Three Dimensional P Wave Velocity Model for the Crust Containing Aftershocks of the Bhuj, India Earthquake

A three-dimensional P wave velocity model has been constructed for the crust in the vicinity of the Mw=7.7 January 26th Bhuj, India earthquake using aftershock data obtained by CERI away teams. Aftershocks were recorded by 8 portable, digital K2 seismographs (the MAEC/ISTAR network) and by a continuously recording Guralp CMG40TD broad-band seismometer. Station spacing is roughly 30 km. The network was in place for 18 days and recorded ground motions from about 2000 aftershocks located within about 100 km of all stations. The 3-D velocity model is based upon an initial subset of 461 earthquakes with 2848 P wave arrivals. The initial 1-D velocity model was determined using VELEST and the 3-D model was determined using the nonlinear travel time tomography method of Benz et al. [1996]. Block size was set at 2 by 2 by 2 km. A 45% reduction in RMS travel time residuals was obtained after 10 iterations holding hypocenters fixed. We imaged velocity anomalies in the range -2 to 4%. Low velocities were found in the upper 6 km and the anomalies follow surface features such as the Rann of Kutch. High velocity features were imaged at depth and are associated with the aftershock hypocenters. High crustal velocities are present at depths exceeding 20 km with the exception of the crust below the Rann of Kutch. The imaged velocity anomaly pattern does not change when different starting models are used and when hypocenters are relocated using P wave arrivals only. The analysis will be extended to an expanded data set of 941 aftershocks.