Spatial correlation of earthquake ground motion and spatial variability of geological characteristics in seismic hazard and risk assessment

在分布式投资组合和空间分布系统(生命线)的地震损失估计中，平均期望损失及其附近的损失分布对于决策和减灾活动是非常重要的。平均损失取决于单个损失的总和，而损失的方差则取决于地震动场的空间相关性。标准危害评估和标准地震动预测不能处理相关性。只有在极少数地点才能从观测到的强震数据中推导出地面运动的相关特性，因为必须有非常高的数据密度。该项目的目标是通过引入一个模型，将地震运动的空间相关性与地质的空间变化联系起来，从而发展一种地震地面运动预测及其空间相关性的方法，用于地震危险性和地震损失评估。具体项目包括：1.利用日本和台湾积累的数据库和现代地震运动预测方程识别与地震动空间相关性相关的因素；2.利用传统的地质统计学技术和先进的Copula方法分析岩土工程地质特征的空间变异性；2.建立地震动空间相关性参数与地质空间变异性特征之间的定量关系；开发由区域地质统计数据预测相关特性的模型(S)；3.在缺乏地震动数据和相关结构信息的地区，论证相关地震动在灾害评估、损失估计、地面运动模拟和摇摆图生成中的相关性。本项目的创新之处有二：(1)目前尚无将地震动相关性与地质变异性联系起来的系统性工作。(2)在保险和金融建模中广泛使用的Copulas也可能被证明是地震工程的有力工具。

In seismic loss estimation for distributed portfolios and spatially distributed systems (lifelines), the average expected loss and the distribution of loss around the mean are very important for decision making and mitigation activities. Whereas the mean loss is determined by the sum of individual losses, the variance of loss depends on the spatial correlation of earthquake ground motion. Standard hazard assessment and standard ground motion prediction cannot deal with correlation. Derivation of correlation properties of ground motion from observed strong motion data is only possible in very few locations as very high data density is mandatory. The objective of the project is the development of a methodology for earthquake ground motion prediction and its spatial correlation for seismic hazard and seismic loss assessment by introducing a model, which relates spatial correlation of ground motion with spatial variation of geology. Specific item are: 1. Identification of factors relevant for spatial correlation of ground motion using the databases accumulated in Japan and Taiwan and modern ground-motion prediction equations; analysis of spatial variability of geotechnical and geological characteristics using traditional geostatistical technique and advanced copula methods; 2. Establishing of quantitative relationships between parameters of spatial correlation of ground motion and characteristics of spatial variability of geology; development of model(s) that predict correlation properties from geostatistical data of regions; 3. Demonstration of relevance of correlated ground motion for hazard assessment, loss estimation, ground motion modeling and Shakemap generation in areas with a lack of ground motion data and information about correlation structure. The innovation by this project is twofold: (1) We do not know of any systematic work to link correlation of ground motion and variability of geology. (2) Copulas used widely in insurance and financial modeling may prove a powerful tool for earthquake engineering as well.