Assessment of Forecast Gains from Global Sea Surface Temperature for Monthly Runoff in Typical Upstream River Basins of China

ZHANG Zewei; ZHU Weike; ZHANG Duanhu; HU Maochuan; PAN Hongzhong; CHEN Peng

doi:null

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Assessment of Forecast Gains from Global Sea Surface Temperature for Monthly Runoff in Typical Upstream River Basins of China

更新时间：2026-06-22

- Assessment of Forecast Gains from Global Sea Surface Temperature for Monthly Runoff in Typical Upstream River Basins of China
- PEARL RIVER Pages: 1-14(2026)
- 作者机构：
  
  1.长江大学资源与环境学院，湖北武汉 430100
  2.广东省水文局韶关水文分局，广东韶关 512028
  3.中山大学水资源与环境研究中心，广东广州 510275
- 作者简介：
- 基金信息：
- DOI：
  CLC： TV21
- Revised：2026-05-29，
  
  Accepted：01 June 2026，
  
  Online First：22 June 2026，
- 稿件说明：
移动端阅览
张泽威，朱维科，张端虎等.全球海表温度对中国典型流域上游月径流的预报增益评估［J］.人民珠江，DOI：10.3969/j.issn.1001-9235.XXXX.XX.001.

ZHANG Zewei,ZHU Weike,ZHANG Duanhu,et al.Assessment of Forecast Gains from Global Sea Surface Temperature for Monthly Runoff in Typical Upstream River Basins of China[J].PEARL RIVER,
张泽威，朱维科，张端虎等.全球海表温度对中国典型流域上游月径流的预报增益评估［J］.人民珠江，DOI：10.3969/j.issn.1001-9235.XXXX.XX.001. DOI：

ZHANG Zewei,ZHU Weike,ZHANG Duanhu,et al.Assessment of Forecast Gains from Global Sea Surface Temperature for Monthly Runoff in Typical Upstream River Basins of China[J].PEARL RIVER, DOI：10.3969/j.issn.1001-9235...001.

摘要

针对现有海表温度—径流研究多局限于单流域、固定海区和样本内关系识别的问题，选取珠江北江上游韶关、黄河上游唐乃亥、长江上游寸滩为对象，构建全球格点相关分析、场显著性检验、稳定海区提取、海温指数构建和样本外检验的框架开展跨流域研究。结果表明：稳定海区表现出跨海盆、多中心分布的特征。韶关、寸滩和唐乃亥分别有9、8和6个月通过场显著性检验，稳定预见期主要集中于冬春季。引入海温因子后，相关的预报增益整体表现为冬春季较强、夏秋季较弱。韶关样本外预报增益提升最显著、寸滩次之、唐乃亥较弱且季节性依赖更明显。海表温度能够提升月径流预报增益，但其增益强度具有显著流域差异和季节差异，所建立的样本外评估框架可为跨流域月径流预报及海温因子筛选提供参考。

Abstract

Existing studies on sea surface temperature (SST)–runoff relationships have often focused on individual basins

predefined oceanic indices

or in-sample relationship identification. This study evaluates whether global SST provides transferable and operationally meaningful forecast gains for monthly runoff in typical upstream river basins in China

and clarifies the basin and seasonal differences in the SST contribution.Three upstream hydrological stations were selected: Shaoguan in the upper Beijiang River of the Pearl River Basin

Cuntan in the upper Yangtze River

and Tangnaihai in the upper Yellow River. Monthly runoff observations from 2003–2022 were combined with ERSST v5 SST fields for 2002–2022 and ERA5 precipitation and temperature data. For each station

runoff month

and lead time

global grid-point Pearson correlations between runoff anomalies and antecedent SST anomalies were calculated. A field-significance permutation test with 1

000 randomizations was used to control the multiple-testing effect. According to the number of significant lead times

stable ocean regions were identified using frequency statistics

bootstrap stability screening

or nested out-of-sample selection

followed by connected-component and size constraints. Area-weighted SST indices and dipole indices were constructed

and six regression models using antecedent runoff

precipitation

temperature

and SST predictors were compared using leave-one-year-out cross-validation. Forecast performance was evaluated using R²

