Estimation of the finite fault source model for the 2018 Mw7.1 Acari, Peru earthquake using a hybrid simulation prediction approach

Abstract

The near-field motion of large earthquakes is generally simulated using finite fault source models (FFSMs). Most of these models use post-earthquake observation data as the constraints for the inversion model, and therefore, it is difficult to generate FFSMs until after the earthquake has occurred. We use parameters from previous earthquakes to generate FFSMs in a hybrid simulation prediction approach that combines the asperity model and the k−2 model. New scaling laws for global and local parameters are used to estimate the asperity model parameters and the coordinates of the fracture initiation point. Using the k−2 model, 30 sets of FFSMs are established using truncated normal distributions and several regional geological structural parameters. After applying the residual evaluation criteria to the various response spectra, we select the average source model. We then apply our methodology to the 2018 Acari earthquake in Peru. Both our average model and the USGS inversion model have two asperities, but their positions are different. The response spectra of the four representative points of our model are similar to those of the USGS model, and the peak difference is between −13.4% and 18.4%. The USGS model is a post-earthquake model, which requires many parameters and post-earthquake observation data. Our FFSM requires few parameters and represents the most representative source model for future earthquakes in this area with similar magnitudes, and it is intended to be supplementary to the inversion model. Therefore, our hybrid approach may be used for the early prediction or rapid estimation of the resulting ground motions.