一种评估地震前兆可信度的方法

A method of evaluating reliability of earthquake precursors

本文提出了一种可以对地震前兆的可信度进行评估的方法.通过对前兆信息的抽取、各类非地震干扰的识别、前兆与地震事件在时间和空间的相关性分析以及前兆的随机检验等一系列手段来评估前兆信息的真伪,提高地震前兆信息的显著性和可信度.该方法及思路也可为其他非地震前兆的可信度评估所借鉴.作为一个具体的例子,本文利用日本气象厅的地震目录以及能定量描述地震活动的时、空、强信息演化特征的RTL(Region_Time_Length)方法对2 0 0 0年10月6日在日本鸟取县西部发生的7 3级地震前周围地区近2 6年的地震活动进行了系统的研究,得到了可能的地震活动前兆信息,并采用本文提出的方法对该前兆进行了评估,证实了上述前兆的显著性和可信性.

One method is presented propose to evaluate reliability of earthquake precursors. The research flow that can improve the significance and reliability of earthquake precursors includes: extracting precursory information from observational data, distinguishing various artificial effects, analyzing the correlation in time and space domains between the detected precursors and the investigated earthquakes, and making stochastic test. The approach of the above evaluation method can also be applied to other non-earthquake precursors. As a case study, I investigate the seismicity changes before the M = 7 3 earthquake in the western region of Tottori prefecture, Japan on October 6, 2000, using the RTL (Region-Time-Length) algorithm. The earthquake data (1975～2000) of the Japan Meteorological Agency (JMA) are chosen for this study. A clear seismic quiescence anomaly is detected before the 2000 western Tottori prefecture earthquake. The close investigation using the above evaluation method indicates that the above quiescence anomaly before the mainshock is significant and reliable.