The Siwaliks of western Nepal

Journal of Asian Earth Sciences 17 (1999) 643±657

The Siwaliks of western Nepal II. Mechanics of the thrust wedge J.L. Mugnier a,*, P. Leturmy a, P. Huyghe a, E. Chalaron b a

Laboratoire de Geodynamique des Chaines Alpines et UPRESA CNRS 5025, rue Maurice Gignoux, 38031, Grenoble, France b Institute of Geological and Nuclear Sciences, New Zealand

Abstract Comparison between numerical models and structural data is used for a better understanding of the evolution of the Siwalik thrust belt of western Nepal. The numerical model involves discontinuities within a critical wedge model, a kinematic forward model of serial cross sections, and a linear di€usion algorithm to simulate erosion and sedimentation. In western Nepal, large Piggy-back basins (Duns) are located above thick thrust sheets that involve more than 5500 m of the Neogene Siwalik Group, whereas Piggy-back basin sedimentation is less developed above thinner thrust sheets (4300 m thick). Numerical model results suggest that thrust sheet thickness and extension of wedge-top basins are both related to an increase of the basal deÂcollement dip beneath the duns. The West Dang Transfer zone (WDTZ) is a N±NE trending tectonic lineament that limits the westward extent of the large Piggy-back basins of mid-western Nepal and is linked to a thickening of the Himalayan wedge eastward. The WDTZ also a€ects the seismotectonics pattern, the geometry of the thrust front, the lateral extent of Lesser Himalayan thrust sheets, and the subsidence of the foreland basin during middle Siwalik sedimentation. Numerical models suggest that the individualisation of the Piggy-back basins at the transition between the middle Siwalik and upper Siwaliks followed the deposition of the middle Siwaliks that induced a geometry of the foreland basin close to the critical taper. As WDTZ induces an E±W thickning of the Himalayan wedge, it could also induce a northward shift of the leading edge of the ductile deformation above the basal detachment in Greater Himalayas of far-western Nepal. Field data locally suggest episodic out-o€-sequence thrusting in the frontal thrust belt of western Nepal, whereas numerical results suggests that episodic out-o€ sequence reactivation could be a general characteristic of the Himalayan wedge evolution often hidden by erosion. # 1999 Elsevier Science Ltd. All rights reserved.

1. Introduction Recent progress in the study of orogenic thrust belts has arisen in: (a) mechanics, from the development of the critical taper model (Davis et al., 1983; Dahlen and Suppe, 1984); (b) kinematics, from the use of balancing procedures (Dahlstrom, 1969) and forward kinematic models (Endignoux and Mugnier, 1990); and (c) syntectonic basin development, from erosion and sedimentation modelling (Flemings and Jordan, 1989). A computer based approach, which incorporates these three aspects is presented and applied to the Neogene

Siwalik Group of western Nepal. Comparison of its predictions with structural ®eld data presented in a companion paper (Mugnier et al., 1999) helps to improve the understanding of the kinematic history of thrusting and the relationships between development of the intramountain basins (duns) and the thrust system geometry. Lesser Himalayan tectonics is also discussed in the light of the implications of the wedge model. 2. Mechanics of the thrust wedge growth 2.1. Coulomb-wedge theory

* Corresponding author. E-mail address: [email protected] (J.L. Mugnier).

The mechanics of fold and thrust belts has been

1367-9120/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 1 3 6 7 - 9 1 2 0 ( 9 9 ) 0 0 0 3 9 - 5

644

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

2.2. Equilibrium of any thrust wedge Any portion of a thrust wedge, de®ned by geometrical and mechanical parameters, can be compared to the critically tapered wedge solution. To determine the ratio of shear and normal stresses along the basal plane, the critically tapered wedge equation is transformed into a ``frictional function'' (Chalaron et al., 1995): mbc …a, b, H, l, S0 , f† ˆ

Fig. 1. (a) Parameters of the Coulomb wedge (see text for the de®nition). (b) E€ects of the conditions applied along the boundaries of a Coulomb wedge. The general pattern of the stable domain is represented in an a±b graph, but is not scaled since it depends upon all the mechanical parameters. The arrows re¯ect, respectively, the following evolution: -1a- thickening increases the surface slope and transforms an undercritical wedge into a stable wedge; -1b- tilting of the whole wedge decreases the surface slope and increases the basal dip by the same value; -2a- basement tilt increases the basal slope whereas continental sedimentation maintains the surface slope horizontal; these two phenomena change an undercritical wedge into a stable wedge; -2b- development of local relief above the frontal ramp momentarily induces a surface slope that dips hinterlandward and locks the displacement along the basal deÂcollement; -3a- erosion decreases the surface slope and changes a stable wedge into an undercritical wedge; -3b- isostatic rebound induced by erosional unloading repartees the decrease in taper angle between surface slope and basal dip.

…a ‡ b†:…1 ‡ …1 ÿ l†:K† ‡ Q:…S0 =rgr†: cot f ÿ b 1ÿl

…1†

where g is the acceleration of gravity, r, S0 and f are respectively density, cohesion and internal friction angle of the wedge sediments, l is the ratio of ¯uid pressure to lithostatic stress (Pf /sn ), H is the thickness, and r and (a+b ) are the cylindrical co-ordinates for the studied portion of the wedge. K and Q are two mathematical coecients (from Eqs. (18a) and (18b) of Dahlen and Suppe, 1984). The value of the ``frictional function'' is compared to the friction coecient of the basal plane (mb) to determine where slippage is likely to occur. This method is close to another former method proposed by Liu and Ranalli (1992). For a critically tapered wedge mb equals mbc (curve of Fig. 1b depicts the solutions in an (a,b ) plot. Brittle deformation occurs within the wedge if mb > mbc and if shear stress along deÂcollement does not exceed the basal friction; this part of the wedge is undercritical (left-lower domain in the (a,b ) plot of Fig. 1b). Slippage is likely to occur along the basal surface if mb < mbc, with no brittle deformation within the wedge; this portion of the wedge is stable (central domain on Fig. 1b) and experiences a rigid displacement along the basal deÂcollement. 2.3. Variations of boundary conditions and evolution of the wedge

described from the Coulomb-wedge theory (Fig. 1a, adapted from Davis et al., 1983). A critically tapered fold and thrust belt is everywhere on the verge of Coulomb failure. The basal deÂcollement (b ), dips gently toward the hinterland and supports an equalstrength shear stress. The topographic slope (a ) and the orientation of the principal stress direction (s1) respectively to the basal deÂcollement (cb) and to the mean topography (c0) are deduced in a critically tapered wedge from an analytical solution proposed by Dahlen and Suppe (1984).

