Multifault rupture rates and uncertainties in PSHA: an NSHM23 case study for the Santa Barbara-Ventura region

Kevin R. Milner, & Edward H. Field

Submitted September 7, 2025, SCEC Contribution #14651, 2025 SCEC Annual Meeting Poster #TBD

Inclusion of multifault ruptures can strongly affect hazard estimates near interconnected fault systems. Despite numerous observations in nature, few data are available to constrain their prevalence for specific fault-pairs or regions. The 2023 update to the National Seismic Hazard Model for the western U.S. (NSHM23-WUS) included segmentation constraints to explore a variety of fault connectivity models: a fully segmented “classic” model, three distance-dependent models, and a fully unsegmented model. Counterintuitively, standard hazard metrics (e.g., peak ground acceleration with 2% probability of exceedance in 50 years) are typically lower for unsegmented models, where more moment is consumed by large and rare multifault ruptures compared to segmented models.

In NSHM23-WUS, we used various Coulomb compatibility metrics to determine multifault rupture plausibility (whether two faults might ever rupture together), but relied solely on jump distance to determine propensity (how often they might rupture together). We explored Coulomb-based segmentation rules but found no improvement over distance-based metrics when compared to results from physics-based earthquake simulators. We also tried empirical metrics based on surface rupture mapping, but found them to be unsuitable for the often-buried and coarsely defined fault network in this sub-region.

We examine the implications of these NSHM23-WUS choices in the Santa Barbara-Ventura region, which is highly connected: of 43 fault sections considered, all have a plausible multifault connection within 2 km (and all but one are within 1 km). Because NSHM23-WUS segmentation rules are largely distance-dependent, only the classic branch substantially limited multifault ruptures here. More advanced physical or empirical segmentation constraints could substantially impact hazard estimates in such regions, and we encourage development of improved models that are robust to uncertain geometric details of complex fault networks.

Key Words
NSHM23, PSHA, ERF, multifault ruptures, segmentation, Santa Barbara, Ventura

Citation
Milner, K. R., & Field, E. H. (2025, 09). Multifault rupture rates and uncertainties in PSHA: an NSHM23 case study for the Santa Barbara-Ventura region. Poster Presentation at 2025 SCEC Annual Meeting.

Related Projects & Working Groups
Earthquake Forecasting and Predictability (EFP)