强震前短水准异常的模糊识别及其特性分析

FUZZY RECOGNITION AND CHARACTERISTIC ANALYSIS OF SHORT-RANGE LEVELLING ANOMALIES BEFORE STRONG EARTHQUAKES

本文利用模糊模式识别的直接方法处理了1975年海城7.3级地震,1976年唐山7.8级地震和1989年大同6.1级地震前震中及其周围地区的短水准观测资料,取得了较好的效果。该方法具有排除干扰和提取地震异常的功能,且物理意义清楚,计算简便,可以作为观测数据处理的一种常用方法。用模糊识别对唐山和大同地震前短水准异常分析后,发现震中位于高异常从属度区附近,且前兆异常的出现与地震活动异常有良好的相关性,震前还有明显的异常频次峰值。

The direct method of fuzzy pattern recognition is applied to treatment and analysis of the shortrange levelling data in epicentral and surrounding areas before 1975 Haicheng M=7.3,1976 Tangshan M=7.8 and 1989 Datong M=6.1 earthquakes. From the obtained results, the noises may be removed better, the precursory anomalies may be recognized,and moreover,the physical nature of observed data may be explained more clearly, the computational method is simple too. Therefore The suggested method could be used as a regular method to treat the short range levelling data.Secondly, through the analysis of the precursory anomalies of short—range levelling before Tangshan and Datong earthquakes with the fuzzy recognition method,it is discovered that the epicenters of these events are located on the vicinities of the regions where the values of membership function of anomalies are higher, and that the appearances of precursory anomalies are in good agreement with the anomalies in seismicity, and that finally, the peak frequency of anomalies is indicated remarkably before each strong earthguake.