论文部分内容阅读
本文仅限于回答再次由 Kagan 和 Jackson(1996)提出的那些观点,尽管对此我们已经在其他对 Kagan(1996)和 Jackson(1996)的答文中作过讨论。例如我们证明他们关于“……大多数 VAN 参数在1987～1989年检验期后作了调整,亦即,反推调整”的观点并不正确。而另一方面我们却同意他们的结论:“这一问题的观点不同,说明地震学界需要一套评估地震预报方法的总方案”,应用这套方案应接着对到目前为止发布的 VAN 预报进行评估(包括从1987～1995的9年时间),而不是仅限于这场争论中讨论的1987～1989年间。Kasan 和 Jackson(1996)还将 Aceves 等(1996)和 Kagan(1996)的检验结果进行了对比并得出这样的结论,即如果目录中包含余震或只是去掉部分余震,则两项检验都表明 VAN 方法优于泊松零假设。另一方面,Kagan 和 Jackson(1996)称,“如果余震被尽可能全部删去”,VAN 方法并不比泊松零假设更为优越。然而我们将阐明,在后一种情况的“减丛集”过程中会出现以下情况:对主震的成功预报同样会被去掉;而且会将成功的预报变成失败的;我们的结论与 Aceves 等(1996)的基本一致。
This article is limited to answering those views again raised by Kagan and Jackson (1996), although we have already discussed them in other replies to Kagan (1996) and Jackson (1996). For example, we prove that they are not correct in the point that “... most of the VAN parameters have been adjusted since the test period of 1987 to 1989, that is, backtash adjustment.” On the other hand, however, we agree with their conclusion: “This issue has a different point of view, indicating that the seismological community needs a set of general scenarios for assessing earthquake prediction methods”. The application of the scheme should then assess the VAN forecasts so far released (Including nine years from 1987 to 1995), not just between 1987 and 1989 as discussed in this debate. Kasan and Jackson (1996) also compare the test results of Aceves et al. (1996) with Kagan (1996) and conclude that if the table contains aftershocks or just a few aftershocks, both tests show that VAN The method is better than Poisson’s null hypothesis. On the other hand, Kagan and Jackson (1996) stated that “VAN methods are no better than the Poisson’s null hypothesis if aftershocks are eliminated as completely as possible.” However, we will clarify that in the latter case, the following cases occur in the process of “diminishing the cluster”: the successful prediction of the main shock will be removed as well; and the successful forecast will become a failure. Our conclusion is similar to that of Aceves et al. (1996) basically the same.