In the last 20 years, the progress of remote sensing techniques and renewed access to many regions of Tibet have yielded new insights into the processes of continental collision. But there is still no consensus on the mechanisms that have led to the growth and rise of the 3.106 km2 wide and 5000 m high plateau.
In one class of models, the whole lithosphere is inferred to deform as a thin viscous sheet. Concurrent thickening of the crust and lithospheric mantle, spread over a vast area, absorbs most of the plate convergence. In the thickest region (Tibet), topographic rebound is subsequently triggered by convective removal of the thickened mantle. This is followed by relief collapse and extension. The rebound has a profound effect on climate. Mostly the final collapse is associated to strike-slip faulting, which is thus seen as a shallow and minor side-effect.
In another class of models, strain localization along shear zones, and decoupling between mantle and crust, are thought to govern the deformation of Asia. Strike-slip faulting and thrusting combine to cause diachronic uplift of Tibet soon after the onset of collision. The crust thickens but the mantle beneath subducts. At any given time, a small number of boundaries extending to the base of the lithosphere absorb much of the convergence.
Testing such competing models rests on mapping and dating Tertiary strain, metamorphism and intrusion, measuring slip- and uplift-rates, and imaging the deep crust and mantle. We review ongoing studies of Tertiary deformation and magmatism in and around Tibet, incoming evidence from seismic tomography experiments, and the kinematic picture emerging from Holocene slip-rate measurements.
New results tend to support the second class of models. Tibet appears to have grown towards the north and east in three distinct stages, as large sinistral strike-slip faults successively sliced farther into the Asian lithosphere, causing mountain growth on oblique thrust splays. The propagation of such faults "pushed" one edge of the plateau inside the continent. Decoupled wedges of crust thickened in conjunction with south-dipping subduction of the Asian mantle along two reactivated suture zones. Overall, the large-scale deformation of Asia thus coinvolved subduction and extrusion of coherent blocks of lithospheric mantle. Major block boundaries, such as the Himalayas, Altyn Tagh and Kunlun, resemble oblique convergent margins, with crustal slip-partionning. Plate Tectonic-like mechanisms may therefore operate beneath the thickened crust. High resolution seismic imaging, and more accurate strain and gravity measurements should help resolve the issue.
A 2D numerical model of mantle convection, which incorporates rigidly moving surface plates, isused to study the effects of the accretion of a 2000 km wide block of continental crust to the margin of a large stationary continent. This continental block is a model representation of the Indian sub-continent. It is carried along with an oceanic plate towards a subduction zone at the leading edge of a much larger stationary continent, meant to represent Asia. Prior to continental collision, subducted oceanic plate material under-rides the margin of the larger continent for several hundred kilometers before peeling away from the surface boundary layer and descending into the mantle as a cold narrow slab. Consequently, once the moving continental block arrives at the continental margin and closes over the original subduction zone the surface location of the continental suture lies several hundred kilometers south of the location of the cold remnant of the sinking slab. Convective flow in the sub continental mantle adjusts to accommodate the change in mechanical boundary condition at the upper surface. We find that once the upper surface is immobilized, the continued descent of the cold slab is inhibited, and the cold remnant remains below the collision zone longer than expected based on 'normal' rates of subduction. Evolution of temperature and flow fields in the former subduction zone result in a relaxation of normal stresses at the upper surface and a consequent slow regional topographic uplift. Although the model is limited to two-dimensions, the observed time scales and orders of magnitude the observed variations are expected to be reasonably accurate.
The mantle under Tibet, India and the adjacent Indian Ocean reveals several zones of relatively high P-wave velocities at various depths. Under the Pamir-Hindu Kush region in northeastern Afghanistan and southern Tajikistan a localized northward-dipping slab is seen in the entire upper 600 km of the mantle and is apparently still attached to the lithosphere of the Indian plate, whereas under northern Pakistan it shows a roll-over structure with the deeper portion overturned and dipping southward. Farther east-southeast (e.g., in the vicinity of Nepal), the deeper (overturned?) portion seems separated from the Indian lithosphere. These upper mantle anomalies are interpreted as the remnants of sub-continental mantle lithosphere that went down when Greater India continued to converge northward with Asia after about 45 Ma.
The deeper high-velocity anomalies under the Indian sub-continent appear clearly separated from the shallower ones as well as from each other, and are inferred to be Tethyan slab remnants overridden by the Indian plate. They occur at depths between 1000 and 2300 km and form three parallel WNW-ESE striking zones. We interpret these zones as remnants of oceanic lithospheric material, which was subducted when the Paleo-and Neo-Tethys oceans closed between India and northern Tibet between Late Jurassic and earliest Tertiary times. The present-day latitudes of the deep slabs under the Indian sub-continent (10 - 30 degrees N) correspond to the paleolatitudes of the Cretaceous to earliest Tertiary southern margin of Asia before its collision with India and the subsequently continued northward movement of India. The slab remnants in the middle mantle occur therefore approximately under the ancient locations where they started their downward journey.
Project INDEPTH (International Deep Profiling of Tibet and the Himalaya) is a collaborative Sino, US, German, Canadian geoscience project aimed at elucidating the structure of the lithosphere beneath Tibet. The project began as a Sino-US collaboration in 1992 with the acquisition of 100 km of CMP and complimentary wide-angle deep reflection data in southern Tibet (INDEPTH 1). The principal result of this work was the imaging of the active Himalayan decollement to 270 km north of the Himalayan thrust front and to a depth of about 50 km. A greatly expanded phase of fieldwork was undertaken during the summers of 1995-95 (INDEPTH 2/GEDEPTH). This phase saw the advent of German and Canadian participation in the project, and included the acquisition of 200 km of additional CMP and wide-angle data, along with complimentary broadband passive seismic, magnetotelluric, and surface geologic data along and adjacent to the INDEPTH 1-2 route. This work produced a suite of observations, which together suggest that the middle crust beneath southern Tibet is partially molten. These include the MT observation that the crust below the INDEPTH 2 transect is electrically conductive below about 15 km, the passive seismic observation of a broadly coincident midcrustal low velocity zone, and the CMP/wide-angle observation of seismic bright spots coincident with the top of the midcrustal conductor. The existence of a midcrustal partial melt layer beneath southern Tibet is counter-intuitive given that the uppermost mantle beneath the region is cool (exhibits relatively high velocity and earthquakes). It is significant in that the existence of such a layer has substantial implications for the way crust deforms in orogenic belts, the way granites form, and on a broader scale, the way continental crust becomes chemically stratified. Other noteworthy results of the INDEPTH 2 work include the discovery of a gently dipping structure in the upper crust that cuts off the Yarlung-Zangbo suture at depth (wide-angle and MT observations), and the tentative observation of a north-dipping structure in the upper mantle beneath the Lhasa terrane that might mark the India/Asia "mantle suture" (receiver function profile).
INDEPTH 3 is currently underway in central Tibet. The program includes a year-long deployment of about 70 broadband, intermediate, and short-period seismic stations for passive seismic studies, a crustal wide-angle reflection/refraction experiment, broadband and long period MT recording, and geological studies. Key goals of the program are to test whether the inferred crustal partial melt layer is a plateau-wide phenomenon, characterize the uppermost mantle beneath the central plateau, and constrain the distribution and magnitude of Cenozoic uppercrustal deformation within the plateau.
In 1994 and 1998, scientists from German institutions joined colleagues from Chinese, American and Canadian institutions in the INDEPTH project in Tibet. In 1994, off-line wide-angle seismic recordings of the shots along the Sino-US reflection line indicated the presence of a reflector at about 20 km depth cutting the Indus-Yarlung suture. A similar structure was also recognized beneath the in-line seismic stations about 40 km to the east. From broadband seismic recordings along the 400 km long profile across the Indus-Yarlung suture in 1994, receiver functions analysis indicates Moho depths of 70-80 km and the presence of a second discontinuity at 50-60 km depth beneath the entire profile. North of the Indus-Yarlung suture at about 10-20 km depth a pronounced low-velocity zone was detected, which together with evidence from other seismic and magnetotelluric data was taken to infer the existence of a partially molten layer in the middle crust beneath much of southern Tibet. On the other hand, the locations of the upper mantle discontinuities at 410 and 660 km depth are in agreement with the IASP91 global reference model. A combined analysis of receiver functions from the 1994 broadband data and data from the 1991-2 Sino-US PASSCAL broadband experiment across Tibet suggests a boundary diverging from the Moho about 100 km north of the Indus-Yarlung suture and dipping northwards in the upper mantle under central Tibet and a south dipping segmented structure in the upper mantle under northern Tibet. A clearer picture of the upper mantle under central Tibet should be provided by broadband seismic data presently being recorded along a 400 km long N-S profile across the Banggong-Nujiang suture in central Tibet. First results from a few broadband seismic recordings between 1997-8 and wide-angle seismic data recorded from large shots in 1998 along this 400 km long profile will also be presented.
Evolutionary models proposed for the Indo-Asian collision usually transcend our ability to test them against geological knowledge of these remote regions. When theory outpaces data, an effort is often later required to modify or reject popular models that have become inconsistent with observations. The Indo-Asian orogen appears to have responded to collision by discrete changes in accommodation mechanisms over time, even though convergence was continuous and the boundary conditions relatively constant. Our traditional inclination to interpret episodic phenomena in the geological record in terms of discontinuous processes (e.g., pulse of the earth) needs revision in light of the growing appreciation that complex physical systems driven by structureless inputs can exhibit highly intermittent, non-linear responses (e.g., turbulence). The Tibetan crust began to thicken upon collision of the Lhasa Block with S Asia producing a fold and thrust belt between 144-110 Ma that remained elevated until the onset of the Indo-Asian collision. The first post-collisional evidence of crustal thickening is thin-skinned thrusting in the Tethyan Himalaya, Eo-himalayan metamorphism, and deformation in the Fenghuo Shan and Nan Shan regions of N Tibet. The latter, initiated between 45-32 Ma, may have led to development of the left-lateral Red River fault system along which ~500 km of left-slip motion occurred from ~35-17 Ma. At 31±1 Ma, crustal thickening began in S Tibet along the Gangdese Thrust, moving southward to the Himalaya shortly thereafter in a series of S-directed thrusts (MCT, MBT, MFT) that sole into a common decollement. Slip along the decollement produced the Himalayan granites between 24-9 Ma. The STDS apears to have ben active between 17-11 Ma. The pattern of foreland thrust propagation was interrupted by the N-directed Renbu Zedong Thrust, active in S Tibet between 19-11 Ma. Tien Shan thrusting and thickening in the W Kun Lun also initiated during the early Miocene. Thickening in NW Tibet, apparently related to transtension along the Atyn Tagh fault, began during the middle Miocene. Following initiation of the MBT, the MCT reactivated between 8-4 Ma producing the classic Himalayan inverted metamorphism. By 9±1 Ma, Tibet had begun to differentially extend E-W via a series of N-S graben that extend across the plateau. Models for the evolving Indo-Asian orogen include those that predict wholesale uplift (mantle delamination, delayed under-plating), progressive growth (Indian under-thrusting, Asian under-thrusting, continental injection,), lateral responses (orogenic collapse, horizontal extrusion), and inheritance of an elevated terrane (multiple collision, intra-arc thickening). The pattern of crustal displacements we infer is inconsistent with most permutations of these single mechanisms but instead requires a specific, time-dependent transfer among several of these processes, often with multiple mechanisms operating simultaneously. The parameters that appear most important in dictating which mechanisms are dominant at any one time are: the location and geometry of pre-existing lithospheric weakness, the distribution of topography before and during the collision, the geometry of the indenter and extruded blocks, the magnitude of boundary stresses, and the age of the lithosphere. The clearest lesson emerging from our study of the Indo-Asian collision is that the continental lithosphere's complex history and geometry exerts a powerful control on continuous plate convergence being manifested in the geological record as episodic phenomena. A corollary of this 'intermittent dynamic' interpretation is that the chain of geologic events we infer to have occurred during the Indo-Asian collision is unlikely to have direct relevance to the understanding other, and perhaps more ancient, orogenic belts.
