Seismic Source Identification Using Self-Organization Technique

This paper propose the ratio of M۱, (Local Magnitude) to Mo (Scaler Seismic Moment) in Eastern Kazakh for ۲۵ earthquakes and ۱۵ explosions, MI. is estimated from the peak horizontal displacements of the short period P waves and the seismic moment Mm is estimated by spectrum measruernents serving as a long-period energy measure (Woods et al. ۱۹۹۳). An artificial neural network (ANN) algortithin is applied on the problem of discriminating between the signals from underground explosions and from earthquakes using these two features (Mr. & KO. The results of preliminary training and testing with a set of local earthquake recordings show that the self-organization network can achieve a good performance in discrimination between earthquakes and artificial events. In overall result, ۹۶% correct identification is achieved by our ANN, we believe that the method presented here is a promising approach to automatic identification of seismic source.