Group B, Poster #190, Ground Motions

Empirical Assessment of Site and Basin Response in the Pacific Northwest via Testing Site-Response Models for Different Tectonic Regimes

Sean K. Ahdi, Brad T. Aagaard, Morgan P. Moschetti, Oliver S. Boyd, Grace A. Parker, & William J. Stephenson
Poster Image: 

Poster Presentation

2023 SCEC Annual Meeting, Poster #190, SCEC Contribution #13238 VIEW PDF
We test the hypothesis that site-response models are independent of earthquake source parameters by analyzing ground-motion records in the Portland, Oregon, region. We assess whether empirical site-response models developed for subduction-zone ground-motion models (GMMs), regionalized for the Pacific Northwest (PNW), perform as well as the site-response model developed for crustal earthquakes in California. This study builds upon our previous regional empirical analyses of basin response for the California Great Valley, the Reno-Sparks basins in Nevada, and the Portland-Tualatin basins in Oregon. Those analyses showed that the Boore et al. (2014) and Campbell and Bozorgnia (2014) NGA-West2 G...MMs improved predictions of basin response in the California Great Valley but did not perform as well in the shallow Reno-Sparks basins or the more geologically complex Portland-Tualatin basins.

We first analyze basin response by compiling and processing regional ground-motion data and performing residuals analyses and mixed-effects regressions. We build upon the study in Portland by (1) improving VS30 characterization of stations that recorded ground motions by using measured VS30 values where available or alternative proxy-based VS30 prediction models other than topographic slope and (2) computing GMM estimates with both an NGA-Subduction empirical site-response model for the PNW and the site-response model built into the NGA-West2 GMMs for shallow crustal events. If the site-response models are shown to be independent of earthquake source type, we can reduce uncertainty in ground-motion predictions by replacing site-response models in ergodic GMMs with region-specific site-response models developed using a subset of potential earthquake sources. This would result in improved ground-motion estimation in regions within national-scale seismic hazard studies that generally employ global ergodic site-response models at all locations within a tectonic regime.