An Overview Study of Deep Learning in Geophysics: Cross-Cutting Research to Advance Geoscience

Abstract

In recent years, Artificial Intelligence technology (AI) has driven rapid advances across various sciences. As a new data-driven technology, Deep Learning (DL) is widely utilized for data processing and adaptive tasks in multiple fields due to its high automation, accuracy, and scalability. DL has garnered widespread attention and developed rapidly in geophysics. DL provides a new power for geophysical exploration and is becoming an essential tool for geophysical data processing, modeling, and analysis. With the proposal and effective application of various new deep learning-based technologies and methods for geophysical data processing, the efficiency and accuracy of geophysical exploration have been significantly improved. This advancement is accelerating the rapid development of geophysics toward intelligent interpretation. This paper reviews the latest research and application status of DL in geophysics, including seismic exploration, electrical prospecting, earthquake science, remote sensing, and other fields. Through systematic analysis of recent literature, it summarizes mainstream technical approaches of DL for addressing diverse geophysical challenges, and the limitations of these technologies in specific application scenarios are discussed. In addition, this paper analyses and prospects for research trends of DL in geophysics. This paper serves as a relevant reference for hobbyists and researchers to understand the latest advances, unresolved issues, and future trends in related fields.

Keywords

• Deep learning
• geophysics
• seismic exploration
• earthquake science
• remote sensing