カルマンフィルターを用いた断層破壊過程の同定

Abstract

The Kalman filter is applied to deconvolute a seismic wave into a system response functionand a train of impulses. Based on several criteria of minimum error, we propose a method toidentify a rupture process of a fault using the deconvoluted train of impulses.A simulated wave and a waveform recorded in a blasting are analized in order to examine thevalidity of the proposed deconvolution method. The records obtained at seven JMA local stationsduring Nihonkai Chubu Earthquake are used to identify its fault mechanism. Since deconvolutedseven trains of impulses contain informations about the fault mechanism, the rupture processdefined by such focal parameters as the rupture velocity, the starting point of rupture, the directionof the rupture propagation and the rise time can be estimated by minimizing the error betweendeconvoluted trains of impulses and theoretically calculated ones by selecting proper values of faultparameters in the rupture process.The identified rupture process compares well with that investigated from different approach.