Rayleigh wave group velocity tomography of Gujarat region, Western India and its implications to mantle dynamics

In the present study, fundamental Rayleigh waves with varying period from 10 to 80 s are used to obtain group velocity maps in the northwest Deccan Volcanic Province of India. About 350 paths are obtained using 53 earthquakes (4.8 M 7.9) recorded by the SeisNetG (Seismic Network of Gujarat). Individual dispersion curves of group velocity of Rayleigh wave for each source-station path are estimated using multiple filter technique. These curves are used to determine lateral distribution of Rayleigh wave group velocity by tomographic inversion method. Our estimated Rayleigh group velocity at varying depths showed conspicuous corroboration with three tectonic blocks [Kachchh Rift Basin (KRB), Saurashtra Horst (SH), and Mainland Gujarat (MG)] in the region. The seismically active KRB with a thicker crust is characterized as a low velocity zone at a period varying from 10 to 30 s as indicative of mantle downwarping or sagging of the mantle beneath the KRB, while the SH and MG are found to be associated with higher group velocities, indicating the existence of the reduced crustal thickness. The trend of higher group velocity was found prevailed adjacent to the Narmada and Cambay rift basins that also correspond to the reduced crust, suggesting the processes of mantle upwarping or uplifting due to mantle upwelling. The low velocities at periods longer than 40 s beneath the KRB indicate thicker lithosphere. The known Moho depth correlates well with the observed velocities at a period of about 30 s in the Gujarat region. Our estimates of relatively lower group velocities at periods varying from 70 to 80 s may correspond to the asthenospheric flow beneath the region. It is interesting to image higher group velocity for the thinner crust beneath the Arabian Sea adjacent to the west coast of Gujarat at the period of 40 s that may correspond to the upwarped or upwelled mantle beneath the Arabian Sea. Our results have better resolution estimated by a radius of equivalent circular averaging area for each period.