Reducing Apparent Variability and Uncertainty and Improving Ergodic and Non-ergodic Ground Motion Forecasts Using Physics-based Data

Yuehua Zeng, Mark D. Petersen, Norman A. Abrahamson, Chih-Hsuan Sung, Oliver S. Boyd, & Arthur J. Rodgers

Submitted September 7, 2025, SCEC Contribution #14547, 2025 SCEC Annual Meeting Poster #TBD

We present a study to reduce large variability and uncertainty in the U.S. National Seismic Hazard Model (NSHM) by integrating physics-based data into the characterization of ground-motion models (GMMs). The approach involves: (1) partitioning the GMM variability into source, path, and site components using random effects techniques; (2) correlating these components with mapped geophysical parameters, enabling region-specific ergodic GMMs with reduced aleatory variability and uncertainty for improved ergodic hazard maps, and (3) exploring further variability reduction by accounting for unexplained but systematic spatial correlations in partially or fully non-ergodic hazard models. Traditional ergodic GMMs used in NSHMs rely on a limited set of predictors, such as tectonic region, moment magnitude, depth to top of rupture, distance metrics, and shallow shear-wave velocities. We incorporate several additional mapped geophysical variables across the U.S., including crustal temperature, shear-wave velocity, stress, density, strain rate, seismicity rate, Bouguer gravity anomalies, and intrinsic attenuation (1/Q), to reduce aleatory variability in ground motion residuals. We present examples in California and Nevada that incorporate these physics-based inputs to reduce variability in ground motion by up to 30%, with an additional 40% reduction achieved through accounting for spatially correlated non-ergodic effects. This work provides a pathway to improve both ergodic and non-ergodic hazard models, thereby enhancing the accuracy of seismic hazard assessments.

Key Words
Ground Motion, Variability, Ergodic, Non-ergodic

Citation
Zeng, Y., Petersen, M. D., Abrahamson, N. A., Sung, C., Boyd, O. S., & Rodgers, A. J. (2025, 09). Reducing Apparent Variability and Uncertainty and Improving Ergodic and Non-ergodic Ground Motion Forecasts Using Physics-based Data. Poster Presentation at 2025 SCEC Annual Meeting.

Related Projects & Working Groups
Ground Motions (GM)