← Back to Sessions

Regional Characterization of Natural and Anthropogenic Seismic Events for Monitoring Efforts With Machine Learning

Description: 

Seismic monitoring and event characterization provides useful information about seismic events to assist in earthquake emergency response and for international monitoring efforts of explosions and nuisance events. In this presentation, we expand on the efforts of Barama et al. (2023) and Kong (2022) for explosion discrimination at regional distances using data from both digital and historical (formerly-analog) seismograms, including the full waveforms, initial seismic phase arrivals, their derivative products like radiated earthquake energy, coda wave envelopes, and phase-ratios. Using both seismograms and physics-based features we trained regional seismic source classifiers, testing approaches of a Convolutional Neural Network (CNN), Random Forest (RF), and Deep Neural Net (DNN) for both per-event (network) and per-station predictions. We anticipate that machine-learning models we developed can be robust tools for both regional characterization and gaining insight on what drives the models’ prediction determinations.

Session: Advancements in Forensic Seismology and Explosion Monitoring - I

Type: Oral

Date: 5/2/2024

Presentation Time: 09:00 AM (local time)

Presenting Author: Louisa

Student Presenter: No

Invited Presentation: 

Authors