Aftershock Identification in Deep Underground Mines Using DBSCAN Clustering

Mining-induced seismicity has become a significant concern in deep underground mining, primarily due to the increasing depths of extraction and the complex stress conditions.Significant seismic events are often followed by aftershock sequences that pose significant risks to worker safety and mine infrastructure, highlighting the importance of reliable aftershock identification and exclusion zone determination for re-entry protocols.The goal is to enhance the accuracy of aftershock identification resulting from underground mining activities and minimize noise in microseismic catalogs, thereby supporting mine safety.This study applies the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to a synthetic microseismic catalog of 1471 events with moment magnitudes ranging from -1.6 Mw to 2.5 Mw.After magnitude of completeness (Mc) filtering, DBSCAN was used to spatially cluster aftershocks, effectively distinguishing them from background seismicity and noise.Results show that large sequences, such as the moment magnitude event 2.5 Mw with 1,471 aftershocks, were reduced to 1,015 clustered events.The clustered events show a more precise spatial concentration than the original distribution, which makes it simpler to tell which aftershocks are related to the mainshock and which are not.Before clustering, the catalog had 1,471 aftershocks for each mainshock with a maximum magnitude of 2.5 Mw.After clustering, 1,015 aftershocks were correctly identified as part of the sequence cluster.