中国大陆地震空间分布模型检验

本文以中国大陆地震目录为基础资料，以泊松模型为零假设模型，并将Neyman-Scott空间丛集过程的各子模型设为检验模型，采用K-function 点过程分析法和最大似然估计法计算各模型参数，并以AIC准侧判定模型的拟合优度，来检验中国大陆地震空间分布模型.检验结果表明：泊松模型的拟合优度最差，说明地震在空间的分布不是完全随机的；广义Thomas模型的拟合优度最好，说明地震的空间分布是丛集的，可用由两个高斯核组成的广义Thomas模型较好地描述.研究结果还表明，同一研究区内，采用不同时段具有不同最小完整起始震级的地震目录计算得到的地震空间分布的丛集尺度几乎不变，这意味着地震空间丛集尺度不受小地震的控制，且可能与研究区的断层规模有关.

Statistical test of spatial distribution model of the Earthquakes in Chinese Mainland

In this paper, we test the spatial distribution model of the earthquakes in Chinese Mainland. We take Poisson model as null hypothesis and test Neyman-Scott and similar point processes. The K-function point processes analysis tool and maximum likelihood estimation method are used to calculate parameters of each model. We use the Akaike Information Criterion (AIC) to compare the goodness-of-fit of these models. The results show that the goodness-of fit of Poisson model is the worst compared with that of any other model, while the goodness-of-fit of General Thomas model is the best. It can be confirmed that the spatial distribution of earthquakes is indeed significantly clustered (General Thomas model) rather than completely random (Poisson model). The best fit is got by the General Thomas model which contains two Gaussian kernels. The scale of cluster of spatial distribution of earthquakes does not change for each seismic sub-catalog in the same research region. This means that the scale of cluster is not controlled by small earthquakes, and that may be related with the scale of faults in research regions.