Student from School of Earth Sciences Conduct Academic Exchange at The University of Texas at Austin

From August 1 to September 30, 2024, Hu Jiachen, an undergraduate student from the Class of 2022 in Geology (Qiushi Honors Program) at Zhejiang University, was invited to participate in a two-month academic research project at the University of Texas at Austin (UT-Austin), focusing on AI-based P-wave first-motion polarity picking methods for microseismic events.

The University of Texas at Austin (UT-Austin) is one of the world's leading research universities. Bureau of Economic Geology (BEG), established in 1909, is the university’s oldest research unit and serves as the Geological Survey of Texas. Over the past century, BEG has driven Texas’ economic development through research on energy, environmental, and economic issues, and has taken a leading position in the field of global earth sciences. Zhejiang University, as one of China’s top comprehensive universities, has increasingly promoted international collaboration in recent years, encouraging students to participate in projects with world-class research institutions. This collaboration not only deepens international academic exchanges in the field of earth sciences, but also provides students with valuable opportunities to engage with and apply cutting-edge research techniques.

Bureau of Economic Geology（BEG）

The primary focus of this research project was the picking of P-wave first-motion polarities for microseismic events. P-wave first-motion polarity is a critical step in determining earthquake focal mechanisms (such as normal faults, strike-slip faults, etc.). However, traditional manual picking methods are time-consuming and limited in precision. In collaboration with the UT-Austin team, Hu Jiachen contributed to the development of an AI-based picking system, which significantly improves the efficiency and accuracy of seismic waveform data processing. This technology has wide-ranging applications in various industries, including shale gas exploration, carbon capture and storage, and geothermal energy production. These activities are often accompanied by microseismic or small earthquake events, and studying their focal mechanisms helps assess potential geological hazard risks.

Night View of UT-Austin Main Campus

During the eight-week collaboration, Hu Jiachen worked closely with UT-Austin professors and their research team. By analyzing large datasets of microseismic events from Texas, Canada, and the Sichuan Basin, the research team optimized the picking algorithm using deep learning strategies based on convolutional neural networks (CNNs) and Transformer models. Results showed that the models exhibited excellent generalization capabilities in polarity picking for seismic data from different regions, particularly for microseismic events, where the accuracy of the models was significantly improved.

The research demonstrated that deep learning techniques achieved over 95% accuracy on microseismic datasets from the Western Canadian Sedimentary Basin and over 90% accuracy on datasets from Sichuan. Compared to traditional methods, the AI-based picking system greatly enhanced the efficiency and accuracy of microseismic data processing. The collaboration between Hu Jiachen and BEG not only resulted in significant breakthroughs in seismology research but also provided new solutions for microseismic monitoring in industrial fields.

Participant Reflection
Hu Jiachen remarked, "This was an incredibly valuable research experience. During my time at UT-Austin, I not only learned the latest techniques in seismic data processing but also gained a deep understanding of the power of interdisciplinary collaboration. Working with a world-class research team has greatly expanded my academic horizons and significantly improved my research skills. This collaboration has filled me with confidence and excitement for future geophysical research."

Tour of the UT-Austin Main Campus Tower