物探化探计算技术2025，Vol.47Issue(5)：623-640,18.DOI:10.12474/wthtjs.20250902-0001
深度学习在地震数据处理中的应用
Application of deep learning in seismic data processing
摘要
Abstract
In recent years,Deep Learning has made remarkable strides in various seismological domains,including earthquake detection and phase picking,event classification,P-wave first-motion polarity determination,and earthquake focal mechanism inversion.It is increasingly emerging as a key driver in the intelligent transformation of seismic data processing.This paper systematically reviews the major advancements in the application of deep learning within seismology,with a particular focus on the development and implementation of large-scale seismological models over the past two years.Furthermore,it summarizes the latest achievements in scientific discovery enabled by artificial intelligence in seismology.On this basis,the paper identifies current technical bottlenecks and challenges faced by deep learning in seismological applications,such as data quality,model architecture,evaluation systems,and the role of AI in facilitating scientific discovery.Finally,it provides an outlook on future trends,aiming to offer systematic reference and guidance for related research.关键词
深度学习/人工智能/地震数据处理/地震学大模型Key words
deep learning/artificial intelligence/seismic data processing/large seismology models分类
天文与地球科学引用本文复制引用
李帅,蒋策,田平,范莉苹,夏登科,房立华..深度学习在地震数据处理中的应用[J].物探化探计算技术,2025,47(5):623-640,18.基金项目
国家自然科学基金(42374081) （42374081）
