Nonlinear site amplification functions for Central and Eastern North America

Site amplification functions are used to modify ground motions from a reference bedrock condition to a surface condition based on the geologic features of the site of interest. Site amplification has been extensively studied and evaluated empirically for seismic regions such as the Western United States where there are abundant ground motion recordings for many sites and earthquake events. In regions of relatively lower or infrequent seismicity, such as Central and Eastern North America (CENA), the lack of ground motion recordings and seismic site properties severely limits the empirical characterization of site amplification. This research uses site response simulations to develop site amplification functions for CENA. The first part of this study is the development of 1,747,278 1-D linear elastic, equivalent linear, and nonlinear site response analyses. Simulations are designed to capture the variability in site conditions in CENA and the uncertainty in soil properties at individual sites. Site profiles are developed for 1,747,278 site response analyses, 582,426 each of linear elastic, equivalent linear and nonlinear 1-D analyses for 70,650 unique site profiles. The database of simulations is the largest of its kind. This study describes the process for generating VS profiles, soil and weathered rock material properties, and ground motions to represent the variability and uncertainty of site conditions in CENA. The second part of this study is the modeling of the site response simulation data with linear and nonlinear site amplification functions for the response spectrum (RS) and Fourier amplitude spectrum (FAS) and a correction factor to convert site amplification from a 3000 m/s CENA hard rock condition to a 760 m/s condition. The modular RS amplification model includes terms for time averaged shear wave velocity in the top 30 m of a site (VS30), site natural period, soil depth, and site nonlinearity which can be coupled with empirically-developed linear empirical amplification models. Including site natural period into the amplification function is shown to greatly improve estimates of site response over models dependent only on VS30. The FAS site amplification model is the first model developed from simulations, and the first to include nonlinear amplification