. Kligfield, 1983, Role of shear in development of the Helvetic fold-thrust belt of Switzerland, Geology, v. 11, p. 439-442.

Rayleigh, C. B., and D. T. Griggs, 1963, Effect of the toe in the mechanics of overthrust faulting, Geol. Soc. Amer. Bull., v. 74, p. 819-830.

Raynolds, R. G., 1980, The Plio-Pleistocene structural and stratigraphic evolution of the eastern Potwar Plateau, Pakistan, Ph. D. thesis, Dartmouth College, Hanover, New Hampshire, 265 pp.

Raynolds, R. G., 1981, Did the ancestral Indus flow into the Ganges drainage?, Geol. Bull. Univ. Peshawar, v. 14, p. 141-150.

Raynolds, R. G., and G. D. Johnson, 1985, Rates of Neogene depositional and deformational processes, north-west Himalayan foredeep margin, Pakistan, in N. J. Snelling, ed., The Chronology of the Geological Record, Geol. Soc. London, Memoir No. 10, p. 297-311.

Rich, J. L., 1934, Mechanics of low-angle overthrust faulting as illustrated by the Cumberland thrust block, Virginia, Kentucky, Tennessee, Amer. Assoc. Petrol. Bull., v. 18, p. 1584-1596.

Rubey, W. W., and M. K. Hubbert, 1959, Role of fluid overpressure in mechanics of overthrust faulting, Geol. Soc. Amer. Bull., v. 70, 98

p. 167-206.

Sarwar, G., and K. A. DeJong, 1979, Arcs, oroclines, syntaxes: the curvatures of mountain belts in Pakistan, in Geodynamics of Pakistan, A. Farah and K. A. DeJong, eds., Geol. Surv. of Pakistan, Quetta, p. 341-349.

Shroder, J. F., in prep., Quaternary history of middle and upper Indus and surrounding valleys, northern Pakistan.

Seeber, L., and J. G. Armbruster, 1979, Seismicity of the Hazara Arc in northern Pakistan: decollement vs. basement faulting, in A. Farah and K. A. DeJong, eds., Geodynamics of Pakistan, Geological Survey of Pakistan, p. 131-142.

Seeber, L., J. G. Armbruster and R. C. Quitmeyer, 1981, Seismicity and continental subduction in the Himalayan Arc, in H. K. Gupta and F. M. Delany, eds., Zagros, Hindu Kush, Himalaya Geodynamic Evolu- tion, Amer. Geoph. Union, Geodynamics Series 3, p. 215-242.

Seeber, L., and K. H. Jacob, 1977, Microearthquake survey of northern Pakistan, preliminary results and tectonic implications, Proc. CNRS Symp. Geol. Ecology Himalayas, Paris, p. 347-360.

Serra, S., 1977, Styles of deformation in the ramp regions of overthrust faults, in E. L. Heisey, E. R. Norwood, P. H. Wach, and L. A. Hale, eds., Wyoming Geol. Assoc. 29th Ann. Field Conf. Guidebook, p. 487-498.

StOcklin, J., 1974, Possible ancient continental margin in Iran: in C. L. Drake, and C. A. Burk, eds., The Geology of Continental Margins, Springer-Velag, New York, p. 889-903.

Suppe, J., 1983, Geometry and kinematics of fault-bend folding, Amer. Jour. Sci., v. 283, p. 684-721.

Suppe, J., 1985, Principles of structural geology: Englewood Cliffs, N. J., Prentice-Hall, 537 p.

Swaminath, J., V. Venkatesh, and R. K. Sundaram, 1964, Role of Precambrian lineaments in the evolution of Cenozoic festoons of the Indian subcontinent, Internat. Geol. Cong. 22nd, New Delhi, 1964, sec. 11, p. 316-333.

Tapponnier, P., and P. Molnar, 1977, Active faulting and tectonics in China, Jour. Geophys. Res., v. 82, p. 2905-2943.

Tapponnier, P., G. Peltzer, A. Y. Le Dain, R. Armijo, and P. Cobbold, 1982, Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine, Geology, v. 10, p. 611-616.

Tauxe, L., and N. D. Opdyke, 1982, A time framework based on magnet- stratigraphy for Siwalik sediments of the Khaur area, northern 99

Pakistan, Palaeogeog., Palaeoclim., Palaeoecol., v. 37, p. 43-61.

Telford, W. M., L. P. Geldart, R. E. Sheriff, and D. A. Keys, 1976, Applied geophysics, Cambridge University Press, 860 p.

Thompson, R. I., 1981, The mature and significance of large "blind" thrusts within the northern Rocky mountains of Canada, in K. R. McClay, and N. J. Price, eds., Thrust and Nappe Tectonics, Geol. Soc. London Spec. Pub. 9, p. 449-462.

Vann, I. R., R. H. Graham, and A. B. Hayward, 1986, The structure of mountain fronts, Jour. Struc. Geol., v. 8, p. 215-228.

Voskresenskiy, I. A., 1978, Structure of the Potwar highlands and Salt Range as a indicator of horizontal movement of the Hindustan Platform, Geotectonics, v. 12, p. 70-75.

