Characterizing geometry and mechanics of structures hosting moderate-to-large earthquakes is essential for seismic hazard assessment, yet remains challenging in extensional environments, where fault systems include multiple segments and bends. In this study, we demonstrate how a short-term array deployment can provide critical insights into seismicity patterns and fault geometry in Southern Apennines, Italy.We integrated data recorded by arrays during a one-year experiment with machine learning methodologies, producing a seismic catalog that enhances the manual catalog for the same period by nearly an order of magnitude, lowering completeness magnitude by one unit. Approximately 65% of the detected events can be accurately relocated, with median uncertainties of ~ 100 m, comparable to those of long-term catalogs. Our results reveal consistent seismicity properties down to decametric earthquake size, with hypocenters and b-value mirroring those from the previous decade. We distinguish a shallow, diffuse seismicity, likely influenced by hydrological loading, from deeper clusters, mostly rupturing patches a few-hundred meters across. Beyond asperity-scale complexity, seismicity follows the boundaries of tomographic anomalies, delineating a 50–60 km-long curving fault, featuring a right-stepping jog several kilometers wide. Dynamic simulations suggest that ruptures nucleating on this fault could propagate through these complexities, potentially generating earthquakes up to magnitude 7.0.
Uccio et al. (2026) studied this question.