Identifying P- and S-Waves Using Artificial Neural Network

An artificial neural network (ANN) is used to identify P- and S- wave arrivals from seismic data. Identification is achieved by utilizing the polarization state as a function of time, input directly into a three-layer neural network. The results demonstrate that an ANN trained with small dataset can satisfactorily identify P- and S-arrivals with a success of 84% and 60% respectively and could be applied to automatic analysis of multicomponent seismic data.