← Back to Sessions

Improvement in Magnitude Estimation Performance with a Combined PGD-PGV Scaling Law for the G-Fast Earthquake Early Warning Module

Description: 

Early estimation of earthquake magnitude is essential for earthquake and tsunami early warning systems and rapid response to these events. We propose a new method to estimate earthquake magnitude using a scaling law relating peak ground velocities (PGVs) and peak ground displacements (PGDs) derived from Global Navigation Satellite Systems (GNSS) data from a global dataset (Crowell et.al, 2023). Peak ground velocities indicate the level of damaging shaking while peak ground displacements are diagnostic of the maximum moment release, so obtaining observations at the highest fidelity and understanding their relationship is crucial for rapid earthquake magnitude estimation through the use of scaling laws. We collected high-rate (1Hz) GNSS observations for earthquakes ranging in magnitude from Mw 4.9 to 9.1. The three-component GNSS velocities were computed with the SNIVEL software package, which uses the time-differenced carrier-phase method, and the three-component GNSS displacements were computed with the precise point positioning code GipsyX from JPL. The peak ground velocities and displacements were computed by taking the maximum of the square root of the sum of each component squared. With this dataset, we explore the evolution of PGV and formally combine it with the PGD scaling law proposed by Crowell et.al, (2013), which forms the core of the G-FAST module in ShakeAlert. We test the new law against the existing performance in G-FAST using the PGD-only approach, with regards to both timing of alerts and accuracy of magnitude estimates, and investigate the downstream impacts on ground motion and tsunami prediction in the context of operational early warning. Additionally, we investigate case studies of large magnitude events with ground motions over large areas to assess the performance of the PGV correction term. We finish by developing a framework for how this additional term would be implemented in the G-FAST earthquake early warning module operationally.

Session: End-to-End Advancements in Earthquake Early Warning Systems - I

Type: Oral

Date: 5/3/2024

Presentation Time: 02:00 PM (local time)

Presenting Author: Jensen

Student Presenter: Yes

Invited Presentation: 

Authors