KEYNOTE
A02 : 3A/11 : G0
The frame of the India-Asia convergence is set by the apparent polar wander paths of both plates. Microplates in between and the northern margin of the Indian Plate have to be considered for reconstruction of crustal shortening and block rotation processes in the collision zone. Primary remanences from former Gondwana blocks north of the IYS suture (e.g. Lin & Watts 1988) provide a rough information on the timing of their relative movement, their stepwise accretion to Asia, and the latitude of the former southern Asian margin before the India-Asia collision. Present studies within the GEODEPTH program should improve the data base in central Tibet. Here we focus on the crustal deformation of the Indian Plate. Klootwijk et al. (1985) introduced an oroclinal bending and rotational underthrusting model to estimate a minimum magnitude of shortening. Only few primary results are available from the key zone for such studies, the Tethyan Himalaya. They controversely indicate shortening between several hundreds of km (Besse et al. 1984) to about 1500 km (Patzelt et al. 1996). The lack of primary data forces concentration on secondary remanences. Stable secondary pyrrhotite remanences are frequently found in the Tethyan Himalaya and were probably acquired during last cooling below 300°C. Quantitative reconstruction of rotational shortening is affected by local and meso-scale block rotations superposed on regional movements. Such effects are clearly seen in the NW, where the Indian indenter creates a complicate rotation pattern on both sides of the IYS, which however, can be still interpreted by tectonic processes. In order to enhance the data base for quantitative modelling of rotational shortening, we presently conduct sampling of low grade metamorphic limestones from the central Tethyan Himalaya between about 77°E and 86°E. Sampling localities are spread in equal distances as much as possible to recognise the regional rotation pattern, with partly closer spacing to detect local variations. Already processed samples indicate that local disturbances exist only within the limit of significance in most areas. Systematic block rotations support a rotational shortening and a more significant quantitative approach seems to be possible when all data are available.
Lin J & Watts DR, Phil Trans R Soc, 327, 239-262, (1988).
Klootwijk CT, Conaghan PJ & Powell CMcA, Earth Planet Sci Lett, 75, 167-183, (1985).
Besse J, Courtillot V, Pozzi JP, Westphal M & Zhou YK, Nature, 311, 621-626, (1984).
Patzelt A, Li H, Wang J & Appel E, Tectonophysics, 259, 259-284, (1996).
The western boundary of the Philippine Sea Plate (PH) with Sundaland (SU) in the Philippines corresponds to a wide deformation zone that includes the stretched continental margin of Sundaland, and the Philippine Mobile Belt (PMB) extending from Luzon to the Molucca Sea. The GPS GEODYSSEA data are used to decipher the present kinematics of this complex area. One of the main results is the quantification of overlapping subductions on both sides of the PMB. To the west, convergence decreases from about 90-100 mm/yr along the Manila trench to a few mm/yr across Mindoro in the zone of collision of Palawan with the PMB. Transfer of 55% of this Manila Trench convergence motion to the Philippine Trench to the east is accommodated principally by rapid counterclockwise rotation of eastern Luzon. Southward, convergence decreases regularly along the Philippine Trench from 54 mm/yr near 13°N to 32 mm/yr east of Mindanao. Inside the PMB, the slip rate of the left-lateral Philippine Fault, which enables complete partitioning to occur at the trench, increases northward from 22 mm/yr at 7°N to 35 mm/yr in the Visayas. As a result, the East Philippine sliver, east of this fault, moves north and transfers this motion to the East Luzon block counterclockwise rotation. West of the Philippine Fault, the Visayas block in central Philippines, is rotating rapidly clockwise, transfering most of the remaining convergence to the Negros/Cotabato trenches on its western border, up to 32 mm/yr near 12°N and 50 mm/yr near 8.5°N. These estimates ignore the amount of convergence absorbed to the west, at the limit of undeformed SU. Two opposite rotations on both sides of the left-lateral Philippine Fault, clockwise to the southwest and counterclockwise to the northeast, illustrate the PMB internal deformation. As a result, the northern part of the Visayas block acts as a cog wheel with respect to the Luzon block. It is remarkable that this system is such that subduction is everywhere perpendicular to the trenches. South of Mindanao, within the Molucca sea, the overlapping divergent subduction system of the PMB is relayed by the convergent Sangihe-Halmahera subduction system. This one accommodates 80 mm/yr of the 105 mm/yr PH/SU convergence and about 15 to as much as 20 mm/yr of the remaining convergence are absorbed far to the west, within the northern Borneo margins.
We present an estimate of the modern erosion of the Himalaya based on riverine fluxes. The dissolved flux related to chemical erosion is estimated on the basis of yearly averages of the rivers. It is 32 and 29*106 ton/yr for Brahmaputra and Ganges respectively. The fluxes of suspended matter measured in Bangladesh are 402 to 608*106 ton/yr for Brahmaputra and 328 to 548 *106 ton/yr for Ganges. Isotopic tracers show that the large majority of these particles is derived from the Himalaya. However the total erosional flux must take into account bedload transport and sedimentation in the floodplain which cannot be simply measured. Their importance can be estimated using the contrast in chemical composition between the suspended load and the bedload or the accumulated sediments due to mineral sorting during transport.
Assuming steady state erosion in Himalaya, the chemical composition of the total erosional flux equals that of the average source rock. This later is relatively well established for silicate formations based on ca. 200 analyses of the Himalayan formations. Compositions of the riverine end members are based on suspended- and bedload sampled in Bangladesh. The composition of sediments accumulated in the floodplain is derived from an average composition of Siwaliks. The principal process driving chemical differenciation is the sorting of phyllosilicates with respect to quartz and feldspars in favor of the suspended load. Si/Al ratios are: 4.2 for the source rock, 2.7 for the suspended load, 6.5 for bedload and 7.8 for the Siwaliks. Mass balance calculation is performed for two extreme case: no bedload or no sediment sequestered in the flood plain. On this basis, the suspended load represents between 40 and 65% of the total erosional flux. Using the fluxes of suspended load mentioned above, total erosional fluxes can be calculated and expressed in term of erosion rate. They are 2.8 ± 1.1 mm/yr for the Brahmaputra Himalaya and 1.8 ± 0.6 for the Ganges Himalaya. The higher erosion rate in the Brahmaputra basin is likely related to its higher runoff.
Using an Atmospheric General Circulation Model, we simulate most of the spatial evolutions of the Asian monsoon only accounting for the changes of paleogeography including continental drift, orogeny and sea level change. The paleogeographic changes modify drastically the climate over the central and southern Asia between the Oligocene and the present day. The retreat of an epicontinental sea (the Paratethys) warms central Eurasia in summer. The heating of this area and the uplifts of the Tibetan plateau and of the Himalayas deepen the Asian low pressure cell and displaces it northwest subsequently shifting precipitation from Indochina towards the southern flank of the Himalayas. The agreement with proxy data is good. Therefore, our modeling studies support a shift and a strengthening of the Asian monsoon during the late Tertiary rather than a real onset. We also investigate the respective impact of the Paratethys shrinkage and of the Tibetan plateau uplift through sensitivity experiments and prove that the Paratethys retreat plays an important role in the monsoon evolution.
Numerous previous studies of Sm/Nd systematics in crustal rocks and sediments have shown the importance of sedimentary recycling for the evolution of the continental crust. The proportion of old sediments vs. granites in the continental crust is a key parameter to constrain if we want to address important questions such as the secular evolution of the continental crust composition or the long term consumption of atmospheric CO2 by the weathering of continental rocks. Recent weathering studies suggest that first cycle weathering is more efficient, in terms of CO2 consumption rates, than the weathering of rocks having passed through several weathering cycles. Thus, the role of sediment recycling in the evolution of climate needs to be considered.
We propose in this study to use the geochemistry of large river erosion products to estimate sedimentary recycling rates. Our study case is the major orogenic zone of East Asia, which is drained by the following large rivers: Huanghe, Changjiang, Xijiang, Hongha and Mekong rivers. The basic idea is to establish mass budgets for the dissolved and solid products of silicate erosion in rivers. In particular, insoluble over soluble element ratios, such as Sm/Na or Th/K ratios are especially sensitive to sedimentary recycling and are used to calculate the (minimum) proportions of recycled sediments involved in present day erosion. These chemical mass budgets are highly sensitive to the amount of sediments transported per year by the rivers (physical denudation rates). For the rivers under consideration, a good agreement exist between the multiyear average of river suspended sediment concentrations and the mass accumulated in the sea by these rivers over the last 2 million years (Métivier and Gaudemer, 1998). We thus consider these rates as reliable estimates of physical denudation.The recycling rates calculated using the different chemical ratios are close to 50% for the Huanghe river (mostly influenced by loess) and range from 70% to 95% for the other river systems. To our knowledge, these values are the first estimates of present day recycling rates and highlight the importance of sedimentary cannibalism in one of the major orogenic zone of the Earth.
On the Northern part of the Tibetan plateau, in a 1300 km long - 300 km wide area roughly parallel to the margin of the plateau, lies Neogene to Quaternary post-collisional potassic magmatism. On Satellite images we have mapped different features of this young volcanism: small cones, large pyroclastic deposits and lava plateaus, and have estimated the volume of volcanic products. In 1997, during a Sino-French mission, we sampled volcanic rocks in the Kokoshili Shan (latitude : N35°30', N36°30', longitude : E88°, E90°), near the western extremity of the Kunlun fault. Field observations of the structure of the edifices and the study of the samples show that this volcanism ranges from mafic to silicic. The more mafic deposits, usually cinder cones, contain deep crustal xenoliths. Thin sections show two types of texture and mineralogy. The mafic rocks contain abundant olivine, some of which may be xenocrystic, plagiocase, pyroxene and oxydes, and the groundmass consists sometimes of glassy matrix with laths of plagioclase and pyroxenes. The more silicic material contains plagioclase, quartz, some alkali feldspar and some biotite in a groundmass of granular texture.