The growth of the regularly shortened Coulombwedge controls the activation of thrust systems. In an undercritical state, the wedge topographic slope is lower than its critical slope and thrusts occur to thicken its internal part; when the wedge reaches the critical shape by structural thickening, the tangential stress ®eld component along the deÂcollement is high enough to propagate deformation toward the external zones (arrow -1a- on Fig. 1b). Flexural tilting of the basement simultaneously increases the basal dip and decreases the topographic slope. It corresponds to an (a,b ) evolution that forms a small angle with the curve that bounds the stability

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

domain (arrow -1b- on Fig. 1b) and causes the wedge to become undercritical. Continental synorogenic sedimentation maintains the surface of foreland basin horizontal, during the foreland tilt (arrow -2a- on Fig. 1b) and increases the basal deÂcollement dip. When the synorogenic sedimented wedge reaches the stable shape, the deÂcollement propagates below this wedge to provide ahead structures. The relief developed above the frontal structure locally induces a surface slope that dips towards the back-stop (arrow -2b- on Fig. 1b) and may locally lock displacement along the basal deÂcollement (Norris and Cooper, 1997). Erosion decreases the surface slope (arrow -3a- on Fig. 1b). Isostatic balance moderates erosional e€ects because its distribution (Turcotte and Schubert, 1982) decreases the taper angle between surface slope and basal dip (Arrow -3b- on Fig. 1b), but does not modify the tendency to limit the forward propagation. Due to all these antagonistic e€ects, the wedge oscillates around its critical equilibrium state, and both frontal displacement and out-of-sequence thrusting occur (Chalaron and al., 1995; DeCelles and Mitra, 1995). To summarize, the evolution of a critical taper is strongly dependent on the variation of the applied conditions along its boundaries, involving: (1) shortening along the trailing edge, (2) sur®cial phenomena over its upper surface; and (3) isostatic movements of its basement. 2.4. Components of the numerical model Numerical development has been performed for a better understanding of the deformation of a frontal thrust system formed by a foreland basin, a tectonic wedge formed by syn-orogenic sediments and by anteorogenic rocks (Fig. 2a) and submitted to variation of the conditions applied along its boundaries. This model is based on: (1) the incorporation of discontinuities within the critical wedge model; (2) a kinematic forward model of serial cross-sections, describing the displacement of thrusts sheets above ¯at and ramp geometry; (3) a tilting of the basement beneath the basal deÂcollement; and (4) a linear di€usion model to simulate sur®cial erosion and sedimentation. 2.4.1. Discontinuities in the critical wedge model Global prismatic shapes are reached by displacement along thrusts both in nature and in sand-box models (Mulugetta, 1988). In order to approach this discontinuous brittle behaviour, the present study simulates the deformation of the wedge by con®ning displacement to several discrete faults (Fig. 2a). If the parameter values are chosen to match speci®c conditions, such as those studied by Huyghe and Mugnier (1992), the stress ®eld calculated in a homogeneous material can nonetheless be applied to a material with localised

645

anisotropic zones (Chalaron et al., 1995). Fig. 2b summarizes how the frictional function (1) is used step by step to determine the tip of the deÂcollement (transition between stable wedge and undercritical wedge) and the location of the active fault. 2.4.2. Kinematic model The evolution of a thrust system has already been described by forward kinematic methods (Endignoux and Mugnier, 1990). Faults are de®ned as a succession of ramps that branch o€ a basal deÂcollement. In this paper, a simple vertical shear model was used (Jones and Linsser, 1986), combined with an incremental displacement applied at the rear part of the thrust system. A regular lattice is used to describe the geometric lateral evolution of the thrusts and is combined with mechanical and erosion models (Chalaron et al., 1996). The amount of displacement between increments Tn and Tn+1 is equal to the size of the elementary cells. Each node of the lattice provides a solution to the frictional function along the basal surface to locate the tip of the basal deÂcollement (Fig. 2b), and displacement then occurs along the outermost ramp. 2.4.3. Tilting of the basement beneath the thrust wedge In fold-thrust belts and foreland basins systems, an elastic ¯exural model approximates the geometry of the continental lithosphere as a bend with a wavelength of several hundred kilometers (Karner and Watts, 1983, Lyon-Caen and Molnar, 1983; Lillie et al., 1987). The in¯uence of the topographic load of the outer wedge is small in comparison to the whole mountain belt (Turcotte and Schubert, 1982) and to the bending torque and the vertical shear forces at the trailing edge of the plate (Lyon-Caen and Molnar, 1983). Thus, isostatic motion beneath the outer wedge can be assumed as a rigid tilting of the basement that mainly depends on tectonic events located in the hinterland part of the fold-thrust belt. As the details of hinterland events are not within the scope of this study, the tilting of the basement beneath the frontal belt is assumed to increase at a constant rate. 2.4.4. Erosion/sedimentation model Numerous 2D and 3D models have already been proposed (Moretti and Turcotte, 1985; Flemings and Jordan, 1989; Beaumont et al., 1992; Sinclair et al., 1991) to consider process of erosion and sedimentation. Todten (1976) proposed that erosion is mainly controlled by the local slope. Flemings and Grotzinger (1996) modelled the erosion, transport and deposition of a bimodal sediment mixture by assigning a di€erent di€usivity to each sediment size. Given the assumption that stream width increases to accommodate increasing shear stress, Flemings and Grotzinger (1996) found that the e€ective di€usivities do not change signi®-

646

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

Fig. 2. Components of the numerical model (see discussion in the text). (a) Geometry of a thrust wedge formed by a stack of 7 thrust sheets which is shortened at the back and is located above a basement which has already been tilted and continues to tilt. (b) Calculation of the tip of the basal deÂcollement and determination of the active ramp. The tip of the deÂcollement (black dot) is located between a stable and an undercritical node (arrow along the deÂcollement represents an undercritical node). Motion (represented by arrow) occurs along a ramp if the branch-node is stable and if the tip of the deÂcollement does not reach the branch-node of a more external ramp. (c) Calculation of the evolution of the topography of the model. Sediment ¯ux J along the topography is proportional (Ks being the transport coecient) to the the local slope … @@ xu †. If the ¯ux decreases between two cells, (i.e., Ji > J0), then sedimentation occurs; if the ¯ux increases, then erosion occurs. Thrust activity deforms (Tn + 2) or translate to the left (Tn + 3) the topography.

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

cantly even as the bed changes from coarse to ®ne in the downstream direction. This result suggests that historical e€orts to approximate erosion processes as a single di€usive process was a reasonable assumption. In the proposed model, sediment horizontal ¯ux (J ) is assumed to be proportional to the topographic gradient … @@ xu †:   @u …2† J ˆ ÿKs @x Ks being the transport coecient (in m2/yr): Considering no change of volume:

3.1. Geological setting

We obtain: @u @ u ˆ Ks 2 @t @x

transports them to the valleys along which they need to be relayed by long-range ¯uvial transport (Beaumont et al., 1992). Nonetheless, hillslope erosion is the less e€ective phenomenon in the succession of processes involved in erosion and controls the solid discharge in the rivers (Souriau, 1995). Our study focuses on ramp anticlines and Piggy-back basins, and the iterative use of the di€usion model gives a ®rst general view for the syntectonic sur®cial transport in these restricted areas.

3. The Siwaliks thrust belt of the Outer Himalayas

@u @J ˆÿ @t @x

2

647

…3†

As the one-dimensional (pro®le) model is expanded to two-dimensions and implemented, a numerical resolution of the di€usion law is applied incrementally at the top of the wedge (Fig. 2c) and allows the calculation of the successive DEM (Digital Elevation Model) induced by the shortened wedge. Beaumont et al. (1992) presented a study of super®cial transports based on several algorithms. The di€usion algorithm that we used (Chalaron et al., 1995) only represents the cumulative e€ects of processes that remove materials from the hill and mountain sides and

The southern portion of the Himalaya is made up of a series of thrust sheets composed of the Neogene Siwalik Group, which is the deformed part of the foreland basin situated above the downwarped Indian plate (Lyon-Caen and Molnar, 1983). The Main Boundary Thrust (MBT) is a major thrust fault system along the entire Himalayan range and separates synorogenic sediments of the Siwalik Group from Lesser Himalayan formations (Fig. 3). Considering the whole scale of the thrust belt, the outer belt faults branch along a major deÂcollement (Raiverman et al., 1983; Gahalaut and Chander, 1992; Biswas, 1994). The diachronous location of this deÂcollement occurs from farwestern Nepal to mid-Western Nepal at the footwall of 13 Ma old sediments (DeCelles et al., 1998; log I on

Fig. 3. Balanced and restored cross-sections through the frontal belt of western Nepal. Location of the cross-section is shown in Fig. 4. Small black dots and lines close to the surface refer to dip measurements. MFT2 is the central segment of MFT on Fig. 4; MDT2 and MDT 3 are two distinct portions of MDT, ID is the roof deÂcollement of a duplex at the footwall of the MBT.

