Crustal Velocity Structure of the Bengal Basin from Regionally Recorded Earthquakes

Tuesday, 16 December 2014
Michael Howe1,2, Won-Young Kim2,3, Syed Humayun Akhter4, Dhiman Ranjan Mondal5, Leonardo Seeber6 and Michael S Steckler6, (1)Columbia University in the City of New York, Palisades, NY, United States, (2)Lamont-Doherty Earth Observatory, Palisades, NY, United States, (3)University of Tokyo, Earthquake Research Institute, Yayoi 1-1, Bunkyo, Tokyo, Japan, (4)University of Dhaka, Dhaka, Bangladesh, (5)CUNY Graduate School and University Center, Earth and Environmental Sciences, New York, NY, United States, (6)Lamont-Doherty Earth Obs, Palisades, NY, United States
The Bengal Basin lies near the junction of the Himalaya and the Burma Arc subduction system. Continental collision is taking place to the north where the Indian Shield converges with the Himalayas with the intervening uplifted Shillong basement block. The Bengal Basin is built over the passive margin of India and has prograded the shelf 300-400 km since the Eocene. The basin is colliding with and subducting eastward under the Burma Arc. Global seismic velocity models do not apply well to this region because of the depositional environment in the Bengal Basin with sediment thicknesses ~20 km surrounded by Precambrian basement and an accretionary prism. Accurate velocity models are needed to improve earthquake locations and focal mechanism solutions in this region of high seismic hazard. While previous studies in this area show diffuse seismicity [Howe et al., 2013] there are several identifiable earthquake clusters. An improved velocity model will improve the accuracy of the hypocenters in known clusters as well as illuminate previously unidentified clusters. Several studies provide useful velocity constraints using well-logs, receiver function analysis, and surface wave dispersion [e.g., Sikder & Alam, 2003; Lindsay et al., 1991; Curray et al., 1991; Mitra et al., 2005,2008]. These approaches, as well as gravity studies [Rajasakhar & Mishra, 2008], support the suggestion of a layer of sediment gradually thickening from northwest to southeast, to more than 15-20 km thick in some places. While receiver function analysis and well-log data are extremely helpful with constrain depth-dependent crustal structure, lateral variations in structure are not well-sampled. In order to obtain better constraints, we use hypocenters and travel times of local and regional earthquakes that have source-receiver raypaths that stay within the crust and upper mantle. An array of seismographs has been in operation in Bangladesh since 2007. This study uses the recordings from this array to provides travel time data for local and regional events which are then used to compute better estimates of the crustal velocity structure in the region.