The age, location and composition of these volcanic rocks provide constraints on their sources and on the degree of interaction with the warm crust of mafic magma that ascends from the mantle. We interpret these results in the framework of a fluid dynamic model of crustal melting induced by the passage of basaltic dykes, and show how the depth and amount of melting depends on the deep thermal structure of this thick continental crust.
Understanding the role of the subcontinental lithospheric mantle in accommodating deformation in collisional regions is fundamental to our understanding of continental tectonics as a whole. The formation of the Tibetan plateau has been linked with the Indo-Asia collision, and it provides an ideal setting in which to investigate these tectonic processes.
Changes in tectonic regime may be reflected by spatial and temporal variations in the associated magmatism. Previous studies have identified small-volume, potassic volcanics across the Tibetan plateau, and these are considered to represent partial melts of the mantle lithosphere (Turner et al, 1996). This broadly supports geodynamical models of convective removal and thinning of the mantle lithosphere that have been proposed to explain the elevation history and active extension of the plateau.
Here we present initial elemental and isotopic data for volcanics collected from southern Tibet. These lavas are predominantly trachyandesites with subordinate dacites and rhyolites. Samples are characterised by extreme enrichment in incompatible elements, and form distinct arrays in La/Yb - Tb/Yb space, implying variations in both source and melting regime. Isotope data shows that samples fall into several broad fields defined by radiogenic 87Sr/86Sr and unradiogenic 143Nd/144Nd. Preliminary laser Ar-Ar ages for these samples show that lavas from southwest Tibet are at least 24 Ma - considerably older than samples from southeast or northern Tibet, which range from 8-14 Ma.
These findings indicate that the geochemistry of these volcanics varies both temporally and spatially across the Tibetan plateau. The implication of these results is that models of Tibetan plateau evolution are considerably oversimplified and that the true picture is far more complex. New geochemical data from the 1998 field season, including a north-south trending lamprophyre swarm, will be included in this discussion.
Turner S P, Arnaud N, Lie J, Rogers N, Hawkesworth C, Harris N, Kelley S, Van Calsteren P & Deng W, J. Petrology, 37, 45-71, (1996).
Recent geophysical data constrain fairly well the deep structure of the Himalayas and Tibet. In particular, they reveal the Indian crust subducted to about 80 km depth. Eclogites discovered in the western Himalayas indicate that the Indian crust has already reached at least this depth at the beginning of the collision (ca. 50 Ma). Since that time, hundreds of kilometres of the Indian crust have been subducted under the Himalayas. The major challenge now is to organise all this information into a physically and geologically consistent evolutionary model, considering that the history of the Himalayas cannot be separated from that of Tibet because they both belong to the same geomechanical system. We started to develop such a model based on the new results of 2-D thermo-mechanical experimental modelling. The principal successive stages of this model are as follows: (1) subduction of the Indian crust to 200-250 km depth followed subduction of the oceanic lithosphere; (2) failure of the subducted crust at ~80 km depth and rapid buoyancy driven uplift of the crustal material from this depth to 20-30 km; (3) break-off of the Indian subducted lithospheric mantle attached to the previously subducted oceanic lithosphere; (4) shallow-angle (horizontal) subduction of the Indian lithosphere under Asia over a distance of a few to several hundreds of kilometres; (5) delamination and roll-back of the mantle layer of the Indian lithosphere until its break-off beneath the Himalayas; (6) failure of the Indian crust in front of the mountain belt (formation of the MCT) and underthrusting of a new portion of the Indian lithosphere under Tibet for a few hundred of kilometres. At the beginning of this stage (6), the crustal slice corresponding to the Crystalline Himalayas undergoes "erosion-activated" uplift and exhumation. The overriding Asian plate is horizontally undeformable in this model and plays a passive role. Its deformation in reality is essentially three-dimensional and thus a 3-D approach will be necessary.
We employ P-to-S converted teleseismic waves recorded by temporary broadband networks for imaging the seismic discontinuities in both the crust and the upper mantle across the whole Tibet. The data used consist of 353 teleseismic receiver functions (RF's) from INDEPTH II/GEDEPTH and 406 RF's from PASSCAL These observations are jointly interpreted together with published teleseismic tomography and gravity models.
A north-dipping interface, called here Zangbo Conversion Boundary (ZCB), is imaged starting 50 km north of the Zangbo suture at the Moho and reaching 200 km depth underneath the Bangong suture. The average amplitude of the converted wave at this boundary (4% of the direct P-wave) is modeled by an S-wave velocity increase downward by 6-8%. Such a velocity change may be explained by a temperature contrast of 500-700°C (considering the effect of anelasticity and chemical reactions) together with chemical differences (a higher Mg content in cratonic mantle). The ZCB probably separates the relatively cold and depleted cratonic lithospheric mantle of India with a temperature of 600-800°C at its top from the lower lithosphere of Asia with a temperature of 1100-1300°C. Teleseismic travel time tomography in this region reveals a North-South P-wave velocity contrast of about 4-5% at a depth of 200-250 km (Witlinger et al., 1996). Most of this contrast can be again explained by a lower temperature and higher Mg content of the Indian lithospheric mantle and the remainder (0-2%), by an upwelling of hot mantle (if any) with a temperature excess of 100-200°C or less. Our model of the gently subducting Indian lithospheric mantle is also consistent with observed gravity and surface topography in southern Tibet (Jin et al.,1996).
Under northern Tibet a less clear south dipping segmented structure is also identified by converted waves. This zone of strong and complicated conversions may indicate an ongoing process of destruction/southward-subduction of the Asian lithospheric mantle which appears to be more complicated than simple convective lithospheric thinning in a viscous media. Together these observations are interpreted as different styles of detachment of the Indian and Asian lithospheric mantles caused by a difference in their composition and buoyancy, where the Indian lithosphere is gently subducting but remains stable and the Asian lithosphere is in the process of destruction.
Jin et al., JGR, 101,11275-11290, (1996).
Witlinger et al., EPSL, 139, 263-279, (1996).
We present new paleomagnetic results obtained at 39 sampling sites from 5 sections of Tertiary redbed formations: 2 Eocene formations from the Qiangtang block of Tibet (Xialaxiu locality; 32.8°N, 96.6°E) and the Xining basin of Qaidam (Xining locality; 36.5°N, 102.0°E), and 3 Neogene formations from the Xining basin (Jungong locality; 34.7°N, 100.7°E) and the Kunlun block (Tuoluo lake and West Yushu localities; 35.3°N, 98.6°E and 33.2°N, 96.7°E respectively). Thermal demagnetization of the samples allowed to obtain a High Temperature Component which we interpret as the primary magnetization in all but one (Tuoluo lake) localities. For the 4 other localities, the paleopoles lie at 52.6°N/ 352°E (dp/dm=6.0°/10.7°) for Xialaxiu, 61.6°N/211.3°E (dp/dm=9.7°/16.1°) for Xining, 66.0°N/228.6°E (dp/dm=3.6°/6.9°) for Jungong and 53.9°N,/205.4°E (dp/dm=5.6°/10.0°) for W. Yushu. As in previous studies of Tertiary formations from Asia, the paleomagnetic inclinations we obtained are systematically shallower (by 22°-26°) than the magnetic field computed from the Eurasian APWP at 10 and 20 Ma for Neogene formations and at 50 and 60 Ma for the Eocene ones. This inclination shallowing, known as the «inclination anomaly», has previously been interpreted as either due to a magnetic field anomaly over Eurasia in the Tertiary, and/or due to compaction processes in sediments. Based on a detailed discussion of these possible causes, and on a compilation of Eocene data from the South China Block, Tibet, Central Asia and Kyrgyzstan, we conclude that these are probably not the main causes of the discrepancy. We propose instead that the «anomaly» arises from an inadequate description of Siberia craton movement based on the Eurasia APWP, the causes of which could be a non rigid behaviour of the Eurasia plate in the Tertiary. Combination of this with intracontinental shortening of Asia under the penetration of India provides a full explanation for the anomaly. Finally, a comparison of our Neogene results with the India APWP demonstrates that the formations we sampled can certainly not be assigned an age as young as Neogene. This addresses the critical problem of dating of the Tertiary redbed formations in Asia.
Thermal models based on simple sliding block kinematics have been successful in reproducing the peak P,T conditions of metamorphism in the Higher Himalayan Crystalline series. However, these models generally assume a constant rate of erosion in time and space and thus may not give realistic exhumation histories. In particular, the presence of aseismically active crustal thrust ramp beneath the high Himalayas north of Kathmandu is suspected and we investigate the possible consequences on the thermal structure and PTt paths. This ramp connects the main seismic decollement to the south with the main ductile decollement to the north and corresponds to a temperature range of 350 to 500°C. We computed temperature fields and PTt paths for steady-state kinematic models with and without ramp and with and without underplating. Uplift above the ramp only causes a small and local perturbation of the geotherms which would probably be undetectable with available geothermometers and geobarometers. However, geochronological data could be radically altered. We examined the example of the ~100 muscovite 40Ar/39Ar ages (corresponding to 350-400°C) and 10apatite fission-track ages (about 100°C) in a section extending across the Himalayan thrust system from ~50 km south to ~ 30 km north of Kathmandu. The muscovite data show a continuous trend of progressively younger ages from south to north, from a maximum of 22 Ma to a minimum of 5 Ma without change in crossing the Main Central Thrust (MCT). Models with a flat decollement fail to reproduce this feature as they give constant (±2 Myr) closure ages over the HHC section but models with a ramp all give significantly younger ages towards the north with an age minimum located above the ramp. We conclude that the crustal ramp is probably a steady state feature that influences the distribution of erosion and exhumation rates since at least 15 Ma. The oldest measured 40Ar/39Ar ages cannot, however, be reproduced with steady-state models and may be the consequence of exhumation along the South Tibetan Detachment or an older geometry of the thrust system not accounted for in our model. In the models, the Lesser Himalayan Series correspond to material continuously underplated at the ramp and possible reactivation of the MCT after underplating is ignored. Although simplified, this mechanism may explain some of the observations, such as why muscovite closure ages appear more or less contiuous across the MCT zone.
