基于 S2S 模式的长江中下游夏季降水次季节预报评估与诊断

Evaluation and diagnosis of subseasonal prediction of summer precipitation in the middle and lower reaches of the Yangtze River by S2S models

摘要

长江中下游是我国夏季降水最为集中的区域之一，提高其次季节预报水平对防汛减灾具有重要意义。基于次季节至季节 (Subseasonal to Seasonal，S2S) 计划里中国气象局 (CMA)、欧洲中期天气预报中心 (ECMWF)、英国气象局 (UKMO) 三套模式 (统称 S2S 模式) 预报资料，系统评估了单模式及多模式集合 (MME) 对长江中下游夏季 1—6 候降水的预报技巧及误差时空结构特征，进一步定量分析了各模式对夏季季节内振荡 (BSISO) 的预报能力及其与降水预报技巧之间的关系。结果表明：(1) 在确定性预报方面，各模式均表现为集合平均优于控制预报，预报技巧随时效延长逐渐衰减；ECMWF 整体最优、UKMO 次之、CMA 相对偏弱。MME 在第 3 候时效以后表现出更稳定且更高的技巧，显示出较高的预报延伸潜力。(2) 概率预报评估显示，ECMWF 在不同降水阈值下均表现最好，且预报技巧随降水阈值升高呈现先增后减特征；MME 在所有时效和所有阈值下均显著优于气候态预报和单模式预报。Brier Score (BS) 评分分解诊断表明，单模式对极端强降水较中部阈值降水表现出更高的可靠性但更低的分辨能力，MME 相较于单模式在可靠性与分辨能力两方面均有明显改进。(3) 各模式对 BSISO 信号的预报技巧与其降水预报技巧具有显著正相关，相关系数达 0.9 以上，BSISO 活动是该区域次季节降水可预报性的关键来源之一。该研究结果有望为各模式在长江中下游夏季降水次季节预报业务中的应用以及预报订正提供参考。

Abstract

The middle and lower reaches of the Yangtze River represent one of the regions in China experiencing heavy summer precipitation. Improving subseasonal prediction skills in this region is therefore of great importance for flood prevention and disaster mitigation. This study systematically evaluates the forecasting skill and error characteristics of both single-model (ECMWF, CMA, and UKMO) and multi-model ensemble systems for summer precipitation in the middle and lower Yangtze River region from lead pentad 1 to 6. We further quantitatively analyze the model ability to forecast BSISO and its relationship with precipitation forecasting skill. The results indicate: (1) In deterministic forecasting, all models show that ensemble mean forecasts outperform control forecasts, with forecast skill gradually decreasing as lead time extends. ECMWF is the best overall, followed by UKMO and CMA. The multi-model ensemble shows the most stable and highest skill after the 3^rd lead pentad, demonstrating greater forecast extension potential. (2) Probability evaluation shows that ECMWF performs the best across all precipitation thresholds, with forecast skill increasing at the beginning and then decreasing as the threshold rises. The multi-model ensemble significantly outperforms both the climate baseline and single-model forecasts across all lead times and thresholds. Brier score decomposition reveals that single-model forecasts for extreme high/low precipitation thresholds exhibit higher reliability but lower resolution compared to mid-range thresholds. In contrast, multi-model ensembles show significant improvements in both reliability and resolution. (3) The skill of models in forecasting BSISO signals is significantly positively correlated with their precipitation forecasting skill, showing correlation coefficients of greater than 0.9 and indicating that BSISO activity is a key source of subseasonal predictability for precipitation over this region. This study provides guidance for application of models in subseasonal precipitation forecasting over the middle and lower Yangtze River, and offers insights for forecast calibration.

关键词

S2S 模式 / 长江中下游 / 夏季降水 / 次季节预报 / 预报技巧

Key words

S2S models / the middle and lower reaches of the Yangtze River / summer precipitation / subseasonal prediction / prediction diagnosis

参考文献

[1] 池艳珍,鲍瑞娟,何芬,等. 2025. WMO/S2S 模式对 2020 年中国东部持续极端高温事件预报评估[J].沙漠与绿洲气象,19(6):71-79.

