Predicting strong motion parameters for the Chamoli earthquake of 28th March, 1999, Garhwal Himalaya, India, from simplified finite fault model

State of Uttaranchal in the northern part of India in the Garhwal Himalaya was hit by the Chamoli earthquake on 28th March, 1999 (GMT). This earthquake was recorded on a strong motion array installed in this region. The maximum peak ground acceleration of 353 cm/sec2 was recorded at an accelerograph located at the Gopeshwar station at an approximate epicentral distance of 14 km. The simplified method of Midorikawa (1993) has been used to model finite fault responsible for causing the Chamoli earthquake. This method is based on the Empirical Green's Function (EGF) technique of Irikura (1986).Modifications in this method have been made to include layered earth model and transmission effects at each boundary by Joshi (2001). Rupture causing the Chamoli earthquake is placed in two structural models of the earth in this work: one is a homogeneous half space and other is the multi layered earth model. Comparison in terms of root mean square error (RMSE) is made between the simulated and actual strong motion parameters like peak acceleration and duration. It is seen that the introduction of multi layered earth system in this simplified technique is capable of significantly reducing the RMSE in observed and predicted strong motion parameters and defining the attenuation rate for peak ground acceleration of this earthquake.