The Himalaya undergoes at present, about one-third of the today convergence between India and Eurasia (58 +/- 4 mm/a). The present-day deformation of the Himalaya is characterised by big earthquakes (nearly half of the chain has ruptured over the last century). The historical sismicity of Nepal indicates the occurrence of big earthquakes in eastern and central Nepal, western Nepal being characterised by a lack of recent big earthquake. To study the present-day deformation of western Nepal a GPS network consisting of 29 sites, was installed in central and western Nepal, measured in 1995, partially remeasured in 1997 and totally in November 1998. Fist data indicate 15 mm/yr of N180° convergence between Higher Himalaya and the Indian border south of Nepal and suggest a 4 mm/year E-W extension. The velocity field is consistent with creep on a dislocation striking N120° and dipping 9° to the north. This dislocation is locked at 17 km deep beneath the Lesser Himalaya and affected by a 19 mm/yr thrust component and a 7.5 mm/yr right lateral component. These results suggest an oblique underthrusting of the Indian crust below the High Himalaya of western Nepal consistent with N-S shortening across the arcuate shape of the Himalaya of Nepal. A N170 contraction and a N80 extension is observed for sites located East of 83°30, where quaternary faults (Darma-Bari Gad fault system and Thakkhola graben) delineate a crustal wedge. This wedge is located in the continuity of the Karakorum fault and may segment the Himalayan thrust belt. During this meeting, the new results obtained from the comparison 1995-1997-1998 will allow to precise the localisation and geometry of the dislocation surface and the amounts of displacement along the presumed active faults.
Processes of exhumation of HP are discussed in the light of the study of the Tso Morari dome (E-Ladakh, NW Himalaya). Petrological, structural and geochronological investigations show that the exhumation of the HP Tso Morari continental unit is a discontinuous process from a minimum depth of 70 km up to the upper crustal level. It is controlled by the major plate tectonics changes that occurred during the India-Asia convergence.
The beginning of the exhumation of the HP Tso morari unit is first characterized by a rapid isothermal and vertical extrusion (> 4 mm.yr-1) from 70 to 40 km depth. This active exhumation occurs during the upper Paleocene (55 ± 7 Ma) while the subduction of the Indian margin is still active. The association of the HP rocks with soft serpentinized mantle rocks and the tectono-metamorphic evolution of the Tso Morari suggest that the beginning of the exhumation is controlled by the return flow developed along the down-going subducting Indian plate. The coaxial D1-D2 deformations related to the vertical extrusion of the dome throughout the mantle are independent of the deformations occurring at the same time close to the surface. Since 48-45 Ma to 30 ± 1 Ma, from the base of the Indian crust (30-40 km) up to upper crustal levels, the slower vertical exhumation (~ 2 mm.yr-1) of the HP Tso Morari dome has been ruled by ductile normal shearing (D3). This second stage of the exhumation is associated with a slight temperature increase, and is contemporaneous with the underthrusting of the High Himalayan Crystallines below the Tso Morari dome. The India-Asia collision governed the second exhumation stage of the HP rocks. It implies a strong horizontal squeezing of the HP unit and the surroundings between the two landmasses, that lead to the pinching of the Tso Morari eclogitic dome and to its upward extrusion close to the surface.
In November 1996 we have carried out a gravity survey in the framework of the Nepalese-French IDYLHIM cooperation program. Using two Scintrex CG3 meters, we collected more than 150 new gravity measurements along two profiles perpendicular to the range, in the vicinity of Kathmandu. Data were tied to the IGSN71 network in Kathmandu. Accurate geographic coordinates were obtained from the Nepalese Geodetic Survey benchmarks and GPS positioning. We estimated inner zones terrain corrections in the field and computed outer zones ones, up to 167 km, using various DTM. The overall accuracy of the resulting complete Bouguer anomaly ranges between 0.5 and 5 mGal depending on the terrain roughness. Our data complemented with those collected during the 1982 French survey in Tibet, with data from the Bureau Gravimetrique International and with previously published ones, allow to get continuous Bouguer anomaly profiles from India to the Tarim Basin. At short wavelengths, these data are used to assess the geometry of the system and specifically the main thrusting faults such as the Main Central Thrust (MCT), the Main Boundary Thrust (MBT) and the Main Frontal Thrust (MFT). Especially, the use of 1D wavelets transforms on these profiles allows us to put quantitative constraints on several geological discontinuities. At long wavelengths, the observed gravity anomalies are consistent with a flexural model of the lithosphere. The modeling is performed using a plane strain finite element model (ADELI) that accounts for thermal and strain dependant rheology and for P,T conditions. We also assess the possibility for petrological changes during the underthrusting; in particular, eclogitization of the lower crust. We find no gravimetric signature for such effects.
Data from 12 broad-band magnetotelluric (MT) soundings, acquired across Central Nepal in November and December 1996, are analysed to determine the electrical structure of the crust of the active Himalayan region. We have estimated the MT impedance tensors by a robust method for frequencies between 0.001 to 500 Hz. A tensor distortion analysis, using the Groom-Bailey decomposition technique, shows that the electrical distortions are weak in the south and strong in the north. This indicates that Central Nepal's present day geoelectrical structure is of growing complexity from south to north, close to one-dimensional behaviour under the Gangetic plain and three-dimensional in the north. Two dimensional joint inversion of TE- and TM- mode data after decomposition reveals from south to north : 1) a depth to the Indian basement of about 5 km under the Siwaliks, in good agreement with seismic data 2) an upper Indian crust more conductive than the lower Indian crust 3) a shallow very resistive body under the Kathmandu klipp extending to depths of about 7 km, likely related to Ordovician granites and underlain by a thin conductor 4) a 20 km deep conductor near the boundary between the Lesser and Higher Himalayas coinciding with the Himalayan crustal ramp and located near a seismically active zone 5) an increasing of the deep conductivity from south to north. Comparison of conductivity models with other geophysical data and thermal modelling will also be presented.
The contact between the Kohistan-Ladakh Arc and the Asian paleomargin (Karakorum) is defined in Indian Ladakh as the Shyok suture zone. Westwards, in the Skardu area (Pakistan) the Arc-Karakorum contact is an imbricate South vergent thrust system, the MKT (Main Karakorum thrust), with thrust contacts underlined by serpentinite strips. Still farther West, North of the Nanga-Parbat /Haramosh spur (consisting of Himalayan high grade gneisses), the MKT is no more a thrust, but a recent vertical and rectilinear fault. In this Western segment, the actual Arc-Asian margin boundary is unclear: geochemical studies suggest that some amphibolitic rocks of the South Karakorum metamorphic pile are fragments of the Ladakh Arc, obducted onto the Asian margin, then metamorphosed and deformed during the early South vergent tectonics (from 67 ± 2 Ma to 35 ± 5 Ma: period of the apparently syn-tectonic Mango Gusar granite emplacement). Late metamorphic evolution along the MKT/Shyok suture is marked by local high temperature anomalies, known in the Shyok suture zone and best studied in Skardu area: here, late metamorphic isogrades define an E-W elongated pattern, oblique to the MKT and the earlier metamorphic pattern. They follow a line of kilometric scale domes, from the Baltoro glacier in the East to the Nanga Parbat Himalayan spur in the West. Various thermochronological data from gneisses in the domes or from syn-doming plutonic cores (Hemasil) indicate young ages: most of them are less than 10 Ma, to the North as well as to the South of the MKT. Thus they can be interpreted in the light of the very recent to present stress and kinematic data. The recent tectonic pattern was investigated using strain trajectories, as marked by the late metamorphic foliation, and the stress directions obtained by inversion of post metamorphic striated planes:- The isogrades pattern, the E-W alignement of domes and the lack of small scale extension shear criteria suggest that the thermal anomalies are linked to a transpressive deformation regime, with along strike and down dip movements on the MKT/Shyok zone,- The MKT, as the dextral Karakorum fault, offset the young metamorphic patterns confirming that the MKT is not the suture, but a very young reactivation of a former contact,- The stress pattern is quite heterogeneous: the principal stress can be vertical in the core of the domes (f.i. the Haramosh dome), or horizontal and either parallel or perpendicular to the MKT (shortening perpendicular and parallel to the belt, very irregular exhumation). As a whole, stress and thermal patterns can be linked to a transpressive deformation regime, with partitioning of the deformation during the oblique Indian-Asian convergence, and control of inherited major crustal discontinuities.
In addition to the magnetotelluric data acquired during the 1996 IDYLHIM survey, we have computed the magnetic field transfer functions, also known as induction vectors, using broad-band data at 12 sites as well as long-period records from four stations. The induction vectors are very small south of Kathmandu, indicating a primarily one- or two-dimensional conductivity structure there. North of Kathmandu however, the vectors are much longer and the opposite of their real parts (Parkinson vectors) generally point towards the Main Central Thrust (MCT): this is particularly spectacular for the Kakani and Pipaltar sites, on either side of the MCT, for which the Parkinson vectors are in near-perfect opposition. Qualitatively, the data show that the MCT does not extend further north than Syabru Bensi, since the Parkinson vectors there point towards the south at all frequencies. This high conductivity associated to the MCT is likely to be related to fluid circulation. To gain more insight into the large-scale conductivity structures around central Nepal, we have computed the response of 3D models for two end-members and present here the first results of our modelling of the magnetic transfer functions. The first end-member is an attempt to quantify the effect of the rough topography, especially along the Trisuli river valley where most of the data were acquired, on the magnetic induction vectors. To do so, we use a 30-second interval digital elevation model and consider an homogeneous earth from which we calculate the response using an integral equation scheme. For the second end-member, we neglect the effects of topography and express lateral variations of conductivity in terms cells of varying integrated conductance. The response is then computed using a thin-sheet algorithm. We show that a satisfying degree of fit can be obtained assuming that the MCT plays an important role in channelling electrical currents in central Nepal. 3D modelling from more realistic intermediate models is currently in progress.
Neodymium model ages are useful tools for deciphering ancient tectonic associations in complex polymetamorphic terrains. Here we present the results of a whole-rock Nd isotopic study of two contrasting regions of the western Himalaya which demonstrate the utility of the approach on the scale of a single orogen. New data for high-grade metasediments from Zanskar yield model ages that are similar to the High Himalayan Crystalline Series (HHCS) (TDM = 1.2-2.0 Ga, (sum)Nd = -6 to -16), and quite distinct from values from the Lesser Himalaya (TDM = 2.3-3.4 Ga, (sum)Nd = -18 to -27). Hence these two lithological sequences can be recognised for 2000 km along strike of the orogen. New Nd data for basement of the Nanga Parbat Haramosh Massif (NPHM) at the western extremity of the Himalaya (TDM = 2.3-2.8 Ga, (sum)Nd = -18 to -30) suggest that these rocks are not equivalent to the HHCS, as previously supposed, but have affinities with the Lesser Himalaya. A thin metasedimentary cover sequence on the margins of the NPHM is isotopically indistinguishable from the HHCS (TDM = 1.6-1.8 Ga, (sum)Nd = -10 to -14). The prior misidentification of the provenance of the NPHM stems from its high metamorphic grade, characteristic of the HHCS, but in this case related to the unique Neogene history of the NPHM, which has exhumed a deeper structural level of the LHF than is seen elsewhere. These findings have implications for the identification of terrains in older orogenic belts, where contrasting P-T-t histories may mask their pre-orogenic correlations.
