.png){kind=icon}
.png){kind=icon}
Return
Improving watershed-scale daily nutrient simulation using a process-model-informed graph attention network with multi-source data integration
.png){kind=icon}
PRE
AI
DOI：10.1016/j.watres.2026.125532.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• A Process-model-informed Graph Attention Network integrates satellite-derived water quality data • Bridging data gaps to reconstruct daily TN fields improves R2 at sparsely observed reaches from 0.14 to 0.59 • Enhancing high-concentration event detection and priority management areas identification • Gains from remote sensing and similarity-guided graph attention are greater in downstream.
Keywords:
water quality
nutrient simulation
graph attention network
remote sensing
watershed-scale modeling
Journal
IF:
12.4
Papers:
3.0W
Citations:
15.7W