论文部分内容阅读
根据对1987～1989年间VAN或SES(地震电信号)的预报进行了统计检验。检验是用Varotsos等(1996)提出的至今还在采用的规则进行的。对希腊的SI-NOA地震目录,虽然VAN方法在形式上是成功的,但高成功率应归因于返回去调整预报规则,或者是地震活动性的非随机性。对震群(前震-主震-余震序列)的一个简单算法,就能得到类似的、甚至更好的预报结果。如果我们去掉目录中的相关事件,那么“预报效果”就变成统计上不显著的了。对PDE(NOAA)目录,本检验表明VAN预报的成功率是偶然造成的。
The statistical tests were carried out based on the forecast of VAN or SES (Seismic Electrical Signal) from 1987 to 1989. The test is carried out using the rules that Varotsos et al. (1996) proposed so far. For the SI-NOA earthquake catalog in Greece, although the VAN method is formally successful, the high success rate should be attributed to the return to adjusting the forecasting rules or the non-randomness of seismicity. A simple algorithm for the swarm (foreshock-mainshock-aftershock sequence) yields similar or better predictions. If we remove the relevant events in the catalog, the “forecasting effect” becomes statistically insignificant. For the PDE (NOAA) catalog, this test shows that the success rate of VAN forecasts is by chance.