Variability in Performance of a Machine-Learning Seismicity Catalog: Central Italy, 2016-2017

TL;DR

This study evaluates the performance of machine learning-based seismic catalogs in Central Italy, showing increased detection sensitivity and spatial variability compared to routine catalogs, with implications for seismic monitoring accuracy.

Contribution

It introduces a station-level detection probability method and assesses the spatial variability of ML catalog performance in seismic detection.

Findings

ML catalogs detect smaller earthquakes and at greater distances.

Magnitude-of-completeness decreases significantly with ML detection.

Greater variability in station performance observed in ML catalogs.

Abstract

Machine learning (ML) catalogs contain many more earthquakes than routine catalogs, but their performance in phase picking and earthquake detection has not been fully evaluated. We develop station-level detection probabilities using logistic regression and combine them across a seismic network to compute spatial magnitude-of-completeness fields. We apply this approach to two catalogs from the 2016-2017 Central Italy sequence that were constructed from the same seismic network, one routine and one ML based. At the station level, the ML picker increases detection sensitivity by identifying smaller magnitude events and detecting earthquakes at greater distances. Spatially, the magnitude-of-completeness decreases substantially, with median values shifting from 1.6 to 0.5 for P waves and from 1.7 to 0.5 for S waves. However, the ML catalog also shows greater variability in station-level performance than the routine catalog. These results demonstrate that ML-based improvements in detectability are widespread but spatially non-uniform, highlighting their benefits, their limitations, and the potential for further improvements.