Nash–Sutcliffe efficiency (NSE)

and root mean square error (RMSE); gains were expressed as absolute differences against the corresponding baseline models. Stable SST-related ocean regions were not confined to a single conventional ENSO area

but exhibited multi-center and trans-basin patterns involving the western Pacific warm-pool vicinity

the Indian Ocean sector

the high-latitude North Atlantic–Arctic Ocean

and mid- to high-latitude regions of the Southern Hemisphere. Shaoguan

Guangdong

and Tangnaihai had 9

and 6 runoff months

respectively

passing the field-significance test. Detectable and stable SST signals were concentrated mainly in winter and spring; no stable lead-time combination was detected for all three stations from July to September. With SST predictors included

simultaneous improvement in R²

NSE

and RMSE occurred in 7 months at Shaoguan

6 months at Cuntan

and 4 months at Tangnaihai. Forecast gains were generally stronger during winter and spring and weaker during summer and autumn. Shaoguan showed the most pronounced out-of-sample improvement

Cuntan showed moderate improvement

and Tangnaihai showed weaker gains with stronger seasonal dependence. These contrasts suggest that SST signals are more readily converted into predictive information in humid monsoon basins through large-scale circulation and moisture-transport pathways

whereas their influence on alpine inland runoff may be weakened by snowmelt

frozen soil

terrain effects

and local land–atmosphere interactions.Global SST contains useful supplementary information for monthly runoff forecasting in upstream basins

but its value is not spatially or seasonally uniform and is not determined simply by the number of significant lead times. The proposed framework integrates field-significance control

stable-region extraction

SST-index construction

and strict out-of-sample validation

thereby reducing the risk of chance correlation and information leakage. It provides a reproducible approach for cross-basin SST predictor screening and for comparing runoff forecast gains under different climatic and hydrological settings.

关键词

Keywords

references

孙周亮，刘艳丽，张建云，等 . 中长期径流预报研究进展与展望［J］. 水资源保护， 2023 ， 39 （ 2 ）： 136 - 144+223 .

Sun Zhouliang ， Liu Yanli ， Zhang Jianyun ， et al . Research progress and prospect of mid-long term runoff prediction ［J］. Water Resources Protection ， 2023 ， 39 （ 2 ）： 136 – 144，223 . （in Chinese）

Hu Feichi ， Yang Qinli ， Yang Junran ， et al . Incorporating multiple grid-based data in CNN-LSTM hybrid model for daily runoff prediction in the source region of the Yellow River basin ［J］. Journal of Hydrology： Regional Studies ， 2024 ， 51 ： 101652 .

李继清，谢宇韬，徐学军，等 . 基于RF-Informer模型的月径流遥相关预报［J］. 水资源保护， 2025 ， 41 （ 3 ）： 39 - 45 . doi: 10.3880/j.issn.1004-6933.2025.03.005 http://dx.doi.org/10.3880/j.issn.1004-6933.2025.03.005

Li Jiqing ， Xie Yutao ， Xu Xuejun ， et al . Monthly runoff teleconnection forecasting based on the RF-Informer model ［J］. Water Resources Protection ， 2025 ， 41 （ 3 ）： 39 - 45 . （in Chinese）

张佳悦，周毓彦，鲁帆，等 . 气候变化下长江黄河上游源区径流丰枯遭遇研究［J］. 水利水电技术（中英文）， 2025 ， 56 （ 11 ）： 32 - 47 .

Zhang Jiayue ， Zhou Yuyan ， Lu Fan ， et al . Study on runoff abundance and depletion encounters in the upper reaches of the Yangtze and Yellow Rivers under climate change ［J］. Water Resources and Hydropower Engineering （Chinese and English）， 2025 ， 56 （ 11 ）： 32 - 47 . （in Chinese）

Chen C J ， Georgakakos P A . Hydro-climatic forecasting using sea surface temperatures：methodology and application for the southeast US ［J］. Climate Dynamics ， 2014 ， 42 （ 11/12 ）： 2955 - 2982 .