The Phenomenon of extension in the India-Asia collisional belt was first noted nearly twenty years ago (Burg et al. 1981). The Southern Tibet Detachment System (STDS), subsequently has been identified all along the Himalaya, with an ever increasing indication that the STDS is a complex system of detachment horizons, and not a simple horizon with a restricted period of activity. This may be expected in a large mountain belt, and the STDS (the cat) has not run away. However, the orogen-wide relationship of the STDS with large bodies of leucogranite (that are ubiquitous along the Himalaya) has not been so easy to catch. Although this relationship is a spatial association, and from the Everest and Manaslu areas (and others) it seemed initially that there was a "specific period" when most of the anatexis and STDS displacement occurred, further mapping and geochronology has allowed the mouse to get away. Khula Kangri suggested that the "specific period" was later in the Bhutan Himalaya, or that there was a second occurrence. A northward younging between the High Himalayan and the "north Himalayan" granites was subsequently proposed and even a slip rate on the STDS was attempted, but the mouse escaped. Structural and metamorphic data have caught the STDS in many parts of the Himalaya, whose associations with subsequently dated melt products show the relationship to have continued since Eocene and Oligocene, with no clear constraint of an end. The INDEPTH results show that the "anatectic front" may have simply pushed on northward. The "anatectic front" (accompanied by large exhumation rates but no "detachment") is also seen at the Himalayan Syntaxis. The Nanga Parbat group results have uncovered a pattern of protracted melting (that includes the Upper Miocene age so common in the main Himalaya). A procession of plutonism interacting with shear zone displacement has a long history within the two main (conjugate) shear zones at Nanga Parbat, and the Indus River, that has maybe focused the deformation, has re-located the sediment to tell the tale. The last twenty years have seen the mouse (the timing of anatexis) run away frequently, however the education is to be able to recognise the signs (and then role) of anatexis interacting with deformation to unravel a long and complex interplay. This melting deformation interaction is a rising issue in studies of granites worldwide, and indeed the Himalaya are frequently turned to for example. But the story goes on and if the mouse is truly caught in the next twenty years, it will be a bad thing indeed.
Burg, J-P, Brunel, M, Chen, GM and Liu, GH, Deformation of the leucogranites of the Crystalline Main Central Thrust sheet in southern Tibet (China) Mitt. Geol. Eidgen. Techn. Hochs. Univ. Zürich, Neue Folge, 239a: 49-51, (1981)
The Higher Himalayan Crystalline (HHC) mainly comprises amphibolite to anchizone grade late Precambrian to early Cambrian metasediments (Haimantas Formation), which have been intruded by Cambro-Ordovician granites. Evidences for a crystalline basement of this huge sedimentary deposits are still missing. This sequence continues through the Paleozoic-Mesozoic up to the Early Tertiary (i.e. Tethyan Zone - TZ) and is locally detached from the HHC along a normal fault system (i.e. South Tibetan Detachment Zone - STDZ). This normal fault system acted synchronously with the Main Central Thrust Zone (MCTZ), which thrust the HHC over the Indian continent in the Early Miocene.
The low-grade metamorphic rocks of the TZ in the Pin Valley (Spiti, N-India) form a SW-vergent fold-thrust belt, with steep, NE dipping axial planes. 40Ar/39Ar ages of newly formed illite scatter between 42 and 45 Ma, indicating that this deformation did not correlate with the MCTZ/STDZ phase but probably formed after the India-Asia collision as a response of crustal thickening. Early Miocene and later brittle deformation is only recorded by a spaced cleavage with shallow dipping axial planes, local sets of extension gashes and minor slickensides, which did not significantly contribute to the overall deformation of this area.
In order to quantify the amount of shortening of the early fold-thrust deformation phase, a 25 km long line-length balanced cross-section through the Pin Valley, essentially perpendicular to the fold-axes and parallel to the thrusting direction, has been drawn and restored. The calculated amount of shortening is about 25% and has been mainly accommodated by break-thrust folding. Depth-to-detachment calculations have been made on the base of several justified assumptions such as constant bed-length, plain strain, no material to leave or enter pin-lines and parallel reference and detachment horizons in order to validate the area balancing by the Chamberlin Method. These calculations suggest a detachment at a depth of 10-11 km, with a calculated décollement dip of 3°.
These results bear important implications for the deformation phase between the India-Asia collision and the well known Early Miocene events: initial crustal thickening was probably partitioned in a more homogeneous thickening at depth, whereas near the brittle-ductile transition zone shortening at more shallow crustal levels was accommodated by the formation of a SW- vergent fold-thrust belt. A hypothetical detachment of the Precambrian to Early Tertiary sedimentary sequence provides an attractive solution to the unknown crystalline basement of the HHC and TZ.
The Structure of the High Himalayan Crystalline zone is dominated by the SW verging folds and thrusts of the Main Central Thrust zone which has been formed by underthrusting of the North Indian margin below Asia (Heim & Gansser, 1939, Frank, 1977, Le Fort, 1986). New investigations in the Lahul and Spiti regions show that these SW verging structures of the MCT are overprinting a stack of older NE directed intracontinental thrusts (Steck et al. 1993, Epard et al. 1995). Such NE-directed thrust structures may be followed over a distance of more than 200 km along the strike of the High Himalayan Crystalline Zone. The NE verging Shikar Beh Nappe stack represents an example of the transition from brittle to ductile conditions and has been formed by a SW-directed intracontinental underthrusting (reactivating an old listric fault) and accretion of Precambrian to Mesozoic sediments of the North Indian margin. The heating of the subducted sediments is responsible for the regional metamorphism of Barrowian type. P-T estimates indicate a thermal field gradient of 25°C/km and according to published geochronological data, thermal peak conditions have been reached between about 40 and 20 Ma. The transition from NE-verging folds in a high tectonic level to a mylonitic NE-directed thrust zone in a deep tectonic level may be observed in the NE-directed Miyar Valley Thrust Zone, located on the southern limb of the High Himalayan Crystalline dome structure of the Zanskar region.
Epard, J-L, Steck, A, Vannay, J-C & Hunziker J, Schweiz. mineral. petrogr. Mitt, 75, 59-84, (1995).
Frank, W, Thöni, M & Purtscheller, F, Colloq. int. CNRS (Paris), 268, 147-160, (1977).
Heim, A & Gansser, A, Mém. Soc. Helv. Sci. nat, 73, 1-245, (1939).
Le Fort, P, Collision tectonics. Geological Society Special Publication (london), 19, 152-172, (1986).
Steck, A, Spring, L, Vannay, JC, Masson, H, Bucher, H, Stutz, E, Marchant, R & Tièche, J-C, Eclogae gel. Helv, 86, 219-263, (1993).
Steck, A, Epard, J-L, Vannay, J-C, Hunziker, J, Girard, M, Morard, A & Robyr, M, Eclogae gel. Helv, 91, 103-121, (1998).
02 : 3P/09 : PO
Higher Himalayan Crystalline (HHC) rocks often show metamorphic zonation from lower green schist facies to migmatites associated with leucogranite intrusions. In the investigated area, the Kinnar Kailas (KK) granite was already described as a concordant low grade intrusion related to the inverted metamorphism developed in the surrounding rocks (Vannay and Grasemann, 1998). The present study based on structural and chronological relationships between the KK granite and surrounding HHC is evidences for a discordant intrusive contact of the KK granite.
The first deformation phase (D1) in basement rocks is reflected by an old poorly preserved schistosity and associated folding. The geometry of the main progressive ductile deformation (D2-D3) results from SW vergent doming and migmatisation. This progressive deformation produces an unique NE-SW stretching lineation. Axial planes of D2-D3 folds are parallel to the main schistosity (S2-S3) which dips slightly towards the NE. The fold axes are oriented NE-SW (D2) and NW-SE for D3 SW vergent folds. The late D4 deformation is specially localized (East of Wangtu Gneisses and West of KK Granite in Sangla Valley). This low grade deformation corresponds to wide angle ductile normal faulting. NW-SE oriented stretching lineations and C/S microstructure testify a lowering of southeastern blocks.
The KK granite crosscuts the high grade deformation structures (D2-D3) but is locally affected by the local late D4 extensional deformation. Moreover, the KK granite is mainly undeformed and contains xenoliths with D2-D3 upper amphibolite facies paragneisses. The geometry of the xenoliths reflects magmatic stoping mechanism. In several localities, the KK granite bears dark microdioritic and hybrid enclaves that show textural evidence for mixing and mingling during the magmatic process. The main body of KK pluton consist of coarse grained biotite rich granite. KF porphyritic granite and fine grained leucogranite are less abundant. KK granite is intruded by fine grained aplitic and leucocratic granitic dykes.
Geochemically, samples of the KK granite show a wide range of SiO2 content (64 to 75%). The KK granite shows characteristics of peraluminous, S-type granite and calc-alkaline affinities in geochemical diagrams. Rb/Sr whole rock ages of =480 Ma with initial 87Sr/86Sr = 0.7201 +.0023 (Kumar, 1986) and Rb/Sr muscovite ages (this study) indicate a pre-Tertiary age for the KK granite and therefore for the high-grade metamorphic deformations D2/D3 in this part of Himachal Himalaya. These petrological and geochemical results on the KK granite reflect a mixing between mantle and deep crustal melts which emphasize the difference with the typical Tertiary Himalayan leucogranites.
Vanney JC & Grasemann B, Schwiz. Mineral. Petrogr. Mitt., 78, 107-132, (1998).
Kumar S, Unpub. Ph.D. Thesis, Phys. Deptt. Panjab University, Chandigarh, India, (1986).
The Shergol-Baltikar blueschist unit (Honegger et al., 1989) of the Indus Suture Zone in Ladakh was investigated for structural, petrological and geochronological aspects. This HP/LT unit, one of only few that occur along the whole Himalayan orogenic belt, is situated between the Ladakh block, a Cretaceous island arc sequence in the north and the Lamayuru Group, Mid-Jurassic to Santonian (Danelian and Robertson, 1997) sediments and volcanics of the Indian passive margin in the south. This whole nappe stack, initially assembled by southward thrusting, is backthrusted to the north, including the post-Eocene Shergol Conglomerate Fm., which rests unconformally on top of the blueschist and the corresponding Indus Suture Melange units.
The blueschist unit consists mainly of basic volcanic and pyroclastic rocks, comprising ash tuffs to agglomerates, and some metasedimentary rocks, quartzites, some cherts and impure marbles. In general, the rocks are only weakly deformed, preserving volcano-sedimentary textures, often. Foliations dip steeply to the SE to S, stretching and mineral lineations trend N-S, rarely E-W. Other deformation features are monomineralic (Na-Ca-amphiboles) shear zones showing similar orientations and lineations. Foliations are sometimes folded around steep axes in a disharmonic manner. Both foliations, defined by Na-Ca-amphiboles (winchite), chlorite and phengite, and the shear zones are overgrown by late lawsonite, indicating that they formed on the burial path within the subduction zone.
