← Back to Sessions

Nonergodic Ground Motion Model for the Central and Eastern United States

Description: 

In this study, using the varying-coefficient model (VCM) and based on Gaussian processes (GPs) regression, we develop a nonergodic ground motion model (GMM) for the Central and Eastern United States (CEUS) based on backbone models. In the VCM method, the coefficients (nonergodic adjustment terms) vary spatially (based on earthquake and station location) and are estimated by Gaussian process regression. We use the GMM developed by Pezeshk et al. (2018) as a backbone model for areas outside the Coastal Plain and the GMM developed by Akhani et al. (2024) as a backbone model for areas inside the Coastal Plain. A benefit of developing nonergodic GMMs based on VCM is that the model can be smoothly extrapolated to large magnitudes and short distances, making the developed nonergodic GMM applicable for any event-station combination. The developed model contains a spatially varying nonergodic source term, a spatially varying nonergodic site term, and a spatially independent nonergodic site term. Individual coefficients are developed for each earthquake and station location, containing the median and the associated epistemic uncertainty of the coefficient. The aleatory variability of nonergodic GMMs is significantly reduced compared to ergodic GMMs. Additionally, in nonergodic PSHA, the median ground motion calculated by ergodic GMMs is affected by nonergodic terms. These changes in median GMMs and aleatory variability significantly impact the results of probabilistic seismic hazard analysis (PSHA) in CEUS, and the calculated hazard might increase or decrease compared to ergodic PSHA.

Session: Accuracy and Variability of Physics-based Ground Motion Modeling [Poster]

Type: Poster

Date: 4/15/2025

Presentation Time: 08:00 AM (local time)

Presenting Author: Mehran

Student Presenter: No

Invited Presentation: 

Poster Number: 82

Authors