基于小波包分析的OLR地震异常信息提取初探  

A Preliminary Study on Extraction of Earthquake-related OLR Anomalies Using Wavelet Packet Decomposition

作　　者：戴勇[1] 丁风和[2] 韩晓明[2] 

机构地区：[1]中国地震局兰州地震研究所,甘肃兰州730000 [2]内蒙古自治区地震局,内蒙古呼和浩特010010

出　　处：《地震》2009年第3期61-66,共6页Earthquake

摘　　要：长波辐射OLR具有可覆盖范围大的特点。本文提取其与地震相关的异常信息,并尝试将所得结果应用于地震预测中,这对于弥补我国部分处于前兆台网监测范围之外地区具有重要意义。文中首先对小波包方法及其原理进行了简要介绍,在此基础上,尝试采用该方法对长波辐射值OLR时间序列进行分解重构,提取异常信息。通过对15个不同震例的研究表明,地震前的10个月内,重构信号均出现能量严重衰减的现象,且其持续时间长短与发震震级之间存在一定相关性。Outgoing Long-wave Radiation （OLR） has the characteristic of large coverage range. Earthquake-related anomaly information of OLR is extracted and applied to earthquake prediction, which can make up the shortage that some areas of China are out of the monitoring range of the current precursory seismic network. In this paper, the wavelet packet method and its principle are introduced briefly, and the time sequence of OLR is decomposed and reconstructed with wavelet packet decomposition in order to extract earthquake-related anomaly information in 15 earthquake cases. The results show that the serious energy attenuations of reconstructed signals often appear during 10 months before the shocks, and there is a significant correlation between the duration of energy attenuation and earthquake magnitude.

关 键 词：长波辐射值(OLR) 小波包法 能量衰减 地震异常 

分 类 号：P315.7[天文地球—地震学]