Seismic imaging of crust beneath the western Tibet-Pamir and western Himalaya using ambient noise and earthquake waveform data

Abstract

Kinematics and dynamics of the evolution of continental collisions and mountain building processes have been a primary area of research to understand the deformation mechanism of continental interiors. The Himalaya-Tibet-Pamir mountain belts, which have an evolutionary history beginning about 50 million years ago with the closure of the Tethys Ocean followed by the collision of the Eurasian and Indian landmass, is a natural laboratory for studying the ongoing large-scale geodynamic processes. Since the early versions of isostasy, proposed by Airy (1855), and Pratt (1859) inferring that a deficit of mass must underlie the Himalayas and Tibet, several models are proposed to explain its large-scale crustal shortening and consequent thickening, and uplift. End member models include thrusting of the Indian continental lithosphere beneath Asia, first proposed by Argand (1924), and/or diffused deformation within Asia that led to crustal thickening and uplift of the orogen. Much of our knowledge about these models are based on geological and geophysical investigations from the central and eastern parts of the Tibet-Himalaya system. These studies have further revealed evidence for the underplating and subsequent eclogitization of the Indian lower crust beneath the Tibetan plateau and the partially molten state of the middle crust of the Tibetan plateau. However, continuity of these features to the western segment of the Himalaya-Tibet system comprising western Himalaya, Ladakh-Kohistan arc, western Tibet and Karakoram, and Pamir Hindu Kush regions, remain speculative due to the absence of experiments on a similar scale as in the central and eastern Tibet.

In this thesis, we present a 3-D crustal architecture, including vertical extent and stratification, and the nature of boundaries separating distinct units laterally and vertically with a lateral resolution of 30−50 km up to a depth of 100 km below sea level beneath the western Himalaya- Ladakh-Pamir-Tibet region, to understand the geological evolution of the region with a focus on defining the geometry of the underthrusting Indian crust and its northern edge, understanding crustal deformation and disparity of evidence for channel flow and mapping the depth reach of the Karakoram fault. The velocity image is produced using ambient noise cross- correlations from about 530 seismological stations along with surface wave observations from 1,261 earthquakes recorded over the seismological network. The velocity image is reconstructed following a two-step procedure. Firstly, we compute Rayleigh wave group velocity maps for the region at 0.5° × 0.5° grid interval from period 5 to 60 seconds using earthquake and ambient noise waveform by the Bayesian Trans-dimensional tree tomography. The dispersion data at each grid node is then converted to shear wave velocity variation with depth by a trans-dimensional, hierarchical Bayesian inversion.

We present the first-order crustal structure in terms of lateral distribution and connectivity of the mid-crustal low-velocity zones (LVZs) and the nature of the Indian lower crust that underplates the Tibetan Plateau. Moho beneath the Himalayas and south Tibet correlates with a velocity of 4.4−4.6 km/s and a reduced velocity of 4.0−4.2 km/s in northern Tibet and Pamir. We used the Moho depth and nature of high-velocity lower crust (Vs > 4.0 km/s) to map the northern limit of the Indian crust that extends beyond the Qiangtang block in western Tibet (77−82°E) from its previously assumed boundary in the Lhasa block and till central Pamir further west. The velocity image reveals discontinuous low-velocity zones (LVZs) (Vs < 3.4 km/s) in the upper and middle crust in western Tibet and Pamir that rarely connect to the high Himalayas as expected for a ductile channel flow. The LVZs in Pamir correlate with the surface distribution of gneiss domes. The lowest velocities (Vs < 3.2 km/s) are observed over Ladakh-Karakoram batholiths and the Nanga Parbat region. The study suggests a continuation of LVZs across the Karakorum Fault at a depth beyond 20 km, indicating the fault’s very shallow (upper crustal) depth extent.

The detailed analysis of the velocity image reveals an unreported nearly arc perpendicular crustal-scale transverse structure along ~77°E that segments western Himalaya longitudinally. East of this transverse structure, the Main Himalaya Thrust (MHT) has a gentle northward dip of 10−16° and reaching a depth of 40−45 km near the Indus Suture Zone (ISZ). In contrast, to the west, the MHT is nearly flat at 15−20 km depth between the Main Frontal Thrust (MFT) and the ISZ with no evidence of a north dipping ramp. The transverse structure is a broad zone with low velocity (Vs < 3.4 km/s) in the mid to lower crust. The eastern limit of this transverse structure coincides with the northward extension of the Simla/Ropar Manali lineament, a pre-Himalayan basement fault in the Indian lithosphere. We propose that the reactivation of this lineament in the down-going Indian plate and subsequent strain propagation onto the surface through the weak mid/lower crust is possibly responsible for the segmentation in the western Himalayan arc.

The velocity image of the region provides a unique opportunity to understand the nature of the Kohistan-Ladakh complex formed as an island arc within the Tethys Ocean in Mesozoic times, thrust southward onto the Indian margin to become ultimately squeezed between the converging Indian and Asian plates. Based on the modeling of seismological data, we present the first evidence for the high-velocity (high-density) lower crustal root under the Kohistan arc that is conspicuously absent in Ladakh. In the upper and middle crust, the Kohistan arc shows a uniform Vs ~ 3.5 km/s underlain by a thick (~25 km) high velocity lower crust (Vs > 4.0 km/s). In contrast, the Ladakh arc crust is largely characterized by a Vs < 3.4 km/s between 15 to 40 km depth, and an absence of the thick mafic basal layer at greater depths.

Finally, we use seismic waveform recorded on 26 closely spaced (~10 km) broadband seismographs across the strike of Garhwal Himalaya to produce a high-resolution image of the MHT geometry from ambient noise tomography. Our model suggests two distinct ramps on the MHT -a gentle dipping (~ 9°) at shallow depth (~ 7−12 km) located 40 km north of the Main Boundary Thrust, and other a steeply dipping (~ 30° to 35°) at ~15−25 km depth beneath the Higher Himalayan front. The inferred double ramp geometry in this study highlights the complex segmentation of MHT in the Garhwal Himalaya, which provides an important constraint in simulating earthquake hazard potential and modeling the growth of Himalayan topography.