A common approach in seismology is to use the nearest neighbours method to identify the spatial and temporal relationships between events and select the closest event pairs to identify the triggering cascade. However, in mining engineering, due to the continuous triggering from dynamic mining development as well as the complex geological conditions (e.g., faults), the mining seismic events may not be purely triggered by one prior event, but by a triggering group of a few different events. In this study, we modify the event-event triggering identification approach applied in longwall seismic events. We find that high-energy events can be triggered by previous low-energy events, with clear foreshadowing events. High-energy events can be related to a few low-energy events whose mechanism is based on mining activities.
Owing to the substantial spatiotemporal features between seismic events, the analysis of their spatial and temporal sequence plays a significant role in seismology (Si et al. 2020; Wang et al. 2021). The spatiotemporal characterizations reveal the essence of seismic events either by the ‘cascade’ or ‘stick slip’ theory, benefiting the understanding of earthquake formation (Davidsen et al. 2021). In longwall mining, the occasionally happened rock bursts and coal bursts accidents, analogously to earthquakes, also indicate a strong spatial cluster and temporal sequence (Holub et al. 2011). Hence, the spatiotemporal analysis of seismicity induced by longwall mining events is essential to workplace safety and mining productivity.
Multiple approaches to analyzing seismic inter-event triggering achieved considerable outcomes in the understanding of seismic temporal cascade (Gu et al. 2013). One of the most important methods is the Omori-Utsu law, whereby the event triggering rate after the source events can be empirically predicted. According to the Omori-Utsu law, the triggering seismic events followed by the source events show an exponential decay, where the triggering rate of seismic events is the highest right after the occurrence of source events and decreases gradually with time (Davidsen et al. 2017). The Omori-Utsu law provides a stable foundation of seismic temporal triggering sequence, whereas the spatial sequence of seismic events also attracts considerable attention. The location of seismic events indicates the breakage of rock mass or the instability (e.g., slippage) of pre-existing discontinuities, which has been proved by extensive research both in laboratory tests or site observations (Wang et al. 2021; Li et al. 2022, 2023). In the laboratory UCS and triaxial tests, the position of the inclined failure plane coming from an intact rock sample is aligned with the cluster of acoustic emission events. In addition, site monitoring systems also indicate that pre-existing faults are more likely to induce seismic events. A quantified analysis of seismic events, namely the fractal dimension, exhibits the fractal features of seismic event clusters. In three-dimensional space, the randomly distributed events illustrate a fractal dimension of three, while a planar event distribution indicates a fractal dimension of two. The change of fractal dimension indicates the variation in the spatial correlation of seismic events.