Wadia, D. N., 1928, The geology of the Poonch State (Kashmir) and adjacent portions of the Punjab, Mem. Geol. Survey of India, v. 51, p. 185-370.

Wadia, D. N., 1931, The syntaxis of the north-west Himalaya: its rocks, tectonics and orogeny, Rec. Geol. Surv. India, v. 65, p. 189-220.

Wells, N. A., 1984, Marine and continental sedimentation in the early Cenozoic Kohat Basin and adjacent northwestern Pakistan, Ph.D. dissertation, University of Michigan, Ann Arbor, 465 pp.

Williams, G., and Chapman, T., 1983, Strains developed in the hanging- walls of thrusts due to their slip/propagation rate: a dislocation model, Jour. Struc. Geol., v. 5, p. 563-571.

Wiltschko, D. V., 1979, A mechanical model for thrust sheet deformation at a ramp, Jour. Geophys. Res., v. 84, p. 1091-1104.

Wiltschko, D. V., and W. M. Chapple, 1977, Flow of weak rocks in Appa- lachain Plateau folds, Am. Assoc. Petrol. Geol. Bull., v. 61, p. 653-670.

Wiltschko, D., and D. Eastman, 1983, Role of basement warps and faults in localizing thrust fault ramps, in R. D. Hatcher, H. Williams, and I. Zietz, eds., Contributions to the Tectonics and Geophysics of Mountain Chains, Mem. 158, p. 177-190.

Windley, B. F., 1983, Metamorphism and tectonics of the Himalaya: Jour. Geol. Soc. Lon., v. 140, p. 849-865.

Woodward, N. B., S. E. Boyer, and J. Suppe, 1985, An outline of balanced cross-sections, notes for a short course on balanced sections sponsored by the Structural Geology and Tectonics Divi- sion of the Geol. Soc. America at annual meeting, Orlando, Fl., Univ. of Tenn. Dept. of Geol. Sci. Studies in Geology 11, 170 pp. 100

Yeats, R. S., in prep., Active faults related to folding.

Yeats, R. S., and A. Hussain, in press, Timing of structural events in the Himalayan foothills of northwestern Pakistan, in prep.

Yeats, R. S., S. H. Khan and A. Akhtar, 1984, Late Quaternary deformation of the Salt Range of Pakistan, Geol. Soc. Amer. Bull., v. 95, p. 958-966.

Yeats, R. S., and R. D. Lawrence, 1984, Tectonics of the Himalayan thrust belt in northern Pakistan, in B U. Haq, and J. D. Milliman, eds., Marine Geology and Oceanography of Arabian Sea and Coastal Pakistan, p. 177-198.

Yeats, R. S., T. Nakata, K.S. Valdiya, M. Pandey, A. Farah, M. A. Mirza, M. Fort, and R. Armijo, in prep., Active fault map of the Himalaya, 1:2,000,000, IGCP 206 Atlas of Worldwide Active Faults.

Zharkov, M. A, 1981, History of Paleozoic salt accumulation: Berlin, Springer-Velag, 308 p. APPENDICES 101

APPENDIX A

DEVELOPMENT OF THE HIMALAYAN FRONTAL THRUST ZONE:

SALT RANGE, PAKISTAN

Dan M. Baker, Robert J. Lillie, and Robert S. Yeats Department of Geology Oregon State University Corvallis, Oregon, 97331

Gary D. Johnson Department of Earth Sciences Dartmouth College Hanover, New Hampshire, 03755

Mohammad Yousuf and Agha Sher Hamid Zamin Oil and Gas Development Corporation 13-N, Markaz F/7 Islamabad, Pakistan

CONTENTS Abstract 102 Introduction 102 Regional Setting 102 Structure of the Salt Range/Potwar Plateau 104 Timing of Ramping 108 Discussion 111 Ramping 111 Calculation of Shortening Rate 111 Conclusions 114 Acknowledgements 115

Submitted to Geology April 10, 1987 102

ABSTRACT

The Salt Range is the active frontal thrust zone of the Himalaya in Pakistan. Seismic reflection data show that a1 km offset of the basement acted as a buttress that caused the central Salt Range/Potwar Plateau thrust sheet to ramp to the surface, exposing Mesozoic and Paleozoic strata at the range front. The frontal portion of the thrust sheet was folded passively as it overrode the subthrust surface on a ductile layer of salt. The hindward portion of the thrust sheet is undeformed, except for minor folding. Lack of internal deformation is due to decoupling of sediments from the basement along this Eocambrian salt layer. The timing of ramping is constrained by stratigraphic evidence, paleomagnetic dating of unconformities and sediment accumulation rates. It is estimated that the thrust sheet overrode the basement offset from 2.1 to 1.6 mya. Cross-section balancing gives a minimum of 19 to 23 km of shortening across the ramp. The rate of Himalayan convergence which can be attributed to underthrusting of Indian basement beneath sediments in the Pakistan foreland is therefore 9-14 mm/yr.

