[1]范娇,曾小团,黄荣成,等.深度学习在降水预报中的研究和应用进展[J].气象研究与应用,2024,45(03):1-11.[doi:10.19849/j.cnki.CN45-1356/P.2024.3.01]
 FAN Jiao,ZENG Xiaotuan,HUANG Rongcheng,et al.Research and application progress of deep learning in precipitation forecasting[J].Journal of Meteorological Research and Application,2024,45(03):1-11.[doi:10.19849/j.cnki.CN45-1356/P.2024.3.01]
点击复制

深度学习在降水预报中的研究和应用进展()
分享到:

气象研究与应用[ISSN:1673-8411/CN:45-1356/P]

卷:
第45卷
期数:
2024年03期
页码:
1-11
栏目:
综述
出版日期:
2024-09-15

文章信息/Info

Title:
Research and application progress of deep learning in precipitation forecasting
作者:
范娇12 曾小团1 黄荣成1 黄增俊2
1. 广西壮族自治区气象台, 南宁 530022;
2. 南宁市气象局, 南宁 530029
Author(s):
FAN Jiao12 ZENG Xiaotuan1 HUANG Rongcheng1 HUANG Zengjun2
1. Guangxi Meteorological Observatory, Nanning 530022, China;
2. Nanning Meteorological Bureau, Nanning 530029, China
关键词:
深度学习降水预报数值天气预报模式耦合
Keywords:
deep learningprecipitation forecastnumerical weather forecastingmodes coupling
分类号:
P456
DOI:
10.19849/j.cnki.CN45-1356/P.2024.3.01
文献标志码:
10.19849/j.cnki.CN45-1356/P.2024.3.01
摘要:
主要总结气象领域常用深度学习算法特征、纯数据驱动的深度学习降水预报技术和深度学习与数值天气预报耦合技术在临近、短时、中短期、气候降水预报中的探索成果,同时对气象业务预报中深度学习的应用进行简单回顾。传统深度学习算法的优缺点不同,在实际应用中,需要根据具体的气象数据特性和业务需求来选择合适的模型和方法。临近预报方面,常使用深度学习算法建立降水预报模型从而进行强降水云团空间特征提取和时间演变分析。长时间序列降水预报,则主要通过初始场资料同化、模式改进、模式预报后处理等方面优化数值模式降水预报效果。深度学习在解决降水预报技术方面有巨大潜力,部分成果已经实现业务应用,尤其在临近降水方面,对于更长时效的降水预测应用还有待进一步研究。
Abstract:
This article summarizes the characteristics of commonly used deep learning algorithms in the meteorological field,as well as the exploration results of pure data-driven deep learning precipitation forecasting technology and deep learning coupled with numerical weather forecasting technology in near, medium-short-term,and extended period precipitation forecasting. At the same time,a brief review is made on the application of deep learning in meteorological operational forecasting. The advantages and disadvantages of traditional deep learning algorithms are different. In practical applications,it is necessary to choose appropriate models and methods based on specific meteorological data characteristics and business needs. In terms of proximity prediction,deep learning algorithms are often used to establish precipitation prediction models for extracting spatial features and analyzing temporal evolution of heavy precipitation cloud clusters; Long time series(medium to short term,extended period)precipitation forecasting mainly optimizes the numerical model precipitation forecasting effect through initial field data assimilation, model improvement,and model prediction post-processing. At present,meteorological departments in various regions have applied deep learning algorithms in business forecasting. In the future,specific problems still need to be addressed,a large amount of related work needs to be carried out to further promote the development of precipitation forecasting.

