一种改进微地震数据极化滤波性能的偏振特征函数设计  

A polarization eigenfunction for improving polarization filteringperformance of microseismic data

作　　者：李林家 李怀良 LI Linjia;LI Huailiang(College of Geophysics,Chengdu University of Technology,Chengdu 610059,China;Key Laboratory of Earth Exploration and Information Technology of Ministry of Education,Chengdu University of Technology,Chengdu 610059,China)

机构地区：[1]成都理工大学地球物理学院,四川成都610059 [2]成都理工大学地球勘探与信息技术教育部重点实验室,四川成都610059

出　　处：《石油物探》2023年第3期471-482,共12页Geophysical Prospecting For Petroleum

基　　金：国家自然科学基金项目(41774118);四川省杰出青年科技人才项目(2021JDJQ0022)共同资助

摘　　要：地震数据极化滤波广泛应用于地震数据处理中噪声压制、震相识别和波形分离且效果很好。在时间域极化滤波中,通常利用协方差矩阵的特征值构造偏振特征函数来设计极化滤波器。针对传统极化滤波器对有效信号和噪声识别不清晰的问题,总结分析了当前应用较为广泛的10种偏振特征函数,讨论了这些函数对地震噪声信号的表征能力。在分析偏振特征函数优、缺点的基础上,使用特征值构造一种新的偏振特征函数表达式,基于该特征函数设计的极化滤波器能够提高地震有效信号和噪声的识别精度,更有效实现噪声衰减。利用合成和实测的微地震数据验证了不同偏振特征函数对极化滤波器性能的影响。结果表明,采用新构建的偏振特征函数设计的滤波器,能有效识别噪声数据段,对初始相位影响较小且滤波后地震数据的信噪比(SNR)、均方根误差(RMSE)和互相关归一化系数(NCC)均优于其它偏振特征函数设计的滤波器。处理后的地震信号矢端曲线图能更好地反映地震信号的偏振方向,更有利于确定微地震初至波的极化方向。Seismic polarization filters are widely used in seismic data processing for noise suppression,seismic phase identification,waveform separation,and first-arrival picking.For the time-domain polarization filter,the eigenvalue of the covariance matrix is typically used to construct the polarization eigenfunction,which is then used to design a polarization filter.Traditional polarization filters cannot clearly identify the effective signal and noise.Therefore,ten popular polarization eigenfunctions were analyzed in this study,and their ability to represent seismic noise signals is discussed.By considering the benefits and drawbacks of existing polarization eigenfunctions,a new polarization eigenfunction is proposed.The polarization filter designed using this new eigenfunction can improve the recognition accuracy of the effective seismic signal and noise,as well as achieve more effective noise attenuation.The impact of different polarization eigenfunctions on the polarization filtering performance was tested using synthetic and real microseismic data.The results showed that filters designed with the newly constructed polarization eigenfunctions can effectively identify noisy data segments with less influence on the initial phase.The results also showed that the filtered seismic data had a higher signal-to-noise ratio,root mean square error,and normalized mutual correlation coefficient than those designed with other polarization eigenfunctions.The hodogram of the processed seismic signal accurately reflects the polarization direction of the seismic signal,which is helpful for determining the polarization direction of the first arrival in microseismic data.

关 键 词：偏振特征函数 地震噪声衰减 极化滤波 微地震 信噪比 矢端曲线图 极化方向 

分 类 号：P631[天文地球—地质矿产勘探]