INTRODUCTION

Approximately 3000 km of seismic reflection profiles (supplied by the Pakistan Oil and Gas Development Corporation and Ministry of Petroleum), along with surface maps and well data, have been used to construct balanced cross-sections of the foreland fold-thrust belt of the Himalaya in Pakistan. The Salt Range, the active thrust front, provides a modern analogue to extinct thrust fronts and, to a lesser degree, internal regions of fold-thrust belts. Recent sedimentological and stratigraphic studies in the Potwar Plateau/Salt Range have used magnetostratigraphy and tephrochronology to temporally constrain horizons in the Siwalik molasse (e.g. Johnson et al., 1986). These data constrain the timing of deformation and the progression of the deformation front across the foreland basin (Raynolds and Johnson, 1985). The rate of shortening at the front of the fold-thrust belt can be determined by using this timing information along with estimates of contraction of the sedimentary cover based on the balanced cross-sections. Ramping of the Salt Range thrust sheet represents a discrete event from which timing and contraction can be determined. The resulting shortening rate can then be used to infer the portion of Indian-Asian plate convergence attributable to underthrusting of the Indian basement beneath sediments at the Salt Range front.

REGIONAL SETTING

The southern margin of the Himalayan collision zone in Pakistan is the Salt Range and Kohat/Potwar Plateau, an active foreland fold-thrust belt formed in response to the underthrusting of cratonic India beneath its own Phanerozoic sedimentary cover (fig. Al). At the Hazara-Kashmir Syntaxis, the fold-thrust belt changes in trend and broadens in 103

Fig. Al: Position of the Salt Range Thrust front and Potwar Plateau (enclosed in box) within the overall Himalayan trend. HKS=Hazara- Kashmir Syntaxis; ISZ=Indus Suture Zone; MBT=Main Boundary Thrust; MFT=Main Frontal Thrust (active); MKT=Main Karakorum Thrust; SH= Sargodha High. Rectangular box denotes area of figure 2. 104

Pakistan. In India west of the Punjab reentrant, there is strong coupling between sediments and basement; consequently the zone of underthrusting is narrow and has a high angle of cross-sectional taper (Davis et al., 1983; Davis and Engelder, 1985). In Pakistan, on the west flank of the Hazara-Kashmir Syntaxis, low-strength evaporites of late Precambrian to Early Cambrian age constitute the zone of decollement (Crawford, 1974). As a result and in contrast to India, the zone of underthrusting extends far out over the foreland (Seeber and Armbruster, 1979), more than 100 km south of the Main Boundary Thrust, and it has a narrow cross-sectional taper. Further evidence for the decoupling properties of the evaporites lies in the large detachment earthquakes which occur beneath the foreland in India, but not in Pakistan (Quittmeyer and Jacob, 1979; Seeber et al., 1981). In northern Pakistan, a southward progression of major deformation has been documented (Raynolds and Johnson, 1985; Johnson et al., 1986; Burbank et al., 1986). At the southern margin of the Peshawar basin, major thrusting occurred prior to 2.8 mya (Burbank, 1982). Farther south, in the northern Potwar Plateau, major deformation ended by 1.9 mya (Raynolds and Johnson, 1985). In the eastern Salt Range, major deformation may be younger than 0.4 mya (Johnson et al., 1979). The deformation front in the central Salt Range may now be shifting 15 km south to the Lilla structure, a thrust-cored anticline in the Jhelum River plain (fig. A2). In India, Acharyya and Ray (1982) showed that progression of the thrust front is accompanied by southward migration of the Siwalik molasse depositional axis. Likewise, Raynolds and Johnson (1985) estimate that the depocenter in Pakistan is migrating southward at 30 mm/yr, and is currently located just south of the Salt Range, along the Jhelum River. South of the fold-thrust belt, sediments slope gently northward from the Sargodha High, an area of moderate seismicity including exposures of Precambrian rocks (Farah et al., 1977). This ridge is parallel to the overall northwest-southeast Himalayan trend and may be related to downflexing of the Indian plate (Seeber et al., 1981; Yeats and Lawrence, 1984; Duroy et al., 1987).

STRUCTURE OF THE SALT RANGE/POTWAR PLATEAU

At the Salt Range front, Eocambrian evaporites and overlying strata override synorogenic fan material and alluvium (Yeats et al., 1984). The central Salt Range is strongly emergent, located between a weakly emergent thrust front at the Surghar Range (J. McDougall, in prep) and a buried thrust front in the easternmost Salt Range (terminology of Morley, 1986). The Salt Range lies approximately 80 km outboard of thrusting in the northern Potwar Plateau; the intervening Soan trough is relatively undeformed (fig. A3). Seismic-reflection profiles in the region (Khan et al., 1986; Lillie et al., in press) show that the Salt Range and southern Potwar Plateau constitute a large slab that is being thrust over the foreland with very little internal deformation. The northern flank of the Salt Range is an eroded monocline that flattens northward into the southern limb of the Soan trough; the monocline is the surface expression of the footwall ramp. Reflection profiles show that lower Siwalik sediments exposed at the surface overlie a fault with a basement offset of one kilometer, down-to-the 105

