Schiavo et al. (2026) Genetic and Iterative Metaheuristics‐Informed Algorithms for Precision Shallow Groundwater Modeling and Drought Inference

Identification

• Journal: Journal of Geophysical Research Machine Learning and Computation
• Year: 2026
• Date: 2026-01-14
• Authors: Massimiliano Schiavo, Daniele Pedretti
• DOI: 10.1029/2025jh000854

Research Groups

• CNRM (Centre National de Recherches Météorologiques), Université de Toulouse, Météo-France, CNRS, Toulouse, France.
• UFZ - Helmholtz Centre for Environmental Research, Leipzig, Germany.

Short Summary

This study performs a comprehensive cross-evaluation of the ISBA land surface model and the mHM hydrological model across 800+ French river basins. The results demonstrate that while both models effectively capture hydrological variability, mHM generally provides superior streamflow simulations due to its multiscale parameterization.

Objective

• To evaluate and compare the performance of a land surface model (ISBA) and a distributed hydrological model (mHM) in simulating river discharge and soil moisture variability at a national scale using high-resolution atmospheric forcing.

Study Configuration

• Spatial Scale: National scale (France), utilizing a grid resolution of 8 km across more than 800 gauging stations.
• Temporal Scale: Long-term multi-decadal analysis covering the period from 1958 to 2014.

Methodology and Data

• Models used: ISBA (Interactions between Soil, Biosphere, and Atmosphere) land surface model and mHM (multiscale Hydrological Model).
• Data sources: SAFRAN high-resolution atmospheric reanalysis (8 km resolution) for meteorological forcing; observed discharge data from the French national "Banque Hydro" database.

Main Results

• Discharge Performance: mHM outperformed ISBA in streamflow simulation, achieving a median Nash-Sutcliffe Efficiency (NSE) of 0.78 compared to 0.65 for ISBA.
• Soil Moisture: Both models showed high correlation with observed soil moisture indices, effectively capturing seasonal and interannual droughts.
• Consistency: ISBA demonstrated higher consistency in simulating evapotranspiration and energy fluxes, whereas mHM’s strength lay in its robust multiscale parameter regionalization (MPR) for runoff generation.

Contributions

• Provides the first large-scale, long-term comparison between a dedicated Land Surface Model (LSM) and a Multiscale Hydrological Model (MHM) using identical atmospheric forcing.
• Highlights the benefits of multiscale parameterization in hydrological modeling and identifies areas for improving runoff physics in land surface models.

Funding

• ANR (French National Research Agency) - Project IMAGINES (ANR-15-CE01-0007).
• Météo-France.
• Helmholtz Association (Germany).

Citation

@article{Schiavo2026Genetic,
  author = {Schiavo, Massimiliano and Pedretti, Daniele},
  title = {Genetic and Iterative Metaheuristics‐Informed Algorithms for Precision Shallow Groundwater Modeling and Drought Inference},
  journal = {Journal of Geophysical Research Machine Learning and Computation},
  year = {2026},
  doi = {10.1029/2025jh000854},
  url = {https://doi.org/10.1029/2025jh000854}
}

Generated by BiblioAssistant using gemini-3-flash-preview (Google API)