资源环境科技发展态势分析平台

Despite being 770 km away from the epicenter, observed ground motions due to the Tohoku earthquake in the Osaka Basin were unexpectedly large, with an amplification of more than a factor of 20 compared to immediately outside the basin, and including 2.7 m peak-to-peak roof displacements at one high-rise building. The local ground motions exceeded expectations based on standard computations of site response by a factor of 3, predicted frequencies of peak acceleration were off by at least 50%, and such discrepancies have not yet been explained quantitatively. Here we show that utilizing semianalytic theory for surface-wave amplification, we are able to accurately predict both the amplitudes and frequencies of large ground amplification in the Osaka Basin using only knowledge of the local one-dimensional structure. Comparison between this simple prediction and observed amplification was not expected to be so favorable and suggests that simple one-dimensional surface-wave site amplification factors can be useful in the absence of full three-dimensional wave propagation simulations. Such surface-wave amplification factors can be included in addition to the standard measures of site-specific site amplification and should help explain strong ground motion variability in future large earthquakes that shake Osaka Basin and elsewhere in the world.

Plain Language Summary Using the local structure to predict surface-wave amplification accurately explains extreme ground motion amplification in Osaka Basin during the 2011 Tohoku earthquake, whereas such ground motions have been unsuccessfully modeled with standard measures of site amplification. The results demonstrate how necessary it is to include a measure of surface-wave amplification if one is to accurately predict earthquake ground motions. The approach can potentially be incorporated into new site-specific earthquake hazards analyses in nearly an identical way to which existing approaches to site amplification are already being used.