Modeling nonlinear site effects in physics-based ground motion simulations of the 2010-2011 Canterbury earthquake sequence

This study examines the performance of nonlinear total stress one-dimensional (1D) wave propagation site response analysis for modeling site effects in physics-based ground motion simulations of the 2010-2011 Canterbury, New Zealand earthquake sequence. This approach explicitly models three-dimensional (3D) ground motion phenomena at the regional scale, and detailed site effects at the local scale. The approach is compared with a more commonly used empirical V-S(30)-based method of computing site amplification for simulated ground motions, as well as prediction via an empirical ground motion model. Site-specific ground response analysis is performed at 20 strong motion stations in Christchurch for 11 earthquakes with 4.7 <= M-W <= 7.1. When compared with the V-S(30)-based approach, the wave propagation analysis reduces both overall model bias and uncertainty in site-to-site residuals at the fundamental period, and significantly reduces systematic residuals for soft or "atypical" sites that exhibit strong site amplification. The comparable performance in ground motion prediction between the physics-based simulation method and empirical ground motion models suggests the former is a viable approach for generating site-specific ground motions for geotechnical and structural response history analyses.