Mining-induced earthquakes have been very frequent in recent years due to increasing mechanized mining. Compared with natural earthquakes, even a small one may cause significant damage to the mine area and its surroundings. Source type identification is important for better understanding the physical processes and is a crucial and fundamental issue for hazard assessment and emergency rescue in the mining environment. The moment tensor (MT) theory plays a pivotal role in distinguishing different source types. In this study, we concentrated on two strong reported “mine collapse earthquakes” in Qufu (ML 3.2, July 13, 2020) and Zoucheng (ML2.9, June 09, 2020), Shandong Province, China. Seismograms from regional seismic stations were utilized to calculate the full moment tensors through low-frequency full-waveform inversion. Our results show that the two studied events exhibit notably different source types. The DC (Double-Couple) component of both events are 5% (Qufu) and 60% (Zoucheng), respectively. The Qufu event which contains approximately 75% closing crack component, is more consistent with the theoretical models of collapse seismic source. However, the Zoucheng event, which exhibits a significant proportion of DC components, demonstrates characteristics typical of shear failure. Focusing on the Zoucheng event, which occurred at the Dongtan Coal Mine, further research was conducted on a local mining scale. Analyzed in conjunction with microseismic sensor data, geologic setting, and mining progress, we illustrated that the source type of Zoucheng event is not a collapse one. The fracture slip of a thick-hard roof due to an overlying load, characterized by a large DC component, is a plausible geomechanical interpretation.
The identification of unnatural earthquake events is one of the tasks of earthquake rapid report. The identification accuracy is of great significance for improving the quality of earthquake catalogue and seismological research. In this work, a seven-layer convolution neural network model was constructed to identify unnatural earthquakes. First, the three-component seismic waveform was used as input to obtain the waveform image classifier, and then the time-frequency spectrum of explosion and collapse was used as input to obtain the time-frequency spectrum classifier. The two classifiers were used to identify earthquake, explosion and collapse. The model was trained and tested using 3386 seismic events of Shandong seismic network from 2017 to 2022. The events identified as explosion events by the waveform image classifier were reidentified by the time-frequency spectrum classifier. Finally, the identification accuracy of natural earthquake, explosion and collapse is 97.50%, 95.87% and 86.84% respectively, with an average identification accuracy of 96.13%. The experimental results show that the two-step convolution neural network can extract the characteristics of seismic signals from different angles, and get a good result in seismic event classification.
1 November 2021 in Izvestiya, Physics of the Solid Earth
The ML 3.8 earthquake occurred in Rushan County On October 1, 2013, Shandong Province, eastern China. After this event, there are already more than 19 600 seismic events recorded by digital seismic seismographs, forming an earthquake sequence. Although three moderate and strong earthquakes occurred successively during this period, the surface rupture undiscovered, it is difficult to diagnose the seismogenic fault from the geological perspective. However, the 3-D velocity structure inversion can provide important information on the location of the moderate earthquakes of the source area, which can help to understand the formation of earthquakes. Compared with source locations, the focal mechanisms can visually reflect the geometric and kinematic characteristics of seismogenic fault rupture, being the basis for inferring the regional tectonic stress field. Moreover, changing the stress field will affect the geometric parameters of the crack and change the shear-wave splitting parameters. Therefore, it is of great significance to explore the processes of earthquakes and the seismogenic mechanisms utilizing the spatiotemporal variation of shear wave splitting parameters. In summary, we comprehensively analyzed the seismogenic fault of the Rushan earthquake sequence (RES) with the help of the above methods. The resluts showed that the seismogenic fault of RES is towards the NWW-SEE direction and dips towards the southwest, dominated by high-angle strike-slip movements with certain thrust dip-slip components.