IDENTIFICATION AND PICKING OF S-PHASE USING AN ARTIFICIAL NEURAL-NETWORK

An artificial neural network (ANN) algorithm has been applied to the a utomatic picking of local and regional S phase. For a set of local thr ee-component seismic data, a variety of features for signal detection and phase identification were analyzed in terms of sensitivity and eff iciency. Comparing the performance of each feature in discriminating t he local S phases, four features were selected as input attributes of the ANN S-phase picker: (1) the ratio between short-term average and l ong-term average, (2) the ratio between horizontal power and total pow er, (3) auto-regressive model coefficients, and (4) the short-axis inc idence angle of polarization ellipsoid. The four attributes were calcu lated in the frequency band of 2 to 8 Hz with a 2.56-sec moving window . This choice of frequency band and window length is appropriate for l ocal microearthquake monitoring. The results of preliminary training a nd testing with a set of local earthquake recordings show that the ANN S-phase picker can achieve a good performance in identification and o nset-time estimation for local S phases. In overall result, 86% correc t rate of phase identification has been achieved by the trained ANN S- phase picker, 74% of them are precisely picked with less than 0.10-sec onset time error. We believe that the method presented here is a prom ising approach to automatic phase identification and onset-time estima tion.