Fig. A2: Geologic map (after Gee, 1980) showing locations of structural features mentioned in text. HR=Hill Ranges; KF=Kalabagh Fault; L=Lilla Anticline; NPDZ=Northern Potwar Deformed Zone; PH=Pabbi Hills Anticline; RF=Riwat Fault; SRFT=Salt Range Frontal Thrust. Cities: I=Islamabad; R=Rawalpindi. Exploration wells: D=Dhariala; L=Lilla. Cross-section traces: A-A' (fig. A3); C-C' (fig. A5), seismic line: B-B' (fig. A4). Magnetostratigraphic sample locations: BN=Baun; KK=Kotal Kund. Solid lines represent structures mapped from Landsat photographs, dashed lines are less prominent features, and dotted lines denote buried structures. 72° 73°

ALLUVIUM, SIWALIK,and CAMBRIAN to CRETACEOUS t RAWALPINDI MOLASSE 0 50km * MAGNETOSTRATIGRAPHIC SITES

111111111 PALEOCENE and EOCENE EOCAMBRIAN SALT RANGE FM. -EXPLORATION WELLS Figure A2 N. SOAN S. SALT POTWAR SYN. POTWAR RANGE C

/1 //\ 11-/ H--

10 I 10

20 TERTIARY MOLASSE H 2 FrACAMBRIAN TO' EOCENE EOCAMBRIAN 30 PRECAMBRIAN BASEMENT ,10 KM V.E. 1.1 40 MOHO

50 100 80 60 40 20 0 -20 DISTANCE FROM CENTER OF SALT RANGE (KM)

Fig. A3: Schematic cross-section of the Himalayan foreland fold-thrust belt in Pakistan (see Lillie et al., in press, for seismic profiles used to construct section). See figure 1for location. 108

north (Lillie and Yousuf, 1986; fig. A4). This ramp and the underlying basement offset correlate with a local steepening of the Bouguer gravity gradient (Farah et al., 1977). The basement offset acted as a buttress which controlled ramping and emplacement of the Salt Range thrust sheet. Passive folding, as the thrust sheet cut upsection, then flattened, is primarily responsible for the broad anticlinal structure of the central Salt Range. The Salt Range, like most anticlines of the Potwar Plateau, is cored by salt which appears to have flowed from beneath synclines (cf. Davis and Engelder, 1985; Fox, 1983), as suggested by the presence of over 2 km of salt at the bottom of the Dhariala well (Gee, 1983; fig. A2). In order to control ramping, the basement offset must have been present prior to the encroachment of the thrust front. Hypotheses for the origin of the down-to-the north basement fault include: (a) extension related to Eocambrian rifting that may have accompanied the formation of evaporite basins on the northwestern margin of the Indian subcontinent, or Mesozoic rifting apart of Gondwana; (b) reverse faulting related to either the latest Cretaceous through Paleocene convergence in the Attock-Cherat Range or the present compressional regime (Yeats and Hussain, in press); or (c) Neogene normal faulting related to flexure as the crust is loaded by thrust sheets from the north (Lillie and Yousuf, 1986; Duroy et al., 1987). Stratigraphic relations suggest that hypothesis (c) is the most likely. The right-lateral Kalabagh tear fault, terminates the Salt Range to the west (Yeats and Lawrence, 1984). In contrast, the eastern termination of the Salt Range is divided into several fault blocks bounded by forward- and hindward-verging thrusts (Johnson et al., 1986). Farther to the north, in the eastern Potwar Plateau, folds and thrusts are aligned en echelon in a SSW-NNE direction. The increase in deformation from the central to the eastern Potwar Plateau may be the result of differential drag at the edge of the salt basin caused by the contrast in basal friction as the detachmentzone salt facies thins eastward (Davis and Engelder, 1985) and as the dip on the basement decreases (Jaume and Lillie, in press).

TIMING OF RAMPING

The timing of thrusting is often bracketed by the youngest rocks involved in deformation and the oldest sediments overlapping these deformed strata. Yeats and others (1984) documented that the Salt Range thrust is young and possibly still active; the initiation of ramping, however, is more difficult to constrain. In the central Salt Range at the Baun locality (Fig. A2), the top of the Dhok Pathan formation, dated as 5.1 mya or slightly younger (Johnson et al., 1982), is overlain disconformably by Upper Siwalik sediments with a conglomerate layer at its base. Dhok Pathan beds are tilted northward and truncated above the footwall ramp (fig. A4), indicating ramping began later than 5.1 Ma. In addition, the Upper Siwaliks are essentially unrotated, while Dhok Pathan beds are rotated 35° counterclockwize (Opdyke et al., 1982). An Upper Siwalik conglomerate bed, deposited about 4.5 mya at the Baun locality, contains clasts of Eocene carbonates and recycled Precambrian clasts from the Permian Tobra Formation (Johnson et al., I 11111111i II II Ill II 11111 1111111111111111111111111111111111111 110

