【摘要】 基于随机poisson分布模型和Gutenberg Richter关系 ,构造出 2 0年尺度内 8组样本量从 1 0 0 0到 1 50 0 0的模拟地震目录数据 ,以讨论样本量条件、最大拟合窗长τmax 等计算条件对地震活动频次序列H值计算结果的影响 .结果表明 ,随着样本量的增加 ,H值整体精度及稳定性增大 .当τmax =3 0 0时 ,表现尤为明显 .实际震例计算结果也表明 ,通过合理选取计算条件 ,可以较好地提取强震前H值的异常变化过程更多还原

【Abstract】 Based on random Poisson distributional model and Gutenberg Richter law, we made 8 groups of simulated seismic catalogues with the sample size of 1 000 to 15 000 within 20 years so that we could discuss the effect on the H value of frequency sequence of seismicity caused by the calculational conditions of sample size and τ max  of maximal fitting window length.The result shows that the whole precision and stability of H value improve with the sample size increasing. When τ max  is equal to 300, it is specially obvious. Case calculations show that it is available to pick up well the process of anomaly change of H value before strong earthquake with selecting reasonable calculational conditions.更多还原