论文部分内容阅读
Stark(1996)所做的一些评论与 Varotsos 等(1996a)的评论基本上是一致的[即,“如果仅当预期的地震震级超过5.0级时我们决定发布一次预报,还有,如果我们的预报规则系统奏效的话,那么,我们预计不能预报出一些震级为5.0以及更小一点的事件(甚至一些较大的事件)”,普遍接受的观点是,“处于自然状态的”地震活动序列不是泊松分布的……”,等等]。不过,在本文中,我们澄清了一些导致 Stark 说”Varotsos 等提出了一些错误的建议”的误解。我们强调,绝大多数的VAN 预报中的容许误差限不是后验校准的,因为这些容许误差限是在1987～1989年这一讨论期间的前一年发表的。在25次结果良好的相关中仅有两次,Δt 值是后验延长的,不过我们强调,这两次预报事先确认为是属于一种新的情况,这种新情况称之为地震电信号的电活动(地震电信号序列),它不同于单一(孤立)的地震电信号情况。我们完全同意 Stark(1996)的建议,按照这一建议,人们“避免了为地震变量规定一种概率分布的必要性,这是一项既有争议又有问题的工作”。
Some of the comments made by Stark (1996) are basically in agreement with those of Varotsos et al. (1996a) [ie, “if we decide to issue a prediction only when the expected magnitude of the earthquake exceeds 5.0, and if our prediction Rules system works, then we do not expect some predictions of magnitude 5.0 and smaller (and even larger ones) to be predicted, ”and the generally accepted view is that the sequence of“ active ”earthquakes is not Poisson Distribution ... “, etc.] However, in this article, we clarify some of the misleading assumptions that led Stark to say” Varotsos et al made some erroneous suggestions. “We emphasize that the vast majority of tolerance limits in VAN forecasts Are not posterior calibrated because these allowable error limits were published in the year prior to the discussion between 1987 and 1989. The Δt value was a posteriori extension only for 25 of 25 good correlations, We emphasize that these two forecasts were previously identified as belonging to a new situation called the electrical activity of seismic electrical signals (seismic electrical signal sequences), which is different from the single (isolated) seismic telecommunications Situation. We fully agree with Stark (1996) suggested, in accordance with this proposal, it ”avoids the necessity to provide for a variable earthquake probability distribution, which is a problem not only controversial but also work."