Simplified geostatistical analysis of earthquake-induced ground response at the Wildlife Site, California, U.S.A

Abstract

Almost all natural soils are highly variable and rarely homogeneous. In this study, the seismic response of the Wildlife Site, Imperial Valley, California, U.S.A., has been analysed to assess the effect of ground heterogeneity on liquefaction assessment in a probabilistic analysis framework. Cone penetration test (CPT) data recorded at the site have been used to identify different lithologies and to estimate elements of soil inherent spatial variability. Monte Carlo simulation has been utilized to obtain several realizations of CPT data that were then implemented into empirical approaches to examine the liquefaction susceptibility of the site. In addition, stochastic analysis of liquefaction-induced surface damage has been carried out through the application of these realizations into damage criteria, such as total liquefaction damage potential and surface settlement. These stochastic analyses have indicated that using mean values in deterministic analysis can be on the unsafe side. As a result, attempts have been made to obtain meaningful representative soil parameters that can be used in simplified deterministic analysis, while continuing to honor detailed ground heterogeneity. In addition, an empirical technique has been developed to compare ground variability of potentially liquefiable sites on a qualitative basis.Key words: liquefaction, spatial variability, stochastic analysis, cone penetration test, damage, characteristic parameters.