648

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

Fig. 4. (a) Simpli®ed structural map of the Siwaliks area of western Nepal. AA' refers to the cross sections of Fig. 3. The Roman numerals and the dotted lines refer to the stratigraphic sections of Fig. 4b. Sur : Surkhet ; Gho : Ghorahi. (b) Simpli®ed log of the Siwalik series involved in the Outer Himalayan belt, showing the respective thickness of Lower (mainly shale), Middle (mainly sandstone) (in white) and Upper Siwaliks (mainly conglomerate) (location on Fig. 4a).

Fig. 4) and beneath 14.6 Ma old sediments (Appel et al., 1991; log. VII on Fig. 4), respectively. The Siwalik Group is informally subdivided into the upper, middle and lower members (Auden, 1935). Satellite images and ®eld data have documented unconformities within the upper Siwalik member of western Nepal (Mugnier et al., 1999). Several north dipping thrusts delineate the tectonic boundaries in the Siwaliks area. A succession of laterally relayed thrusts called the Main Dun Thrusts (MDT) by HeÂrail and Mascle (1980), Delcaillau et al. (1987) and Mugnier et al. (1992) die out in laterally propagating folds or branch and relay on lateral transfer zones. All of these structures give a general festooned pattern to the Main Frontal Thrust (MFT) (Fig. 4). One of the major transfer zones is the West Dang Transfer zone (WDTZ) which is bounded by strike-slip tear faults, sigmoidal folds, and reverse faults (Huyghe et al., 1998). The frontal belt varies in width from 25 km west of WDTZ to 40 km east of WDTZ. The northward extent of the WDTZ forms: (1) the boundary between the zone of intense micro-seismicity in western Nepal and the gap of seismic activity of mid-western Nepal (DMG, 1997); (2) the eastward boundary of the relatively low (i.e., less than 4000 m high) Karnali

drainage area in the Lesser and Greater Himalayas; (3) the eastern boundary of the Greater Himalayan nappes (Upreti and Le Fort, 1999). Large Piggy-back basins (Duns) have developed in mid-western Nepal (Fig. 4). The Dun of Dang is one of the largest and forms a nearly 1000 km2 plain that collects a local drainage basin of less than 3000 km2. The Deukhury valley forms a nearly 600 km2 plain that collects a drainage basin of 6100 km2. The thermal evolution deduced from the maturity of organic matter (Mugnier et al., 1995) indicates a geothermal gradient near 208/km in the frontal belt and supports a brittle deformational regime (Kirby, 1985) above a depth of 10±15 km in most of the frontal and Lesser Himalayan wedge. 3.2. The frontal Himalayan thrust wedge 3.2.1. The taper geometry Studies of focal mechanisms for large earthquakes in the Himalaya (Ni and Barazangi, 1984) suggest that the basal deÂcollement dips gently (i.e. 48 2 28) northward (Fig. 5a adapted from Cotton et al., 1996). Aftershocks are distributed over the thrust wedge and suggest a pattern of distributed brittle deformation.

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

649

Fig. 6. Mechanical parameters estimation: (l, m, mb) solution for a critical wedge de®ned by a 1.38 topographic slope and a 48 basal deÂcollement dip (r=2500 kg/m3, C0=5 MPa). l, m, mb are respectively the ratio of ¯uid pressure to lithostatic stress, m the coecient of friction along the basal deÂcollement and mb the internal coecient of friction. The hatched domain of the grid refers to the values that induce a great deviation from the horizontal from the principal stress direction. Fig. 5. (a) Basal deÂcollement geometry (adapted from Cotton et al., 1996); (b) Mean topographic pro®le at E868218. The standard deviation of the height along the Himalayan arc re¯ects the relief (adapted from Bilham et al., 1997). Same horizontal scale for (a) and (b); no vertical exaggeration for (a), and vertical scale exaggerated by 50 for the topographic pro®le on (b).

Topographic studies of the Himalaya show a mean slope about 1.38 in the frontal and Lesser Himalayas and about 3.78 in the Greater Himalayas (Fig. 5b, modi®ed from Bilham et al., 1997). These recent works are consistent with the assumptions of Davis et al. (1983) which is to consider the whole frontal and Lesser Himalayas as a simple critical wedge. The in¯exion in the mean Himalayan pro®le at nearly 2000 m elevation (Fig. 5b) could be related to a deep crustal ramp (Pandey et al., 1995) or may re¯ect the leading edge of ductile deformation above a still brittle detachment, as suggested by Williams et al. (1994) to explain a similar pattern in the Andean belt. 3.2.2. The orientation of the stress ®eld Steep faults produced by back-tilting of pre-existing thrusts above younger thrusts show a normal motion component along the Surkhet±Ghorahi segment of the MBT (see Fig. 4 for location) and this pecular pattern is used to calibrate the mechanical characteristics of the Coulomb wedge. A normal component of displacement along the mentioned steep faults occurs when the dip of the principal stress direction exceeds the dip of the normal to these faults (Fig. 1a, adapted from Mugnier et al., 1994). The deÂcollement level, located in lower Siwalik ®ne-grained rocks (Huyghe et al., 1999a), exhibits a weak lithologic contrast with respect to the thrust wedge. This weak mechanical contrast de-

viates the major principal stress orientation within the frontal Himalayan thrust wedge by more than 208 from the horizontal (Davis and Lillie, 1994). 3.2.3. Mechanical parameters for the Outer Himalayas The density of Tertiary syntectonic sediments typically ranges from 2100 to 2500 kg/m3 (Lyon-Caen and Molnar, 1983; Lillie et al., 1987) and we used the latter value. Internal cohesion of the sedimentary rocks in the wedge ranges from 5 to 20 MPa (Dahlen and Suppe, 1984). Herein the lowest value, 5 MPa, is used. Other parameter values (m, mb, l ) are considered as a function of both topographic slope and basal deÂcollement dip, adapted from the analytical solutions of Dahlen and Suppe (1984). The problems of where (m, mb, l )=F(a,b ) is obviously underconstrained, and the sets of solutions depict a surface in a (m, mb, l ) space. The sets of values of l, m and mb which satisfy the critical wedge of the frontal Himalayas (topographic slope of 1.38 and deÂcollement dip of 48) are located on the grid shown on Fig. 6. Herein, a very poor contrast of strength between the basal deÂcollement and rocks within the wedge body and a high pore-¯uid pressure ratio (close to 0.8±0.9) (Mugnier et al., 1994) induces a strong deviation of the principal stress from the horizontal. The solutions are therefore located in the hatched domain of Fig. 6. 3.2.4. In¯uence of erosion and shortening parameters Concomitant shortening and erosional rates (Avouac and Burov, 1995) in¯uence the topography of a collision belt. To reduce the fundamental problem of the location of drainage pattern and of the small scale