参考文献/References:

[1] 董良鹏,张萍萍. 2020年湖北省两次低涡型极端降水气象因子异常特征对比[J]. 气象与环境学报,2022,38(6): 20-28.
[2] 张霞,杨慧,王新敏,等.“21·7”河南极端强降水特征及环流异常性分析[J]. 大气科学学报,2021,44(5): 672-687.
[3] 刘国忠,覃月凤,覃卫坚,等. 2022年广西极端"龙舟水"暴雨过程环境场特征分析[J]. 气象研究与应用, 2023,44(1): 7-13.
[4] 方楠,谢国权,阮小建,等. 长短期记忆神经网络(LSTM)模型在低能见度预报中的应用[J]. 气象与环境学报,2022,38(5): 34-41.
[5] 周维勋. 基于深度学习特征的遥感影像检索研究[J]. 测绘学报,2023,52(1):167.
[6] 滕琳,王斌,冯前进. 头颈癌放疗计划剂量分布的预测方法:基于深度学习的算法[J]. 南方医科大学学报, 2023,43(6): 1010-1016.
[7] 张萌,贾朋群,王小光. 人工智能技术在大气科学领域的应用及其发展态势[J]. 科学观察,2020,15(1): 22-29.
[8] 张梦涵,魏进,卞海丁. 基于机器学习的边坡稳定性分析方法—以国内618个边坡为例[J]. 地球科学与环境学报,2022,44(6): 1083-1095.
[9] 杨玲,魏静,许子伏. 基于平滑先验法-麻雀搜索算法-支持向量机回归模型的滑坡位移预测——以三峡库区八字门和白水河滑坡为例[J]. 地球科学与环境学报, 2022,44(6): 1096-1110.
[10] 黄兴友,马玉蓉,胡苏蔓. 基于深度学习的天气雷达回波序列外推及效果分析[J]. 气象学报,2021,79(5): 817-827.
[11] 唐伟,周勇,董昊,等. S22-气象应用人工智能的现状和影响分析[C]//中国气象学会年会.第35届中国气象学会年会S22供给侧结构性改革与气象高质量发展.2018.
[12] SHI X J,CHEN Z,WANG H,et al. Convolutional LSTM network: A machine learning approach for precipitation nowcasting[C]//Advances in Neural Information Processing Systems. 2015: 802-810.
[13] RAVURI S,LENC K,WILLSON M,et al. Skilful precipitation nowcasting using deep generative models of radar[J]. Nature,2021: 672-677.
[14] 方巍,齐媚涵. 基于深度学习的高时空分辨率降水临近预报方法[J]. 地球科学与环境学报,2023,45(3): 706-718.
[15] HE K M,ZHANG X,REN S,et al. Deep residual learning for image recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition.2016.
[16] RONNEBERGER O,FISCHER P,BROX T. U-Net: Convolutional networks for biomedical image segmentation [C]//International Conference on Medical image computing and computer-assisted intervention. 2015.
[17] HAN L,CHEN M X,CHEN K K,et al. A deep learning method for bias correction of ECMWF 24-240 h forecasts [J]. Advances in Atmospheric Sciences,2021,38(9): 1444-1459.
[18] 张延彪,陈明轩,韩雷,等. 数值天气预报多要素深度学习融合订正方法[J]. 气象学报,2022,80(1): 153-167.
[19] HOCHREITER S,SCHMIDHUBER J. Long short-term memory[J]. Neural Computation,1997: 1735-1780.
[20] VASWANI A,SHAZEER N,PARMAR N,et al. Attention is all you need[C]// Advances in Neural Information Processing Systems.2017.
[21] WANG Y B,LONG M S,WANG J M,et al. Predrnn: Recurrent neural networks for predictive learning using spatiotemporal lstms[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems(NeurIPS). 2017.
[22] WANG Y B,GAO Z F,LONG M S,et al. PredRNN++: Towards a resolution of the deep-in-time dilemma in spatiotemporal predictive learning[C]//Proceedings of the 35th International Conference on Machine Learning. 2018.
[23] GOODFELLOW I J,POUGET-ABADIE J P,MIRZA M, et al. Generative adversarial nets[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems.2014.
[24] WANG X T,YU K,WU S X,et al. ESRGAN: Enhanced super-resolution generative adversarial networks[C]//Proceedings of the 15th European Conference on Computer Vision.2018.
[25] WHITE B L,SINGH A,ALBERT A.Downscaling numerical weather models with GANs[C]//Proceedings of the 9th International Workshop on Climate Informatics. 2019.
[26] JING J R,LI Q ,PENG X,et al. HPRNN: A hierarchical sequence prediction model for long-term weather radar echo extrapolation[C]//IEEE International Conference on Acoustics,Speech,and Signal Processing. 2020.
