Global River Forecasting with a Topology-Informed AI Foundation Model
Hancheng Ren, Gang Zhao, Shuo Wang, Louise Slater, Dai Yamazaki, Shu Liu, Jingfang Fan, Shibo Cui, Ziming Yu, Shengyu Kang, Depeng Zuo, Dingzhi Peng, Zongxue Xu, Bo Pang · 2026
Read Global River Forecasting with a Topology-Informed AI Foundation Model by Hancheng Ren, Gang Zhao, Shuo Wang, Louise Slater, Dai Yamazaki, Shu Liu, Jingfang Fan, Shibo Cui, Ziming Yu, Shengyu Kang, Depeng Zuo, Dingzhi Peng, Zongxue Xu, Bo Pang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-22293.txt.
Description
Reliable river forecasts protect water, food, and energy, but sparse gauges have confined machine learning to isolated basins rather than connected river systems. Hancheng Ren and fourteen co-authors introduce GraphRiverCast, a topology-informed foundation model for reach-level prediction across the global river network. It exploits river links and energetics, training on sparse gauges to predict daily water flows with no prior river conditions.
The model outperforms strong physics-based and machine-learning rivals. Tests of what drives it show it rebuilds river states from topology and forcing, learning the flows themselves rather than leaning on time recurrence like standard autoregressive models. Water managers gain a global, connected forecasting tool that inherits rivers' self-correcting behavior.
Catalog record
| CREATOR | Hancheng Ren, Gang Zhao, Shuo Wang, Louise Slater, Dai Yamazaki, Shu Liu, Jingfang Fan, Shibo Cui, Ziming Yu, Shengyu Kang, Depeng Zuo, Dingzhi Peng, Zongxue Xu, Bo Pang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-22293