Generalized Nearest-Neighbor Distance and Hawkes Point Process Modeling Applied to Mining-Induced Seismicity

Abstract Modeling seismic activity rates and clustering plays an important role in studies of induced seismicity associated with mining and other resource extraction operations. This is critical for understanding the physical and statistical characteristics of seismicity and assessing the associated hazard. In this work, we introduce the generalization of the nearest-neighbor distance (NND) method by incorporating an arbitrary distribution function for the frequency–magnitude statistics of seismic events. Operating within a rescaled hyperspace that includes spatial, temporal, and magnitude domains, the NND method provides an effective framework for examining seismic clustering. By integrating a mixture of the two tapered Pareto distributions, the generalized NND approach accommodates deviations from standard frequency–magnitude scaling when studying the clustering properties of seismicity. In addition, the application of the temporal Hawkes process to model the mining seismicity rate reveals that the seismicity is primarily driven by external factors and lacks pronounced interevent triggering. A case study from a potash mine in Saskatchewan is presented to illustrate the application of the generalized NND method and the Hawkes process to estimate the clustering properties and occurrence rates of induced microseismicity. The implications of observed temporal variations and clustering behavior are discussed, providing insights into the nature of induced seismicity within mining environments.