基于概率统计的地震事件公报评估算法  

Seismic event bulletin evaluation based on probability analysis

作　　者：王海军[1] 王燕[1] 王伟[1] 

机构地区：[1]禁核试北京国家数据中心和放射性核素实验室,中国北京100085

出　　处：《地震学报》2009年第3期342-346,共5页Acta Seismologica Sinica

摘　　要：给出一种基于概率统计的地震事件公报评估算法.选择事件的发震时间、震中经纬度构造特征函数.当特征函数为正态分布时计算事件匹配概率,选择匹配概率大于90%的事件为匹配事件.将该算法应用于国际数据中心(IDC)自动处理与人工分析公报的比较分析.相对于人工分析结果,IDC数据自动处理的匹配率约为79%,误检率约为52.3%.该算法可应用于地震事件公报的日常对比分析、台网监测能力分析、数据自动处理结果评估等领域.A seismic event bulletin evaluation algorithm based on probability analysis was proposed. An eigenfunction is constructed with origin time, latitude and longitude of the event. Suppose the eigenfunction obeys a normal distribution, then the event matching probability could be calculated. The events with matching probability greater than 90% were selected as the matching events. Initial application in comparison with IDC SEL3 and IDC REB shows that for automatic processing the event detection ratio is nearly 79% and false detection ratio is about 52.3%. The agorithm developed in this study can be used for routinely comparing event bulletins, and can also be applied to network monitoring capability analysis and automatic data processing evaluation.

关 键 词：地震事件 事件公报 评估 概率 

分 类 号：P315.63[天文地球—地震学]