10:45 AM - 12:15 PM
[SSS03-P06] Rapid Earthquake Detection and Location for Dense Array Data in the Guye area of Tangshan Based on Deep Learning
Keywords:Dense array, Deep learning, Earthquake location
The detection and localization of seismic events is a hot spot and difficult point in seismological research. In fault zone development areas, it is of great significance to effectively identify earthquakes and obtain accurate source information to understand regional seismicity and characterize fault information. With the development of seismic monitoring technology and the large-scale deployment of dense arrays, seismic data is growing exponentially, and it is very difficult to rely on manual processing alone, and the accuracy is not guaranteed, so the demand for high-precision automated processing is very urgent. Many deep learning algorithms are designed with this aspect in mind, and often more data yields better results. In recent years, the dense array seismic detection and localization method based on deep learning can directly extract the seismic phase characteristics from the original seismic waveform to identify seismic signals, and has been successfully applied in real-time monitoring of the network, construction of seismic catalog, and morphological analysis of fault zones. This paper draws on the ideas of previous research, uses the short-period dense array data in the Guye area of Tangshan, extracts the seismic phase to time and seismic wave information from the continuous waveform data based on the deep learning method RNN, uses the REAL algorithm to determine the number of earthquakes, the time of seismic origin, and the rough location, uses Hypoinverse absolute positioning, and finally uses HypoDD relocation to accurately calculate the seismic time, latitude and longitude and depth information of seismic events.