Spatiotemporal extraction of slow slip information considering signal oscillation characteristics
Aiming at the problem of the false information removal or false signal extraction using the traditional filtering and fixed function fitting methods in detection of slow slip signals,the paper proposed a method for detecting slow slip spatiotemporal information based on signal oscillation characteristics:the coordinate sequence was decomposed by multi-channel singular spectrum analysis;and then the signal properties and start-end time were determined according to the displacement characteristics of slow slip;finally,the direction and intensity of the spatial response were determined based on the oscillation direction and the amplitude normalization of the signal.Simulational results showed that the proposed method would have obvious advantages in detecting the start-end time,spatial response direction and signal intensity of slow slip compared with principal component analysis and independent component analysis;moreover,in practice,the method had been successfully used to detect the slow slip event in Manawatu,New Zealand,where the spatial responses on both sides of the line between the station WANG and PNUI were opposite in directions,and the intensity was large,which could lead to a high seismic risk,however,the spatial responses detected by independent component analysis and principal component analysis might be affected by common mode errors,and the start-end time of slow slip could not be determined by time responses.
万方数据
维普