Alekseev G ， Soldatenko S ， Glok N ， et al . Tropical Sea Surface Temperature and Sea Level as Candidate Predictors for Long-Range Weather and Climate Forecasting in Mid-to-High Latitudes ［J］. Climate ， 2025 ， 13 （ 5 ）： 84 - 95 .

Canchala T ， Escobar C Y ， Morales A W ， et al . Seasonal influence of tropical Pacific and Atlantic sea surface temperature on streamflow variability in the patia river basin ［J］. Theoretical and Applied Climatology ， 2024 ， 155 （ 6 ）： 5417 - 5431 .

彭梅香，葛朝霞，王怀柏 . 黄河上游径流与太平洋海温场关系及其预测应用［J］. 水科学进展， 2000 （ 3 ）： 272 - 276 .

Peng Meixiang ， Ge Chaoxia ， Wang Huaibai . Relationships between the Runoff in the Upper Yellow River and SST Field in the Pacific and Its Application ［J］. Advances in Water Science ， 2000 （ 3 ）： 272 - 276 . （in Chinese）

庞轶舒，秦宁生，罗玉，等 . 秋季热带印度洋偶极子年际振荡对长江上游径流量多寡的影响分析［J］. 高原气象， 2021 ， 40 （ 2 ）： 353 - 366 .

Pang Yishu ， Qin Ningsheng ， Luo Yu ， et al . Analysis on Effect of Tropical Indian Ocean Dipole Inter-Annual Oscillation on Annual Variation of Runoff in the Upper Reaches of Yangtze River ［J］. Plateau Meteorology ， 2021 ， 40 （ 2 ）： 353 - 366 . （in Chinese）

冯鑫，刘艳菊，童宏福，等 . 基于海表温度和随机森林的珠江流域干旱预报模型研究［J］. 人民珠江， 2024 ， 45 （ 5 ）： 96 - 102 .

Feng Xin ， Liu Yanju ， Tong Hongfu ， et al . Drought Prediction Model of Pearl River Basin Based on SST and Machine Learning ［J］. Pearl River ， 2024 ， 45 （ 5 ）： 96 - 102 . （in Chinese）

Umirbekov A ， Peña-Guerrero M D ， Didovets I ， et al . The value of hydroclimatic teleconnections for snow-based seasonal streamflow forecasting in central asia ［J］. Hydrology and Earth System Sciences ， 2025 ， 29 （ 14 ）： 3055 - 3071 .

Yin Xin ， Wu Xiaohui ， Niu Hailin ， et al . Inter-Model Spread in Representing the Impacts of ENSO on the South China Spring Rainfall in CMIP6 Models ［J］. Atmosphere ， 2024 ， 15 （ 10 ）： 1199 .

Chang Jianbo ， Yan Baowei ， Sun Mingbo ， et al . Integrated forecasting of monthly runoff considering the combined effects of teleconnection factors ［J］. Journal of Hydrology： Regional Studies ， 2025 ， 58 ： 102206 .

魏显贵，潘红忠，王现勋，等 . 基于海洋表面温度的龙羊峡入库径流中长期预报［J］. 人民黄河， 2021 ， 43 （ 1 ）： 29 - 34 .

Wei Xiangui ， Pan Hongzhong ， Wang Xianxun ， et al . Mid-Long Term Inflow Forecast of Longyangxia Reservoir in Upstream of Yellow River Based on Sea Surface Temperature ［J］. Yellow River ， 2021 ， 43 （ 1 ）： 29 - 34 . （in Chinese）

陈丽娟，王悦颖，李维京，等 . 海温异常对中国汛期降水的影响及预测应用［J］. 应用气象学报， 2024 ， 35 （ 2 ）： 129 - 141 . doi: 10.11898/1001-7313.20240201 http://dx.doi.org/10.11898/1001-7313.20240201

Chen Lijuan ， Wang Yueying ， Li Weijing ， et al . Review of the Influence and Application of SST Anomaly to Flood Season Precipitation in China ［J］. Journal of Applied Meteorological Science ， 2024 ， 35 （ 2 ）： 129 - 141 . （in Chinese）

Wu Chuanhao ， Huang Guoru ， Liu Zhiyu ， et al . Prediction of extreme floods based on CMIP5 climate models： a case study in the beijiang river basin， south China ［J］. Hydrology and Earth System Sciences ， 2015 ， 19 （ 3 ）： 1385 - 1399 .