Mineral parageneses allow to constrain metamorphic conditions at about 400-440°C and 10.5-12.5 kbar based on the petrogenetic grid of Will et al. (1998). The stability of albite + lawsonite places maximum P-T conditions of 12.9 kbar and 440°C. Their coexistence with Na-amphibole indicates maximum temperatures of 405°C (at 10.5 kbar), in disagreement with minimum temperatures of 430°C and a minimum pressure of 12.2 kbar for the paragenesis garnet + Na-amphibole + lawsonite. However, garnet could be stabilised to lower temperatures by the addition of Mn, not considered in the calculations of Will et al. (1998). These P-T conditions are in agreement with preliminary geothermobarometric investigations. Part of the prograde, compressional path is recorded in mineral zonations (garnet, phengite, Na-amphibole, Na-pyroxene) and their order of crystallisation, but there are no indications for any retrograde overprint related to exhumation of the rocks (except the transformation of aragonite to calcite). The latter points to rapid cooling and exhumation after the blueschist metamorphic event.
Ar/Ar and Rb/Sr dating of phengite (single and multiple grains), which are currently in progress (and will be presented) should allow to constrain the timing of the blueschist metamorphic event and thus of subduction and exhumation of the leading edge of the colliding Indian plate.
Danelian T & Robertson AHF, Mar. Micropaleont, 30, 171-195, (1997).
Honegger K, Le Fort P, Mascle G & Zimmermann JL, J. Met. Geol, 7, 57-72, (1989).
Will T, Okrusch M, Schmädicke E & Chen G, Contr. Min. Petrol, 132, 85-102, (1998).
Occurrences of medium-temperature eclogites are now recognized as a distinctive feature of many continental collision zones. In the Himalayas, eclogites, dated as late Eocene (Tonarini et al., 1993), were first described from the Higher Himalayan Crystallines of the upper Kaghan valley in Pakistan (Pognante & Spencer, 1991). Glaucophane eclogites, first reported from the Kaghan HHC (Pognante, 1992), were recently described in the North Himalayan Tso Morari Dome of Ladakh (De Sigoyer et al., 1997).
Equilibration conditions in the Kaghan glaucophane eclogites have been estimated at T = 600 ± 30°C and P > 1.3 GPa (Lombardo et al., 1998b). This estimate is close to that of De Sigoyer et al. (1997) for the eclogites of the Tso-Morari Dome (T = 580 ± 60°C, P > 1.6 GPa).
In contrast with the Tso Morari and Kaghan eclogites, which both record isothermal decompression (or cooling) after the eclogite peak, the eclogites recently found in the Kharta region of southern Tibet at the top of the Main Central Thrust zone (Lombardo et al., 1998a) record decompression under increasing temperature.
Reaction textures and geothermobarometry suggest that in the Kharta eclogites an early mineral assemblage, formed at metamorphic T of 600-650°C and minimum P between 1.2 and 1.4 GPa (garnet-omphacite-rutile-phengite), was overprinted by granulite facies compatibilities at medium P (0.55-0.65 GPa) and high T (750-770°C), followed by recrystallization in the amphibolite facies at low P (0.4 GPa) and high T (700°C ).
A simple tectonic model is proposed to explain the metamorphic evolution recorded in the Kharta eclogites: a) formation of an orogenic prism during the first phases of continental collision, with eclogitization of basic rocks in the deepest parts of the prism (Early Himalayan event); b) self-warming of the prism through radioactive heating, and extrusion of the Higher Himalayan Crystallines along zones mechanically weakened by partial melting and formation of leucogranite melts (Late Himalayan event).
At the regional scale, the main difference between the E Himalaya eclogites and those of the NW Himalaya lies probably less in equilibration conditions than in the different P-T paths they followed during their exhumation, the P-T path followed by the Kharta eclogites implying rapid extrusion of lower crustal rocks in a relatively narrow orogenic belt.
De Sigoyer J, Guillot S, Lardeaux JM & Mascle G, Eur. J. Mineral., 9, 1073-1083, (1997).
Lombardo B, Pertusati P, Rolfo F & Visonà D, Mem. Sci. Geol. Padova, 50, 67-68, (1998a).
Lombardo B, Rolfo F & Compagnoni R, Geol. Bull. Univ. Peshawar, 31, 116-118, (1998b).
Pognante U, Geodinam. Acta, 6, 5-17, (1992).
Pognante U & Spencer DA, Eur. J. Mineral, 3, 613-618, (1991).
Tonarini S, Villa IM, Oberli F, Meier M, Spencer DA, Pognante U & Ramsay JG, Terra Nova, 5, 13-20, (1993).
The initial India-Asia contact that occurred during the Upper Paleocene and the Eocene is related to collisional processes. However, constraints from stratigraphy, paleo-magnetism, geochronology and tectonophysics in NW Himalaya allow to dis- criminate during the India-Asia convergence three successive events : (1) the Indian continental subduction (2) the progressive thickenning of the Indian Crust and (3) the steady-state collision.
The Indian continental subduction : a major decrease in plate velocity (from 18 cm.yr-1 to 10 cm.yr-1) occurring from 55+ Ma to 52 Ma is interpreted as the first effects of the India-Asia collision. At the same time, onset of the India-Asia contact at the Paleocene/Eocene boundary (55 ± 0.5 Ma) is recorded by shoaling of the marine sediments in Zanskar, and related to the obduction of the Asian accretionary prism onto the Indian margin. It is followed by the end of the marine sedimentation and the beginning of the erosion of the first Indian reliefs in the Indus suture during the Late Ypresian (< 52 Ma). During this period, part of the subducted Indian continental margin was exhumed within the mantle wedge at the base of the Indian crust. These results suggest that during the Paleocene/Lower Eocene period, the Indian margin records the progressive ending of the cold subduction of the Indian plate and the beginning of the thickening of the Himalayan wedge by under-plating of continental units and superficial nappe emplacement. This period leads to the progressive apparition of reliefs in the suture zone.
The progressive thickenning: from ~50 Ma to 40 Ma the intracontinental subduction of the High Himalayan Crystallines allow the progressive thickening and warming of the Himalayan orogenic wedge, that leads to the establishment of the Eo-Himalayan metamorphism. Similarly, it corresponds to the progressive decrease of the plate velocity down to 5 cm.yr-1. A the surface, this period corresponds to the formation of perisutural basins.
The steady-state collision: since 40 Ma, the India-Asia convergence is stabilized at 5 cm.yr-1. It corresponds to the progressive exhumation and erosion of the metamorphic units and the global deformation of the whole units from the suture zone to the Himalayan front.
Study of the structural and geochronological evolution of the Ladakh Batholith in NW India tightly constrains the timing of continental collision in that part of the Himalayan orogen. The batholith is a multipluton calc-alkaline body, which developed above a subduction zone on the southern margin of Eurasia between 100 and 50 Ma ago. The sudden end of calc-alkaline magmatic activity is related to the collision of India, and disturbance of the subduction process. Dyke orientation, penetrative foliation, shear zones and microstructural features suggest that the 61 Ma Gyamsa Pluton records syn-magmatic deformation unrelated to collisional deformation. The Gyamsa Pluton solidified and hardened and did not record the younger deformation recorded by the Leh Pluton. The 49 Ma Leh pluton records syn-magmatic deformation analogous to young, collision-related, regional structures developed in softer country rocks. Thus, collision-related deformation must have started between 61 and ca. 50 Ma. An independent and tighter constrain on the time of collision is given by the magmatic activity and cooling history of the batholith. The Leh pluton is the youngest major magma pulse in eastern Ladakh. Its intrusion was followed by rapid cooling and the intrusion of a few small and narrow subvolcanic dykes at 46±1 Ma. A number of published Ar-Ar and K-Ar cooling ages indicate that when the Leh pluton crystallized at ca. 50 Ma, the Ladakh Batholith in general was undergoing cooling, reaching temperatures around 350°C in most places before 45 Ma. Thus, I conclude that 50 Ma marks disturbance of the magmatic cycle at the continental margin caused by collision. However, this does not date the exact timing of collision. Generalized cooling and syn-magmatic deformation of the Leh Pluton indicate that collision was already under way at 50 Ma and that it may have started somewhat before that, possibly in the Upper Ypresian (ca. 52 Ma), as concluded from the study of sedimentary sequences in NW Himalayas.
The conventional dislocation model of strain accumulation is a convenient approach to explain most of the observed displacements during interseimic phase. However, this approach have many shortcomings including the assumption of homogeneous elastic lithosphere and failure to approximate deformation associated with aseismic displacement. Furthermore this model cannot describe the plastic deformation that accumulates over many seismic cycles to form the geologically observed deformation.
We have developped another approach in order to model the wide range of informations available on this long-term deformation as well as on interseismic deformation in the Himalaya of Nepal. We use a plane strain finite element model (ADELI). We consider thermal and strain dependant rheologies for the crust and mantle derived from laboratory experiments.In addition surface processes are modeled from a 1-D diffusion equation, assuming flat deposition in the foreland. In all our experiments a section of lithosphere, initially loaded by the topography is submitted to 20 mm/yr horizontal shortening. We study the effect of varying erosion rate, rheological properties, thermal regime and boundary conditions to derive some constraints on these parameters.
Interseismic vertical velocity can be matched only if erosion is accounted for. Moreover the fact that most of the shortenig is accomodated by slip along the MFT (Main Frontal Thrust) requires a low effective friction coefficient on the seismic decollement. The observed long-term deformation depends mostly on geometry of the crustal thrust and on rheology. The interseismic deformation is also sensitive to rheology and is independant on the assumed goemetry of the decollement and ramp system. The best results are obtained assuming a quartz dominated crustal composition with high radioactive production in the upper crust.
The striking increase of the 187Os/188Os ratio of seawater during the Cenozoic Era (Pegram et al., 1992) has been attributed, by some authors, to enhanced weathering provoked by the Himalayan collision. In order to test this hypothesis, we analyzed river bedloads and bedrocks from Central Nepal and Bangladesh. The Himalayan Range consists of three main units. Moving progressively downstream (ie, from N to S), these are: (1) the Tethyan Sedimentary Series (TSS) composed of sediments of Cambrian to Eocene age. The TSS samples are unradiogenic with 187Os/188Os from 0.6 to 1.4 for the bedrocks and from 0.9 to 1.2 for bedloads. The Os concentration varies from 20 to 200 ppt for the bedrocks and from 30 to 100 ppt for bedloads. (2) The High Himalayan Cristallines (HHC), highly metamorphosed paragneisses and leucogranites, are the principal formation of the Range. The bedrocks have 187Os/188Os between 0.8 and 2.0 and bedloads with an 187Os/188Os ratio of 1.7 represent an average composition of the formation. Os concentrations of both bedloads and bedrocks are lower (3 to 50 ppt) than the upper continental crust concentration (50 ppt) (Esser & Turekian, 1993). (3) The Lesser Himalaya (LH) composed of variably metamorphosed bedloads with very old Nd model ages, displays very large variations in 187Os/188Os ratios (0.7 to 3.0 for bedloads, 1.4 to 8.3 among the bedrocks). The Os concentration also varies widely, ranging from 7 to 500 ppt. The highest Os concentrations are associated with the highest Os isotopes ratios and are found in black shales.
