← Back to Gallery

An H/V Geostatistical Approach to Account for Spatial Variability in 1D Seismic Site Response

The subsurface spatial variability of shear-wave velocity (V_S) has a significant impact on seismic site response. Common procedures used to account for spatial variability in 1D site response analyses, such as stochastic randomization of Vs, have been shown to yield site response predictions that are different than observations at well-characterized borehole array sites. These differences have provided motivation to develop a more rational, site-specific approach to incorporate spatial variability in 1D seismic site response analyses. We show, using observations from recorded low amplitude ground motions at Treasure Island (TI) and Delaney Park (DP) downhole arrays, that seismic site response is likely influenced by site-specific subsurface conditions at significant distances (as much as 1 km) away from the arrays. To rationally account for this spatial variability in Vs and fundamental frequency (f₀), we propose a geostatistical approach based on a single borehole Vs profile and spatially distributed horizontal-to-vertical spectral ratios (H/V) of noise measurements. Using irregularly-spaced f₀ estimates from H/V measurements (f_0,H/V) and Gaussian geostatistical regression, a map with uniformly-predicted f_0,H/V across an area of interest can be obtained. Then, using a V_S profile measured in close proximity to the site and the predicted f_0,H/V map, a spatially variable V_S model may be obtained on a uniformly-sampled grid. This is a cost effective, time efficient and feasible approach for developing a pseudo 3D subsurface model that captures the key aspects of spatial variability that influence site response. By implementing this approach at TI and DP downhole arrays, we are capable of more accurately predicting the measured site response at the borehole array sites. Hence, the proposed geostatistical approach, which relies mainly on f_0,H/V measurements and a single borehole Vs profile, provides a promising framework to account for site-specific spatial variability in Vs using 1D site response without the need to perform advanced 3D site response analyses.

Presenting Author: Mohamad M. Hallal

Additional Authors