S042-05
Using Dynamic Rupture Simulations to Assess Megathrust Earthquake Behaviors in the Cascadia Subduction Zone

Friday, 11 December 2020: 16:18
Virtual
Marlon Dale Ramos1, Yihe Huang1, Thomas Ulrich2, Duo Li, Alice-Agnes Gabriel2, Amanda Thomas4 and Marine Denolle5,6, (1)University of Michigan Ann Arbor, Ann Arbor, MI, United States, (2)Ludwig Maximilians University of Munich, Munich, Germany, (3)University of Oregon, Department of Earth Sciences, Eugene, OR, United States, (4)Harvard University, Earth and Planetary Sciences, Cambridge, MA, United States, (5)Harvard University, Department of Earth and Planetary Sciences, Cambridge, MA, United States
Recorded Presentation
Abstract:
The Cascadia subduction zone dominates the earthquake hazard in the United States Pacific Northwest. It is oft-cited that the probability of a magnitude ~9 (M9) event occurring in the coming decades is 10 - 14% (Peterson et al., 2002). Outstanding questions for its next megathrust earthquake include where it will initiate, what conditions are dynamically favorable for an M9 event to develop, and how much slip may occur offshore. We present a suite of dynamic rupture simulations to address these questions by combining recurrence intervals (T) estimated by paleoseismic data with geodetic coupling models to constrain dynamic stress-drop on the megathrust. Uncertainty in offshore megathrust coupling is explored with different geodetic slip-rate deficit models and variable T. Our simulations solve for elastodynamic rupture propagation using the open-source software package SeisSol.

We document the sensitivity of nucleation location, stress-drop amplitude, and fault geometry to the final seismic moment and coseismic coastal subsidence amplitudes. Megathrust segmentation is suggested from along-strike differences in T, which directly maps to a heterogeneous stress-drop distribution. We find that the final earthquake size depends on where it is nucleated; if rupture is nucleated in the northern or southern Cascadia regions, it is difficult to propagate through the central region unless there is a relatively higher dynamic stress-drop there. Furthermore, we note that T > 320 years produce subsidence amplitudes that can exceed the 1700 A.D. measurements by at least a factor of 2. This suggests that if geodetic coupling models are sufficiently close to the true state of slip-deficit, then the most recent event in Cascadia was unlikely to have exhibited complete stress-drop along the entirety of the megathrust. Our physically consistent source models show that dynamic stress-drop amplitude in the central region determines if margin-wide rupture will occur, whereas the frictional behavior in the transition zone along-dip controls the down-dip rupture limit. These results can be used to inform earthquake and tsunami hazard assessment for not only Cascadia, but other young and thermally warm subduction zones such as southern Chile or southwest Japan.

<< Previous Abstract | Next Abstract >>