Factors influencing soil surface seismic hazard curves

Abstract

Performance-based seismic design of important structures requires design ground motions from probabilistic seismic hazard analysis (PSHA) that incorporate the effects of local site conditions. Seismic hazard curves incorporating site-specific soil conditions can be generated through the convolution of rock hazard curves with statistical models for site-specific ground motion amplification factors (AF). The AF relationships are developed from a series of site response analyses. The goal of this study is to evaluate how the AF relationships and the resulting surface hazard curves are influenced by different approaches in the site response analysis, specifically the time series (TS) vs. random vibration theory (RVT) approaches, and by different levels of shear wave velocity variability introduced in the site response analysis. The results show that the median AF relationships derived from TS and RVT analyses are similar, except at periods near the site period, where RVT analysis may predict larger AF. Including the effect of shear wave velocity variability reduces the median AF and increases the standard deviation associated with the AF relationship (σlnAF). Generally, the soil hazard curve derived by the AF relationship with the largest σlnAF generates the largest ground motions, and this effect is most significant at small annual frequencies of exceedance. The effect of σlnAF on soil hazard curves is larger than the effect of different median AF relationships. The value of σlnAF is influenced significantly by the variability in the shear wave velocity and therefore proper specification of this variability is critical when developing soil hazard curves.