Introduction

The understanding of 3D basin structure is very important in the estimation of damaging strong shaking. Since Nakamura (1989) first used ambient noise to estimate the site amplification, several techniques to delineate 3D basin structure using background noise have been developed. The advantage of the background noise approach is that ambient seismic noise is ubiquitous and continuous. It means that we don't need to wait for earthquakes or detonate expensive explosions for studying structure. Dolenc and Dreger (2005) showed that the frequency of the dominant H/V spectral peak (hereafter referred to as FDP) due to microseisms correlates with the thickness of the Santa Clara Valley (SCV). However, the synthetic FDPs for one dimensional models could not explain the observed shifts of the FDP. This result indicates that 2D and 3D wave propagation effects should be considered for the deep basins when their thickness is comparable to their size. In this study, we developed a simple method to compute FDP due to microseisms for 3D basin models and applied the new method to predict observed FDPs during Santa Clara Valley Seismic Experiment (SCVSE)(Lindh et al., 1999).