Gao Chao ， Su Buda ， Krysanova V ， et al . A 439-year simulated daily discharge dataset （1861–2299） for the upper yangtze river， china ［J］. Earth System Science Data ， 2020 ， 12 （ 1 ）： 387 - 402 .

Ji Guangxing ， Wu Leying ， Wang Liaodong ， et al . Attribution analysis of seasonal runoff in the source region of the Yellow River using seasonal budyko hypothesis ［J］. Land ， 2021 ， 10 （ 5 ）： 542 - 555 .

Huang Boying ， Thorne P W ， Banzon V F ， et al . Extended Reconstructed Sea Surface Temperature， Version 5 （ERSSTv5）： Upgrades， Validations， and Intercomparisons ［J］. Journal of Climate ， 2017 ， 30 （ 20 ）： 8179 - 8205 .

Huang Shaozhe ， Yu Lei ， Luo Wenbing ， et al . Runoff Prediction of Irrigated Paddy Areas in Southern China Based on EEMD-LSTM Model ［J］. Water ， 2023 ， 15 （ 9 ）： 1704 - 1715 .

黄翀，张强，肖名忠 . ENSO、NAO、IOD和PDO对珠江流域降水的影响研究［J］. 中山大学学报（自然科学版）， 2016 ， 55 （ 2 ）： 134 - 142 .

Huang Chong ， Zhang Qiang ， Xiao Mingzhong . Influences of ENSO，NAO，IOD and PDO on precipitation regimes in the Pearl River basin ［J］. Acta Scientiarum Naturalium Universitatis Sunyatseni ， 2016 ， 55 （ 2 ）： 134 - 142 . （in Chinese）

马有绚，张调风，温婷婷，等 . 黄河上游地区径流量的转折性变化及其成因［J］. 气象科技进展， 2023 ， 13 （ 4 ）： 87 - 92，113 .

Ma Youxuan ， Zhang Diaofeng ， Wen Tingting ， et al . The Transitional Change of Runoff in the Upper Reaches of Yellow River and Its Causal Analysis ［J］. Advances in Meteorological Science and Technology ， 2023 ， 13 （ 4 ）： 87 - 92，113 . （in Chinese）

李福威，包爱美，疏杏胜，等 . 基于水文-气象因子的综合多模型长期径流预报研究［J］. 中国农村水利水电， 2022 （ 11 ）： 6 - 12 .

Li Fuwei ， Bao Aimei ， Shu Xingsheng ， et al . Long-term Runoff Prediction for Huanren River Basin Based on Multiple Models ［J］. China Rural Water and Hydropower ， 2022 （ 11 ）： 6 - 12 . （in Chinese）

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Drought Prediction Model of Pearl River Basin Based on SST and Machine Learning

Application of Grey Self-Memory Model Based on Smoothing Method Without Seasonal Fluctuations on Monthly Runoff Forecast

Research on Monthly Runoff Forecast in Beijiang River Basin Based on Multi-model Ensemble Method

Application of Hydrological Models Coupling Interpretable Machine Learning in Flood Simulation Error Correction

Research on Dam Deformation Prediction Model Based on STL-Informer

Related Author

FENG Xin

LIU Yanju

TONG Hongfu

QIAN Shuni

WAN Yang

YANG Haiyan

TANG Jingjing

ZHONG Yixuan

Related Institution

School of Civil Engineering and Transportation, The South China University of Technology

Comprehensive Technology Center of Pearl River Water Resources Commission of Ministry of Water Resources

Qingyuan Hydrology Sub-Bureau of Guangdong Province

Guangzhou Water Planning Survey and Design Institute Co., Ltd.

HNAC Science & Technology Co. ,Ltd.

AI问答

⁰