1986; Burbank and Raynolds, in press). These clasts are important stratigraphic evidence constraining the timing of either ramping or basement faulting. If the source area of Upper Siwalik clasts is the Salt Range, then conglomerate deposition in combination with diminished sedimentation rates and differential rotation may indicate incipient Salt Range thrusting by 4.5 mya (Johnson et al., 1986; Burbank and Raynolds, in press). Alternatively, Eocene and Permian clasts are interpreted as derived from the south, from the up-thrown basement fault block. Diminished sedimentation rates, erosion and basement faulting may have occurred when this portion of the crust was at the crest or along the northern flank of the peripheral bulge (now underneath the Sargodha High; fig. Al). This interpretation is supported by the Shell #1 Lilla well (fig. A2), which penetrated Cambrian rocks directly beneath Siwaliks. Erosion of intervening strata, including Eocene carbonates and the Lower Permian Tobra Formation, is therefore demonstrated south of the basement offset. If Upper Siwalik conglomerates preserved at Baun were derived from the up-thrown basement fault block, then the down-to-the north, basement offset (figs. A3, A4) was formed or enhanced by Neogene faulting related to flexure, as suggested by Lillie and Yousuf (1986; see also Duroy et al., 1987). The age of thrust ramp initiation in the central Salt Range appears poorly constrained by stratigraphic evidence alone. Initiation of ramping predates deposition of the Potwar Silt, which is believed to have been ponded behind the rising Salt Range.Raynolds (1980) found only normal magnetization in the Potwar Silt, indicating a Brunhes age of less than 0.7 Ma, while Shroder (in press) cites evidence suggesting that the silt is only as old as 170,000 yr. To better resolve the onset of ramping, information on the timing of specific events has been compiled from the eastern margin of the Potwar Plateau/Salt Range, where deformation is more evenly distributed, rather than concentrated at the frontal thrust (Johnson et al., 1986; Burbank et al., 1986). As a result, a more complete record of sedimentation is exposed there, where the progressive south-southeastward onset of deformation can be traced from the Riwat fault to the Pabbi Hills anticline (Johnson et al., 1986; fig. A2). The age of initiation of folding within the central portion of the Soan trough is poorly constrained. In the ramp region to the south, initial deformation may have been expressed between 3.0 and 2.2 mya, by a change in sediment accumulation throughout the eastern Potwar Plateau. A major structural event then occurred as the Salt Range thrust sheet overrode the basement offset between 2.1-1.6 mya. The timing of this event is suggested by the abrupt termination of folding in the northern Potwar Plataeu at about 1.9 mya, termination of sedimentation over folds in the eastern Potwar, cessation of sedimentation in the Kotal Kund region at 1.6 mya (Johnson et al., in press) and accelerated rotation of the Soan syncline after 2 Ma (Burbank et al., in press). Initiation of ramping between 2.1 and 1.6 mya also fits the chronology of the southeastward-propagating deformation front when the position of the ramp is projected along strike to the east.At this time, there was a transfer of strain release from the northern Potwar Plataeu to the Salt Range ramp, implying the Soan trough/Salt Range region began to move as one coherent thrust sheet. 111

DISCUSSION

RAMPING: In Fig. A5, the Salt Range is interpreted as a hanging-wall ramp anticline. The shape of the central Salt Range differs from a typical fault-bend fold due to its lack of a frontal, forward dipping limb and the presence of thickened salt at the base of the thrust sheet. A frontal limb forms as a thrust flattens after cutting up section (Suppe, 1983). Therefore, a plausible explanation is that the Salt Range thrust does not flatten substantially, and has a relatively smooth trajectory. The gentle slope to the surface from the basement offset is consistent with the thrust cutting upsection through relatively incompetent molasse, and the thrust sheet climbing over its own debris. Loading by the thick molasse sequence within the Soan trough is believed to cause ductile flow of salt toward the south (Gee, 1983). The basement fault acting as a buttress may have localized the growth of an elongate salt pillow or dome as the overlying competent strata buckled against it. Noting Wiltschko and Eastman's (1983) work on potential thrust fault orientations associated with basement faults, it is likely that the normal fault plane would be reactivated during compression in strata overlying the basement (fig. A5). As salt growth continued, Cambrian and younger strata on the north side of the fault would be lifted until they could be pushed across the basement offset. During this later process, the steep initial ramp angle would be abandoned in favor of a shallower trajectory in conjunction with salt thickening in the lower ramp region of the upper plate (eg. mode 2 of Serra, 1977). This mechanism explains the lack of footwall deformation normally associated with smooth thrust trajectories (see Cooper and Trayner, 1986) and thickened salt in the allochthon (e.g. Dhariala well; Gee, 1983), while avoiding an unreasonably sharp initial interlimb ramp angle. Therefore, the southern flank of the central Salt Range is not only an eroded thrust fault scarp (Yeats et al., 1984), but may also be close to the preserved portion of the downthrown normal fault block (fig. A5). Due to its limited lateral extent (fig. A5), the basement offset does not satisfactorily explain the emplacement of the entire Salt Range. A thrust fault surface will propagate along strike into areas where there is no other mechanical cause for failure (Wiltschko and Eastman, 1983). In the western Salt Range, ramping was probably facilitated by a basement warp interpreted by Leathers (1987) as the northeast flank of the Sargodha basement ridge converging with the overthrust belt. In the eastern Salt Range, a southward thinning of the detachment zone salt facies may have aided thrust ramping. Therefore, it appears that a combination of ramping mechanisms was responsible for the emplacement of the entire Salt Range.