[27] SHI X J,GAO Z H,LAUSEN L,et al. Deep learning for precipitation nowcasting: A benchmark and a new model [C]//Conference on Neural Information Processing Systems.2015.
[28] 吴昆,申妍燕,王书强. 基于3D卷积神经网络的区域降雨量预报[J]. 图像与信号处理,2018,7(4): 200-212.
[29] ZAMO M,BEL L,MESTRE O. Sequential aggregation of probabilistic forecasts — application to wind speed ensemble forecasts[J]. Journal of the Royal Statistical Society Series C-applied Statistics,2021: 202-225.
[30] SONDERBY C K,ESPEHOLT L,HEEK J,et al. MetNet: A neural weather model for precipitation forecasting [J/OL]. 2020.
[31] AYZEL G,SCHEFFER T,HEISTERMANN M. RainNet v1.0: a convolutional neural network for radar-based precipitation nowcasting[J]. Geoscientific Model Development,2020,13(6): 2631-2644.
[32] 周康辉,郑永光,韩雷,等. 机器学习在强对流监测预报中的应用进展[J]. 气象,2021,47(3):274-289.
[33] 董琪如,邱晓滨,王莹,等. 循环同化雷达资料对一次飑线系统临近预报的改进作用[J]. 气象与环境学报, 2021,37(3): 1-11.
[34] 郑永光,周康辉,盛杰,等. 强对流天气监测预报预警技术进展[J]. 应用气象学报,2015,26(6):641-657.
[35] 孙继松,戴建华,何立富,等. 强对流天气预报的基本原理与技术方法[M]. 北京:气象出版社,2014.
[36] 王丹,王改利,刘黎平,等. 基于雷达回波外推和中尺度模式预报的短时降水对比分析[J]. 高原气象,2014, 33(3): 811-822.
[37] 龙清怡,刘海文,顾建峰,等. 雷达资料与中尺度数值预报的融合方法研究及其在临近预报中的应用[J].气象,2014,40(10): 1248-1258.
[38] 顾建峰,周国兵,刘伯骏,等. 人工智能技术在重庆临近预报业务中的初步研究与应用[J]. 气象,2020,46(10): 1286-1296.
[39] 张亚萍,张焱,翟丹华,等. 深度学习方法在强降水临近预报技术中的应用及其改进思考[J]. 暴雨灾害, 2022,41(5): 506-514.
[40] AGRAWAL S,BARRINGTON L,BROMBERG C,et al. Machine learning for precipitation nowcasting from radar Images[J/OL]. 2019.
[41] 郭瀚阳,陈明轩,韩雷,等. 基于深度学习的强对流高分辨率临近预报试验[J]. 气象学报,2019,77(4): 715-727.
[42] 曹伟华,南刚强,陈明轩,等. 基于深度学习的京津冀地区精细尺度降水临近预报研究[J]. 气象学报,2022, 80(4): 546-564.
[43] 胡家晖,卢楚翰,姜有山,等. 深度学习模型TAGAN在强对流回波临近预报中的应用[J]. 大气科学,2022,46(4): 805-818.
[44] CAO Y,CHEN L,ZHANG D C,et al. Hybrid weighting loss for precipitation nowcasting from radar images[C]// IEEE International Conference on Acoustics,Speech, and Signal Processing.2022.
[45] 夏侯杰,肖安,聂道洋. 基于观测的短时强降水深度学习预报模型[J]. 高原气象,2023,42(4): 1005-1017.
[46] 林青,潘宁,刘铭. 光流法以及结合机器学习在临近预报中的应用[J]. 海峡科学,2017(12):17-20.
[47] WOO W C,WONG W K. Operational application of optical flow techniques to radar-based rainfall nowcasting[J]. Atmosphere,2017.
[48] 张智察,罗玲,陈列,等. 卷积门控循环单元神经网络与光流法在临近预报中的适用性研究[J]. 气象,2022, 48(11):1361-1372.
[49] WANG Y B,WU H X,ZHANG J J,et al. PredRNN: A recurrent neural network for spatiotemporal predictive learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(2): 2208-2225.
[50] 杨绚,代刊,朱跃建. 深度学习技术在智能网格天气预报中的应用进展与挑战[J]. 气象学报,2022,80(5): 649-667.
[51] 马雷鸣,林红,储海,等. 上海强对流智能预报业务新技术研究进展[J]. 地球科学进展,2023,38(2):111-124.
[52] 马雷鸣. 天气预报中的人工智能技术进展[J]. 地球科学进展,2020,35(6): 551-560.
[53] 李扬,刘玉宝,许小峰. 基于深度学习改进数值天气预报模式和预报的研究及挑战[J]. 气象科技进展,2021, 11(3):103-112.
[54] GEER A J. Learning earth system models from observations: Machine learning or data assimilation?[J]. Philosophical Transactions of The Royal Society.Mathematical, Physical and Engineering Sciences,2021.
[55] CINTRA R ,VELHO H F D C. Data assimilation by artificial neural networks for an atmospheric general circulation model [M]. 2018.
[56] BRAJARD J,CARASSI A,BOCQUET M,et al. Combining data assimilation and machine learning to emulate a dynamical model from sparse and noisy observations: a case study with the Lorenz 96 model[J]. Journal of Computational Science,2020.