[2] Chi Y Z, Bao R J, He F, et al. 2025. Assessment of WMO/S2S in simulating persistent extreme heat events over Eastern China in 2020 [J].Desert and Oasis Meteorology,19(6):71-79 (in Chinese). doi:10.12057/j.issn.2097−6801.2406.21002

[3] 任宏利,沈雨旸. 2016. MJO 对我国天气气候影响的新事实[J].气象科技进展,6(3):97-105.

[4] Ren H L, Shen Y Y. 2016. A new look at impacts of MJO on weather and climate in China [J].Advances in Meteorological Science and Technology,6(3):97-105 (in Chinese). doi:10.3969/j.issn.2095−1973.2016.03.013

[5] 孙晨,杨青青,吴碧琼,等. 2025. BCC—CPSv3 次季节降水预测在长江流域汛期的检验评估[J].气象,51(6):724-734.

[6] Sun C, Yang Q Q, Wu B Q, et al. 2025. Performance evaluation of BCC—CPSv3 subseasonal precipitation prediction in the flood season of the Yangtze River basin [J].Meteorological Monthly,51(6):724-734 (in Chinese). doi:10.7519/j.issn.1000−0526.2025.011301

[7] 孙可可,吴小飞. 2024. FGOALS—f2 气候预测系统对西南地区持续性极端降水的次季节预测评估[J].高原山地气象研究,44(1):21-30.

[8] Sun K K, Wu X F. 2024. Sub—seasonal prediction of the regional persistent extreme precipitation in Southwest China by GOALS—f2 climate prediction system [J].Plateau and Mountain Meteorology Research,44(1):21-30 (in Chinese). doi:10.3969/j.issn.1674−2184.2024.01.003

[9] 王熙乔,李健颖,李双林. 2023.北半球夏季 30~60 天季节内振荡年代际变化及其对中国东部降水和气温的影响[J].大气科学,47(6):1665-1679.

[10] Wang X Q, Li J Y, Li S L. 2023. Decadal variation of boreal summer 30—60—day intraseasonal oscillation and its influence on precipitation and temperature over Eastern China [J].Chinese Journal of Atmospheric Sciences,47(6):1665-1679 (in Chinese). doi:10.3878/j.issn.1006−9895.2204.21228

[11] 吴佳,高学杰. 2013.一套格点化的中国区域逐日观测资料及与其它资料的对比[J].地球物理学报,56(4):1102-1111.

[12] Wu J, Gao X J. 2013. A gridded daily observation dataset over China region and comparison with the other datasets [J].Chinese Journal of Geophysics,56(4):1102-1111 (in Chinese). doi:10.6038/cjg20130406

[13] 许乐心,张人禾,齐艳军. 2017.长江中游和下游夏季降水季节内振荡的差异[J].大气科学,41(6):1125-1140.

[14] Xu L X, Zhang R H, Qi Y J. 2017. Differences in intraseasonal summer rainfall oscillation between the middle and lower reaches of the Yangtze River [J].Chinese Journal of Atmospheric Sciences,41(6):1125-1140 (in Chinese). doi:10.3878/j.issn.1006−9895.1703.17112

[15] 杨梦兮,胡瑞金,刘梅,等. 2025.江苏梅雨期极端小时降水时空变化特征[J].暴雨灾害,44(4):473-482.

[16] Yang M X, Hu R J, Liu M, et al. 2025. The spatiotemporal variation characteristics of extreme hourly precipitation during the Meiyu period in Jiangsu Province [J].Torrential Rain and Disasters,44(4):473-482 (in Chinese). doi:10.12406/byzh.2024−033

[17] 章大全,郑志海,陈丽娟,等. 2019. 10~30 d 延伸期可预报性与预报方法研究进展[J].应用气象学报,30(4):416-430.

