Seismic Event Identification: Context to Confidence

We present a probabilistic framework to automate event-type labeling of seismic event catalogs via contextual data fusion. Our framework fuses information derived from seismic waveform features with geospatial context such as the location of known earthquakes, nuclear test sites, and mining operatio...

Full description

Saved in:
Bibliographic Details
Main Authors: Richard Alfaro-Diaz, Jonas A. Kintner, Josh D. Carmichael
Format: Article
Language:English
Published: Seismological Society of America 2025-04-01
Series:The Seismic Record
Online Access:https://doi.org/10.1785/0320250005
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a probabilistic framework to automate event-type labeling of seismic event catalogs via contextual data fusion. Our framework fuses information derived from seismic waveform features with geospatial context such as the location of known earthquakes, nuclear test sites, and mining operations to infer event-type labels and associated uncertainty. This framework mirrors expert analyst reasoning by simultaneously considering both the seismic waveform characteristics and the geospatial context of each event to formally combine multiple lines of evidence. We integrate these disparate observations via Bayesian hierarchical modeling to structure data and exchange of information across data categories. Unlike standard approaches with deterministic seismic event-type labels, our framework provides a quantitative measure of uncertainty for informed decision-making. We demonstrate this technique on a comprehensive catalog of analyst-derived ground-truth earthquake and explosion event types located within the Great Basin in the western United States.
ISSN:2694-4006