[57] PEYRON M,FILLION A,GUROL S,et al. Latent space data assimilation by using deep learning[J]. Quarterly Journal of the Royal Meteorological Society,2021,147(740):3759-3777.
[58] STENSRUD D J. Parameterization Schemes: Keys to understanding numerical weather prediction models[M]. Cambridge:Cambridge University Press,2007.
[59] SCHNEIDER T,TEIXEIRA J,BRETHERTON C S,et al. Climate goals and computing the future of clouds[J]. Nature Climate Change,2017.
[60] PAL A,MAHAJAN S,NORMAN M R. Using deep neural networks as cost-effective surrogate models for superparameterized E3SM radiative Transfer[J]. Geophysical Research Letters,2019,46(11): 6069-6079.
[61] WANG J L,BALAPRAKASH P,KOTAMARTHI R. Fast domain-aware neural network emulation of a planetary boundary layer parameterization in a numerical weather forecast model[J]. Geoscientific Model Development, 2019.
[62] SCHER S. Toward data-driven weather and climate forecasting: Approximating a simple general circulation model with deep learning[J]. Geophysical Research Letters, 2018.
[63] RASP S,PRITCHARD M S,GENTINE P,et al. Deep learning to represent subgrid processes in climate models [J]. Proceedings of the National Academy of Sciences of the United States of America,2018,115(39): 9684-9689.
[64] BRENOWITZ N D,BRETHERTON C S. Spatially extended tests of a neural network parametrization trained by coarse-graining[J]. Journal of Advances in Modeling Earth Systems,2019,11(8): 2728-2744.
[65] SCHAUMANN P,HESS R,REMPEL M,et al. A calibrated and consistent combination of probabilistic forecasts for the exceedance of several precipitation thresholds using neural networks[J]. Weather and Forecasting, 2021,36(3): 1079-1096.
[66] LIU Y Q,ZHANG J P,CHEN L,et al. SSAS: Spatiotemporal scale adaptive selection for improving bias correction on precipitation[J]. IEEE Transactions on Cybernetics, 2022,52(11): 12175-12188.
[67] 陈锦鹏,冯业荣,蒙伟光,等. 基于卷积神经网络的逐时降水预报订正方法研究[J]. 气象,2021,47(1): 60-70.
[68] SAYEED A,CHOI Y,JUNG J,et al. A deep convolutional neural network model for improving WRF simulations[J]. IEEE Transactions on Neural Networks and Learning Systems,2023,34(2): 750-760.
[69] GHAZVINIAN M,ZHANG Y,SEO D J,et al. A novel hybrid artificial neural network-parametric scheme for post-processing medium-range precipitation Forecasts [J]. Advances in Water Resources,2021.
[70] SCHEUERER M,SWITANEK M B,WORSNOP R P,et al. Using artificial neural networks for generating probabilistic subseasonal precipitation forecasts over california [J]. Monthly weather review,2020,148(8): 3489-3506.
[71] MISRA S,SARKAR S,MITRA P. Statistical downscaling of precipitation using long short-term memory recurrent neural networks[J]. Theoretical and Applied Climatology, 2018,134(3/4): 1179-1196.
[72] PAN B X,HSU K,AGHAKOUCHAK A,et al. Improving precipitation estimation using convolutional neural network[J]. Water Resources Research, 2019, 55(3): 2301-2321.
[73] BANO-MEDINA J,MANZANAS R,GUTIERREZ J M. Configuration and intercomparison of deep learning neural models for statistical downscaling[J]. Geoscientific Model Development,2020,13(4): 2109-2124.