650

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

Fig. 7. (a) Map of the bulk elevation (in m) of the frontal hills west of the Dang area (top of the map is the west side). Elevation of white domain is more than 850 m, elevation of light grey domain is less than 650 m; elevation of heavy grey is less than 850 m and more than 650 m. Black line is the MFT trace. (b) Forward modelling of a single ramp anticline (deformation described by a vertical simple shear) a€ected by erosion (hillslope sur®cial processes modeled by a linear di€usion law). (c) Elevation of the relief as a function of horizontal velocity and transport coecient. One curve is calculated for a value of Ks (expressed in 10ÿ4 mÿ1). The range of typical values is shown in grey. (d) Time-dependent parameters estimation: normalised elevation of the ramp anticline (ratio of eroded elevation and non-eroded elevation) versus V/Ks (ratio of horizontal shortening velocity and transport coecient). The case of the frontal Himalayan structure is shown by a grey dot.

erosional processes, we only considered the bulk topography, that is as an interpolated surface through the divide lines for rivers reaching the plains. The bulk elevation of the frontal relief nicely mimics the structures and outlines their trends (Fig. 7a). The respective in¯uences of rates of erosion and shortening have been studied in the case of a simple tectonic system constituted by a single ramp anticline a€ected by a linear downslope di€usion of the material along the surface (Fig. 7b from Leturmy et al., 1995). It has been shown that the elevation of the ramp culmination is an increasing function of the horizontal velocity and a decreasing function of the transport coecient (Fig. 7c). Furthermore, the bulk elevation is not dependent on the amount of shortening and the elevation of the frontal anticline remains constant over time: its topography is stationary (Delcaillau, 1997) and the maximum erosional rate nearly equals the maximum rate of uplift. Nonetheless, the eroded volume and the elevation of the relief mainly depend on the ratio of the horizontal velocity versus the transport parameter. For a given thrust geometry and displacement value, this ratio is estimated by the use of a graph that plots the normalized height of the anticline (real elevation versus elevation that would be reached without erosion) versus this ratio (Fig. 7d). Such a

procedure applied to the relief of the Mohand anticline in western India shows that the present height is only 1/3 of the height of a non-eroded anticline that involved 4000 m thick sediments, and consequently the ratio of shortening rate/transport parameter ranges from 5  10ÿ4 mÿ1 to 1  10ÿ4 mÿ1 (Leturmy et al., 1995). Care has to be taken that the elementary cell size used to describe the whole structure in¯uences the value of the transport parameter (Leturmy et al., 1995). This scaling e€ect agrees with the self-similarity properties of topography and implies that Ks is not an intrinsic physical characteristic, but rather includes a length scale component (Turcotte, 1992). We used a value of 2.510ÿ4 mÿ1 to model the Siwaliks thrust system of western Nepal with a 480 m cell size. 3.2.5. Shortening and transport coecient estimation The above discussion shows that the absolute values of the time-dependent parameters are not essential for model runs. For an easier understanding, a 16 mm/yr shortening rate is used. This estimation of the Himalayan rate shortening is proposed by Molnar (1987) and is con®rmed by the recent geodetic measurements of Bilham et al. (1997). A similar rate was calculated for Neogene (12±2 Ma) shortening in the Lesser Himalaya (DeCelles et al., 1998) and in the

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

651

Fig. 8. Geometry of Piggy-back basins in an accretionary wedge: Comparison between steep (68, upper cross-section) and gentle (48, lower crosssection) initial basal dip. Mentioned lengths are in meters; (l, m, mb) are determined from Fig. 6, ratio shortening rate/transport coecient is 2.510ÿ4 mÿ1; tilting velocity: 15 ' per Ma. The geometries refer to runs 2 and 5 in Table 1. (a) Initial (dotted lines) and ®nal geometry. Synorogenic sediments are shown in black (vertical scale=horizontal scale). (b) Propagation sequence diagram. The number of displaced sheets is on the vertical axis, the X-axis is the time. Results of the calculation run are plotted for every increment. Regular variation in the number of displaced thrust sheets represents a migration of the front of the thrust system, horizontal line refers to a permanent displacement along a single ramp, and ``teeth'' pattern refers to out-of sequence reactivation.

frontal belt (Mugnier et al., 1999). A transport coecient of 60 m2/yr is inferred from the above shortening rate and ratio of shortening rate/transport parameter. As slope varies drastically along cross-sections of the frontal belt, erosion rate varies the same manner: it reaches 10±15 mm/yr at the hanging-wall of ramps (LaveÂ, 1997, Hurtrez et al., 1999), but is nearly null above the ¯at-ramp transition. 3.3. Numerical modeling of the frontal himalayan belt Numerical runs have been performed to study the evolution and dynamics of the frontal part of the thrust belt of western Nepal. The aim of this modeling process is not to ®t the whole geometry of the frontal structures, but rather to understand the e€ects of the dynamic feedback between tectonic, erosional, and sedimentary processes; thus the initial and ®nal geometry of the models do not exactly match the ®eld struc-

tures. Therefore the parameter values estimated above give a reasonable scaling for the processes involved in the wedge evolution. The following comments refer to Fig. 8 but are also supported by other runs. Table 1 summarizes the e€ects of varying time-scaled parameters (Chalaron et al., 1995) with the upper and lower values estimates, i.e., 15' or 30' per Ma for the rate of lithosphere tilting, 10 or 20 mm/yr for the displacement velocity and 40 or 200 m2/yr for the sur®cial transport parameter (i.e., 0.5 or 510ÿ4 mÿ1 for the ratio shortening rate/ transport parameter). 3.3.1. Initial geometry Restoration of balanced cross-sections (companion paper and Fig. 3) suggests an initial length in the plane of tectonic motion greater than 50 km for the Siwaliks belt. Therefore the initial state is de®ned by a regular lattice of 125  125 cells to represent a

5000 2500 2000 8450 9750 8750 8750 26 26 54 40 40 68 40 10 10 0 10 10 0 10 3 3 3 7 7 3 7

10 7 0 3 3 0 3

Fault No. 3 Fault No. 4

3 7 3 7 7 0 7 7 7 3 13 10 0 10 40 40 36 20 23 29 23 15' 15' 15' 15' 15' 15' 30' 5.00E-04 2.50E-04 5.00E-05 5.00E-04 2.50E-04 5.00E-05 2.50E-04 m m m m m m m 48 48 48 68 68 68 68 1 2 3 4 5 6 7

3800 3800 3800 5700 5700 5700 5700

Fault No. 5 Fault No. 6 Fault No. 7 Tilt rate (1/Myr) V/Ks (1/m) b

Thickness at the back

60  60 km surface of a future thin-skinned thrust belt. Thrust sheet displacement is parallel to the lateral boundaries of the model. The basal deÂcollement dip varies continuously from 48 to 68 to study the e€ects of the basal deÂcollement dip variations. As a result, the thickness of the thrust sheets is not constant. Faults branch o€ from the basal deÂcollement. Their initial dip is 258, a value which is in a good agreement of those predicted by the Mohr±Coulomb±Anderson isotropic criteria considering fresh fractures within a thrust wedge (Dahlen and Suppe, 1984). A regular spacing between the ramps is used and its value matches the smallest spacing observed in the ®eld. With such a choice, all the ramps are not involved for some runs (see for example runs 3 and 6 on Table 1), leading to a greater spacing of the e€ectively involved ramps in the ®nal geometry. The model run calculated with the parameters discussed above is represented by two crosssections (Fig. 8a), one in the thinnest part (initial thickness 3800 m, west of the WDTZ) and one in the thickest part (initial thickness 5700 m, east of the WDTZ). 3.3.2. Results of the modeling

Run No.

