Articles | Volume 29, issue 18
https://doi.org/10.5194/hess-29-4457-2025
https://doi.org/10.5194/hess-29-4457-2025
Research article
 | 
17 Sep 2025
Research article |  | 17 Sep 2025

Interrogating process deficiencies in large-scale hydrologic models with interpretable machine learning

Admin Husic, John Hammond, Adam N. Price, and Joshua K. Roundy

Viewed

Total article views: 5,534 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
4,628 764 142 5,534 353 150 203
  • HTML: 4,628
  • PDF: 764
  • XML: 142
  • Total: 5,534
  • Supplement: 353
  • BibTeX: 150
  • EndNote: 203
Views and downloads (calculated since 13 Nov 2024)
Cumulative views and downloads (calculated since 13 Nov 2024)

Viewed (geographical distribution)

Total article views: 5,534 (including HTML, PDF, and XML) Thereof 5,534 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 

Cited

Saved (final revised paper)

Latest update: 19 May 2026
Download
Short summary
We used interpretable machine learning to evaluate the accuracy of two continental-scale hydrologic models. We analyzed a suite of catchment attributes and found that soil water content had the biggest impact on model performance, especially in dry areas. Key thresholds for variables like precipitation and road density were identified, which could guide future improvements in these models. Our findings highlight the potential of data-driven methods to inform process-based models.
Share