Abstract: The first-break picking accuracy greatly affects the seismic events discriminant and source localization.At present，none of the first-break picking algorithms proposed is ideal to collect the low SNR seismic data.In order to solve the noise problem (improve noise resistance performance)，an improved first arrival picking method of micro-seismic event based on power ration and wavelet transform is proposed.Firstly，the characteristics function is introduced to the traditional energy ratio method wavelet transform method so as to search the range of the P-wave first break by adopting the improved energy ratio method.Then，the first break time of P-wave is positioned precisely by using wavelet transform.Finally，the seismic model data and actual measured seismic data are used to conduct experiment to analyze the performance of the method proposed in this paper.The results show that，the method based on energy ratio and wavelet transform is the first-break automatic picking method with the characteristics of noise resistance and stability.Besides that，it can identify the P-wave first arrival time from the data with a low signal to noise ratio accurately.

Key words: P-wave, First arrival picking, Energy ratio method, Wavelet transform, Characteristics function