Ning et al. (2026) From empirical to physical constraints: Revisiting the structure of monthly water balance models with global evaluation

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-01-04
• Authors: Zhongrui Ning, Jianyun Zhang, Nan Wu, Xie Kang, Yuli Ruan, Yueyang Wang, Cuishan Liu, Guoqing Wang
• DOI: 10.1016/j.jhydrol.2026.134916

Research Groups

• The National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing 210098, China
• Yangtze Institute for Conservation and Development, Nanjing 210098, China
• Research Center for Climate Change, Ministry of Water Resources, Nanjing 210029, China
• Nanjing Hydraulic Research Institute, Nanjing 210029, China

Short Summary

This study revises the RCCCWBM by incorporating physical constraints (Budyko-based hydrothermal balance and storage capacity curve) to develop two improved monthly water balance models (RM-M and RM-G). Globally evaluated across 2,003 catchments, the revised models significantly enhance parameter interpretability, simulation accuracy, and cross-regional robustness, particularly in arid and semi-arid regions.

Objective

• To revise the conceptual structure of monthly water balance models (MWBMs) by incorporating physical constraints to enhance parameter interpretability and cross-regional robustness.
• To develop improved lumped (RM-M) and distributed (RM-G) versions of the RCCCWBM without increasing tunable parameters.
• To globally evaluate the performance, parameter interpretability, and robustness of the revised models across a large sample of catchments.

Study Configuration

• Spatial Scale: Global, across 2,003 catchments worldwide.
• Temporal Scale: Monthly water balance, assessing long-term variations.

Methodology and Data

• Models used: RCCCWBM (original), RM-M (revised lumped version), RM-G (revised distributed version).
• Data sources: Large-sample assessment across 2,003 catchments worldwide (implies observational hydrological and meteorological data for these catchments).

Main Results

• The revised models (RM-M and RM-G) substantially reduced the systematic underestimation of evapotranspiration (mean absolute error decreased from 14.4% to 5.4%).
• Runoff simulation accuracy and robustness were improved (median KGE increased from 0.60 to 0.68–0.69).
• Long-term water balance bias was controlled within 10%.
• Sensitivity analysis and SHAP analysis revealed a clearer hierarchy of parameter sensitivity, with evapotranspiration efficiency, storage capacity, and underground runoff coefficient as key controls.
• Model parameters exhibited stable, physically consistent relationships with catchment climate, vegetation, topography, and soil attributes.
• The improved structure offers significant advantages in arid and semi-arid regions.
• Critical thresholds of catchment attributes influencing model performance were identified.

Contributions

• Enhanced physical interpretability, simulation accuracy, and cross-regional robustness of monthly water balance models (MWBMs).
• Provided methodological and insight advances for global hydrological modeling.
• Developed valuable tools for large-scale water resources assessment and climate change adaptation studies.
• Introduced a Budyko-based hydrothermal balance framework and a storage capacity curve to reconstruct the runoff–evapotranspiration chain in MWBMs without increasing model complexity.

Funding

• Not specified in the provided text.

Citation

@article{Ning2026From,
  author = {Ning, Zhongrui and Zhang, Jianyun and Wu, Nan and Kang, Xie and Ruan, Yuli and Wang, Yueyang and Liu, Cuishan and Wang, Guoqing},
  title = {From empirical to physical constraints: Revisiting the structure of monthly water balance models with global evaluation},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.134916},
  url = {https://doi.org/10.1016/j.jhydrol.2026.134916}
}