CALCULATION OF SHORTENING RATE: In the western Himalaya at longitude 73° east, the northward motion of the Indian plate relative to the Asian plate is calculated as 42 mm/yr (Minster and Jordan, 1978). Convergence is taken up by: (1) shortening of sedimentary strata in the foreland as India underthrusts its cover rocks; (2) contraction within the collision zone via uplift in the high Himalaya and reactivation of 112

Fig. A5: Balanced cross-section and restorations demonstratinga possible mode of origin of the Salt Range. Formation contacts are dashed above upward limit of preserved strata.Erosion occurred during Salt Range thrusting. (a) Stratigraphic sequence, at the end of platform deposition and prior to molasse deposition (approx. 35 mya). The thinner Cambrian to Eocene sequence south of future basement offset is interpreted to result from erosion during subsequent basement faulting. (b) restoration after bulk of molasse deposition, prior to compressional deformation (approx. 2 to 5 mya). Normal fault formed as Indian lithosphere flexed downward due to thrust emplacement to the north. (c) Present configuration. The thrust sheet encountered the basement offset and ramped above a thickening layer of salt. Salt has flowed into this section from the north. Molasse removed from the autochthon during thrusting is not shown. Dashed detachment surface is schematic. Vertical faults were mapped as "normal or probably normal" by Gee (1980). I I I c' \/.\ / I I // / / 7. \ / \ \I \ / I ' I k \ ". I I - . /, / I \ /11 / /, \I/ I//11, / \ '-1- \11/1 1-; /-

/\,\A1/1\/\"\'\-/\JI;7\2/\'\,\\,/,)I /\\1\\__A 1 \\/ ' \ /\\ \/ I-I\- /\ /, \,i\/\/\ i_\ /. \ / \ I \ I 1, ,i\i \ I /, s, 7,\1\1 , /\)\; \--71\ /,\ / \I \/\--i'/--- \/- I/-\//1 \ _\ I / ./1\ 1. _\1 1-- / \/0/(\/ I \I; \I BI

00 01 o001 POST SIWALIK SIWALIK TERTIARY MOLASSE RAWALPINDI ..... I 1i I . i CAMBRIAN TO EOCENE o , i I EOCAMBRIAN t I

I"",' PRECAMBRIAN BASEMENT 0 3.4 km I 9 kmyolasse 10 km Molasse

V.E. 1:1 / I\/1 \--N/TrIF\--117-(1'/\13-1-1 / \ A \---1/N v_i/N/0/11\A-- Irr / 7\- / I r1C,)/1/\\,(- -I / -1/:1! ZI/ //i- 1 _ii/1\ k I I_ I 7-, /\ _ \ \

FigureA5 114

thrust faults; and (3) crustal shortening via thickening of the crust and escape block tectonics along strike-slip faults to the north of the collision zone (Molnar and Tapponnier, 1975; Tapponnier and Molnar, 1977). Balanced sections through the central Salt Range suggest a minimum of 19 to 23 km of horizontal shortening of the competent Cambrian through Eocene section (e.g. fig. A5). The calculated shortening is a minimum because erosion from the Salt Range front has not been accounted for. The apparent shortening is less across reentrants where erosion and deposition are greater (Baker, in prep). Therefore, the true amount of shortening across the frontal ramp is thought to be closer to 23 km. Additional factors such as a reinterpretation of the orientations of high-angle faults (see fig. A5) only have a minor effect on the measured amount of shortening. Allowing for 1 km of error, 18-24 km is used as the range of shortening values. Assuming the thrust sheet overrode the ramp between 2.1 and 1.6 mya, the minimum rate of shortening between the Indian basement and the Phanerozoic sedimentary cover is between 9 and 14 mm/yr. These rates are comparable to the 10-15 mm/yr convergence absorbed by underthrusting at the Himalaya as estimated by Lyon-Caen and Molnar (1985) for northern India. Our balanced cross-section extended to the north indicates 20% shortening for a 90 km length of the fold-thrust belt between the Lilla structure and intense deformation in the northern Potwar Plataeu. This is a relatively low degree of shortening compared to values obtained in other fold-thrust belts around the world, probably because the thrust front only recently jumped the large distance from the northern Potwar Plateau to the Salt Range. The calculated shortening rate represents about 20-35% of the convergence rate between the Indian and Asian plates. Therefore, most convergence is absorbed north of the frontal thrust zone by contraction within the collision zone and escape-block tectonics. The calculated rate and percentage of shortening should be compared to the Indian portion of the Himalayan foreland fold-thrust belt, where a salt detachment zone is lacking. A comparison of results could be used to evaluate the decoupling effects of salt and the origin of the Hazara-Kashmir Syntaxis.

CONCLUSIONS

Surface geology, reflection, and drilling data provide insight into the structure, timing, and mode of evolution of the Salt Range. A pre-existing, down-to-the north basement fault controlled ramping of the encroaching thrust front.The buttressing effect of the basement offset apparently caused a local buildup of salt; the Salt Range thrust sheet subsequently overrode the ramp as competent strata were passively folded into the present anticlinal form. The thrust sheet north of the ramp is only mildly deformed due to the decoupling properties of salt.

Changing sediment accumulation rates used in conjunction with stratigraphic evidence suggest that Salt Range ramping occurred about 2 mya. Balanced cross-sections suggest that a minimum of 19 to 23 km of shortening occurred across the ramp at a shortening rate of 9-14 mm/yr. 115

Therefore, frontal thrusting in the Pakistan foreland accounts for part of but not the majority of the total plate convergence.

