Evaluating rock instability risk areas in deep mining using microseismic monitoring-based risk fields

Microseismic (MS) monitoring is an efficacious technology for the detection and early warning of rock mass instability in deep mining operations. However, existing MS-based early warning methodologies, which primarily rely on single-parameter analysis or local clustering techniques, typically lack quantitative risk assessment capabilities. Inspired by the concept of a rock instability risk field, this study proposes a spatially continuous methodology for risk evaluation. Specifically, the Empirical Green's Function (EGF) method is employed to estimate failure probabilities, while a Cloud Model is introduced to quantify potential severity. By integrating these components, a three-dimensional risk field function is formulated, using spatial coordinates as independent variables. Regional risk is then assessed through surface integral formulations, enabling a detailed characterization of its spatial distribution. The proposed methodology was applied to a lead-zinc mine in Northwest China. The calculated regional risk values demonstrate strong spatial consistency with energy density clusters identified using the DBSCAN algorithm. Quantitative comparisons reveal a significant positive correlation between the risk field outputs and independent MS event clusters, confirming the method's efficacy in capturing the spatiotemporal concentration of instability potential. These findings indicate that representing rock instability risk as a continuous field enhances the regional interpretability of MS data and offers a quantitative foundation for dynamic hazard zonation in deep underground excavations.