The 187Os/188Os of river bedloads collected at the outflow of the Range reaches 3.25. This radiogenic value can be explained by addition of a small fraction (<5%) of organic-rich rocks with high Os concentration from the LH to less radiogenic sediments from the HHC and the TSS. The mixing relationship is also apparent in a Sr vs Os isotopic diagram. The 187Os/188Os ratios of river bedloads in the Indo-Gangetic Plain (Ganges) are controlled by a single lithology (Black Shale), which is of minor occurence.
Bedloads from the Indo-Gangetic Plain were also analyzed. Samples from the Ganges and Tista (Central Nepal) are radiogenic with a 187Os/188Os ratios of 2.3-2.6. These ratios are consistent with those of bedloads collected at the outflow of the Range. In contrast, two samples from the eastern part of the Bangladesh (Brahmaputra) are less radiogenic than the Ganges (187Os/188Os = 1.5 and 0.8) and appear to reflect seasonal variations.
Pegram WJ, Krishnaswami S, Ravizza GE & Turekian KK, Earth Planet. Sci. Lett, 113, 569-576, (1992).
Esser BK & Turekian KK, Geochim. Cosmochim. Acta, 57, 3093-3104, (1993).
We present an estimate of the modern erosion of the Himalaya based on riverine fluxes. The dissolved flux related to chemical erosion is estimated on the basis of yearly averages of the rivers. It is 32 and 29*106 ton/yr for Brahmaputra and Ganges respectively. The fluxes of suspended matter measured in Bangladesh are 402 to 608*106 ton/yr for Brahmaputra and 328 to 548 *106 ton/yr for Ganges. Isotopic tracers show that the large majority of these particles is derived from the Himalaya. However the total erosional flux must take into account bedload transport and sedimentation in the floodplain which cannot be simply measured. Their importance can be estimated using the contrast in chemical composition between the suspended load and the bedload or the accumulated sediments due to mineral sorting during transport.
Assuming steady state erosion in Himalaya, the chemical composition of the total erosional flux equals that of the average source rock. This later is relatively well established for silicate formations based on ca. 200 analyses of the Himalayan formations. Compositions of the riverine end members are based on suspended- and bedload sampled in Bangladesh. The composition of sediments accumulated in the floodplain is derived from an average composition of Siwaliks. The principal process driving chemical differenciation is the sorting of phyllosilicates with respect to quartz and feldspars in favor of the suspended load. Si/Al ratios are: 4.2 for the source rock, 2.7 for the suspended load, 6.5 for bedload and 7.8 for the Siwaliks. Mass balance calculation is performed for two extreme case: no bedload or no sediment sequestered in the flood plain. On this basis, the suspended load represents between 40 and 65% of the total erosional flux. Using the fluxes of suspended load mentioned above, total erosional fluxes can be calculated and expressed in term of erosion rate. They are 2.8 +/- 1.1 mm/yr for the Brahmaputra Himalaya and 1.8 +/- 0.6 for the Ganges Himalaya. The higher erosion rate in the Brahmaputra basin is likely related to its higher runoff.
The growing set of data on fault rates and finite offsets permits to envisage a step-by-step backwards restoration of the intracontinental deformation between India and Asia. We here assume that the major part of the deformation is localized along a few majors faults separating less deformed lithospheric blocks. The motion of such blocks can thus be modelled by moving microplates on the sphere. To account for crustal thickening or extension at the limits of the blocks, we allow changes in the surface of these blocks. Finally, minor deformation of the blocks interiors (bending, local rotation...) permits to minimize gaps or overlaps due to motion along faults. Tectonic and geophysic data used to constrain these reconstructions are: present-day rates and finite offsets along fault (strike-slip, thrusts); amount of crustal thickening or extension; paleomagnetic rotations; kinematics of large plates. We use Euler poles and rotation parameters determined for India, Tarim, Tibet, China, and Indochina. The present reconstruction is done by steps corresponding to the timing of major changes in the active fault pattern during the collision evolution. Such changes correspond to the dextral reactivation of the Red River fault at 5 Ma, the activation of the Kunlun fault at 10 Ma, the end of the sinistral motion on the Red River fault at 15 Ma, and the end and beginning of motion along the Wang Chao and 3 Pagodas faults at 30 and 40 Ma. The recent contours of the blocks are drawn along major faults of Central Asia, whose segmented traces and junctions have been simplified. As we step back in time, the size of the area of deformation shrinks and the block contour pattern gets simpler. At each time step we propose a overall solution for the position of all the blocks of the deformation zone, in agreement with the data set available for this period. We finally obtain a synthetic evolution of the collision, back to its onset.This backward reconstruction allows us to test the validity of the assumption of coherent lithosperic blocks and permits to make kinematic and mass balances. This model confirms the importance of the extrusion mechanism that plays a role comparable to crustal thickening to absorb plate convergence.
Recent total convergence between Indian and Eurasian plates is at very high rate of about 45 mm/a (DeMets et al., 1990). Crustal shortening across the Tien Shan is found at a rate of about 20 mm/a (Abdrakhmatov et al., 1996), implying that convergence is transferred across the entire Pamir range. Questions arise, where and how the crustal deformation is distributed across the Tien Shan and Pamir region from the time of India-Eurasia collision until today. To constrain the propagation of deformation and to estimate how crustal shortening might be translated into uplift, we collected structural data and geochronological samples along a N-S traverse from S-Tien Shan to N- and Central Pamirs. Sampling was concentrated on E-W striking magmatic belts, which are associated to Late Paleocoic, Triassic-Jurassic and Early Creataceous suture zones. Ar-Ar biotite ages from granitoids in the Rushan Pschart zone (Early Creataceous suture zone) demonstrate regional cooling below 300° C in Late Cretaceous time. The apatite fission-track (FT) ages cluster around 11 Ma. The track-length distribution shows shortened tracks, indicating a longer time in the PAZ (partial annealing zone) or thermal overprint. Granites in the Muzkol metamorphic core complex of the Central Pamirs show very fast cooling in Miocene time (Ar-Ar hornblende: 15-22 Ma; Ar-Ar biotite: 15-17 Ma; zircone FT: 16 and 17 Ma; apatite FT: 14-17 Ma) reflecting fast exhumation of the dome. The emplacement of the granites was determined by a U/Pb zircon age of 31 Ma. In the northern Pamirs (Lake Karakul area) granitoids show the following apatite FT age distribution: in the south of the lake the oldest apatite FT age is around 52 Ma with track lengths partly shortened. To the north the ages decrease to 36 respectively 24 Ma. The track-length distributions indicate undisturbed cooling of the rocks. The batholith north of the lake yielded apatite fission track ages of 17 and 18 Ma with a narrow track-lengths distribution (mean 15 µm), expressing a much faster cooling in Miocene time. The age pattern and track-length distributions in the northern Pamirs may be interpreted in terms of block rotation above a listric fault, thrusting over a ramp or local back-thrusting.
DeMets C, Gordon RG, Argus DF & Stein S, Geophys. J. Int., 101, 425-478, (1990).
Abdrakhmatov KY, et al, Nature, 384, 450-453, (1996).
The Chuya basin in Gorno-Altai (South Siberia) and the Zaisan depression in South-Altai (East Kazakhstan) contain the most complete sedimentary sections of the Altai mountain range in the Former Soviet Union. Stratigraphic, morphotectonic, fault kinematic, remote sensing and paleomagnetic studies allowed to highlight the dynamic evolution of these basins during the Cenozoic, and also provide some time constraints on the mountain building process in Altai.
The Zaisan depression was initiated during the Permian and was active during most of the Mezosoic. It was reactivated during the Cenozoic. The Chuya basin is superimposed on a Devonian depocenter, and was initiated only in the Paleogene. In the Late Cretaceous-Paleocene, a vast peneplain developed in Central Asia in a warm and humid climate. The area of the present Chuya depression, now in the middle of the Altai massif, had a short marine incursion during the Santonian-Campanian, while sedimentation in the Zaisan depression remained continental.
The basins developed first in a transtensional setting during Oligocene-Miocene, with fluvio-lacustrine and lacustrine sedimentation. In the Pliocene, at about 3 Ma, strong transpression caused disappearance of the lacustrine environment, disruption of the lake beds, coarse clastic sedimentation, occasional block rotation and the formation of oblique thrusts at the basin margins. This strong transpressional deformation occurred approximately during a relatively short time span (2 Ma), until the early-middle Pleistocene. Similar intensification of tectonic movement (both compressional and extensional) has also been observed in other areas of the Tian-Shan mountains and in the Baikal Rift Zone. It can be expected that a significant part of the Late Cenozoic structures of Altai was formed during the Late Pliocene-Early Pleistocene. The cause for this tectonic event seems not straighforwardly related to the dynamics of the Asian plate boundary process.
From the Middle Pleistocene to Holocene, transpressional deformation continued, but in a less vigorous way, and was accompanied by rapid vertical movements. Asymmetric uplift of the mountain ranges bordering both basins on the southern side may have been stimulated by a modification of the weathering, erosion and transport conditions due to climate change.
Mainly based on paleomagnetic results from South and North China blocks (SCB & NCB), this eastern part of the Asian mosaic is generally considered to have been accreted to Siberia prior to the Cretaceous. In order to check this hypothesis, we performed in the summers 1995 and 1997 two field trips during which we collected Jurassic and Cretaceous effusives north and south of the Mongol-Okhotsk suture which separates the Mongol block to the south from Siberia to the north. South of the Mongol-Okhotsk suture, i.e. in the Mongol block, we have collected one Jurassic locality in the Transbaikal region along the Unda river at (51.7°N, 117.4°E, basaltes, 12 sites), and three lower Cretaceous effusive formations: the first one in Amur region around the town of Taldan (53.8°N, 124.5°E, andesites), and the two others in the Transbaikal at the Kriemlovka peak locality (51.8°N, 117.5°E, basalts, 12 sites) and the Torey lakes locality (50.2°N, 115.8°E, basalts, 14 sites). North of the Mongol-Okhotsk suture, i.e. in the Siberian continent, two Cretaceous localities have been sampled. The first one along the Ingoda river at (51.2°N, 112.3°E, trachy-basalts, 12 sites) and the second in the arounds of Bitchura town (50.6°N, 107.6°E, trachy-basalts, 10 sites). The Jurassic locality and three out the five Cretaceous localities yielded consistent results deserving attention, which can be summarized as follows:
1) We have determined an upper Jurassic and two lower Cretaceous poles for the Mongolian block, and a new Cretaceous pole for the Siberian block. 2) When compared on an equal area projection, the Cretaceous poles from the north and the south of the Mongol-Okhotsk suture lie on small circle containing the Cretaceous poles of the reference apparent polar wander path of Eurasia through the error bars demonstrating that Mongolia and Siberia were accreted by the lower Cretaceous. 3) The North China block-South China block-Mongolia-Siberia assemblage was already formed in the Cretaceous and did not suffer any relative N-S motion since that time. 4) Important relative rotations between the north (clockwise) and the south (anticlockwise) of the suture could indicate that it has acted as left lateral shear zone allowing a post-Cretaceous eastward extrusion of Mongolia. 5) Based on the preliminary Jurassic pole of Mongolia, we can not do better than estimate the width of the the Mongol-Okhotsk ocean to between 2500 and 4000 km. 6) The Mongol-Okhotsk ocean seems to be entirely closed between the upper Jurassic and the lower Cretaceous.
