基于震情会商实时异常分类记录的综合异常指数研究

为进行异常特征或异常判据的定量分析，在多种预测手段(方法)的异常隶属度转换的基础上生成单项异常指数时间序列，再通过将多种预测手段(方法)单项异常指数时间序列逐时段累加取均值，得到综合异常指数时间序列，实现前兆信息的定量化。研究中根据新疆台站分布、历史地震及地震构造背景等，选定7个研究区进行应用研究。各研究区综合异常指数时间进程曲线分析表明，台站相对密集的北天山3个研究区综合异常指数在中等地震前出现较明显的短期异常，而台站相对稀疏的南天山地区几个研究区除柯坪块体的阿克苏-乌什-阿合奇地区的综合异常指数在巴楚-伽师6．8级地震前出现明显的短期异常外，其它地区综合异常指数时间进程曲线稳定性及异常对应效果较差。

Study on Synthetic Anomaly Index Based on the Real-time Anomaly Classified Recording of Earthquake Consultation

In order to perform the quantitative analysis to the anomaly features and anomalous criteria, on the basis of transformation of anomaly grade of membership of multi observations (methods), the time sequence of single anomaly index has been created. The time sequence of the synthetic anomaly index is obtained through the averaging value got from segment summation of the time sequence of single anomaly index of multi observations (methods). So the precursor information is quantified. In the study, 7 studying areas are selected based on the station distribution, historical earthquakes and seismic structural settings. The analysis of time history curve for the synthetic anomaly index in every studying area shows that obvious short-term anomalies appear in the 3 studying areas of north Tianshan before mid-strong earthquake, where the stations are relatively dense. The time history curve for the synthetic anomaly index has less stability and less effect in the studying areas of south Tianshan, where the stations are relatively rare, except for the obvious anomaly before Bachu-Jiashi M6.8 earthquake appeared in Akesu-Wushi-Aheqi region in the Keping block.