% of total displacement along each fault

Fault No. 2

Fault No. 1

Width basin (m)

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657 Table 1 In¯uence of parameter values (initial thickness at the back; eciency of tectonics respectively to erosion: V/Ks; lithospheric tilt rate) on the partitioning of total shortening along the di€erent thrusts (see Fig. 8 for numbers of the thrusts). The total displacement is 15 km in all runs and occurs in 1 Ma. The ratio is the % of displacement along fault No. i/total displacement. The width of the basin is the surface expression of the Piggy-back basin after 1 Ma

652

3.3.2.1. Sequence of thrusting. The displacement sequence along the ramps is shown by the propagation sequence diagram (Fig. 8b) where the vertical axis is the number of displaced sheets and the horizontal axis is time. The results of the calculations are plotted for every elementary displacement and allow the recognition of thrust sequences as classically de®ned (McClay, 1992). At the beginning, the more external thrust is active (thrust 7 of Fig. 8). This location is related to the shape of the initial sedimentary wedge which formed a nearly stable taper. Development of a relief above the active frontal thrust (arrow 2b on Fig. 1b) and tilting of the basement (arrow 1b on Fig. 1b) induce a hinterland dipping of the topography and promote a backward thrust sequence (seq. 1 of Fig. 8b). This ®rst sequence of thrusting is followed by a forward sequence (seq. 2 on Fig. 8b). The ``teeth'' pattern of the propagation sequence diagram (seq.3 of Fig. 8b) represents out-of-sequence reactivations. Outof-sequence reactivation of the thrusts occurs to balance erosion and tilting e€ects (arrows -1b- and -3bon Fig. 1b) and are characteristic of a steady-statetype evolution (Chalaron et al., 1995). This steadystate evolution is found for every run, but displacement partitioning is strongly dependent on the dip of the basal deÂcollement: 15% of the displacement along the basal deÂcollement is linked up with out-of sequence reactivations in the case of a wedge located above a steep deÂcollement, and 30% for a wedge deformed above a gently dipping deÂcollement (runs 2 and 5 of Table 1). This is due to the steeper surface slope that

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

enhances erosion in the case of a gently dipping deÂcollement. Comparisons between runs (see runs 1, 3 and 4, 6 of Table 1) demonstrate that the increase in the super®cial transport eciency (expressed by a diminution of V/Ks from 510ÿ4 to 510-5 mÿ1) promotes displacement along the innermost thrusts: the displacement along the innermost thrust (fault No. 1 on Table 1) is twice in the case of a 48 basal deÂcollement and increases by 70% in the case of a 68 deÂcollement. The frontal and median thrust sheets are less in¯uenced by the variations of sur®cial transport as sediments are alternatively trapped or cleared away above these structures during the development of wedge-top basins. Furthermore, the interrelationship between an increase in sur®cial transport eciency and the motion along the frontal thrust is complex, and according to the runs either accentuates or decreases the displacement. Variation in the rate of basement tilting from 15'/ Ma to 30'/Ma does not a€ect the results (compare runs 5 and 7 on Table 1). This weak sensitivity indicates that the (a,b ) evolution induced by the progressive basement tilting is nearly parallel to the domain of stability of the wedge (arrow -1b- on Fig. 1b), and that the change of the dip of the basal deÂcollement during frontal tectonics is weak compared to the dip of the ¯exured lithosphere (in the order of 15' compared to 58, i.e., 5%). 3.3.2.2. Piggy-back basin development. All the runs show the development of wedge-top basins in the hollow between the frontal culmination at the hanging wall of the frontal ramp and the large thrust sheet at the trailing edge. Deposition of sediments occurs: (a) in a ®rst stage, in a Piggy-back basin which is progressively deformed by a backward nucleation of ramps. Deformation in this basin is diachronous from the external part to the internal one; (b) during the end of the ®rst sequence, (activation of the internal sheet), the basin is passivelly ®lled and previous deformations are sealed; (c) during the forward sequence of thrusting and the out-o€ sequence reactivations, the basin is a€ected by a complex deformation. The depth and extension of the Piggy-back basins are nonetheless strongly in¯uenced by parameter values. The volume of sediment in the Piggy-back basin increases when the ratio of shortening rate/transport coecient decreases. Nonetheless, when a high erosional rate promotes internal reactivations (arrow 3 on Fig. 1b), the rear part of the basin is underthrusted beneath the innermost wedge and the global extent of its outcrops remains stable or even decreases. The horizontal extent of the Piggy-back basins (table I, right column) decreases from 5000 to 2500 m when V/

653

Ks diminished from 510ÿ4 to 510ÿ5 mÿ1 above a 48 basal deÂcollement (runs 1 and 3 of Table 1). Large (from 8450 to 9750 m from Table 1) and regular basins are located above the steeper basal deÂcollement (Fig. 8a), while small (from 2000 to 5000 m from Table 1), asymmetrical and strongly deformed basins are located above the gentler dipping basal deÂcollement. The shape of the Piggy-back basins therefore re¯ects the deep geometry of the thrust system. This last result suggests a qualitative tool to estimate along-strike variation in the basal deÂcollement dip of the Himalayan Outer belt. 4. Geometry and kinematics of the frontal belt of western Nepal as an expression of the mechanics of the Himalayan thrust wedge 4.1. Piggy-back basins 4.1.1. Lateral variation of the frontal belt substratum in Western Nepal Few published data concern the structure of the frontal Himalayas substratum in Nepal. A large uncertainty still a€ects the estimate of the dip of the basal deÂcollement. For example, Bilham et al. (1997) give the prudent estimate of 08 to 88 for this dip value. Reconstruction of the geometry of the foreland basin now incorporated in the frontal belt of Himalayas nonetheless provides valuable information concerning lateral variation of the deep structures. A large scale map of the subsurface structure of the Gangetic plain (Raiverman et al., 1983) suggests lateral variations of the foreland basin geometry and our ®eld work shows that the stratigraphic thickness of the more external thrust sheets increases from nearly 4300 m in the western part of the study area to more than 5500 m in the eastern part (Fig. 4b). This thickening is mainly a result of the increased thickness of the middle Siwaliks. A basement scarp that abruptly modi®es the thickness of the middle Siwalik series is supposed to be located beneath the west boundary of the duns of Deukhury and Dang valley and may control di€erences in the ¯exural subsidence pattern. Furthermore, a variation of 28 in the dip of the substratum of a 60 km wide foreland basin would induce a variation of the thickness from 3800 to 5700 m (dotted lines on Fig. 8) and is sucient to explain the stratigraphic thickness variations. 4.1.2. Extent of Piggy-back basins of mid-western Nepal Model runs show that a minor (on the order of 28) steepening of the basal deÂcollement is sucient to account for the large extent of Piggy-back basins, even if no change in the shortening velocity, deformation style, super®cial transport conditions or mechanical

