Coseismic slip on the southern Cascadia megathrust implied by tsunami deposits in an Oregon lake and earthquake‐triggered marine turbidites

Abstract

We test hypothetical tsunami scenarios against a 4,600‐year record of sandy deposits in a southern Oregon coastal lake that offer minimum inundation limits for prehistoric Cascadia tsunamis. Tsunami simulations constrain coseismic slip estimates for the southern Cascadia megathrust and contrast with slip deficits implied by earthquake recurrence intervals from turbidite paleoseismology. We model the tsunamigenic seafloor deformation using a three‐dimensional elastic dislocation model and test three Cascadia earthquake rupture scenarios: slip partitioned to a splay fault; slip distributed symmetrically on the megathrust; and slip skewed seaward. Numerical tsunami simulations use the hydrodynamic finite element model, SELFE, that solves nonlinear shallow‐water wave equations on unstructured grids. Our simulations of the 1700 Cascadia tsunami require >12–13 m of peak slip on the southern Cascadia megathrust offshore southern Oregon. The simulations account for tidal and shoreline variability and must crest the ∼6‐m‐high lake outlet to satisfy geological evidence of inundation. Accumulating this slip deficit requires ≥360–400 years at the plate convergence rate, exceeding the 330‐year span of two earthquake cycles preceding 1700. Predecessors of the 1700 earthquake likely involved >8–9 m of coseismic slip accrued over >260 years. Simple slip budgets constrained by tsunami simulations allow an average of 5.2 m of slip per event for 11 additional earthquakes inferred from the southern Cascadia turbidite record. By comparison, slip deficits inferred from time intervals separating earthquake‐triggered turbidites are poor predictors of coseismic slip because they meet geological constraints for only 4 out of 12 (∼33%) Cascadia tsunamis.