← Back to Sessions

Incorporating Stress Drop into Non-ergodic Ground Motion Models

Description: 

The ground motion of an earthquake is determined by three components: the source, the path, and the site. While path and site effects have been extensively studied and modeled, source effects are often simplified in ground motion models (GMMs). Stress drop, a critical source parameter that reflects the energy release during fault rupture, significantly influences ground motion but is oversimplified or overlooked in GMMs. In this study, we utilize stress drop datasets derived from small earthquakes to provide consistent non-ergodic source effects in ground motion models. Using a Bayesian Gaussian Process framework, we extract robust spatial variations in stress drop from highly variable data. We analyze more than 5,000 small-magnitude earthquakes in the San Francisco Bay area and compare processed stress drop values with non-ergodic source terms derived from peak ground acceleration (PGA). Our results reveal consistent patterns between stress drop and PGA, emphasizing the key role of stress drop in capturing source effects in ground motion. Additionally, we show that incorporating processed non-ergodic stress drop reduces uncertainties in ground motion predictions, as shown using the DesignSafe dataset (Ji et al., 2022). This study highlights the potential of using stress drop datasets from small earthquakes to enhance seismic hazard assessments and improve non-ergodic ground motion models for seismic hazard predictions.

Session: Challenges and Opportunities in Constraining Ground-motion Models from Physics-based Ground-motion Simulations [Poster]

Type: Poster

Date: 4/17/2025

Presentation Time: 08:00 AM (local time)

Presenting Author: Shiying

Student Presenter: No

Invited Presentation: 

Poster Number: 45

Authors