高阶统计量及AIC方法在区域地震事件和直达P波初动识别中的应用

Detection of regional seismic events by high order statistics method and automatic identification of direct P-wave first motion by AIC method

利用高阶统计量(偏斜度和峰度)与赤池信息量准则(简称AIC)相结合,进行区域地震事件实时检测和P波初至精细识别的新方法研究,通过处理山东地震台网记录的地震波资料,结果表明:应用高阶统计量(偏斜度和峰度,尤其是峰度)能够有效识别地震事件,降低地震事件的错误报警率和漏报率;与人工识别震相到时结果相比,根据Ske-AIC、Kur-AIC震相自动识别方法得到的震相到时的平均绝对值误差小。

Basing on high order statistics and AIC method,we put forward new methods for real-time detection of regional earthquake event and automatic identification of direct P-wave first motion,and apply it to process seismic data recorded by Shandong Seismic Network.The results show as follows：①The high order statistics method （skewness and kurtosis,kurtosis especially） effectively detect earthquake events,and may effectively reduce false alarm and missing report rates;②Compared with phase arrival time results in manual identification,average absolute error of phase arrival time in automatic identification based on Kur-AIC and SkeAIC method are （0.09±0.08）s and （0.06±0.14）s,respectively.