Research On Microseismic Data Time Picking Method Based On Shearlet Transform And Akaike Information Criterion

At present,the real-time dynamic monitoring of oil and gas fields using microseismic exploration has attracted much attention at home and abroad.With the huge consumption of oil and gas,unconventional oil and gas exploitation has become to a new hot issue.Low-permeability fields,microseismic exploration has gradually replaced traditional petroleum exploration and became the most important method of oil exploration.Microseismic monitoring technology is usually divided into three steps: pre-processing,first arrival picking and source positioning.The micrseismic monitoring technology needs real-time,automatic,and efficient interpretation of the high-density microseismic exploration data received by the geophone to achieve high-precision positioning of the source.Time picking is a crucial step in microseismic data processing,the picking results have great influence on orientation of hypocenter location.For guaranteeing the accuracy of microseismic data,the create of an automatic,real-time,high speed and reliable time picking method is a crucial step to improve the processing efficiency.The field microseismic record valid signal is of high frequency domain,weak energy strength,short time period and severe interference caused by complex random noise,resulting in extremely low signal to noise ratio.Manual picking is of low efficiency,easy introduction of manual errors,and inability to adapt to actual condition.Automatic microseismic traditional time picking methods,for instance,the energy ratio picking algorithm,the Akaike Information Criterion(AIC)method and the wavelet transform have good performance in picking high signal-to-noise ratio(SNR)microseismic data,but when the SNR of data is low,it is difficult to obtain arrival times accurately with these conventional approaches.This article studies the automatic time picking method of the microseismic data in order to guarantee the accuracy of picking results in low SNR environment.There is no remarkable characteristic difference between the valid signal and background noise in low SNR condition,which makes it hard to achieve the arrival time effectively in time domain.Thus the Shearlet transform is introduced into microseismic time picking area in this dissertation,by which the noisy microseismic record is mapped from time domain to Shearlet domain.According to the parameter differences between the valid signal and noise at the fine scales,the signal points can be differentiated from background noise.Compared with the traditional picking methods such as the short and the long time average(STA/LTA)algorithm and the AIC algorithm,the algorithm based on the Shearlet transform can efficiently raises the picking accuracy of the low SNR microseismic data.By analyzing the Shearlet transformation,the feasibility and effectiveness of distinguishing the microseismic signal from the noise in the Shearlet domain are clarified.In order to achieve better picking performance under low SNR environment,this paper propose a new time picking approach based on the Akaike Information Criterion(AIC)and Shearlet transform named the Shearlet-AIC algorithm,it can accurately pick the arrival times.According to the parameter differences between the effective signal and background noise at fine scales,the signal points can be preliminary identified from noise.Then,we operate the data by the AIC algorithm and regard the minimum AIC value as the arrival times.To verify the reliability of the proposed method,we conduct it on both synthetic and field data sets that were recorded with a vertical array of receivers.The experimental results show that our method can precisely pick the arrival times compare with the STA/LTA algorithm,the AIC and wavelet based AIC method and confirm the reliability and stability of the Shearlet-AIC algorithm.It is an of high accuracy and practical first arrival picking algorithm and it can provide exact and credible arrival times for low SNR microseismic records.