Spatiotemporal Clustering of Microseismic Signals in Mining Areas

Abstract

Identifying the spatiotemporal clustering of microseismic signals in mining areas is crucial for understanding underground structures, geological activities, and the potential risks of geological disasters. However, existing cluster identification methods often ignore the spatiotemporal distribution and occurrence mechanisms of events. This study investigated the factors that trigger microseismic events in mines across various spatiotemporal levels, using a specific mine as a case study. By applying a spatiotemporal clustering approach, this study analyzed the microseismic activity generated by mining operations. Initially, the frequency of microseismic events was quantified over a given timeframe to uncover the temporal patterns and active phases of microseismic activity. The integrity of microseismic data across various magnitude ranges was then assessed to determine the minimum magnitude required for reliable analysis. Subsequently, the Gutenberg–Richter frequency–magnitude relationship was employed to examine the magnitude distribution of microseismic events. Finally, by evaluating the spatial distribution of mining activities and utilizing a combined K-means and Gaussian mixture model clustering method, the causes of microseismic events were categorized into blasting, rock drilling, noise, ore falling and transportation, and ground stress redistribution. These findings align with on-site production activities, validating potential microseismicity triggering factors during specific mining operations and offering theoretical support for identifying the spatiotemporal triggers of microseismic events in mines.

Keywords: Mining activities, Microseismic events, Spatial distribution, Cluster analysis, K-means clustering, Gaussian mixture model