前兆观测数据基本特征的震前差异及异常判别

本文通过实例阐述前兆手段观测数据基本特性(平稳性、周期性和正态性)在地震前的差异,说明连续观测数据中包含平稳随机过程和非平稳随机过程;亦包含常态下的年变周期和非年变周期,非平稳随机过程和非年周期变化可能产生在地震前兆发育期,其特征与地震孕育进程有关,可用非参数轮次法和方差分析法加以检验。随机过程的特征量正态分布的标准离差用以衡量过程的分散程度,异常时段与正常时段的标准差,相差可达数量级,用统计参量值X~2能在一定信度下检验其是否超出了限界。可见它们是前兆识别的基础依据,在其基础上可进一步选择科学计算方法剖析资料。同时在莱以特准则(3σ)判别异常的基础上,讨论了格罗布斯准则和狄克松准则在常熟地震(M_s5.1)和南黄海地震(M_s6.2)短期和中短期异常判别上的应用。

DIFFERENCE OF BASIC CHARACTERISTICS OF PRECURSORY OBSERVATION DATA BEFORE EARTHQUAKES AND DISCRIMINATION OF THEIR ANOMALIES

This paper explained the difference of the basic characteristics (stability, periodicity and normal distribution) of observed data of precursory data, before earthquakes through examples. It is shown that the successive observation data include the stable random process, also include annual and nonannual variation periods. The instable random process and nonannual period variation are probably produced in the seismically precursory development phase, its feature is related to the earthquake preparation process, and it can be examined using non-parametric ring method and various analysis method. The standard deviation of the normal distribution of the characteristic quantity of the random process is used to judge the dispersion degree of the process. The standard deviation between the abnormal time interval and the normal time interval may reach one order of magnitude. The statistic parametric value X2 can be used to examine whether the deviation exceeded the limit under a. certain confidence level. It is obvious that they are basis for distinguishing the precursors, based on this, the scientific computation method to analyze the data can be further chosen. Simultaneously, on the basis of Laiyite Crierion (3) judging, anomalies, the paper discussed the application of the Grobus and Dickenson criteria in judging the short-term and mid-short-term anomalies of Changshu earthquake with Ms = 5.1 and south Yellow Sea earthquake with Ms = 6.2.