Introducing Geodetic Locking to Stochastic Slip Rupture Models: An Example Application to Tsunami Hazard Analysis in Cascadia

Session: Understanding Non-Traditional Seismic Tsunami Hazards
Type: Oral
Date: 4/28/2020
Time: 05:15 PM
Room: 120 + 130
Description: The Cascadia subduction zone is quiescent; the last great earthquake ruptured in 1700 as a Mw 9.0. Previously determined, the Cascadia subduction zone has ruptured in great earthquakes (>Mw8.5) cyclically every 250-500 years. To best understand the hazard potential of a great Cascadia megathrust earthquake that will rupture in the future, an accurate assessment of tsunami hazards is crucial. Within tsunami modeling, ruptures have primarily been modeled using a homogeneous slip finite fault model; however, this inherently introduces biases towards smaller tsunami potentials and models not representative of true megathrust ruptures. More recently, the use of stochastic slip models has been explored to produce more realistic sources. Here we will show how including geodetic locking models as a proxy for slip deficit budget can add considerable a priori information into the stochastic rupture characteristics. Using the MudPy open access code, 2800 rupture models are created using two different decade-scale locking models for Cascadia for events between Mw 7.8 to 9.2. This method for determining ruptures implements a scaling law of rupture geometry for magnitudes based on the Von Karman correlation. Tsunami models are then determined for each rupture using the open source Geoclaw code. We will show that tsunami hazard varies significantly between both models, with larger tsunami amplitudes noticed in the coupling model that has shallow strong locking. For implications in hazards assessment, along shore potential statistical means and probabilities are determined. The stark variations in tsunami hazards created by introducing the different coupling models shows the necessity of further research and implementation of seafloor geodesy to elucidate the ambiguities between potential locking models.

Presenting Author: David Small

Authors