← Back to Gallery

Session: Overall Prediction Accuracy and Simulation Validation for Real-World Applications [Poster]

Type: Poster

Date: 10/12/2023

Time: 07:00 AM

Room: Stanley Park Ballroom

Adjusting Cascadia Ground Motion Models based on M9 Cascadia Earthquake Simulations

Seismic hazard in the Cascadia Subduction Zone (CSZ) is in part due to the potential for a future megathrust earthquake. However, no instrumental recordings are available to evaluate subduction zone earthquake ground motion models for such a scenario. Therefore, we evaluate the behavior of the “M9 Project” simulated ground motions, including the estimated basin effects for the Seattle basin (Frankel et al., 2018). The M9 Project amplification factors of 2-5 for reference basin depths of 3-5 km at periods of 1.5-10 s in the Seattle basin are higher than the empirically derived basin terms for Cascadia for two of the NGA-Subduction GMMs (KBCG20 and PSBAH20). This suggests a need for simulation-based basin-depth scaling that can be incorporated into empirical ground motion models to account for the high epistemic uncertainty in basin amplification in this region. We expand the methodology used to develop the M9 basin-depth scaling model presented in Sung and Abrahamson (2022) to the other two NGA-Subduction Cascadia GMMs and consider the effect of the functions chosen to represent the Vs₃₀-Z_2.5 correlation in the GMMs. We also evaluate the effect of incorporating site-specific Vs₃₀ into the residuals. We propose an M9 adjustment factor for Cascadia interface events based on the difference in the source term between the simulations and the empirical GMMs (C_sim). This centering adjustment is meant to apply to all sites in Cascadia, whereas the basin-depth scaling only applies for sites located on sedimentary basins. We compare the C_sim adjustment to alternative ways of adjusting the empirical GMMs to the simulation results, such as using the bias. These adjustments, in general, lead to higher ground motions when used in hazard applications and are one way to incorporate physics-based simulation results into empirical GMMs.

Presenting Author: James Smith

Additional Authors