|
|
Post-processing of numerical precipitation forecast based on spatial-temporal deep learning model |
Chao-hao ZHENG1( ),Zhi-wei YIN2,Gang-feng ZENG3,Yue-ping XU1,Peng ZHOU1,Li LIU1,*( ) |
1. College of Civil Engineering and Architecture, Zhejiang University, Hangzhou 310058, China 2. Taizhou Water Resources Bureau, Taizhou 318000, China 3. Taizhou Water Resources & Hydropower Survey Designing Institute, Taizhou 318000, China |
|
|
Abstract In order to improve the accuracy and resolution of precipitation forecasts, taking Jiaojiang river basin in Zhejiang province as a case study, a precipitation post-processing model with the deep learning (CNN-LSTM) was proposed by using CMA-CMORPH precipitation grid dataset and ECMWF numerical precipitation predictions. The accuracy of precipitation forecast before and after the post-processing for different lead times were explored and the ability of precipitation forecasts to predict typical rainstorm events was evaluated. Results showed that CNN-LSTM can significantly improve the accuracy of the raw precipitation forecasts, the root mean square error decreasing from 6.0 mm to 3.0 mm, and the coefficient correlation was improved from 0.6 to 0.9. After the post-processing of two typhoon events, the precipitation forecast error of 6-hour accumulated areal rainfall forecasts after post-processing in Jiaojiang river basin were confined within 10%. The TS score during rainy season after post-treatment was generally near 0.90. The TS score of light rain was increased from less than 0.80 to 0.91, and the TS score of moderate rain increased from less than 0.50 to 0.60.
|
Received: 14 November 2022
Published: 16 October 2023
|
|
Fund: 公共安全科技关键技术、装备研发及应用示范-基于大数据和人工智能的流域性洪水灾害预防预警关键技术和应用示范(2021C03017);浙江省自然科学基金“基于智能化参数分区和定量降水预报的椒江流域集合洪水预报研究”(LQ22E090004);基于多源信息和深度学习的台风暴雨洪水分布式预报预警研究(2023M733117) |
Corresponding Authors:
Li LIU
E-mail: 22012252@zju.edu.cn;li_liu@zju.edu.cn
|
基于时空深度学习模型的数值降水预报后处理
为了提高降水预报的精度和分辨率,以浙江省椒江流域为研究对象,使用CMA-CMORPH降水网格数据集和ECMWF数值降水预报产品,提出基于深度学习的降水后处理模型CNN-LSTM. 探讨在不同预报时效的后处理前后降水预报的精度变化,评估降水预报对典型暴雨事件的预报能力. 结果表明:CNN-LSTM能够显著提升原始降水预报的精度,均方根误差从6.0 mm下降为3.0 mm,相关系数从0.6上升至0.9. 2起台风事件后处理的降水预报在椒江流域逐6 h面雨量误差均不超过10%. 经过雨季后处理的TS评分集中于0.90;并且在各降水等级表现均好于后处理前,小雨TS评分从不足0.80提升至0.91,中雨的TS评分从不足0.50提升至0.60.
关键词:
ECMWF,
时空后处理模型,
深度学习,
降水预报
|
|
[1] |
BLACUTT L, HERDIES D, GONCALVES L, et al Precipitation comparison for the CFSR, MERRA, TRMM 3B42 and combined scheme datasets in Bolivia[J]. Atmospheric Research, 2015, 163: 117- 131
doi: 10.1016/j.atmosres.2015.02.002
|
|
|
[2] |
DUAN Z, LIU J Z, TUO Y, et al Evaluation of eight high spatial resolution gridded precipitation products in Adige Basin (Italy) at multiple temporal and spatial scales[J]. Science of the Total Environment, 2016, 573: 1536- 1553
doi: 10.1016/j.scitotenv.2016.08.213
|
|
|
[3] |
陈超辉, 王勇, 杜钧 欧洲业务集合预报系统进展[J]. 气象科技进展, 2020, 10 (2): 19- 29 CHEN Chao-hui, WANG Yong, DU Jun Overview of the European operational ensemble prediction systems[J]. Advances in Meteorological Science and Technology, 2020, 10 (2): 19- 29
doi: 10.3969/j.issn.2095-1973.2020.02.004
|
|
|
[4] |
LI W, DUAN Q, WANG J Factors influencing the performance of regression-based statistical postprocessing models for short-term precipitation forecasts[J]. Weather and Forecasting, 2019, 34 (6): 2067- 2084
doi: 10.1175/WAF-D-19-0121.1
|
|
|
[5] |
郝福亮, 王旭 数值集合降雨预测的校正后处理方法研究[J]. 水资源于水工程学报, 2022, 33 (2): 460- 465 HAO Fu-liang, WANG Xu Post-correction method of numerical ensemble rainfall prediction[J]. Journal of Water Resources and Water Engineering, 2022, 33 (2): 460- 465
|
|
|
[6] |