[74] KUMAR B,CHATTOPADHYAY R,SINGH M,et al. Deep learning-based downscaling of summer monsoon rainfall data over Indian region[J]. Theoretical and Applied Climatology,2021,143(3/4): 1145-1156.
[75] SHA Y K,GAGNE Ⅱ D J,WEST G,et al. Deeplearning-based gridded downscaling of surface meteorological variables in complex terrain. Part Ⅰ : Daily maximum and minimum 2-m temperature[J]. Journal of Applied Meteorology and Climatology,2020,59(12): 2057-2073.
[76] SHA Y K,GAGNE D J II,WEST G,et al. Deep-learningbased gridded downscaling of surface meteorological variables in complex terrain. Part Ⅱ : Daily precipitation[J]. Journal of Applied Meteorology and Climatology,2020,59(12): 2075-2092.
[77] SHA Y K,GAGNE Ⅱ D J,WEST G,et al. A hybrid analog-ensemble-convolutional-neural-network method for post-processing precipitation forecasts[J]. Monthly Weather Review,2022.
[78] LEINONEN J,NERINI D,BERNE A. Stochastic superresolution for downscaling time-evolving atmospheric fields with a generative adversarial network [J]. IEEE Transactions on Geoscience and Remote Sensing,2021, 59(9): 7211-7223.
[79] PRICE I,RASP S. Increasing the accuracy and resolution of precipitation forecasts using deep generative models [C]//International Conference on Artificial Intelligence and Statistics. 2022.
[80] HERNANDEZ E,SANCHEZ-ANGUIX V,JULIAN V, et al. Rainfall prediction: A deep learning approach[C]// Proceedings of the 11th International Conference on Hybrid Artificial Intelligence Systems.2016.
[81] 张鹏程,张雷,王继民. 一种基于深度网络的多环境因素降水量预报模型[J]. 计算机应用与软件,2017,34(9): 241-245,261.
[82] PATHAK J,SUBRAMANIAN S,HARRINGTON P,et al. FourCastNet: A global data-driven high-resolution weather model using adaptive fourier neural operators [J/OL]. 2022.
[83] 周佩,黄颖婕,胡冰逸,等. 基于LSTM深度学习的ENSO预测及其春季预报障碍研究[J]. 北京大学学报(自然科学版),2021,57(6):1071-1078.
[84] 张星,关悦,党鑫鑫,等. 基于遗传算法-长短期记忆神经网络的月降水量预测研究[C]// 2021年(第七届) 全国大学生统计建模大赛获奖论文集(二). 2023.
[85] HAM Y G,KIM J H,LUO J J. Deep learning for multiyear ENSO forecasts[J]. Nature,2019.
[86] 潘乐,郑子华,游立军,等. 基于改进U-Net模型的次季节降水预报订正[J]. 福建师范大学学报(自然科学版),2023,39(6):63-71,92.
[87] 张丽婷,李鹏飞,庞文静,等. 基于季节性自回归积分滑动平均与深度学习长短期记忆神经网络的降水量预测[J]. 科学技术与工程,2022,22(9): 3453-3463.
[88] WANG X,HAN Y,XUE W,et al. Stable climate simulations using a realistic GCM with neural network parameterizations for atmospheric moist physics and radiation processes[J]. Geoscientific Model Development,2021.
[89] 金荣花,代刊,赵瑞霞,等. 我国无缝隙精细化网格天气预报技术进展与挑战[J]. 气象,2019,45(4): 445-457.
[90] ZHOU K H,ZHENG Y G,LI B,et al. Forecasting different types of convective weather: A deep learning approach [J]. Journal of Meteorological Research,2019.
[91] 张小玲,杨波,盛杰,等. 中国强对流天气预报业务发展[J]. 气象科技进展,2018,8(3): 8-18.
[92] 孙健,曹卓,李恒,等. 人工智能技术在数值天气预报中的应用[J]. 应用气象学报,2021,32(1): 1-11.
[93] 唐伟,周勇,董昊,等. 我国气象领域应用人工智能技术的现状和国际对比[J]. 气象科技进展,2019,9(5): 55-56,62.
[94] BI K F,XIE L X,ZHANG H H,et al. Accurate mediumrange global weather forecasting with 3D neural networks [J]. Nature,2023.
[95] ZHANG Y C,LONG M S,CHEN K Y,et al. Skilful nowcasting of extreme precipitation with nowcastnet[J]. Nature,2023.
[96] WILLARD J,JIA X,XU S,et al. Integrating physicsbased modeling with machine learning: A survey[J]. ACM Computing Surveys,2021,55(4): 1-37.
[97] 庄潇然,郑玉,王亚强,等. 基于深度学习的融合降水临近预报方法及其在中国东部地区的应用研究[J].气象学报,2023,81(2): 286-303.

