摘要
为了从跨断层形变观测资料中获取中强地震的前兆异常信息 ,运用二进小波变换方法和噪声与信号的定量识别方法 ,分析了不同尺度的小波变换的细节信号变化特征。研究结果表明 :( 1)不同尺度小波变换分解的细节信号显示出不同的时变特征。当尺度为 1时 ,细节信号主要为噪声 ;尺度为 2、4、5时 ,细节信号主要为非平稳信号 ;尺度为 3时 ,细节信号主要为平稳周期信号 ;( 2 )每月观测一次的跨断层形变资料在尺度为 3时的细节信号有较好的年周期特征 ;( 3)当尺度为 3时的细节信号变化超出 2倍均方误差时 ,测点周围 2 0
In order to catch the seismic anomalous informations from deformation observation before moderately strong earthquake, we applied wavelet transform method and quantitative recognition method of noise and signal to analyze the characteristics of the detailed part of signal variation after wavelet decomposition. The research results indicate that: (1)the detailed decomposed signal characteristics of wavelet were different when the scale was different:The detailed decomposing signal of wavelet contains mainly noise corresponding with scale 1; the detailed decomposing signal is non stationary when scale are equal to 2,4 and 5,and the signal is quasi periodic when scale is equal to 3;(2)The detailed signal of wavelet of monthly cross fault deformation data have annually periodic feature when scale is equal to 3;(3)It is considerably possible that a moderately strong earthquake will occur in the range of about 200 km from the observation point when the detailed signal variation exceeds the double mean square error and the scale is 3.
出处
《大地测量与地球动力学》
CSCD
2004年第3期73-78,共6页
Journal of Geodesy and Geodynamics
基金
山东省自然基金项目 (Y2 0 0 0E0 8)
2 0 0 4年山东省地震局合同制项目
关键词
小波变换
跨断层形变观测
地震前兆异常
地震预测
wavelet transform,cross-fault deformation data,the separation of signal from white noise, precursory anomaly of earthquake,earthquake prediction
