【摘要】 本文对1976年7月28日唐山地震(M=7.8)和其它一些破坏性地震震前若干年内频度—震级关系N=exp(A—bM)中的b值作统计分析,说明在正确处理观测资料和运用统计方法的前提下,可以为合理的时空区间求得客观的、可信的b值.在对频度—震级关系唯象讨论的基础上,按照茂木清夫和肖尔茨的观点论证b值不仅是一个统计分析参数,而且具有直接的物理意义:它反映一个地区承受平均应力和接近强度极限的程度.因此,对b值变化进行动态时空扫描,就越出了单纯统计预报的范畴,成为观察应力集中和转移监视破坏性地震孕育过程的一种手段. 本文提出对频度—震级的指数分布直接作非线性最小二乘拟合.这是一种有效的估值方法,它能给出比较确切的误差估计.更多还原

【Abstract】 Statistical analysis of the b-value in the frequency-magnitude relation N = exp(A - bM) was performed for several years before the Tangshan Earthquake on July 28, 1976 (Ms = 7.8) and before some other catastrophic earthquakes. It was shown that objective and statistically significant b-values could be always obtained provided seismic data being treated correctly and statistical methods being used properly. On the basis of a phenomenological discussion of the frequency-magnitude relation in the view of Mogi[1]and Scholz[2], it was concluded that fr-value is not only a parameter in statistical analysis, but also has its own physical meaning. It reflects the average stress being applied to the region considered and the extent of approach to its strength limits. Therefore, dynamical scanning of b-values in time and space goes beyond the scope of simple statistical prediction and becomes a method to monitor the concentration and migration process of stress and to monitor the development of catastrophic earthquakes.A direct procedure for non-linear least-squere fitting to the exponential frequency-magnitude distribution was suggested in this work. This is an effective method for estimation which does give more exact limits for statistical errors.更多还原