利用小波变换识别新疆地震台网震相  

SEISMIC PHASE IDENTIFICATION WITH WAVELET TRANSFORMATION IN XINJIANG SEISMIC NETWORK

作　　者：冯雪玲[1] 魏斌[1] 刘东亚[1] 闫新义 王宝柱[1] 上官文明[1] FENG Xue-ling;WEI Bin;LIU Dong-ya;YAN Xin-yi;WANG Bao-zhu;SHANGGUAN Wen-ming(Earthquake Agency of Xinjiang Uygur Autonomous Region,Urumqi 830011,Xinjiang,China)

机构地区：[1]新疆维吾尔自治区地震局,新疆乌鲁木齐830011

出　　处：《内陆地震》2020年第4期340-346,共7页Inland Earthquake

基　　金：中国地震局“三结合”课题(3JH-202001097);新疆地震科学基金(202004)。

摘　　要：结合新疆地震台网记录的地震震相特征,针对噪声干扰大、深度震相sPn记录不清晰和重叠地震等问题,采用sym8小波变换进行处理。发现该方法能够有效抑制地震信号中的噪声干扰,提高地震震相识别精度。对于振幅弱,初动不清的深度震相sPn,当选择合适的尺度-频率特征重构后也能清晰分辨出该震相初至,并明显降低了sPn震相到识别误差,sPn震相的优势频带范围为0.1~1 Hz。对于同源的重叠地震,采用某个分解尺度上对应的小波系数进行重构,即可有效识别重叠地震震相初至。分析认为,小波变换是一种识别新疆地区地震震相的有效方法。Combined with the seismic phase characteristics recorded by the Xinjiang Seismic Network, the sym8 wavelet transform is used to deal with the problems of large noise interference, unclear deep seismic phase sPn records and overlapped earthquakes. It is found that this method can effectively suppress the noise interference in seismic signal and improve the accuracy of seismic phase recognition. For the deep seismic phase sPn with weak amplitude and unclear initial motion, the initial moving position of the seismic phase can also be clearly distinguished after the reconstruction of the appropriate scale-frequency characteristics. The dominant frequency band range of the sPn phase is 0.1~1 Hz. For homologous overlapping earthquakes, the phase initiation position of overlapping earthquakes can be effectively identified by using the corresponding wavelet coefficients on a decomposition scale.It is considered that wavelet transform is an effective method to identify seismic phases in Xinjiang.

关 键 词：小波变换 震相识别 干扰 SPN震相 重叠地震 

分 类 号：P315.7[天文地球—地震学]