654

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

properties occur. This modelling suggests that the Dang and Deukhury valleys of mid-western Nepal, that developed above a thick Siwalik imbrication, are located above a steeper deÂcollement level than the Outer belt of far-western Nepal. The WDTZ that forms the western border of midwestern Nepal duns appears as a complex structural zone related to a substratum fault that: (a) modi®es the stratigraphic thickness of the Siwalik members; (b) increases the depth of the deÂcollement level from West to East in the thin-skinned thrust belt; (c) locates tear faults that transfer displacement from one thrust to the other. 4.1.3. Development of Piggy-back basins of mid-western Nepal Tilting of the substratum and over®lled sedimentation in a foreland basin are very ecient processes that propagate the thrust front toward the foreland and delineate a Piggy-back basin. The great thickness of the middle Siwaliks in mid-western Nepal could re¯ect such a tilt and could induce a forward propagation of the frontal thrust (arrows a and b on Fig. 1b, and top of Fig. 8). It would then promote an early development of the wedge-top basins of mid-western Nepal. This early activation of the frontal thrust is evidenced by the presence of unconformities in the Upper Siwaliks deposited above the ramp anticline of MFT and above WDTZ (Fig. 4a and companion paper). Model runs suggest that such an early stage of development would be followed by out-of-sequence thrusting. The complex relationships between the Quaternary sediments of the Dang valley and the underlying Siwalik beds are partly due to the presence of back-thrusts (Fig. 4a) and intermediate deÂcollements between Dang valley and Deukhury valley (companion paper), but could agree with such out-of sequence tectonics. Furthermore, out-of sequence tectonics is already evidenced during Holocene in Deukhury valley (Leturmy et al., 1997). In contrast, the small Piggyback basin of Surkhet (Far-western Nepal, Sur on Fig. 4) developed at the top of a wedge (bottom of Fig. 8) is controlled by an evolution depicted by the arrow 1a and 1b on Fig. 1b. 4.1.4. Controls of Piggy-back basins of Himalayas A comparison of Piggy-back basin evolution (Huyghe et al., 1999b) and the above numerical runs show that the development of Piggy-back basins depends on the sediment supply, geometry of the thrust system, and mechanical parameters (Chalaron et al., 1995). In the Dang and Deukhury valleys cases, sediment supply does not greatly in¯uence their development because their rivers catchment are localized. The mechanical parameters are fairly constant in western Nepal since there is no evidence of major di€er-

ences between the deformation of the thrust zones of mid-western and those of far-western Nepal (Huyghe et al., 1999a). They always show a superposition of cataclastic deformation and pressure-solution deformation in a clay matrix. The geometry of the foreland basin sediments and the geometry of the deÂcollement appears to be the predominant control for the Piggyback basins of mid-western Nepal. This simple geometric control is not applicable to every Piggy-back basin of the Outer Himalayas. For example, the wide basins of Pakistan are mainly controlled by the Cambrian evaporitic levels (Burbank and Beck, 1989) that form a very weak basal deÂcollement (Davis and Lillie, 1984). The basins of Dehra Dun (Kumar and Ghosh, 1994) in Western India or Narayanghat (Chalaron et al., 1995) in central Nepal show similarities with the Deukhury valley. Nonetheless, care has to be taken that the sediment supply of the rivers that cross these two basins are greater than those of the Rapti river that crosscuts the Deukhury valley, and could be a factor as important as the geometric control. 4.2. Deformation of the frontal and lesser Himalayas thrust wedge 4.2.1. Lateral evolution of the deformation The geology of western Nepal is characterised by subtle distinctions between the mid-Western and the far-Western part. Some of them have been already presented in the above paragraphs concerning the extension of the Piggy-back basins and the role of the WDTZ in the Outer belt. A comparison of two topographic pro®les through the lesser Himalaya of farwestern and mid-western Nepal (Fig. 9a) also outlines this distinction. The mean slope of the frontal belt varies from 2.38 in far western to 1.38 in mid-western Nepal. The basal deÂcollement dip variation, suggested beneath the frontal Himalayan belt in the above paragraphs in¯uences the topographic pro®les of the wedge (Fig. 9b). This dip variation could also be considered for the basal detachment beneath the Lesser Himalayas. Following this assumption, the Lesser Himalayan wedge would be thicker in mid-western Nepal, and considering a constant thermal gradient, the brittle-ductile transition would be shifted toward the south. The northward prolongation of the WDTZ forms in the Lesser Himalayas: (1) the boundary between the high micro-seismicity zone of western Nepal and the gap of seismic activity of mid-western Nepal (DMG, 1997); (2) a north±northeast corridor that a€ects the extent of the outcrops of the Mahabarat nappes and High Himalayas crystalline klippen (Upreti and Le Fort, 1999); (3) the eastward boundary of the relatively low Karnali drainage area in the Lesser and Higher Himalayas. This last feature

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

Fig. 9. (a) Comparison of a topographic pro®le through far-western Nepal and a topographic pro®le through mid-western Nepal (location on Fig. 4). The thin lines are the mean altitudes calculated on a windows 10 km large; the thick lines are the smoothed topographic pro®les, using a polynomial interpolation of degree 4. The continuous lines are far-western Nepal pro®les, the pecked lines are the mid-western Nepal pro®les. (b) Topographic pro®les of numerical experiments of Fig. 8. The pecked line is the pro®le of the wedge located above the steep decollement (upper part of Fig. 8) while the continuous line is the pro®le of the wedge located above the gentle deÂcollement (lower part of Fig. 8).

therefore could partly re¯ect the deep geometry of the Himalayan thrust wedge. 4.2.2. Reactivation of thrusts within the wedge Out-of sequence thrusting has already been inferred in the frontal Himalayas (Delcaillau et al., 1987; Burbank and Beck, 1989) and in the Lesser Himalayas (Schelling and Arita, 1991; Arita and Paudel, 1997). One of the major thrusts (MDT) of the Outer belt of western Nepal show several splays at the hanging-wall of the MDT, and two key features have been observed (Mugnier et al., 1998): (a) thrusts are bevelled by erosional unconformities and overlapped by sediments; (b) these sediments are overthrust by inner faults

655

Fig. 10. Schematic representation showing the relationships between the sequence boundaries and the thrusts forming the MDT Splay. Vertical exaggeration: 2. Faults Ia, Ib, II, III are splays which branch from the Main Dun Thrust (MDT) in the Karnali area. Ia and Ib dip 45±508 to the NE. II, III dip 30±348 to the NE. (a), (b), (c), (d), (e) refer to the stratigraphic sequences that overlap above the thrust sheets. The analysis of the section leads to the following succession for tectonic and sedimentary events : (1) large displacement along thrust Ia. Thrust Ib, is possibly active at that time; (2) a major relief culmination developed at the hanging wall of the thrust Ia; (3) erosion reached the Lower Siwaliks; (4) sedimentary sequences (a) and (b) cover over the thrusts Ia and Ib; (5) thrust II propagated through the previously tilted Lower Siwalik strata and through the upper part of sequence (b); (6) more than 40 m of displacement occurred along this thrust and erosion a€ected its hanging wall; (7) top of sequence (c) covers over the thrust II; (8) thrust III propagated and more than 40 m displacement occurred along this thrust; (9) regional recent terrace overlapped above the hanging wall of the thrust III. The huge Lower Siwalik boulders at the bottom of sequence (c) and (d) are related to erosion of hanging wall culminations developed respectively during motion along the thrusts II and III.

(Fig. 10) and indicate a break-back sequence of thrusting where periods of quiescence separated several 40 m displacement pulses that correspond in depth to reactivation of the MDT. These ®eld data indicate an episodicity of the out-of sequence reactivation, and model runs evidence that periodic out-of sequence reactivations consume 15±30% of the total shortening. The good agreement between both scarce ®eld data and the numerical results suggests that episodic out-o€ sequence reactivations could be a main feature of the Himalayan wedge deformation, although geological evidence is generally erased by erosion. To summarize, the western Nepal thrust belt is sensitive, as is any thrust belt, to the variations of the conditions applied along its external boundaries, including sur®cial transport (erosion-sedimentation) along its topography and isostatic motions of its base-

656

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657

ment. Erosion partly controls the long term partitioning of shortening in the wedge and long-lasting average rate of displacement is depressed along the deÂcollement from the inner zones of the wedge to the frontal thrust. For a duration shorter than the quiescence lapses of motion along the Piggy-back thrusts (and greater than the seismic-aseismic cycle), the rate of displacement along the frontal thrust is nonetheless equal to the displacement of the inner zones of the wedge.