ACKNOWLEDGEMENTS

The Oregon State University project in northern Pakistan is supported by NSF grants INT-8118403, EAR-8318194, INT-8609914, and EAR-8608224. NSF grants INT-8517353 and EAR-8407052 support ongoing studies by Dartmouth College and Peshawar University. Additional support for the construction of balanced cross sections was provided by Chevron International, CONOCO Inc. and Amoco Production Company. We are grateful to the Oil and Gas Development Corporation and the Ministry of Petroleum and Natural Resources of Pakistan for the use of reflection and gravity data discussed in this study. The work could not have been undertaken without scientific collaboration from individuals within O.G.D.C., the Geological Survey of Pakistan, and the Centre of Excellence in Geology, University of Peshawar. Special appreciation is extended to R. A. K. Tahirkheli. 116

APPENDIX B

RESTORATION RESULTS AND IMPLICATIONS

Paleobasement dips (i.e. flexure), the original geometry of salt deposits, timing of deformational events, and the age of molasse north of the Salt Range were all used as constraints on restored sections. The great amount of salt deposited north of the basement offset (compared to relatively minor amounts to the south), in combination with the independent balancing assumption (see methods), helped to reduce the retrodeformation of salt to two models. Here, "salt" is used to represent all strata above the basement and beneath prominent platform reflectors. The first model, deemed the tabular salt model (fig. 21a) consists of redistributing salt into two rectangular shaped blocks, south and north of the basement offset (fig. 13). By summing the salt area and then dividing by the length over which it must be redistri- buted, the thickness of the salt blocks were found to be about 0.3 and 1.3 kms, respectively. In the model, the assumption of a discontinuity in thickness across the basement offset implies Eocambrian basement faulting (fig. 21a). The second model represents the retrodeformation of salt to a wedge shape (fig. 13). The shape of the wedge was found by assuming no change in thickness across the basement offset, and by using the thickness of autochthonous salt south of foreland deformation (i.e. south of the Lilla anticline, fig. 5) as a starting point. Knowing this thickness, the length of the wedge, and the total salt area, a value of constant taper was obtained. This taper was found to be 0.20°, resulting in an initial salt thickness of 0.33 km at the basement offset. Similarly, using 0.33 km as the thickness of salt at the south end of the second wedge, a taper of 0.93° was found. At this taper the salt- wedge thickens to 3.0 km at the northern end of the restoration. The second model requires a change in taper but not thickness at the basement offset. The restored basement dip also changes at this point. Because Paleogene strata across the Potwar Plateau are primarily shallow marine limestone and shale the top surface at the end of Eocene deposition is assumed to be horizontal. The absence of Paleogene strata south of the basement offset is assumed to be due to post depositional erosion (fig. 22b), because widespread Paleogene deposition is indicated (Khan et al., 1986). This assumption, combined with salt redistribution models, resulted in an increase in basement dip north of the basement offset at the end of Eocene (fig. 13). Basement dip is therefore not entirely due to Neogene flexure during Himalayan orogen- esis, a consideration that is consistent with a possible miogeoclinal setting of the northwestern Indian margin during the Paleozoic and Mesozoic. The assumption of a southward topographic slope toward the basin depocenter during molasse deposition was derived by comparison to present day Himalayan foreland basin topography. South of the foreland basin axis, a reversal in depositional slope is expected where rivers drain northward from the Indian shield. Because the thickness of Siwaliks for restoration are fairly well known from seismic and well control, the requirement of southward topographic slope supplies a 117 valuable constraint on the degree of post 4.5 m.y. flexure. To attain a southward topographic slope upon restoration of the tabular salt model, an essentially unflexed basement was required, similar to the basement structure which existed at the end of Eocene time. In addition, a substantial amount of unpreserved Upper Siwaliks in the Potwar Plateau would be required to compensate for potential southward salt flowage. For the salt-wedge model (fig. 13), the proper direction of topographic slope is attained unless the basement is completely flexed into its present shape. Potential southward salt flowage could be accommodated by unflexing the basement by the necessary amount to retain the propper direction of surface slope. A special problem existed in attempting to restore molasse south of the basement offset because the age of these sediments are unknown. However, a valuable constraint was supplied because northward transport of conglomerate clasts into the Baun syncline is interpreted at about the same time as the 4.5 mya restoration. As a result, the restorations, after northward retrodeformation of the Baun syncline, required 1.7 km of molasse above the up-thrown basement block for this area to be topographically higher than the Baun syncline (fig. 13). Two restoration models (i.e. tabular salt and salt-wedge models) encompassing: (1) the age of basement faulting, (2) redistribution of salt, (3) timing of flexure, and (4) ramping style, have been considered. In the tabular salt model, Upper Siwalik conglomerate deposition at the Baun site would have to be from an uplifted footwall source (see fig. 22a; Burbank and Raynolds, in press), because the basement is not yet flexed, and therefore autochthonous Paleozoics should not be exposed. The wedge shaped salt model requires southward directed salt flow prior to ramping, in order to emplace the thick salt section now located beneath the Salt Range. Unlike the previous model a plane of weakness in platform strata would be available for reactivation during salt growth and subsequent ramping. Finally, instead of an allochthonous source area, a source terrane of Tobra Formation clasts along the flexural bulge is reasonable in accordance with basement flexure prior to 4.5 mya. 118

APPENDIX C

FACTORS INFLUENCING SHORTENING CALCULATIONS

In the central Salt Range, erosion is thought to be the biggest cause of varying shortening values. Rivers debauch from mountain fronts at reentrants, increasing the potential for erosion. Smaller shortening values were obtained where transects cross the range front within the 30 km wide re-entrant in the central Salt Range (figs. 5 and 16). A greater amount of range front retreat within this reentrant would explain differential shortening values. The smallest amount of shortening was measured along one of the easternmost sections which crosses the range front at a gorge where erosion is concentrated via funneling of runoff and salt diapirism. Since shortening at the toe of the central portion of the Salt Range thrust sheet increases eastwards, erosion resulting in an equal distance of range front retreat would remove strata representing a greater amount of shortening compared to areas to the west (fig. 17). The central Salt Range frontal thrust is not exposed except near Kattha Masral, due to burial by recent sediments. Burial is exemplified by topographic profiles across the range front (figs. 13 and 17), showing wedges of fanglomerates emanating from the Salt Range. Rivers discharge their loads at reentrants since they represent the shortest dispersal path between the hinterland and the foreland basin (Eisbacher et al., 1974). The effects of burial parallel those of erosion. Maxi- mum burial should be at reentrants and gorges, where sediments are debouched. Neglecting burial also minimizes shortening values. Cross-sections crossing tear faults may not balance (Woodward and Boyer, 1985). Balancing assumes that reversing fault displacements will fit pieces back together that were once adjacent. However, strike-slip displacements move pieces in and out of the plane of section. The problem occurs where zones of different shortening accommodation are juxtaposed by strike-slip movements. Whether shortening is over or underestimated depends on the original distribution of internal shortening. In part, these factors are minimized since Salt Range strike-slip faults apparently developed prior to or during ramping and associated internal deformation (fig. 19; e.g. Dubey, 1979). Close inspection of figure 16 reveals that potential juxtaposition of different regimes of shortening should not significantly effect shortening values. A potentially significant factor is the assumption of matching footwall and hanging-wall ramps. Slight hanging-wall ramp steepening was used (fig. 13 and 17) to bring sections into balance. Ramp shapes may not match due to folding prior to or during ramping, resulting in a smooth rather than a ramp-flat fault trajectory (Cooper and Trayner, 1986). The interpreted ramping style, which utilizes a plane of weakness, suggests that ramps should approximately match, unlike fault- propagation folds in the eastern SR/PP. Restoration of the hanging- wall ramp could have a large effect on shortening. The position of the ramp cutoff was selected so as to minimize shortening values. The assumption of approximately matching ramp shapes together with line length balancing helped to constrain the shapes of both hanging-wall 119 and footwall ramps. Folds cannot simply be unrolled above areas of exposed ductile flow; overlying beds may have undergone elongation (Ramsay, 1969). Therefore, unpreserved extensional features above salt uplifts would result in an overestimation of shortening. This is not thought to be a major factor since cross-section traces do not cross wide outcrops of salt and the largest shortening values were obtained along transects which crossed little salt (fig. 17). The orientations of high angle faults are generally not discernable on seismic lines. There are two major considerations when interpreting fault orientations during balancing. First, a predominance of normal faults decreases measured shortening while reverse faults would increase calculated shortening. This influence decreases with increasing fault dip. To minimize shortening most of the high angle faults on the frontal balanced sections were interpreted as normal faults (fig. 17). In addition, fault orientation control on shortening was minimized by maximizing fault dips within reason.The second factor con- cerns the cross-sectional profile of faults. Fault plane changes are a major tool for bed length adjustments. Reorientation of straight line fault plane profiles does not change relative bed lengths. However, curved fault plane profiles (e.g. listric normal faults) do change relative bed lengths. For example, a downward concave reverse fault lengthens upper relative to lower beds while an upward concave reverse fault has the opposite effect. This principle was used in balancing frontal sections (fig. 17c). In this way, bed length balancing can constrain the orientations of major faults. Failure to restore clipped footwall strata will throw a section out of balance and decrease measured shortening. For this reason, horses of clipped footwall strata are tacked on to the range front, even though they may lie beneath the Salt Range (fig. 17). 120

APPENDIX D

PETROLEUM GEOLOGY

Petroleum exploration in the Pakistan foreland is in its first phase; only large structural traps have been tested (see Khan et al., 1986). Statistics (Baker et al., 1984) suggest that great reserves remain in untested structural and stratigraphic traps. For example, truncated Paleozoic and Mesozoic strata beneath the pre-Tertiary unconformity across the Potwar Plateau represent potential stratigraphic traps. The recent discovery of the Dhurnal field beneath an upper-level detachment, and its structural similarity to the Canadian foothills indicate that large reserves may remain undiscovered in the NPDZ. The lack of success farther south (e.g. Lilla well; fig. 3) points to the problem of sufficient petroleum migration into young structural traps. For these reasons, the timing of deformation throughout the foreland as well as the structural style in the NPDZ are important priorities.