Deep Structure of the Himalaya and Tibet from Gravity and Seismological Data

Most mountain ranges are flanked by foredeep basins, and in general neither the ranges nor the basins are in local isostatic equilibrium. The basins are overcompensated whereas the ranges themselves are usually undercompensated, I describe a method of analysing gravity anomalies over mountain ranges assuming that the topography is supported by the elastic strength of the continental lithosphere which is flexed down by the weight of both the overthrust mountain and the sediments in the adjacent foredeep. Using Bouguer gravity data from one profile across the Himalaya and Ganga Basin, I present a detailed study of the effects of each of the parameters (flexural rigidity, extent to which the elastic plate underlies the range, density contrasts between the crust and mantle and between the sediments and crust) on the configuration of the elastic plate and the gravity anomalies. Deviations from local isostatic equilibrium on five profiles across the Himalaya and the Ganga Basin can be understood if the strong Indian plate underthrusts the Lesser Himalaya. However the load of the High Himalaya is too large to be supported solely by the elastic stress in the Indian plate if the flexural rigidity of the plate is constant and if no other external forces act on the plate. The increase of the gradient in Bouguer gravity anomalies from about 1 mGal/km over the Ganga Basin and the Lesser Himalaya to about 2 mGal/km over the High Himalaya implies that the Moho dips more steeply (10*-15') beneath the High Himalaya than beneath the Lesser Himalaya (2*-3*). This steepening is interpreted to be due to a weakening of the plate. Even with a weak plate beneath the High Himalaya, the weight of the mountains depresses the plate too much unless an external force system helps support the weight of the High Himalaya. Both the reduced flexural rigidity and the bending moment and force applied to the plate can be understood if part or all of the Indian crust has been detached from India's mantle lithosphere. The magnitudes of the bending moment and force are compatible with their source being gravity acting on part of the stripped mantle lithosphere. The tectonic implications of these results are studied by means of a series of idealized balanced cross sections, from the collision of India with Eurasia to present, that reproduce several important features of the geology of the Himalaya and predict an amount of eroded material comparable to that in the Ganga Basin and the Bay of Bengal. They also predict rapid uplift only in the High Himalaya and at the foot of the Lesser Himalaya. The cross sectional shape of the Ganga Basin is controlled by the deflection of the Indian plate. iii Thus, if the Ganga Basin is a steady state feature, the age of the basal sediments in a given locality should be proportional to the distance of that locality from the southern edge of the basin. If the rate of convergence of India and the Himalaya were constant, that rate should equal the distance divided by the corresponding age. A rate of 10 to 15 mm/a for the last 15 to 20 Ma is found, which is consistent with a large part of the 50 mm/a rate of convergence between India and Eurasia being absorbed by the eastward extrusion of parts of Tibet. Profiles of Bouguer gravity anomalies show only a small peak or plateau over the southern edge of the Lesser Himalaya, implying that the boundary between the light sediments of the Ganga Basin and the heavier crustal rocks of the Lesser Himalaya is not sharp and that there exists some light material beneath the range. We infer that some sediment deposited in the Ganga Basin has been underthrust beneath the Lesser Himalaya, but the quantity is small; most of this sediment probably is scraped off the Indian plate to make the foothills of the range. Gravity anomalies across the western part of the Tarim Basin and the Kunlun mountain belt show also that this area also is not in local isostatic equilibrium. These data can be explained if a strong plate underlying the Tarim Basin extends southwestward beneath the belt at least 80 km and supports part of the topography of northwest Tibet. This corroborates an earlier inference that late Tertiary crustal shortening has occured in this area by southward underthrusting of the Tarim Basin beneath the Kunlun, and it places a lower bound on the amount of underthrusting. Travel times and waveforms of SH waves recorded at distances of 10* to 30' and some SS waveforms are used to constrain the upper mantle shear wave velocities down to a depth of 400 km beneath both the Indian Shield and the Tibetan Plateau. The uppermost mantle shear velocity beneath both the Indian Shield and the Tibetan Plateau is high and close to 4.7 km/s. The Indian Shield has a fairly thick lid and the mean velocity between 40 and 250 km is between 4.58 and 4.68 km/s. In contrast, S wave travel times and waveforms, as well as a few SS waveforms, show that the mean velocity between 70 and 250 km beneath the central and northern part of the Tibetan Plateau is slower by 4% or more than that beneath the Indian Shield and probably is between 4.4 and 4.5 km/s. No large differences below about 250 km are required. These results show that the structure of Tibet is not that of a shield and imply that the Indian plate is not underthrusting the whole of the Tibetan Plateau. Thesis Supervisor: Peter Molnar Title: Professor of Geophysics