BPMF: A Backprojection and Matched-Filtering Workflow for Automated Earthquake Detection and Location

• DOI: 10.1785/0220230230
• Publication Date: 2023-12-04T17:00:09Z
• Authors: Eric Beaucé; William B. Frank; Léonard Seydoux; Piero Poli; Nathan Groebner; Robert D. van der...; Michel Campillo
• Keywords: Python

Abstract We introduce BPMF (backprojection and matched filtering)—a complete fully automated workflow designed for earthquake detection location, distributed in a Python package. This enables the creation of comprehensive catalogs with low magnitudes completeness using no or little prior knowledge study region. uses seismic wavefield backprojection method to construct an initial catalog that is then densified filtering. integrates recent machine learning tools complement physics-based techniques, improve location earthquakes. In particular, offers flexible framework which detectors can be harmoniously combined, effectively transforming single-station into multistation detectors. The modularity grants users ability control contribution within workflow. computation-intensive tasks filtering) are executed C CUDA-C routines wrapped code. leveraging low-level, fast programming languages graphic processing unit acceleration efficiently handle large datasets. Here, we first summarize methodology describe application interface. illustrate BPMF’s capabilities characterize microseismicity 10 yr long Ridgecrest, California area. Finally, discuss workflow’s runtime scaling numerical resources its versatility across various tectonic environments different problems.