Journal of Computer Applications ›› 2026, Vol. 46 ›› Issue (7): 2318-2326.DOI: 10.11772/j.issn.1001-9081.2025070874

• Frontier and comprehensive applications • Previous Articles    

Survey of research on applications of explainable deep learning in tropical cyclone forecasting

Miaogen LING1,2, Rui JING1,3, Wei FANG1,3,4,5()   

  1. 1.School of Computer Science,Nanjing University of Information Science and Technology,Nanjing Jiangsu 210044,China
    2.Jiangsu Provincial Key Laboratory for Computer Information Processing Technology (Soochow University),Suzhou Jiangsu 215006,China
    3.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (Nanjing University of Information Science and Technology),Nanjing Jiangsu 210044,China
    4.Key Laboratory of Basin Heavy Precipitation of China Meteorological Administration / Hubei Key Laboratory for Heavy Rain Monitoring and Warning Research (Wuhan Institute of Heavy Rain,China Meteorological Administration),Wuhan Hubei 430205,China
    5.State Key Laboratory of Disaster Weather (Chinese Academy of Meteorological Sciences),Beijing 100081,China
  • Received:2025-08-04 Revised:2025-09-11 Accepted:2025-09-11 Online:2025-11-05 Published:2026-07-10
  • Contact: Wei FANG
  • About author:LING Miaogen, born in 1987, Ph. D., associate professor. His research interests include computer vision, deep learning.
    JING Rui, born in 1999, M. S. candidate. Her research interests include deep learning.
  • Supported by:
    General Program of National Natural Science Foundation of China(42475149);National Natural Science Foundation of China(62202235);Open Fund of Key Laboratory of Basin Heavy Precipitation of China Meteorological Administration(2023BHR-Y14);Open Project of State Key Laboratory of Severe Weather(2024LASW-B19)

可解释性深度学习在热带气旋预报中的应用研究综述

凌妙根1,2, 井瑞1,3, 方巍1,3,4,5()   

  1. 1.南京信息工程大学 计算机学院,南京 210044
    2.江苏省计算机信息处理技术重点实验室(苏州大学),江苏 苏州 215006
    3.江苏省大气环境与装备技术协同创新中心(南京信息工程大学),南京 210044
    4.中国气象局流域强降水重点开放实验室/暴雨监测预警湖北省重点实验室(中国气象局武汉暴雨研究所),武汉 430205
    5.灾害天气国家重点实验室(中国气象科学研究院),北京 100081
  • 通讯作者: 方巍
  • 作者简介:凌妙根(1987—),男,浙江乐清人,副教授,博士,CCF会员,主要研究方向:计算机视觉、深度学习
    井瑞(1999—),女,河北沧州人,硕士研究生,CCF会员,主要研究方向:深度学习
  • 基金资助:
    国家自然科学基金面上项目(42475149);国家自然科学基金资助项目(62202235);中国气象局流域强降水重点开放实验室开放研究基金资助项目(2023BHR-Y14);灾害天气国家重点实验室开放课题(2024LASW-B19)

Abstract:

Deep learning offers a new path for tropical cyclone forecasting, yet its black-box characteristic hinders its application in critical fields. Therefore, a review was conducted on the application research of explainable deep learning in tropical cyclone forecasting. First, the latest advances of deep learning research on tropical cyclone forecasting were classified systematically. Second, the advantages and disadvantages of traditional methods of forecasting were reviewed. Third, attention is focused on the applications of Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM) Network, and Transformer for forecasting tasks, and the frontier explorations of methods such as Generative Adversarial Network (GAN) were discussed. On this basis, the importance of explainable deep learning in tropical cyclone forecasting was analyzed in a deep way, the specific application of explainable deep learning to tropical cyclone forecasting was explored in three dimensions: feature importance interpretation, spatio-temporal mode interpretation, and uncertainty quantification, and its potential in enhancing model transparency and user confidence was pointed out. Finally, to address current challenges faced by explainable research, future research prospects of combining physical mechanisms, multi-granularity visualization, and causal inference were introduced.

Key words: tropical cyclone, deep learning, explainability, tropical cyclone forecasting, black-box characteristic

摘要:

深度学习为热带气旋预报研究提供了新途径,但其黑盒特性影响了它在关键领域的应用。因此,对可解释性深度学习在热带气旋预报中的应用研究进行综述。首先,系统梳理深度学习在热带气旋预报中的最新研究进展;其次,回顾传统预报方法的优势与局限;然后,聚焦卷积神经网络(CNN)、循环神经网络(RNN)、长短期记忆(LSTM)网络和Transformer等模型在不同预报任务中的应用,并探讨生成对抗网络(GAN)等方法的前沿探索。在此基础上,深入分析可解释性深度学习在热带气旋预报中的重要性,从特征重要性解释、时空模式解释以及不确定量化这3个维度探讨可解释性深度学习在热带气旋预报中的具体应用,并指出它在提升模型透明度和用户信任度方面的潜力。最后,针对当前可解释性研究中面临的挑战,提出未来融合物理机制、多粒度可视化及因果推理的研究展望。

关键词: 热带气旋, 深度学习, 可解释性, 热带气旋预报, 黑盒特性

CLC Number: