USE OF ARTIFICIAL INTELLIGENCE FOR THE CHARACTERIZATION OF A SEISMIC SEQUENCE IN THE SOUTHERN PART OF THE SAN CLEMENTE FAULT SYSTEM
Seismicity in northwestern Mexico is dominated by the interaction between the North American and Pacific tectonic plates, resulting in continuous seismic activity in the state of Baja California. On November 22, 2022, a magnitude Mw 6.1 earthquake occurred off the Pacific coast, near the city of San Quintín. This earthquake has been associated with the San Isidro Fault, in the southern part of the San Clemente Fault System. Historically, only one earthquake of magnitude greater than 6, which occurred in 1954 (USGS, 2023), has been recorded related to this part of the fault system, making the 2022 earthquake an unusual event for this region.
The 2022 earthquake presented foreshocks and triggered a significant number of aftershocks. This seismic sequence has not yet been fully characterized, partly due to the low coverage of seismic stations in that region. For cases like this, with a low station coverage, artificial intelligence has been proven to be very effective and in some cases even superior to traditional methods for phase identification (Mousavi, Ellsworth, Zhu, et al., 2020; Mousavi & Beroza, 2023; Münchmeyer et al., 2022; Woollam, Münchmeyer, Tilmann, et al., 2022; Wollam, van der Heiden, Rietbrock, et al., 2022).
In this work, we use novel detection techniques to build a catalog of the 2022 seismic sequence, that allow us to better understand the faults and seismicity in this area. We present preliminary work on earthquake detection using Earthquake Transformer (Mousavi, Ellsworth, Zhu, et al., 2020) via Seisbench (Woollam, Münchmeyer, Tilmann, et al., 2022) for phase identification, PyOcto (Münchmeyer, 2024) for association, and nonlinloc (Lomax et al., 2000; Lomax, Michelini & Curtis, 2014) for source relocation.