利用小波变换方法分析形变观测资料的正常背景变化特征  

ANALYSES ON NORMAL BACKGROUND CHARACTERISTICS ABOUT DEFORMATION OBSERVATION DATA ON THE BASIS OF WAVELET TRANSFORM METHOD

作　　者：李杰[1] 刘希强[1] 李红[1] 毛玉华 郑树田 

机构地区：[1]山东省地震局,中国济南250014 [2]泰安地震基准台,中国山东泰安271600 [3]安丘地震局,中国山东安丘282100

出　　处：《地震学报》2005年第1期33-41,共9页Acta Seismologica Sinica

基　　金：山东省自然基金项目(Y2000E08);2002年中国地震局三结合项目;2003年山东省地震局合同制项目共同资助.

摘　　要：　为定量刻画数字化形变观测资料中背景信息和噪声的时 频分布特征, 本文应用二进小波变换方法, 通过对小波分解的主模特征和随机白噪声识别因子变化特征分析, 剖析了山东数字化形变观测资料的正常动态背景和噪声变化规律. 结果表明, 当尺度取 2, 3 和 4 时,分解后的细节部分存在着1/4日波、半日波、日波和半月波等准循环周期信号, 其中尤以尺度为3时的信号波幅最大; 尺度取1和5时的细节部分主要包含着噪声; 通过分析和追踪指定尺度的数字化形变观测资料小波变换的非震异常特征变化, 可望捕捉到与强地震孕育过程有关的前兆异常信息.Wavelet transform method is applied to measure time frequency distribution characteristics of digital deformation data and noise. Based on the characteristics of primary modulus and stochastic white noise discrimination factor of wavelet decomposition, we analyze the variation rule of normal background and noise data from Shandong digital deformation observation data. The research results indicate that: a) 1/4 daily wave, semi diurnal tide wave, daily wave and half lunar wave and so on quasi periodic signal exist in the detail decomposing signal of wavelet when scale are equal to 2, 3 and 4; b) The amplitude of detail decomposing signal is the biggest when scale is equal to 3; c) The detail decomposing signal contain mainly noise corresponding to scale 1 and 5, respectively; d) We may trace the abnormal precursory which is related to earthquake by analyzing non earthquake detail wavelet decomposing signal whose scale is specified from digital deformation observation data.

关 键 词：小波变换 数字化形变资料 信号与噪声分离 地震前兆识别 

分 类 号：P315.31[天文地球—地震学]