5. Conclusion We have presented a model where the evolution of the fold and thrust belt of the Outer Himalayas is considered as a brittle Coulomb wedge with a very poor contrast in strength between the basal deÂcollement and rocks in the wedge body. Its internal deformation is mainly located along major fault zones whereas its external bulk geometry is close to a static critical geometry. The proposed model gives a dynamic clue for ®eld data presented in the companion paper: (a) the development of Piggy-back basins, that are located above thick thrust sheets of mid-western Nepal, is related to a steeper basement favouring the deposition of a thick Siwalik pile in the foreland basin, the geometry of which is a wedge close to the critical taper; (b) the early individualization of the Piggy-back basin (shown by unconformities above the frontal ramp anticline of the MFT and the major lateral transfer zone of WDTZ) followed an episode of rapid subsidence at the end of middle Siwalik deposition and was followed by out-of sequence reactivation that deformed the Piggyback basin, an evolution in agreement with the results of our model; (c) episodic out-of sequence reactivations are characteristic of the Lesser and Outer Himalayas and maintain the critical taper of the thrust wedge a€ected by lithospheric ¯exure and erosion.

Acknowledgements Constructive critical reading of the manuscript by P. DeCelles Y. Philippe and K. Burke is gratfully acknowledged. We thank the Institut National des Sciences de L'Univers (CNRS) and Institut Franc° ais du PeÂtrole for the their ®nancial supports.

References Appel, E., Rosler, W., Corvinus, G., 1991. Magnetostratigraphy of the Miocene-Pleistocene SuraõÈ Khola Siwaliks in West Nepal. Journal of Geophysics International 105, 191±198. Arita, K., Paudel, L.P., 1997. Thrust tectonics in central Nepal and

its signi®cance on the Himalayan uplift. 2th Nepal Geological Society meeting, Katmandou, p. 38. Auden, J.B., 1935. Transverse in the Himalayan. Geological Survey of Indian Record 69, 123±167. Avouac, J.C., Burov, E., 1995. Erosion as a driving mechanism of intracontinental mountain growth. Journal of Geophysical Research 101, 17,747±17,769. Beaumont, C., Fullsack, P., Hamilton, J., 1992. Erosional control of active compressional orogens. In: McClay, K.R. (Ed.), Thrust tectonics. Chapman & Hall, London, pp. 1±18. Bilham, R., Larson, K., Freymueller, J., Project IDYLHIM members, 1997. GPS measurements of present-day convergence across the Nepal Himalaya. Nature 386, 61±64. Biswas, S.K., 1994. Status of exploration for Hydrocarbons in siwalik Basin of India and Future trends. In: Kumar, R., Ghosh, S.K., Phadtare, N.R. (Eds.), Siwalik Foreland Basin of Himalaya: Dehra Dun, India. IBH, Oxford, pp. 283±301. Burbank, D., Beck, R.A., 1989. Early pliocene uplift of the salt range. In: Lawrence, L., Malinconico Jr, Lillie, R. (Eds.), Tectonics of the western Himalayas, special paper AGSA, 232, 113±128. Chalaron, E., Mugnier, J.L., Mascle, G., 1995. Control of thrust tectonics in the himalayan foothills: a view from a numerical model. Tectonophysics 248, 139±163. Chalaron, E., Mugnier, J.L., Sassi, W., Mascle, G., 1996. Tectonic, erosion and sedimentation in a overthrust system: a numerical model. Computer and Geosciences 22 (2), 117±138. Cotton, F., Campillo, M., Deschamps, A., Rastogi, B.K., 1996. Rupture history and seismotectonics of the 1991 Uttarkashi, Himalaya earthquake. Tectonophysics 258, 35±51. Dahlen, F.A., Suppe, J., 1984. Mechanics of fold and thrust belts and accretionary wedges. Cohesive Coulomb theory. Journal of Geophysical Research 89, 10,087±10,100. Dahlstrom, C.D.A., 1969. Balanced cross sections. Canadian Journal of Earth Sciences 6, 743±757. Davis, D.A., Lillie, R.J., 1994. Changes in mechanical response during continental collision: active example from the foreland thrust belts of Pakistan. Journal of Structural Geology 16, 21±34. Davis, D., Suppe, J., Dahlen, F.A., 1983. Mechanics of fold-andthrust belts and accretionary wedges. Journal of Geophysical Research 88, 1153±1172. DeCelles, P.G., Mitra, G., 1995. History of the Sevier orogenic wedge in terms of critical taper models, northeast Utah and southwest Wyoming. Geological Society Bulletin 107, 454±462. DeCelles, P.G., Gehrels, G.E., Quade, J., Ojha, T.P., Kapp, P.A., 1998. Neogene foreland basin deposits, erosional unroo®ng and the kinematic history of the Himalayan fold-thrust belt, Western Nepal. Geological Society of America Bulletin 110, 2±21. Delcaillau, B., 1997 Les fronts de chaõÃ nes actives, Memoir of Habilitation, 335 pp. Delcaillau, B., Herail, G., Mascle, G., 1987. Evolution geÂomorphostructurale de fronts de chevauchements actifs: le cas des chevauchements intra-Siwalik du NeÂpal central. Zeitschrift fuÈr Geomorphologie Neue Forschung 31, 339±360. DMG 1997. (Departement of Mines and Geology, Ministry of Industry His Majesty's Government of Nepal). Microseismic EpiCenter Map of Nepal Himalaya and adjoining region. Endignoux, L., Mugnier, J.L., 1990. The use of forward kinematic model in the construction of balanced cross sections. Tectonics 9, 1249±1262. Flemings, P.B., Jordan, T., 1989. A synthetic stratigraphic model of foreland basin development. Journal of Geophysical Research 94, 3851±3866. Flemings, P., Grotzinger, J., 1996. Strata: freeware for analyzing classic stratigraphic problems. GSA Today 6 (12), 1±7. Gahalaut, V.K., Chander, R., 1992. On the active tectonics of Dehra

