论文部分内容阅读
为了从一给定地震目录中建立空间连续的地震活动性模型(地震概率分布),提出了核估计法。对于具有不同空间分布特征的地震震源分布,具有全局(空间恒定)带宽的核估计处理不好。例如,一个典型的地震目录具有几个高活动性(成丛)区及几个低地震活动背景区。另一可采用的方法是自适应核估计,该方法使用一种在空间上变化的带宽参量。把它与空间恒定带宽的核估计法进行比较时,情况表明,(离散的)地震分布要求进行不同程度的局部光滑处理,以得到有用的空间地震活动性模型。通过使用自适应核估计法,可对任何地震概率分布的(局部)时间离散指数进行评估,并将其用于模拟主震的空间概率分布。这些方法在新西兰和澳大利亚地震目录上的应用表明,主震发生处的空间特征(震群)在整个观测期间一直呈现一种合理的稳定状态。纵观整个观测期间,活动较高的地区一直持续高活动性,无一处降至背景活动状态。这表明,在接下来的几年或几十年中,这些地区将会继续呈现出发生中到大地震事件的更高危险性。此外还观测到,浅源地震多半都是时间序列的组成部分(例如余震或群震),而在观测期间,在新西兰消减带之内的地震却只表现出小的时间变化。
In order to establish a spatially continuous model of seismicity (seismic probability distribution) from a given seismic catalog, a nuclear estimation method is proposed. For seismic focal distribution with different spatial distributions, the kernel estimation process with global (spatially constant) bandwidth is not good. For example, a typical seismic catalog has several high activity (clumps) zones and several low seismic activity background zones. Another available method is adaptive kernel estimation, which uses a spatially varying bandwidth parameter. Comparing this with the nuclear estimation of space constant bandwidth, the situation shows that the (discrete) seismic distribution requires varying degrees of local smoothing to obtain useful models of spatial seismicity. By using the adaptive kernel estimation method, the (local) temporal dispersion index of any seismic probability distribution can be evaluated and used to simulate the spatial probability distribution of the main shock. The application of these methods to the earthquake catalogs in New Zealand and Australia shows that the spatial characteristics of the mainshock (swarm) have been shown to be reasonably stable throughout the observation period. Throughout the observation period, high activity areas continued high activity, no one fell to the background activity. This shows that in the coming years or decades, these areas will continue to present a higher danger of a medium to large earthquake. In addition, it has also been observed that shallow earthquakes are mostly part of the time series (eg aftershocks or group earthquakes), whereas earthquakes within the New Zealand abatement zone show only small temporal variations during the observation period.