A simple method for simulating microseism H/V spectral ratio in 3D structure

Abstract

The understanding of detailed 3D basin structure is very important for seismic hazard analysis because ground shaking can be both amplified and prolonged due to basin effects and causing significant damage. Background seismic noise is ubiquitous and continuous and these characteristics make ambient seismic noise useful to study velocity structure because neither infrequent earthquakes nor expensive explosions are needed. The horizontal to vertical (H/V) spectral ratio method is a common technique to study the structure of the basin using background seismic noise. Many observations of the H/V ratio recorded at seismic stations over basins display systematic decreases in the frequency of the dominant H/V peak with increasing basin thickness. However, some observations cannot be fully explained by theoretical simulations of the H/V ratio based on 1D velocity profiles beneath the stations. In order to study the precise relationship between the H/V ratio and the basin structure, wavefield simulation of background noise for 3D velocity structure can be quite useful. However, this simulation is still far from being routine because our knowledge of the noise source is quite poor. In this paper, we propose a simple technique to directly simulate the H/V ratio instead of the wavefield of background noise. To evaluate the method, we performed numerical experiments for the Santa Clara Valley and the results show that the synthetics can predict the observations well. We also performed several sensitivity tests for the source and the velocity structure of the basin and found that Rayleigh type wave propagation and a minimum velocity contrast between the basin and background media are required to make the H/V ratio sensitive to the basin structure. Although we also found that the H/V ratio technique has a limitation in determining small scale basin structure at deeper depth, this technique can still be used to evaluate pre-existing velocity models and give some constraints on the development of new velocity models for the basins.