代刊, 朱跃建, 毕宝贵 集合模式定量降水预报的统计后处理技术研究综述[J]. 气象学报, 2018, 76 (4): 493- 510 DAI Kan, ZHU Yue-jian, BI Bao-gui The review of statistical post-process technologies for quantitative precipitation forecast of ensemble prediction system[J]. Acta Meteorologica Sinica, 2018, 76 (4): 493- 510
doi: 10.11676/qxxb2018.015
|
|
|
[7] |
ZHAO T, BENNETT J, WANG Q, et al How suitable is quantile mapping for postprocessing GCM precipitation forecasts?[J]. Journal of Climate, 2017, 30 (9): 3185- 3196
doi: 10.1175/JCLI-D-16-0652.1
|
|
|
[8] |
ANDREW S, ZHAO T, QUAN J, et al A Bayesian modelling method for post-processing daily sub-seasonal to seasonal rainfall forecasts from global climate models and evaluation for 12 Australian catchments[J]. Hydrology and Earth System Sciences, 2018, 22 (2): 1615- 1628
doi: 10.5194/hess-22-1615-2018
|
|
|
[9] |
ROBERTSON D, SHRESTHA D, WANG Q Post-processing rainfall forecasts from numerical weather predic-tion models for short-term streamflow forecasting[J]. Hydrology and Earth System Sciences Discussions, 2013, 10 (5): 6765- 6806
|
|
|
[10] |
BAEZ V, ZAMBRANO B, BECK H, et al RF-MEP: a novel random forest method for merging gridded precipitation products and ground-based measurements[J]. Remote Sensing of Environment, 2020, 239: 111606
doi: 10.1016/j.rse.2019.111606
|
|
|
[11] |
ROMILLY T, GEBREMICHAEL M Evaluation of satellite rainfall estimates over Ethiopian River basins[J]. Hydrology and Earth System Sciences, 2011, 15 (5): 1505- 1514
doi: 10.5194/hess-15-1505-2011
|
|
|
[12] |
QIU M H, ZHAO P L, ZHANG K, et al. A short-term rainfall prediction model using multi-task convolutional neural networks [C]// 2017 IEEE International Conference on Data Mining. New Orleans: IEEE, 2017: 395–404.
|
|
|
[13] |
KRASNOPOLSKY V, LIN Y A neural network nonlinear multi-model ensemble to improve precipitation forecasts over continental US[J]. Advances in Meteorology, 2012, 649450
|
|
|
[14] |
皇甫江, 胡志群, 郑佳锋, 等 利用深度学习开展偏振雷达定量降水估测研究[J]. 气象学报, 2022, 80 (4): 565- 577 HUANGFU Jiang, HU Zhi-qun, ZHENG Jia-feng A study on polarization radar quantitative precipitation estimation using deep learning[J]. Acta Meteorologica Sinica, 2022, 80 (4): 565- 577
doi: 10.11676/qxxb2022.046
|
|
|
[15] |
NI L L, WANG D, SINGH V P, et al Streamflow and rainfall forecasting by two long short-term memory-based models[J]. Journal of Hydrology, 2020, 583: 124296
doi: 10.1016/j.jhydrol.2019.124296
|
|
|
[16] |
韩明星, 印萍, 褚忠信, 等 椒江流域水沙变化特征及人类活动影响[J]. 海洋地质前沿, 2022, 38 (12): 26- 39 HAN Ming-xing, YIN Ping, CHU Zhong-xin, et al Characteristics and influencing factors of water and sediment changes in Jiaojiang River Basin[J]. Marine Geology Frontiers, 2022, 38 (12): 26- 39
doi: 10.16028/j.1009-2722.2022.168
|
|
|
[17] |
SHEN Y, ZHAO P, PAN Y, et al A high spatiotemporal gauge-satellite merged precipitation analysis over China[J]. Journal of Geophysical Research: Atmospheres, 2014, 119: 3063- 3075
doi: 10.1002/2013JD020686
|
|
|
[18] |
KRIZHEVSKY A, SUTSKEVER I, HINTON G ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60 (6): 84- 90
doi: 10.1145/3065386
|
|
|
[19] |
LECUN Y, BENGIO Y, HINTON G Deep learning[J]. Nature, 2015, 521: 436- 444
doi: 10.1038/nature14539
|
|
|
[20] |
WU H, YANG Q, LIU J, et al A spatiotemporal deep fusion model for merging satellite and gauge precipitation in China[J]. Journal of Hydrology, 2020, 584: 124664
doi: 10.1016/j.jhydrol.2020.124664
|
|
|
[21] |
郭换换, 智协飞, 段明铿 数值天气预报中的不一致性问题综述[J]. 气象科学, 2016, 36 (1): 134- 140 GUO Huan-huan, ZHI Xia-fei, DUAN Ming-keng A review of inconsistencies in numerical weather forecasting[J]. Journal of the Meteorological Sciences, 2016, 36 (1): 134- 140
doi: 10.3969/2014jms.0063
|
|
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|