相似文献/References:

[1]梁振清,陈生.基于深度学习和雷达观测的华南短临预报精度评估[J].气象研究与应用,2020,41(01):41.[doi:10.19849/j.cnki.CN45-1356/P.2020.1.09]
 Liang Zhenqing,Chen Sheng.Accuracy evaluation of nowcasting in South China based on deep learning and radar observation[J].Journal of Meteorological Research and Application,2020,41(03):41.[doi:10.19849/j.cnki.CN45-1356/P.2020.1.09]
[2]金龙,黄颖,姚才,等.人工智能技术的热带气旋预报综述(之二)——流形学习、智能计算及深度学习的热带气旋预报方法[J].气象研究与应用,2020,41(04):5.[doi:10.19849/j.cnki.CN45-1356/P.2020.4.02]
 Jin Long,Huang Ying,Yao Cai,et al.Summary of tropical cyclone forecasting based on artificial intelligence technology (part 2)——tropical cyclone forecasting methods based on manifold learning, intelligent calculation and deep learning[J].Journal of Meteorological Research and Application,2020,41(03):5.[doi:10.19849/j.cnki.CN45-1356/P.2020.4.02]
[3]黄友菊,韦强,罗恒,等.基于SAR影像的广西东北地区2022年“龙舟水”洪涝智能监测[J].气象研究与应用,2023,44(01):94.[doi:10.19849/j.cnki.CN45-1356/P.2023.1.16]
 Huang Youju,Wei Qiang,Luo Heng,et al.Intelligent monitoring of flooding of a dragon-boat precipitation process in northeast Guangxi in 2022 based on SAR images[J].Journal of Meteorological Research and Application,2023,44(03):94.[doi:10.19849/j.cnki.CN45-1356/P.2023.1.16]
[4]黄颖,陆虹,黄小燕,等.基于EOF和LSTM的广西月降水量预测模型研究[J].气象研究与应用,2023,44(02):20.[doi:10.19849/j.cnki.CN45-1356/P.2023.2.04]
 Huang Ying,Lu Hong,Huang Xiaoyan,et al.Study on monthly precipitation prediction model in Guangxi based on EOF and LSTM[J].Journal of Meteorological Research and Application,2023,44(03):20.[doi:10.19849/j.cnki.CN45-1356/P.2023.2.04]

备注/Memo

备注/Memo:
收稿日期:2024-2-15。
基金项目:广西自然科学基金项目(2022GXNSFAA035482)、广西气象科研计划项目(桂气科2022QN05、桂气科2020QN03)
作者简介:范娇(1993-),工程师,主要从事数值天气预报研究。E-mail:742895059@qq.com
通讯作者:曾小团,正高级工程师,主要从事天气预报技术研究和系统开发工作。E-mail:158083890@qq.com
更新日期/Last Update: 2024-09-15