Study on the characters of the aftershocks of Beiliu 5. 2 earthquake using machine learning method and dense nodal seismic array

In this study, a machine learning method is used to detect the aftershocks of the Guangxi Beiliu 5. 2 earthquake based on the 30 days continuous records of dense nodal seismic array. 441 reliable events are detected, which is around 34 times the number from permanent seismic network in the same period. P and S wave arrival time information is used to obtain an earthquake catalog with 299 events. In addition, the focal mechanism solutions of 65 aftershocks with high SNR waveforms are acquired by using the gCAP method. The seismogenic structure of strong earthquakes in this region is discussed based on the spatial distribution of aftershocks and focal mechanism solution characteristics. The results show that the aftershocks of the Beiliu earthquake are mainly concentrated in a range of about 1 similar to 3 km northwest of the main shock, and the focal mechanism solutions mostly are closed to the foreshock. The seismogenic fault of the mainshock is the Shiwo fault with a strike of NWW-SEE and a dip angle of nearly 70 degrees. In addition, we speculate that there is a NEE-SWW fault in this region, which may be the seismogenic fault of the foreshock. The mainshock is triggered by the foreshock, and then the two faults are both active and cause a series of aftershocks with different focal mechanism solutions. Some aftershocks are also detected at the northern end of the Jiaolin fault and the southern end of the Shiwo fault. The focal mechanism solutions of these events are mostly of thrust type and strike-slip type with poor consistency, which proves that many faults in these areas began to be active after the Beiliu earthquake. However, the specific trigger mechanism is still indistinct.