A Machine-Learning-Enhanced Catalog of California Statewide Seismicity over Two Decades

Bo Rong, Weiqiang Zhu, Ian W. McBrearty, & Gregory C. Beroza

Submitted August 30, 2026, SCEC Contribution #15268, 2026 SCEC Annual Meeting Poster #TBD

Microseismicity provides important insights into fault geometries at seismogenic depths, and its spatiotemporal characteristics can provide constraints on the mechanics of earthquake nucleation and rupture. The detection of small earthquakes, though challenging due to their weak seismic signals, has been greatly advanced by the development of machine-learning (ML) methods, which have been successfully applied in a wide range of tectonic settings. In California, the complex network of active faults and dense seismic networks highlight the necessity and opportunity for an enhanced earthquake catalog with temporal continuity. Here, we present an ML-based earthquake catalog of California from 2000 to 2025, utilizing continuous seismic data from the Northern and Southern California Earthquake Data Centers. Our earthquake detection workflow, QuakeFlow, includes phase picking, phase association, absolute location, and cross-correlation-based relative relocation. We increase the number of detected earthquakes by a factor of about five relative to the routine catalog, and lower the magnitude of completeness. Through an augmented view of microseismicity across California, we aim to constrain the geometry of active faults in detail. The improved spatial and temporal resolution of seismicity will advance our understanding of fault mechanics, earthquake interactions, and foreshock-aftershock behavior.

Citation
Rong, B., Zhu, W., McBrearty, I. W., & Beroza, G. C. (2026, 08). A Machine-Learning-Enhanced Catalog of California Statewide Seismicity over Two Decades. Poster Presentation at 2026 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology