论文部分内容阅读
事实表明,一次M6级或更大的地震发生后,另一次相似或更大的事件发生于附近地区比发生于较远地方的可能性更大(指每单位面积内的发生率)。我们特别要指出的是,第一次事件的余震活动提供了评估邻近地区(几度的范围内)下一次大事件发生概率的信息。也就是,如果第一次事件的余震活动与正常衰减相比变得相对平静,在主震后的10年内,附近地区大地震的发生率就会比正常余震活动情况下高几倍。为了精确地测量这种现象,我们要模拟实际的余震活动。实际上,许多余震序列远比修正的大森公式所描述的简单负幂衰减模型复杂。传染型余震序列(ETAS)模型是修正的大森公式的一般形式,它对各种类型的余震序列都拟合得很好,包括非火山型的震群。用这种模型研究了日本及其邻近地区过去3/4世纪所发生的76次主震的余震序列。重点放在客观地检验一次余震序列中是否存在一个地震活动偏离传染型余震序列模型预测的明显变点,在该点后是相对的平静,即地震活动显著降低。任何水平的地震活动都可以发生相对的平静。如果存在这种平静,它的长度就可以根据估计的传染型余震序列模型预测的发生率的累积数与发生转换时间的图上看出来。这就表明,相对的平静是判别余震活动异常的一项有用因子。这可以用来预测在第一次事件之后的短期内(大约6年内)邻近地区(3°范围内)是否更可能发生大地震。
Facts have shown that another similar or larger event after a M6 or greater earthquake is more likely to occur in the vicinity than in other places (ie, incidence per unit area). In particular, we point out that the aftershock activity of the first event provides information for assessing the probability of the next major event in the vicinity (within a few degrees). That is, if the aftershock activity of the first event becomes relatively calm compared to normal attenuation, the occurrence of major earthquakes in the vicinity will be several times higher than normal aftershock activity within 10 years after the main shock. In order to accurately measure this phenomenon, we want to simulate the actual aftershock activity. In fact, many aftershock sequences are far more complex than the simple negative exponential decay model described by the modified Omori formula. The Infectious Aftershock Sequence (ETAS) model is a generalized version of the modified Omori formula that fits well for all types of aftershock sequences, including non-volcanic swarm. Using this model, the aftershock sequence of 76 major earthquakes that occurred in Japan and its neighboring regions in the past three to four centuries has been studied. An emphasis is placed on objectively verifying whether there is an obvious change point in a aftershock sequence that is predicted by a deviation of the seismic activity from the model of an aftershock-based aftershock sequence, after which point it is relatively calm, ie the seismic activity is significantly reduced. Any level of seismic activity can occur relatively calm. If this calm exists, its length can be seen from the cumulative number of occurrences predicted by the estimated infectious aftershock sequence model and the time at which the transition occurred. This shows that the relative calm is a useful factor to judge the abnormal aftershock activity. This can be used to predict whether a large earthquake is more likely to occur in the immediate vicinity (within the range of 3 °) in the short term (about 6 years) after the first event.