J.L. Mugnier et al. / Journal of Asian Earth Sciences 17 (1999) 643±657 Dun region from observations of ground elevation changes. Journal of the Geological Society of India 39, 61±68. HeÂrail, G., Mascle, G., 1980. Les Siwalik du Nepal central: structures et geÂomorphologie d'un pieÂmont en cours de deÂformation. Bulletin Association Geographique Franc° aise 431, 259±267. Hurtrez, J-E., Lucazeau, F., Lave, J., Avouac, J-P., 1999 Investigation of the relationships between basin morphology tectonic uplift and denudation from the study of an active belt in the Siwalik Hills (Central Nepal), Journal of Geophysical Research 104, 12779±12796. Huyghe, P., Mugnier, J.L., 1992. Short-cut geometry during structural inversions: competition between faulting and reactivation. SocieÂte GeÂologique de France, Bulletin 163, 691±700. Huyghe, P., Mugnier, J.L., Leturmy, P., 1998. Segmentation of the Siwalik thrust belt of Western Nepal and localisation of Piggyback basins. 13th Himalayan Karakorum Tibet International Workshop, Peshawar, 20±22 April 1998. Huyghe, P., Galy, A., Mugnier, J.L. 1999a. Micro-structures, clay mineralogy and d18O isotopes in the shear zones of the Siwaliks of western Nepal. Nepal Geological Society Bulletin 18, 239±248. Huyghe, P., Mugnier, J.L., Griboulard, R., Deniaud, E., Gonthier, E., Faugeres, J.C., 1999b Review of the tectonic controls and sedimentary patterns in Late Neogene Piggyback basins on the Barbados Ridge Complex, Caribbean Basins. In: Mann, P. (Ed.), Sedimentary Basins of the World, Elsevier, Vol. 4. pp. 367±386. Jones, P.B., Linsser, H., 1986. Computer synthesis of balanced cross sections by forward modelling. American Assiociation of Petroleum Geologists Bulletin 70, 605 (abstract). Karner, G.D., Watts, A.B., 1983. Gravity anomalies and ¯exure of the lithosphere. Journal of Geophysical Research 88, 449±477. Kirby, S.H., 1985. Rock mechanics observations pertinent to the rheology of the continental lithosphere and the localization of strain along shear zones. Tectonophysics 119, 1±27. Kumar, R., Ghosh, S.K., 1994. Evolution of the Moi-Pliocene alluvial fan system in the Siwalik foreland basin, Dhera Dun, India. In: Kumar, G., Phadtare (Eds.), Siwalik foreland basin of Himalaya, Himalayan Geology, 15, 143±160. LaveÂ, J. 1997. Tectonique et eÂrosion ¯uviale: la contribution des rivieÁres aÁ l'eÂtude sismo-tectonique de l'Himalaya du NeÂpal, Ph D thesis, Univ. Paris VII, Paris, 250 pp. Leturmy, P., Mugnier, J.L., Chalaron, E., 1995. Erosion et seÂdimentation au voisinage d'un anticlinal de rampe: apport d'un modeÁle numeÂrique. SocieÂte GeÂologique de France, Bulletin 6, 783±795. Leturmy, P., Mugnier, J.L., Delcaillau, B., Mascle, G., Vidal, G. 1997. Uplift rates linked to the out-o€-sequence reactivation of the outer thrust belt of western Nepal. 12th Himalaya± Karakorum±Tibet Workshop, Rome, April 1997. Lillie, R.J., Johnson, G.D., Yousuf, M., Zamin, A.S.H., Yeats, R.S., 1987. Structural development within the Himalyan foreland fold and thrust belt in northern Pakistan. In: Beaumont, C., Tankard, A. (Eds.), Sedimentary basins and basin forming mechanisms. Canadian Society of Petroleum Geologists, 12, 379±392. Liu, J.Y., Ranalli, G., 1992. Stresses in an overthrust sheet and propagation of thrusting: an airy stress function solution. Tectonics 11, 549±559. Lyon-Caen, H., Molnar, P., 1983. Constraints of the structure of the Himalaya from analysis of gravity anomalies and a ¯exural model of the lithosphere. Journal of Geophysical Research 88, 8171±8191. McClay, K.R., 1992. Glossary of thrust tectonic terms. In: McClay, K.R. (Ed.), Thrust tectonics. Chapman & Hall, London, pp. 419± 433. Molnar, P., 1987. Inversion pro®le of uplift rate for the determination of the geometry of dip-slip faults at depth, with examples from the Alps and the Himalaya. Annales de GeÂophysiques 5, 663±670.

657

Moretti, I., Turcotte, D.L., 1985. A model for erosion, sedimentation, and ¯exure with application to New Caledonian. Journal of Geodynamics 3, 155±168. Mugnier, J.L., Mascle, G., Faucher, T., 1992. La structure des Siwalik de l'Ouest NeÂpal: un prisme d'accreÂtion intracontinental. SocieÂte GeÂologique de France, Bulletin 163, 585±595. Mugnier, J.L., Huyghe, P., Chalaron, E., Mascle, G., 1994. Recent movements along the main Boundary Thrust of the Himalayas: normal faulting in an over-critical thrust wedge? . Tectonophysics 238, 199±215. Mugnier, J.L., Chalaron, E., Mascle, G., Pradier, B., Herail, G., 1995. Structural and thermal evolution of the Siwaliks of Western Nepal. Bulletin of the Geological Society, NeÂpal 11, 171±179. Mugnier, J.L., Delcaillau, B., Huyghe, P., Leturmy, P., 1998. The break-back thrust splay of the main dun thrust: evidence of an intermediate displacement scale between earthquake slip and ®nite geometry of thrust systems. Journal of Structural Geology 20, 857±864. Mugnier, J.L., Leturmy, P., Mascle, G., Huyghe, P., Chalaron, E., Vidal, G., Husson, L., Delcaillau, B. 1999. The Siwaliks of western Nepal. I. Geometry and kinematics. Journal of Asian Earth Sciences 17, 629±642. Mulugetta, G., 1988. Modelling the geometry of Coulomb thrust wedges. Journal of Structural Geology 10, 847±860. Ni, J., Barazangi, M., 1984. Seismotectonics of the Himalayan Collision Zone: geometry of the underthrusting Indian plate beneath the Himalaya. Journal of Geophysical Research 89 (B2), 1147±1163. Norris, R., Cooper, F.A., 1997. Erosional control on the structural evolution of a transpressional thrust complex on the Alpine Fault, New Zealand. Journal of Structural Geology 19, 1323± 1342. Pandey, M.R., Tandukar, R.P., Avouac, J.P., LaveÂ, J., Massot, J.P., 1995. Evidence for recent interseismic strain accumulation on a mid-crustal ramp in the Central Himalayas of Nepal. Geophysical Research Letters 22, 751±754. Raiverman, V., Kunte, S.V., Mukherjee, A., 1983. Basin geometry, Cenozoic sedimentation and hydrocarbon prospects in northwestern Himalaya and Indo-Gangetic plains. Petroleum Journal of Asia 6, 67±92. Schelling, D., Arita, K., 1991. Thrust tectonics, crustal shortening, and the structure of the far-eastern Nepal Himalaya. Tectonics 10/5, 851±862. Sinclair, H.D., Coakley, B.J., Allen, P.A., Watts, A.B., 1991. Simulation of a foreland basin stratigraphy using a di€usion model of erosion of mountain belt uplift an erosion: an example of the central Alps, Switzerland. Tectonics 10, 599±620. Souriau, M., 1995. Les processus d'eÂrosion meÂcanique aÁ l'inteÂrieur des grands bassins ¯uviaux. SocieÂte GeÂologique de France, Bulletin 6, 763±781. Todten, H., 1976. A mathematical model to describe surface erosion caused by overland ¯ow. In: Ahnert, F. (Ed.), Quantitative slope model, vol. Supplementband 25. Zeitschrift fuÃr Geomorphologie, Berlin, pp. 89±105. Turcotte, D.L., Schubert, G., 1982. Geodynamics, Appplication of continuum Physics to Geological problems. Wiley, New York, 450 pp. Turcotte, D.L., 1992. fractals and Chaos in Geology and Geophysics. Cambridge University Press, Cambridge, pp. 221. Upreti, B.N., Le Fort, P., 1999 Lesser Himalayan crystalline nappes of Nepal: problems of their origin, Memoirs of the Geological Society, America special paper 328, 225±238. Williams, C.A., Conners, C., Dahlen, F., Price, E., Suppe, J., 1994. E€ects of the brittle-ductile transition on the topography of compressive mountain belts on Earth and Venus. Journal of Geophysical Research 99, 19,947±19,974.