Stochastic finite-fault method controlled by the fault rupture process and its application to the Ms 7.0 Lushan Earthquake

Fast and precise estimation of strong ground motion is important for seismic hazard responses and engineering purposes. Stochastic finite-fault methods are effective ground-motion time-series simulation methods using stochastic models of point-source spectra and considering physical parameters on finite faults. We propose two main modifications of current stochastic finite-fault methods to control the simulations. We replace the predetermined time function with different source time functions for each subfault to improve low-frequency results including the envelopes of accelerations and velocities. We replace the uniform stress drop with a distribution of non-uniform stress drops to reduce the dependence of simulated results on a single stress drop. The two modifications are applied to simulate ground motions of the 2013 Ms 7.0 Lushan earthquake. The simulation results agree with the observations in terms of acceleration, velocity, and pseudo-spectral acceleration (PSA). Damage estimates based on the simulated PSA also agree with the post-earthquake disaster investigation.