The Impact of GNSS Derived Finite-Fault Rupture Models on Ground Motion Predictions Based on Cascadia Megathrust Rupture Scenarios

Session: Advances in Real-Time GNSS Data Analysis and Network Operations for Hazards Monitoring [Poster]
Type: Poster
Date: 4/30/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Geodetic data from regional Global Navigation Satellite System (GNSS) stations offer a beneficiary role in the near real-time monitoring of moderate to large destructive earthquakes and also overcome the limitations of regional near-field seismic observations. In this study we concentrate on showcasing our methodologies for using GNSS data to accurately compare and quantify results from earthquake ground motion prediction equations (GMPE) to observed recordings using information from finite-fault (FF) rupture models. The Cascadia subduction zone has the potential to produce destructive megathrust earthquakes in the future but regional GNSS stations have not yet recorded a significant earthquake in the region. Lacking in geodetic observations, we use a set of 1300 simulated earthquakes (termed Fakequakes) between M7.5-M9.5 at over 400 stations distributed throughout the Cascadia margin. We compute geodetic finite-fault rupture models from the Fakequakes dataset consisting of synthetic high-rate displacement waveforms using the Geodetic First Approximation of Size and Timing (G-FAST) algorithm (Crowell et al., 2016). We test the original G-FAST implementation, using the CMT solution fault orientations for our slip inversions, as well as a new set of predefined rupture planes that approximate the subduction zone geometry. As a result, a suite of varying fault discretization models is used to invert for coseismic slip and determine FF models. For ground motion predictions, we use the Next Generation Attenuation Model (NGA) presented in Chiou and Youngs (2008) for estimating horizontal ground motion amplitudes from the FF solutions. Based on Peak Ground Acceleration (PGA), predicted Modified Mercalli Intensity (MMI) values on average agree with observed MMI values within few minutes of the mainshock rupture. We further summarize the uncertainties and impact of GNSS based ground motion predictions in a real-time monitoring context.

Presenting Author: Kevin B. Kwong

Authors