[18] Zhang D Q, Zheng Z H, Chen L J, et al. 2019. Advances on the predictability and prediction methods of 10—30 d extended range forecast [J].Journal of Applied Meteorological Science,30(4):416-430 (in Chinese). doi:10.11898/1001−7313.20190403

[19] 张立凤. 2024.暴雨数值预报若干关键技术发展的回顾与思考[J].暴雨灾害,43(3):243-254.

[20] Zhang L F. 2024. Review and thinking on the development of several key technologies for heavy rainfall numerical weather prediction [J].Torrential Rain and Disasters,43(3):243-254 (in Chinese). doi:10.12406/byzh.2023−230

[21] 张铭明,李建,甘玉婷,等. 2021.基于 GWR 模型的中国中东部降水与海拔高度关系特征分析[J].暴雨灾害,40(1):1-11.

[22] Zhang M M, Li J, Gan Y T, et al. 2021. Analysis of the relationship between precipitation and altitude over central and eastern China based on the Geographically Weighted Regression Model [J].Torrential Rain and Disasters,40(1):1-11 (in Chinese). doi:10.3969/j.issn.1004−9045.2021.01.001

[23] Buizza R, Milleer M, Palmer T N. 1999. Stochastic representation of model uncertainties in the ECMWF ensemble prediction system [J].Quarterly Journal of the Royal Meteorological Society,125:2887-2908. doi:10.1002/qj.49712556006

[24] Buizza R, Palmer T N. 1998. Impact of ensemble size on ensemble prediction [J].Monthly Weather Review,126:2503-2518. doi:10.1175/1520−0493(1998)126<2503:IOESOE>2.0.CO;2

[25] Ding Y, Chan J. 2005. The East Asian summer monsoon: an overview [J].Meteorology and Atmospheric Physics,89:117-142. doi:10.1007/s00703−005−0125−z

[26] Ding L, Li T, Sun Y. 2021. Subseasonal and synoptic variabilities of precipitation over the Yangtze River Basin in the summer of 2020 [J].Advances in atmospheric sciences,38:2108-2124. doi:10.1007/s00376−021−1133−8

[27] Fan Y, Lyu Y, Zhu S, et al. 2024. Monthly prediction on summer extreme precipitation with a deep learning approach: Experiments over the mid—to—lower reaches of the Yangtze River [J].Earth and Space Science,11:e2024EA003926. doi:10.1029/2024EA003926

[28] Ferro C A T, Richardson D S, Weigel A P. 2008. On the effect of ensemble size on the discrete and continuous ranked probability scores [J].Meteorological Applications,15:19-24. doi:10.1002/met.45

[29] Guo L, Wu J, Li Q, et al. 2025. Advantages of the multimodel ensemble approach for subseasonal precipitation prediction in China and the driving factor of the MJO [J].Advances in Atmospheric Sciences,42:551-563. doi:10.1007/s00376−024−4107−9

[30] Han J Y, Kim S W, Park C H, et al. 2023. Ensemble size versus bias correction effects in subseasonal—to—seasonal (S2S) forecasts [J].Geoscience Letters,10:37. doi:10.1186/s40562−023−00292−9

[31] Hsu P C, Lee J Y, Ha K J. 2016. Influence of boreal summer intraseasonal oscillation on rainfall extremes in southern China [J].International Journal of Climatology,36:1403-1412. doi:10.1002/joc.4433

[32] Kanamitsu M, Ebisuzaki W, Woollen J, et al. 2002. NCEP—DOE AMIP—II reanalysis (R—2) [J].Bulletin of the American Meteorological Society,83:1631-1644. doi:10.1175/BAMS−83−11−1631

[33] Kikuchi K. 2021. The boreal summer intraseasonal oscillation (BSISO): A review [J].Journal of the Meteorological Society of Japan. Ser. II, 99:933-972. doi:10.2151/jmsj.2021−045

