China runoff-field forecasting based on cross-scale gating and basin-topology attention

Aiju Li, Xiang He, Kai Yang, Weiya Ge, Jian Hua · 2026

Read China runoff-field forecasting based on cross-scale gating and basin-topology attention by Aiju Li, Xiang He, Kai Yang, Weiya Ge, Jian Hua (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/plos-pone-0350218.txt.

Description

Aiju Li and colleagues built a forecasting system for river runoff across China, the numbers that decide flood warnings and reservoir releases. Rivers remember at several speeds at once, quick storm pulses and slow seasonal swells, and water flows downhill through connected basins, so a model must respect both clocks and maps. Their design fuses time scales with a gating step and teaches spatial attention the true upstream-downstream links.

Tested on a national gridded benchmark built from a public global flood dataset, the method beat existing approaches overall and kept its forecasts spatially consistent with the river network. For agencies watching monsoon floods, forecasts that honor basin shape, not just local history, could mean earlier and better-placed warnings. The work is a modeling study, so its proof rests on benchmark data rather than a live season.

Catalog record

CREATORAiju Li, Xiang He, Kai Yang, Weiya Ge, Jian Hua
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE84070
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0350218

Browse the full library · About the Tower of Babel