Bitew et al. (2026) HYDRO-XM: a computationally efficient, externally coupled, and flexible watershed model

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-07-02
• Authors: Menberu Meles Bitew, A. T. M. Sakiur Rahman, Jiřı́ Šimůnek, Hoori Ajami, Carl L. Unkrich, Juan Sebastian Acero Triana, Jason Goulding, Scott A. Bradford
• DOI: 10.1016/j.jhydrol.2026.135954

Research Groups

• USDA-ARS, Sustainable Agricultural Water Systems, Davis, CA, USA
• Department of Environmental Sciences, University of California, Riverside, CA 92521, USA
• USDA-ARS, Southwest Watershed Research Center, Tucson, AZ, USA

Short Summary

HYDRO-XM is a computationally efficient, externally coupled, and flexible watershed model that integrates surface flow, vadose-zone processes, and groundwater dynamics. It enables stable, continuous simulation of overland flow, channel flow, infiltration, soil moisture redistribution, and groundwater dynamics across full watershed domains.

Objective

• Develop and implement an externally coupled HYDRO-XM modeling framework that integrates adaptive temporal synchronization, dimensionality reduction, parallelized execution, KINEROS2 state serialization, dynamic time stepping, adaptive boundary-condition switching, and multi-scale coupling.
• Evaluate model performance against benchmark cases and ParFlow simulations and assess applicability across homogeneous, heterogeneous, and structurally complex hillslopes.

Study Configuration

• Spatial Scale: Watershed scale, with domain discretization and definition of exchange intervals at each coupling step.
• Temporal Scale: Nested time stepping strategy, with ΔtMF5 > ΔtH1D > ΔtK2 (Eq. (4)).

Methodology and Data

• Models used:
  • KINEROS2 (K2) for overland flow and erosion
  • HYDRUS-1D (H1D) for variably saturated flow in the unsaturated zone
  • MODFLOW-5 (MF5) for saturated groundwater flow
• Data sources: Satellite, observation, reanalysis data

Main Results

• The model showed strong agreement with ParFlow outputs and field observations.
• Performance metrics ranged from NSE = 0.62–0.98, KGE = 0.24–0.93, and R2 = 0.85–0.99 compared to observed streamflow.
• HYDRO-XM ran up to 28.4 times faster than a comparable ParFlow configuration.

Contributions

• The framework provides a tractable tool for evaluating managed aquifer recharge, predicting hydrologic responses, and quantifying watershed-scale impacts of land-use change and climate variability.
• It enables emergent watershed-scale feedback representation, including spatial redistribution of recharge, temporally variable stream–aquifer interactions, groundwater-mediated soil moisture dynamics, and transient transitions between infiltration-excess and saturation-excess runoff generation under evolving hydrologic conditions.

Funding

• This research was funded by the USDA-ARS and the University of California, Riverside.

Citation

@article{Bitew2026HYDROXM,
  author = {Bitew, Menberu Meles and Rahman, A. T. M. Sakiur and Šimůnek, Jiřı́ and Ajami, Hoori and Unkrich, Carl L. and Triana, Juan Sebastian Acero and Goulding, Jason and Bradford, Scott A.},
  title = {HYDRO-XM: a computationally efficient, externally coupled, and flexible watershed model},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.135954},
  url = {https://doi.org/10.1016/j.jhydrol.2026.135954}
}