[34] Lee J Y, Wang B, Wheeler M C, et al. 2013. Real—time multivariate indices for the boreal summer intraseasonal oscillation over the Asian summer monsoon region [J].Climate Dynamics,40:493-509. doi:10.1007/s00382−012−1544−4

[35] Liang P, Lin H. 2018. Sub—seasonal prediction over East Asia during boreal summer using the ECCC monthly forecasting system [J].Climate Dynamics,50:1007-1022. doi:10.1007/s00382−017−3658−1

[36] Lyu Y, Zhi X, Zhu S, et al. 2025. Subseasonal prediction of weekly precipitation anomalies over Southern China during early summer based on predictability analysis and machine learning [J].Atmospheric Research,315:107927. doi:10.1016/j.atmosres.2025.107927

[37] Murphy A H. 1973. A new vector partition of the probability score [J].Journal of Applied Meteorology and Climatology,12:595-600. doi:10.1175/1520−0450(1973)012<0595:ANVPOT>2.0.CO;2

[38] Palmer T N, Alessandri A, Andersen U, et al. 2004. Development of a European multimodel ensemble system for seasonal—to—interannual prediction (DEMETER) [J].Bulletin of the American Meteorological Society,85:853-872. doi:10.1175/BAMS−85−6−853

[39] Robertson A W, Kumar A, Peña M, et al. 2015. Improving and promoting subseasonal to seasonal prediction [J].Bulletin of the American Meteorological Society,96:ES49-ES53. doi:10.1175/BAMS−D−14−00139.1

[40] Tian D, Wood E F, Yuan X. 2017. CFSv2—based sub—seasonal precipitation and temperature forecast skill over the contiguous United States [J].Hydrology and Earth System Sciences,21:1477-1490. doi:10.5194/hess−21−1477−2017

[41] Vitart F, Ardilouze C, Bonet A, et al. 2017. The subseasonal to seasonal (S2S) prediction project database [J].Bulletin of the American Meteorological Society,98:163-173. doi:10.1175/BAMS−D−16−0017.1

[42] Vitart F, Robertson A W. 2018. The sub—seasonal to seasonal prediction project (S2S) and the prediction of extreme events [J].npj Climate and Atmospheric Science,1:3. doi:10.1038/s41612−018−0013−0

[43] Vitart F, Robertson A W, Brookshaw A, et al. 2025. The WWRP/WCRP S2S project and its achievements [J].Bulletin of the American Meteorological Society,106:E791-E808. doi:10.1175/BAMS−D−24−0047.1

[44] Wilks D S. 2019. Statistical methods in the atmospheric sciences [M]. 4rd Edition. Elsevier, ISBN: 978—0—12—815823—4.

[45] Wu J, Li J, Zhu Z, et al. 2023. Factors determining the subseasonal prediction skill of summer extreme rainfall over southern China [J].Climate Dynamics,60:443-460. doi:10.1007/s00382−022−06326−w

[46] Yang J, Zhu T, Gao M, et al. 2018. Late‐ July barrier for subseasonal forecast of summer daily maximum temperature over Yangtze River Basin [J].Geophysical Research Letters,45:12610-12615. doi:10.1029/2018GL080963

[47] Zhu S, Zhi X, Ge F, et al. 2021. Subseasonal forecast of surface air temperature using superensemble approaches: Experiments over Northeast Asia for 2018 [J].Weather and Forecasting,36:39-51. doi:10.1175/WAF−D−20−0096.1

[48] Zhu Z, Wu J, Huang H. 2024. The influence of 10—30—day boreal summer intraseasonal oscillation on the extended—range forecast skill of extreme rainfall over southern China [J].Climate Dynamics,62:69-86. doi:10.1007/s00382−023−06900−w

基金资助

全国暴雨研究开放基金项目(BYKJ2024Q17，BYKJ2025Q18，BYKJ2025Q25)

国家自然科学基金项目(42205162，42275164)

江苏省自然科学青年基金项目(BK20231110)