Validating Ground Motion Predictions from Peak Ground Displacement Scaling in Cascadia Simulations

Session: Earthquake Early Warning: Current Status and Latest Innovations [Poster]
Type: Poster
Date: 4/29/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Peak ground displacement (PGD) scaling using high-rate Global Navigation Satellite System (GNSS) data is capable of robustly determining magnitude for earthquakes M > 6.5 without going off scale. This information is currently being added to both earthquake and tsunami early warning systems around the world. However, while this method is capable of determining magnitude to an accuracy of ~0.3 magnitude units, little work has been done to show how this impacts the ground shaking forecasts for an area. In this study, we use high-rate displacement waveforms from 1300 simulated Cascadia megathrust earthquakes between M7.5-M9.5 at over 400 stations throughout the Cascadia margin. We run these displacement waveforms through G-FAST’s PGD module to determine magnitude and look at the evolution of these magnitude results with respect to the true magnitudes. We take the point source magnitude estimates and run them through ground motion models commonly used in the Pacific Northwest and compare these predictions with the simulated recordings. We find 80+% true positive and negative rates, in as little as 20 seconds after the earthquake origin times, and we analyze the evolution of these results over time. Finally, we investigate the impact of these ground motion intensity predictions in major population centers to show how useful they are to operational earthquake early warning.

Presenting Author: Brendan W. Crowell

Authors