M≥6．0强震后地震序列类型的模糊聚类分析
|
摘要
在充分了解地震序列特征的基础上，选出２６项地震序列的原始特征。根据这些原始特征的Ｗａｌｌｅｎ评分，精选出１１项能突出表现地震序列的特征参数作为指标。采用ＩＳＯＤＡＴＡ模糊聚类方法对地震序列进行分类，Ⅰ类为Ｍ≥６．０强震后三个月内还有震级为Ｍ≥５．０的强震发生；Ⅱ类为Ｍ≥６．０强震后三个月内无Ｍ≥５．０的地震。对中国大陆地区１９６６年以来的２２次Ｍ≥６．０强震序列进行聚类分析，建立预报震后趋势判别模型，然后以８次强震序列为例进行了外推检验。结果表明，该方法对地震序列类型的判别效果较好。
Based on the full understanding of characteristics of earthquake sequences,26 items oforiginal features of earthquake sequences are selected from distinct views of physics back-ground.According to the Wallen scores of these original features,11 items of original fea-tures which can prominently represent the characteristics of earthquake sequences are care-fully chosen as indexes.Using ISODATA method of fuzzy clustering,earthquake sequencesare classified into two categories:Type-Ⅰ sequences are these in which earthquakes withM≥5.0 occur in the same area within 3 months after an earthquake with M≥6.0;Type-Ⅱ sequences are these in which no M≥5.0 earthquakes occur within 3 months.The dis-criminating model of post-earthquake tendency has been built by using 22 known earth-quake sequences occurred in the China continent since 1966.Then,extrapolation tests aretaken by using other 8 earthquake sequences occurred in China.The results indicate thatthis method has good effects on discriminating the type of earthquake sequences.
Based on the full understanding of characteristics of earthquake sequences,26 items oforiginal features of earthquake sequences are selected from distinct views of physics back-ground.According to the Wallen scores of these original features,11 items of original fea-tures which can prominently represent the characteristics of earthquake sequences are care-fully chosen as indexes.Using ISODATA method of fuzzy clustering,earthquake sequencesare classified into two categories:Type-Ⅰ sequences are these in which earthquakes withM≥5.0 occur in the same area within 3 months after an earthquake with M≥6.0;Type-Ⅱ sequences are these in which no M≥5.0 earthquakes occur within 3 months.The dis-criminating model of post-earthquake tendency has been built by using 22 known earth-quake sequences occurred in the China continent since 1966.Then,extrapolation tests aretaken by using other 8 earthquake sequences occurred in China.The results indicate thatthis method has good effects on discriminating the type of earthquake sequences.
引文
1韩渭宾，王虹，曾键，席敦礼．中强以上地震的震后趋势早期综合判断方法的研究．地震学报，1993，15（1）．2曾健，韩渭宾，王虹，席敦礼．震后趋势的模糊聚类分析判别．四川地震，1991，（3）．3曾健，韩谓宾，席敦礼．鲜水河地震带和松潘、龙门山地震带可能发生5级以上地震的模糊聚类分析．四川地震，1988，（3）．4冯德益，等．模糊数学方法与应用．北京：地震出版社，1985．5冯德益，等．模糊数学和地震预报．地震，1982，（2）．6冯德益，等．模糊聚类分析在以震报震研究中的应用．地震，1982，（4）．7陈荣华，冯德益．模糊聚类分析及其在地震研究中的应用．地震研究，1985，8（5）．8顾谨平，虞雪君，钱家栋．地震活动性的综合评判．地震学报，1989，11（2）．9虞雪君，等．地震形势的模糊综合评判．地震学报，1990，12（3）．
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心