The seismic wave consists of many seismic phases, which contain rich geophysical information from the hypocenter, medium of seismic wave passing through and so on. It is very important to detect and pick these seismic phases for understanding the mechanism of earthquake, the Earth structure and property of seismic waves. In order to reduce or avoid the loss resulted from the earthquake, one of the important goals of seismic event detecting is to obtain its related information before and after it occurs. Because of the particularity of P wave and S wave, the seismic event detecting focuses on distinguishing P and S waves and picking their onset time, it has been becoming one of the research hotspots for many geophysicists to pick the P and S wave arrival accurately and effectively. Because the wavelet transform is a very good time- frequency analysis method to deal with the non-steady and nonlinear signals, and the shape of wavelet is much similar to that of seismic wave and they match each other well (WANG, 2004), wavelet transform is more and more widely used to analyze seismic phase in recent years. At present, on the basis of the wavelet transform, the analysis method of detecting the seismic phase and picking their onset time mainly includes these steps, that is transforming seismic signal by wavelet, reconstructing characteristic function by virtue of existing detecting algorithm, and finally picking onset time according to some criterions. Kiyoshi (1994) used wavelet transform to detect simulation seismic phase and obtained good results, which is important on applying wavelet transform (WT) to studying seismic phase. Anant and Dowla (1997) decomposed seismic signals using multi- scale wavelet and then calculated the covariance matrix of wavelet coefficient on each scale, finally determined the onset time of P and S waves by the linear polarization and energy ratio on the transformation plane. Oonincx (1998) also improved the detecting method. In virtue of the automatic detection seismic phase method and some problem occurring during the analysis process, by combining the wavelet transform with the traditional methods, he divided seismic signals into different frequency range by WT according to the frequency of P and S wave. Owing to the arrival time difference of P and S wave and the difference of the local and non-local earthquake frequency range, the best and most decomposing scale can be obtained. According to this scale, the onset time of P and S waves will be detected by the traditional ones. LIU et al (2000) also picked the arrival time of P wave and S wave by using the similar method to Oonincx' s. The analysis result is very good on some seismic events using the above mentioned methods, however, it is difficult to get the satisfied results on most seismic events by the orthogonal wavelet transform due to complexity of seismic signal. The biorthogonal wavelet introduces the linear phase and keeps weak orthogonal by giving up its partial orthogonality, so it is advantageous to obtain seismic phase characteristic and improve precise and efficiency of picking seismic phase arrival time. In this paper, the method is proposed to construct the characteristic function and pick P wave onset time based on the B-spline biorthogonal wavelet. In the first section, it gives the principle and merits of the biorthogonal wavelet, then in the second section, it shows the method and steps of constructing the characteristic function indetails. The compared results are described between our analysis and the other ones based on the real earthquakeevents in the third section, and in the last section our conclusion, main research results and some related questions are discussed.
