A semi-empirical method using a hybrid of stochastic and deterministic fault models: Simulation of strong ground motions during large earthquakes.

Abstract

A new semi-empirical method is developed to estimate strong ground motions, which takes into account fault plane irregularities in the prediction of strong ground motions during large earthquakes. According to this method, the displacement on the fault plane of a large earthquake is divided into two parts. One is the average displacement Do over the fault plane and the other is the deviation ΔDij from Do. The seismic radiation SL(t) due to Do and the seismic radiation SS(t) due to ΔDij are synthesized separately using the observed seismograms for a small element earthquake. A parameter SD is introduced as the standard deviation of ΔDij divided by the displacement of the element earthquake, which shows the degree of inhomogeneity of displacement on the fault plane. The new method is applied to synthesis of strong ground motions at 18 stations during 8 large earthquakes with magnitude M of 6.7 to 7.9 occurring in and around Japan. The synthesized records are in good agreement with the observed ones except for the records obtained at the stations on soft soil deposits, if SD is chosen to be about 1.0. These results indicate the applicability of the new method for predicting strong ground motions on relatively stiff ground which Does not show non-linear behavior during strong ground shaking.