← Back to Sessions

Precision and Accuracy of Earthquake Locators: Insights From a Synthetic 2019 Ridgecrest Sequence Experiment

Description: 

By solving the inverse problem, various earthquake location programs serve as important tools for the precise determination of earthquake hypocenters. However, the variety of algorithms and misfit norms raises questions about the efficiency and accuracy of different methods. With limited knowledge of the true velocity structure and sparse seismic network coverage, the efficacy of these tools for accurately pinpointing the true locations remains uncertain. Some programs incorporate uncertainty analysis; however, due to the inherent nonlinearity of the problem, the reliability of the resultant confidence measures of earthquake locations deserves some scrutiny. This study embarks on an evaluation of widely used earthquake location methods through a controlled synthetic experiment to compare the outputs of seven programs (HypoDD, GrowClust, HypoInverse, VELEST, XCORLOC, NonLinLoc, HypoSVI) using the 2019 Ridgecrest earthquake sequence. We calculate arrival times by the Fast-Marching Method using a 3D velocity model extracted from the SCEC Community Velocity Model with a von Karman perturbation superimposed to represent unmodeled small scale structure, including elevation effects. We mimic a realistic seismic network monitoring scenario by introducing picking errors, phase availabilities, and phase outliers into the synthetic traveltimes. Our comparative analysis of the recovered locations utilizing the uniform traveltime dataset and 1D velocity structure reveals the superior precision of relative location methods compared to absolute location methods (except VELEST). We demonstrate that near source S phases are important for constraining the focal depth; however, our study shows the limited ability of consistently recovering depths under a 1D velocity structure approximation. As established uncertainty techniques frequently fall short of encompassing true hypocenters, our findings should motivate revisting earthquake location uncertainty assessment.

Session: Network Seismology: Recent Developments, Challenges and Lessons Learned - V

Type: Oral

Date: 5/2/2024

Presentation Time: 02:00 PM (local time)

Presenting Author: Yifan

Student Presenter: Yes

Invited Presentation: 

Authors