The Pamir mountain building is bounded on the east and on the west by narrow depressions filled with undeformed Quaternary sediments. They cut out from main orogen the marginal moun-tain chains that composed of the Tertiary molasse (Darvaz) or of the Paleozoic metamorphics (Kun-Lun). Together with basins located on the front of the Pamir Mountains, these boundary depressions reveal kinematics of recent structuring in the Central Asian region. A basement of the Darvaz depressions includes the faulted Mesozoic and Cenozoic formations. Pre-orogenic clastics and carbonates present shelf and lagoon facies those thickness strongly decreases eastward. Orogenic sediments (more 7000 m in thickness) comprise the lower red-colored Oligocene - middle Pliocene molasse that follows continuously the quasi-platform sequences and upper gray-colored post-orogenic molasse. The upper Pliocene - lower Pleistocene conglomerates are enveloped by two regional unconformities; uppermost middle Pleistocene to Holocene sediments are thin and located only within river valleys. Structure of the Darvaz arc de-velops in transpressive setting. The echeloned faults of the region consist left-lateral strike-slip zone, in which the faults rotate gradually toward the north from NS to NE, and become in the same time more flat transforming consequently from subvertical strike-slips to low-angle thrusts. Depressions are integrated in this structure and expressed by almost isolate linear valleys embedded between overstepping strike-slips, so they compose echeloned chain of oblique or Freund's pull-aparts.
Configuration of the East Pamir depressions is almost mirrorlike in respect to the Darvaz valleys. The local basins constitute right-lateral row and each of the basins is clearly segmented. Basin closures coincide with the fractures of the Karakorum fault zone that are unfolded southward fanlikely from WE to SE. The depressions is separated from Kun-Lun structures with large-scale normal faults (Brunel et al.,1994), and the Quaternary clactics within them lack remarkable deformations. Thus, we consider the East Pamir depressions as structures of strike-slip related extension (normal or Burchfield's pull-aparts). Thermochronological data establish that concordant uplift of bordering mountains and opening of basins started ~ 2 m.y. ago (Arnaud et al. 1993), i.e. synchronously with development of the Darvaz depressions. Transtensional evolution of the Kun-Lun border of the Pamir microplate versus transpression on the Darvaz border follows non-parallelism in the Pamir displacement to orientation of the bordering strike-slip zones reflected complex many-block interaction controlling dynamics of this region.
The work is supported by RFFI, grant 96-05-65521
The western boundary of the Philippine Sea Plate (PH) with Sundaland (SU) in the Philippines corresponds to a wide deformation zone that includes the stretched continental margin of Sundaland, and the Philippine Mobile Belt (PMB) extending from Luzon to the Molucca Sea. The GPS GEODYSSEA data are used to decipher the present kinematics of this complex area. One of the main results is the quantification of overlapping subductions on both sides of the PMB. To the west, convergence decreases from about 90-100 mm/yr along the Manila trench to a few mm/yr across Mindoro in the zone of collision of Palawan with the PMB. Transfer of 55% of this Manila Trench convergence motion to the Philippine Trench to the east is accommodated principally by rapid counterclockwise rotation of eastern Luzon. Southward, convergence decreases regularly along the Philippine Trench from 54 mm/yr near 13°N to 32 mm/yr east of Mindanao. Inside the PMB, the slip rate of the left-lateral Philippine Fault, which enables complete partitioning to occur at the trench, increases northward from 22 mm/yr at 7°N to 35 mm/yr in the Visayas. As a result, the East Philippine sliver, east of this fault, moves north and transfers this motion to the East Luzon block counterclockwise rotation. West of the Philippine Fault, the Visayas block in central Philippines, is rotating rapidly clockwise, transfering most of the remaining convergence to the Negros/Cotabato trenches on its western border, up to 32 mm/yr near 12°N and 50 mm/yr near 8.5°N. These estimates ignore the amount of convergence absorbed to the west, at the limit of undeformed SU.Two opposite rotations on both sides of the left-lateral Philippine Fault, clockwise to the southwest and counterclockwise to the northeast, illustrate the PMB internal deformation. As a result, the northern part of the Visayas block acts as a cog wheel with respect to the Luzon block. It is remarkable that this system is such that subduction is everywhere perpendicular to the trenches. South of Mindanao, within the Molucca sea, the overlapping divergent subduction system of the PMB is relayed by the convergent Sangihe-Halmahera subduction system. This one accommodates 80 mm/yr of the 105 mm/yr PH/SU convergence and about 15 to as much as 20 mm/yr of the remaining convergence are absorbed far to the west, within the northern Borneo margins.
GPS measurements acquired over Southeast Asia in 1994 and 1996 in the framework of the GEODYSSEA program revealed that stations located in Vietnam, Malaysia, Thailand, Borneo, Eastern Sumatra and Java have no resolvable relative motions. A large piece of continental lithosphere comprising Indochina, Sunda shelf and part of Indonesia thus behaves as a rigid "Sundaland" platelet. A direct adjustment of velocity vectors obtained in an Eurasian frame of reference shows that Sundaland block is rotating clockwise with respect to Eurasia around a pole of rotation located south of Australia. Although the GEODYSSEA network was carefully attached to the ITRF-94 reference frame using IGS stations, an internal systematic misadjustment cannot be ruled out from GPS data alone, raising the possibility that the derived Sundaland/Eurasia motion is contaminated by some whole network rotation artifact. We present here an original and totally independent check by using earthquakes slip vectors at Sunda and Philippines trenches. Seven stations of the GEODYSSEA network are close to the trenches and not separated from them by large active faults (two at Sumatra trench, three at Java trench and two at the Philippines trench). The difference between the vector at the station and the adjacent subducting plate vector defines the relative subduction motion, and should thus be aligned with the subduction earthquakes slip vectors. We thus first derive a frame-free solution that minimizes the upper plate (or Sundaland) motion. When corrected for Australia-Eurasia and Philippines-Eurasia NUVEL1A motion, the misfit between GPS and slip vectors azimuths is high and scattered (-10° to 15° for Sunda trenches stations, -45° and 22° for the two Philippines trench stations), indicating that the upper plate has significant motion with respect to Eurasia. We then determine by least squares inversion the best rigid upper plate rotation that minimizes this azimuth misfit. The misfit between GPS and slip vectors drops to about 5° for Sunda trenches, and about 10° for the Philippines trench. This best fit solution is close to the initial fit obtained through direct adjustment of the Eurasian frame GEODYSSEA solution. A series of theoretical modelling shows that elastic coupling at trenches introduces only a minor correction. We conclude that Sundaland motion is essentially correct and discuss some of the geodynamic implications.
Recent data obtained by the 600 permanent GPS sites of the Geographical Survey Institute of Japan in 1995 indicate that full mechanical coupling prevails in the Japan trench and the Nankai trough along the Japanese Island Arc (Le Pichon et al., 1998). We use the 1996 and 1997 GPS data to compute the strain rate tensors that we invert to estimate the average subduction velocity loading the central Nankai and Japan Trench seismogenic zones, independently of the data reference frame. The results show that both subduction planes are fully locked on a regional scale.We then use the GPS velocity vectors in order to look in greater details at the geometry and local coupling of the seismogenic zones. Defining partial coupling along a fault segment as the association of smaller scale fully locked and free-slip zones, we define small portions of the Pacific and Philippine slab surfaces that we assume to be either fully locked or fully unlocked. We then invert the GPS data for the distribution of free-slip and locked areas along both seismogenic zones. This method confirms that the Central Nankai trough and the Japan trench slabs are fully locked, with small consistent areas of free-slip. These area coincides, in the Japan trench, with a series of recent large earthquakes (M 7.4 1989, M 7.0 1992, M 7.7 1994), and in the Nankai trough with a zone of divergence between the isodepth and isotherm contours. We study the spatial and temporal variation of these free-slip zones from 1995 to 1997. We conclude that, at least around Japan, trenches are either fully locked or slipping and that "partial coupling" does not exist there.
Le Pichon, X, Mazzotti, S, Henry, P, Hashimoto, M, Geophysical Journal International, 134, 501-514, (1998).
Fieldwork along several segments of the Altyn Tagh Fault, between 85 and 95 °E, confirms that it ranks as one of the most active faults of Asia. In the East, near Aksay, the active fault trace offsets numerous stream channels, terrace risers and fans tens to hundreds of meters. 14C dating of organic remains and charcoal within terrace gravels, and 10Be and 26Al cosmogenic dating of surface pebbles, still in progress, indicate that the principal terraces were emplaced 2.5 and 5.5 Ka ago, implying a left slip rate between 2 and 3 cm/yr. Large mole tracks attest to the occurrence of great earthquakes. Even larger mole tracks are found north of Lenghu, within the Altun Shan push-up, a 6000 m-high range in a restraining bend of the fault, now sliced by its most active strand. North of Huatougou, at the transition between another push-up mountain and a recent pull-apart basin, a spectacular sequence of five flat-floored, hanging channels, beheaded by the fault from a unique source in the mountain, have been horizontally displaced by up to 1250 m. Cosmogenic dating of the abandonment of these channels is in progress. Several km to the west, pressure ridges exceeding 10 m in height across a large young fan, imply the repeat of several great earthquakes in a relatively short time span.Large mole tracks and hectometric cumulative offsets of post-glacial fans and terrace risers, sampled for cosmogenic dating in several localities, are also visible for tens of kilometers along the fault east and west of Tura. The cumulative pressure ridges of great earthquakes are preserved at five thousand meters, above the permafrost line, implying that they are due to fairly recent events. 10Be dating of quartz cobbles from a lateral moraine offset several kilometers by the fault yields exposure ages on order of 40 Ka. This suggests a slip rate in excess of 3 cm/yr, to be further tested by the dating, in progress, of other moraines. The consistency of our observations over a length of nearly 900 km confirms that the Altyn Tagh fault ruptures mostly with M*8 earthquakes and slips at a rate of several cm/yr.