《计算机应用》唯一官方网站 ›› 2023, Vol. 43 ›› Issue (6): 1958-1968.DOI: 10.11772/j.issn.1001-9081.2022050745

• 前沿与综合应用 • 上一篇    下一篇

深度学习在天气预报领域的应用分析及研究进展综述

董润婷1, 吴利1(), 王晓英1, 曹腾飞1, 黄建强1, 管琴2, 吴洁瑕3   

  1. 1.青海大学 计算机技术与应用系,西宁 810016
    2.青海省气象台,西宁 811300
    3.北京弘象科技有限公司,北京 100195
  • 收稿日期:2022-05-24 修回日期:2022-09-07 接受日期:2022-09-12 发布日期:2023-06-08 出版日期:2023-06-10
  • 通讯作者: 吴利
  • 作者简介:董润婷(1998—),女,河南郑州人,硕士研究生,主要研究方向:人工智能
    吴利(1992—),女,安徽铜陵人,助教,硕士,主要研究方向:人工智能、高性能计算Email:wuli_qhu@163.com
    王晓英(1982—),女,吉林大安人,教授,博士,主要研究方向:智能电网、高性能计算、体系结构
    曹腾飞(1987—),男,湖北钟祥人,副教授,博士,主要研究方向:智能网络优化、网络攻防
    黄建强(1985—),男,陕西西安人,副教授,博士,主要研究方向:高性能计算、大数据处理
    管琴(1974—),女,江苏如皋人,正高级工程师,硕士,主要研究方向:高原灾害性天气机理研究与预报
    吴洁瑕(1986—),女,浙江杭州人,博士,主要研究方向:气象数值预报。
  • 基金资助:
    国家自然科学基金资助项目(62162053);清华大学-宁夏银川水联网数字治水联合研究院横向课题(SKL-IOW-2020TC2004-01);青海省自然科学基金资助项目(2020-ZJ-943Q);青海省科技厅应用基础研究项目(2022-ZJ-701)

Review of application analysis and research progress of deep learning in weather forecasting

Runting DONG1, Li WU1(), Xiaoying WANG1, Tengfei CAO1, Jianqiang HUANG1, Qin GUAN2, Jiexia WU3   

  1. 1.Department of Computer Technologies and Applications,Qinghai University,Xining Qinghai 810016,China
    2.Qinghai Meteorological Observatory,Xining Qinghai 811300,China
    3.Beijing PRESKY Technology Company Limited,Beijing 100195,China
  • Received:2022-05-24 Revised:2022-09-07 Accepted:2022-09-12 Online:2023-06-08 Published:2023-06-10
  • Contact: Li WU
  • About author:DONG Runting, born in 1998, M. S. candidate. Her research interests include artificial intelligence.
    WANG Xiaoying, born in 1982, Ph. D., professor. Her research interests include smart grid, high performance computing, computer architecture.
    CAO Tengfei, born in 1987, Ph. D., associate professor. His research interests include intelligent network optimization, network attack and defense.
    HUANG Jianqiang, born in 1985, Ph. D., associate professor. His research interests include high performance computing, big data processing.
    GUAN Qin, born in 1974, M. S., professorate senior engineer. Her research interests include plateau severe weather mechanism research and forecast.
    WU Jiexia, born in 1986, Ph. D. Her research interests include numerical weather prediction.
  • Supported by:
    National Natural Science Foundation of China(62162053);Horizontal Project of Tsinghua University-Ningxia Yinchuan Digital Water Governance Joint Research Institute(SKL-IOW-2020TC2004-01);Natural Science Foundation of Qinghai Province(2020-ZJ-943Q);Application Basic Research Project of Science and Technology Department of Qinghai Province(2022-ZJ-701)

摘要:

随着传感器网络和全球定位系统等技术的进步,兼有时间与空间特性的气象数据体量呈爆炸式增长,针对时空序列预测(STSF)的深度学习模型研究得到了迅猛发展。然而,长期以来用于天气预报的传统机器学习方法在提取数据的时间相关性与空间依赖性方面的效果往往并不理想。与此同时,深度学习方法通过人工神经网络自动提取特征,可以有效提高天气预报的准确度,并且在编码长期空间信息的建模方面有相当优秀的效果。同时,由观测数据驱动的深度学习模型与基于物理理论的数值天气预报(NWP)模型结合的方式可以构建拥有更高预测精度与更长预报时间的混合模型。基于这些,将深度学习在天气预报领域的应用分析及研究进展进行了综述。首先,将天气预报领域的深度学习问题与经典深度学习问题从数据格式、问题模型与评价指标这3个方面进行了对比研究;然后,回顾了深度学习在天气预报领域的发展历程与应用现状,并总结分析了深度学习技术与NWP结合的最新进展;最后,展望了未来的发展方向和研究重点,为天气预报领域的深度学习研究提供参考。

关键词: 深度学习, 天气预报, 时空序列预测, 数值天气预报

Abstract:

With the advancement of technologies such as sensor networks and global positioning systems, the volume of meteorological data with both temporal and spatial characteristics has exploded, and the research on deep learning models for Spatiotemporal Sequence Forecasting (STSF) has developed rapidly. However, the traditional machine learning methods applied to weather forecasting for a long time have unsatisfactory effects in extracting the temporal correlations and spatial dependences of data, while the deep learning methods can extract features automatically through artificial neural networks to improve the accuracy of weather forecasting effectively, and have a very good effect in encoding long-term spatial information modeling. At the same time, the deep learning models driven by observational data and Numerical Weather Prediction (NWP) models based on physical theories are combined to build hybrid models with higher prediction accuracy and longer prediction time. Based on these, the application analysis and research progress of deep learning in the field of weather forecasting were reviewed. Firstly, the deep learning problems in the field of weather forecasting and the classical deep learning problems were compared and studied from three aspects: data format, problem model and evaluation metrics. Then, the development history and application status of deep learning in the field of weather forecasting were looked back, and the latest progress in combining deep learning technologies with NWP was summarized and analyzed. Finally, the future development directions and research focuses were prospected to provide a certain reference for future deep learning research in the field of weather forecasting.

Key words: deep learning, weather forecast, SpatioTemporal Sequence Forecasting (STSF), Numerical Weather Prediction (NWP)

中图分类号: