摘要
多年来的观测表明,青岛地电的主要干扰因素是地下水位和温度。它们对地电的影响并不是简单的线性关系,而是带有滞后记忆效应的非线性关系。为了较好地消除地下水位和温度的影响,我们采用了非线性的褶积滤波方法,同时采用多元回归法。对地电资料进行综合处理。处理后的地电3个分向的资料都在震前11～12天开始出现明显的短临异常,3个分向的资料的一阶差分曲线也在震前11～12天出现明显的短临异常。
Over years observations showed that the interference factors for Qingdao geoelectricity are mainly from groundwater level and temperature variation. Their affection on geoelectricity observation are not simply in linearity but in nonlinearity with lag memory effects . To e-liminate the mentioned above affections,the nonlinear convolution filtering method is used with multiple regression to deal with the geoelectric data. After that,three subdirections ofgeoelectricity showed obvious anomalies 11-12 days immediately before the earthquakes.Meanwhile, the first order difference on the three subdirections had anomaly at the same time.
出处
《内陆地震》
1993年第4期344-349,共6页
Inland Earthquake
关键词
地震前兆
地电
短临异常
地震烈度
Geoelectricity Ms 5. 3 earthquake in Yellow Sea Convolution filtering Multiple regression
