Refining Time-Dependent Probabilities of Full-Margin Rupture Along the Cascadia Subduction Zone Using Recurrence Intervals Derived from Sediment Deposition Rate and Thickness

Chris Milliner, Michael E. Oskin, Jessica Velasquez, & Jochen Woessner

Submitted August 30, 2026, SCEC Contribution #15394, 2026 SCEC Annual Meeting Poster #TBD

Paleoseismic and historic records indicate full-margin Cascadia Subduction Zone (CSZ) megathrust ruptures (Mw 9+) depart from a Poisson, memoryless process, motivating a renewal-based, time-dependent probability model. In this study, we investigate the sensitivity of time-dependent modeling choices on Mw 9+ full margin rupture probabilities. In addition, we present a new approach to estimate the relative inter-event times (IET) of full margin Mw 9+ events using the known sediment core thickness and estimated sedimentation rate that are less sensitive to absolute age uncertainties. Here we quantify how epistemic and aleatory uncertainties propagate into 50-year earthquake probabilities by (1) comparing alternative interpretations of the paleoearthquake chronology (Goldfinger et al., 2025 update; Nelson et al., 2021; Walton et al., 2021), (2) testing renewal distributions (Brownian passage time, Weibull, lognormal, Gaussian) and mixtures using a tempered maximum-a-posteriori mixed-model approach, and (3) imposing a priori physical constraints through (i) removal of implausibly short inter-event times for M9+ events (<100 yr) and (ii) geodetic locking-based bounds on mean recurrence.

For Mw 9+ full-margin events, we find the distributional choices have minimal influence on hazard rate because the ~300-year open interval since 1700 CE places the system far beyond short-recurrences where distribution tails differ most (and just beyond the mean ~200 yr). Excluding short repeating IETs increases 50-year conditional occurrence probability by only ~1-2%. Alternative locking models likewise produce minor changes. In contrast, hazard rate is most sensitive to the varying interpretations of the earthquake-timing history: 50-year conditional rupture probabilities differ by up to ~6-7% across the three published chronologies. Lastly, we will present preliminary findings of a new approach to estimate the relative IET of full-margin ruptures which could help further improve time-dependent hazard rate estimates. This involves using the ratio of turbidite-bounded unit thickness, taken directly from the core to the corresponding OxCal-derived sedimentation rate. This is advantageous in that it avoids differencing the wide absolute age-PDFs assigned to event horizons, and we anticipate should yield relative IET’s with substantially narrower uncertainty.

Key Words
Time-dependent hazard, paleoseismology, Cascadia

Citation
Milliner, C., Oskin, M. E., Velasquez, J., & Woessner, J. (2026, 08). Refining Time-Dependent Probabilities of Full-Margin Rupture Along the Cascadia Subduction Zone Using Recurrence Intervals Derived from Sediment Deposition Rate and Thickness. Poster Presentation at 2026 SCEC Annual Meeting.

Related Projects & Working Groups
Earthquake Geology