GEODYNAMIC MODEL OF THE HIMALAYAS ON THE BASIS OF EARTHQUAKE TOMOGRAPHY

Abstract

Travel time tomography was carried out for the entire Himalayas and adjacent NE India and Indo Burma ranges to its southeast and Pamir-Hindu Kush region to its northwest to evolve a comprehensive geodynamic model of this tectonically active and complicated area. The area was divided in to three parts for carrying out this work to make data analysis possible. Brief report of results and interpretation follows. In northeastern part the Indian lithosphere gently underthrusts under the Himalayas and steeply subducts below the Indo–Burma Ranges. In between these two ranges the Indian lithosphere is in a vice like grip. The tomographic image shows heterogeneous structure of lithosphere depicting different tectonic blocks. Though the results are limited to shallower depth (0–90 km), it matches well with the deeper continuation of lithospheric structure obtained in an earlier study. A low velocity structure all along the Eastern Himalayas down to ~70 km depth is observed, which may be attributed to deeper roots/thicker crust developed by underthrusting of Indian plate. Parallel to this low velocity zone lies a high velocity zone in foredeep region, represents the Indian lithosphere. The underthrusting Indian lithosphere under the Himalayas as well as below the Indo–Burma Ranges is well reflected as a high velocity dipping structure. The buckled up part of bending Indian plate in study region, the Shillong Plateau–Mikir Hills tectonic block, is marked as a high velocity structure at shallower depth. The Eastern Himalayan Syntaxis, tectonic block where the two arcs meet is identified as a high velocity structure. The Bengal Basin, tectonic block to the south of Shillong Plateau shows low velocity due to its thicker sediments. Based on the tomographic image, a schematic model is presented for the study region. Variations of the crustal thickness was estimated beneath the Nepal Himalayas based on tomographic inversion of local and regional earthquake data. A low-velocity anomaly in the upper part of the model down to depths of ~40 to 80 km represents crust. Lower limit of this anomaly represents variations of Moho depth. The obtained variations of crustal thickness match fairly well with free-air gravity anomalies: thinner crust patterns correspond to lower gravity values and vice versa. There is also some correlation with magnetic field: higher magnetic values correspond to major areas of thicker crust. I propose that elevated magnetic values can be associated with more rigid segments of the incoming Indian crust which cause more compression in the thrust zone and lead to stronger crustal thickening. Rotation of Indian plate after collision may also lead to variation in crustal thickness along the tectonic trend of the Himalayas. I have also corroborated estimated seismic velocity structure and crustal iv thickness with the recent Nepal earthquakes (magnitude ≥7) occurred on April 25, 2015 and May 12, 2015 in the eastern Nepal. It observed that the first earthquake of magnitude >7 initiated in a zone where crustal thickness is relatively lower and the rupture propagated eastward towards a region where both crustal thickness and S-wave velocity is higher, i.e. towards a more rigid part of the crust. This may have led to stress loading in the rigid part that subsequently led to occurrence of second magnitude >7 earthquake in that area within a month. The velocity and crustal thickness maps estimated in this study nicely explains why two earthquakes of magnitude >7 occurred within a short span of time within such close distance range. Different geotectonic units for the third segment, viz. Northwest Himalayas, Western Syntaxis, Pamir and Hindu Kush regions and Salt ranges are well imaged. The tomographic image is well resolved up to ~150 km the depth in the Western Himalayas and up to ~300 km depth in the Pamir and Hindu Kush region. The top low velocity anomaly imaged up to ~80 km depth correlates well with the thicker crust with deeper low density roots under the high mountains in the northwest Himalayas as well as in the Pamir and Hindu Kush region. The thickness of the crust is varying all along and across the tectonic trend of the northwest Himalayas which I interpret as due to episodes of clockwise and anticlockwise rotations of the Indian plate and thrusting, shortening, buckling and exhumation during Indo-Asia collision. The Indian lithospheric slab is imaged as a gently underthrusting high velocity anomaly under the northwest Himalayas and subducted Indian lithospheric slab which follows the trend of intermediate depth seismicity under the Pamir and Hindu Kush region. On the other hand beneath the Pamir-Tien Shan the dipping high velocity anomaly which follows the trend of intermediate depth seismicity, represents the remnant of the southward subducted Asian slab. In the southwest of Hindu Kush the Indian lithospheric slab rolls over and overturns at a depth of ~250 km and dips southward. The Delhi-Haridwar ridge and Salt ranges orthogonal to the strike of the Himalayas are well imaged as a high velocity structures. The Delhi-Haridwar ridge is butting against the northwest Himalayas that led to ramming and locally buckling of the crust therein. I have collated the results and geodynamic interpretations derived from the study conducted in three different segments and provided a unified model of this tectonically important and unique region. I have also presented the velocity structure in cross-sections taken all along the tectonic trend of the Himalayas and the Indo-Burma Ranges. I have found that the crustal thickness shows variation also along the tectonic trend of the Himalayas and tectonic trend of the Indo-Burma Ranges. This is interpreted as a manifestation of stresses encountered v by this region due to collision between Indian and Eurasian plates and rotation of the Indian plate during this process. I suggest that this is an effect of first collision between Indian and Eurasian plates in the NW and then subsequent anticlockwise rotation of Indian plate, leading to crumpling of the crust. This could also be due to variable thickness of more rigid portion of the incoming crust of Indian plate. The Indian lithospheric slab is divided into two parts one towards its south east where it underthrusts under the Eastern Himalayas and buckled up below the northeast India region and then subducts below the Indo-Burma Ranges where it breaks off at a depth ~90 km and the broken part is sinking into the mantle with steep angle. The other part in the northwest of the study region represents the subducted part of the Indian lithospheric slab beneath the Pamir-Hindu Kush region towards north-northwest. The intermediate depth of seismicity following the Benioff zone is dominant beneath the Indo-Burma Ranges towards east and Pamir-Hindu Kush region towards northwest of the study region. The comprehensive tomographic image estimated in this study provides further insight into the geodynamics of the whole study region that helps to understand the earthquakes generating mechanisms.