Development of a quantitative strong motion prediction technique based on simple site effect evaluation methods with sufficient accuracy

点击翻译按钮获取中文摘要

The purpose of the project is to propose a quantitative strong motion prediction technique based on simple site effect evaluation methods with sufficient accuracy in order to use these predicted strong motions as inputs to structures for aseismic design. The strong motion prediction technique based on a physical modeling of source, path, and site effects may have wide applicability so that it is suitable for a scenario type evaluation of strong ground motions. However, it requires many physical parameters, especially information on local site conditions, in order to produce quantitative synthetics. In this project we propose various methods that yield quantitative site effects.To this end we use strong motion data observed by K-NET in Kyushu and separate source, path, and site effects to generate statistical Green' s functions. Combining them with ordinary wave summation technique we show that we can successfully reproduce observed strong motions during the Hyogo-ken Nanbu earthquake, the Kagoshima-ken Hokuseibu earthquake, and the Tottori-ken Seibu earthquake. Then we use site amplification factors to invert S-wave velocity structures underneath by Genetic Algorithm. At the same time we obtain horizontal-to-Vertical spectral ratios (HVRs) for P-wave part, S-wave part, and Coda part and compare them with observed site amplification factors, theoretical S-wave amplification factors, and theoretical HVRs for Rayleigh wave. We found that observed HVRs show quite good correlation with observed site amplification factors. Based on these information we propose a highly accurate technique for strong ground motions that can be sued for aseismic design of structures.