Page 27 - 2025年第56卷第9期

/
［ 76］ CHEN R T Q，RUBANOVA Y，BETTENCOURT J，et al. Neural ordinary differential equations［C］/ Proceedings
of the 32nd International Conference on Neural Information Processing System. Red Hook，NY，US：Curron Associ⁃
ates InC. 2018.
［ 77］ SONG W， JIANG S， CAMPS-VALLS G， et al. Towards data-driven discovery of governing equations in geosci⁃
.
ences［J］ Communications Earth & Environment，2024，5（1）：589.
［ 78］ WAGEMANN J，FIERLI F，MANTOVANI S，et al. Five guiding principles to make jupyter notebooks fit for earth
observation data education［J］ Remote Sensing，2022，14（14）：3359.
.
［ 79］ WILLETT D S， BRANNOCK J， DISSEN J， et al. NOAA open data dissemination： petabyte-scale earth system
.
data in the cloud［J］ Science Advances，2023，9（38）：eadh0032.
［ 80］ GORELICK N， HANCHER M， DIXON M， et al. Google earth engine： planetary-scale geospatial analysis for
.
everyone［J］ Remote Sensing of Environment，2017，202：18-27.
［ 81］ MILES B. Small-scale residential stormwater management in urbanized watersheds：a geoinformatics-driven ecohy⁃
drology modeling approach ［D］ North Carolina，United States：University of North Carolina at Chapel Hill Gradu⁃
.
ate School，2015.
［ 82］ PECKHAM S D，HUTTON E W H，NORRIS B. A component-based approach to integrated modeling in the geosci⁃
.
ences：The design of CSDMS［J］ Computers & Geosciences，2013，53：3-12.
［ 83］ KNOBEN W J M， CLARK M P， BALES J， et al. Community workflows to advance reproducibility in hydrologic
modeling： separating model-agnostic and model-specific configuration steps in applications of large-domain hydro⁃
.
logic models［J］ Water Resources Research，2022，58（11）：e2021WR031753.
［ 84］ TUFANO M，AGARWAL A，JANG J，et al. AutoDev：automated AI-driven development［J］ arXiv，2024. doi：
.
10. 48550/arXiv. 2403. 08299.
.
［ 85］ LU C，LU C，LANGE R T，et al. The AI scientist：towards fully automated open-ended scientific discovery［J］
arXiv，2024. doi：10. 48550/arXiv. 2408. 06292.
［ 86］ 冯钧，吕志鹏，范振东，等 . 基于大语言模型辅助的防洪调度规则标签设计方法［J］ 水利学报，2024，55
.
（8）：920-930.

Exploration and consideration of AI-driven hydrological forecasting and reservoir operation

1 1 1 1，2 1 1

OUYANG Wenyu ，ZHANG Chi ，MA Haoran ，YE Lei ，WANG Ze ，LÜ Heng

（1. Department of Hydraulic Engineering Dalian University of Technology，Dalian 116024，China；

2. NingBo Institute of Dalian University of Technology，Ningbo 315016，China）

Abstract：The rapid development of artificial intelligence （AI）technology has brought new opportunities to hydro⁃
logical forecasting and reservoir operation research. This paper focuses on the mining of watershed hydrological pro⁃
cess regularities and the decision-making in reservoir flood control scheduling as entry points. It provides an overview
of AI application case studies，analyzing the advantages of AI in identifying complex common patterns in watershed
hydrology and in learning from real-world reservoir operation experiences to support future decision-making. Further⁃
s
more，he paper discusses current challenges in AI applications，uch as the insufficiency of data to describe common
t
a
hydrological patterns and the inadequate digital representation of real-world conditions in scheduling. Based on this，
pathway for the collaborative development of data，algorithm and computing power is proposed to enhance data com⁃

pleteness，strengthen algorithm capabilities，and promote practical engineering applications，highlighting the integra⁃

tion of remote sensing big data with ground observation systems，he deep integration of domain knowledge with AI
t
algorithms，and the research and application of platform-based software products. This aims to provide references for

research and practice in related fields.

Keywords：artificial intelligence；hydrological forecasting；reservoir operation；algorithm；computing power

（责任编辑：耿庆斋）
— 1131 —