Xu et al. (2026) Improvement and application of the Green-Ampt model integrating mechanism correction and intelligent parameter prediction

Identification

• Journal: Advances in Water Resources
• Year: 2026
• Date: 2026-07-27
• Authors: Longsheng Xu, Yuqin Gao, Ming Wu, Chenyu Yuan, Zhihong Liang
• DOI: 10.1016/j.advwatres.2026.105439

Research Groups

College of Water Conservancy and Hydropower Engineering, Hohai University

Short Summary

This study presents an integrated infiltration modeling framework that improves the Green-Ampt model by introducing a wetting front suction correction factor and a saturated hydraulic conductivity correction factor, enabling decoupled characterization of gravity-driven, preferential-flow, and capillary-fringe controls. The modified model reduces cumulative infiltration RMSE by an order of magnitude relative to both the classical GA and approximate Richards equation solutions.

Objective

• Investigate the limitations of the classical Green-Ampt (GA) model in capturing nonlinear decay of infiltration rates under heterogeneous soil conditions.
• Develop a modified GA model that integrates mechanism correction and intelligent parameter prediction.

Study Configuration

• Spatial Scale: Watershed scale, with application to the Shaying River Basin during the 2010 flood season.
• Temporal Scale: Daily time steps, with focus on short-term infiltration dynamics.

Methodology and Data

• Models used:
  • Dual-factor modified Green-Ampt model
  • Random Forest regression for predictive parameterization
  • Genetic Algorithm, Particle Swarm Optimization (PSO), and Differential Evolution (DE) for optimization
• Data sources:
  • HYDRUS-1D dataset for benchmarking
  • Shaying River Basin data for distributed hydrological simulations

Main Results

• The modified model reduces cumulative infiltration RMSE by an order of magnitude relative to both the classical GA and approximate Richards equation solutions.
• Applied to distributed hydrological simulations, the modified model corrects the systematic "peak-clipping" bias in runoff generation.

Contributions

• Original value: The study presents a novel integrated infiltration modeling framework that improves the Green-Ampt model by introducing mechanism correction and intelligent parameter prediction.
• Contribution to existing literature: The study addresses the limitations of traditional hydrological models, including the Green-Ampt module of HEC-HMS, the Green-Ampt Mein-Larson option of SWAT, and the infiltration scheme of the Variable Infiltration Capacity (VIC) model.

Funding

• This research was funded by the National Natural Science Foundation of China (Grant No. 51979215).

Citation

@article{Xu2026Improvement,
  author = {Xu, Longsheng and Gao, Yuqin and Wu, Ming and Yuan, Chenyu and Liang, Zhihong},
  title = {Improvement and application of the Green-Ampt model integrating mechanism correction and intelligent parameter prediction},
  journal = {Advances in Water Resources},
  year = {2026},
  doi = {10.1016/j.advwatres.2026.105439},
  url = {https://doi.org/10.1016/j.advwatres.2026.105439}
}