Sivelle et al. (2025) Investigating hydrological modeling uncertainties in the Mediterranean region by combining precipitation and soil moisture products

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Authors: Vianney Sivelle, Christian Massari, Yves Tramblay, Paolo Filippucci, Hamouda Dakhlaoui, Pere Quintana‐Seguí, Roger Clavera-Gispert, Hamouda Boutaghane, Tayeb Boulmaiz, Raphael Quast, Mariette Vreugdenhill
• DOI: 10.1016/j.ejrh.2025.103015

Research Groups

• Research Institute for Geo-Hydrological Protection, National Research Council (CNR), Perugia, Italy.
• HSM (HydroSciences Montpellier), Univ Montpellier, CNRS, IRD, Montpellier, France.
• Espace-Dev, Univ. Montpellier, IRD, Montpellier, France.
• LMHE, Ecole Nationale d’Ingénieurs de Tunis, Université Tunis El Manar, Tunisia.
• Observatori de l’Ebre (OE), Ramon Llull University, CSIC, Spain.
• Department of Geodesy and Geoinformation, TU Wien, Vienna, Austria.
• Laboratory of Hydraulics and Hydraulic Constructions, Badji Mokhtar University, Algeria.

Short Summary

This study evaluates the impact of various satellite-derived precipitation and soil moisture products on hydrological modeling performance across five Mediterranean catchments. It demonstrates that while gauge-based data remains superior, merged precipitation products and Sentinel-1 soil moisture data significantly reduce predictive uncertainties, especially when forcing models with satellite-only rainfall.

Objective

• To investigate how the combination of different precipitation forcing agents and soil moisture observation products affects streamflow simulation and parametric uncertainty in mesoscale Mediterranean catchments.

Study Configuration

• Spatial Scale: Five mesoscale catchments in the Mediterranean region: Arga (Spain, 2755 $km^2$), Herault (France, 952 $km^2$), Tiber (Italy, 929 $km^2$), Medjerda (Tunisia, 2414 $km^2$), and Tafna (Algeria, 324 $km^2$).
• Temporal Scale: Calibration and analysis period from 01/08/2015 to 31/12/2021 at a daily time step.

Methodology and Data

• Models used: Four lumped parameter hydrological models: GR4J, HBV, HYMOD, and MILc.
• Precipitation Data: 13 datasets including gauge-based (CPC, E-OBS), reanalysis (ERA5-Land), satellite-derived (IMERG-LR GPM, CHIRPS, PERSIANN-CDR, SM2RAIN-ASCAT), and a merged 1 km product (CPC-GPM-ASCAT).
• Soil Moisture Data: Four products: SMAP-SMP, TUWIEN-RT1 (Sentinel-1), SMAP-Sentinel, and GLEAM-SMroot.
• Approach: A Bayesian inference scheme (Markov Chain Monte Carlo) was used to estimate posterior parameter distributions for every combination of model and data. Performance was evaluated using the Nash-Sutcliffe Efficiency (NSE) and wavelet multi-resolution analysis (MRA).

Main Results

• Sensitivity: Hydrological model performance is more sensitive to the choice of precipitation forcing than to the specific model structure.
• Precipitation Performance: Gauge-based (E-OBS, CPC) and reanalysis (ERA5-Land) products generally outperformed satellite-only products. However, the merged CPC-GPM-ASCAT product outperformed standard satellite products and, in some cases (Tiber and Medjerda), even the ERA5-Land reanalysis.
• Soil Moisture Value: Incorporating soil moisture into parameter estimation improved model consistency and predictive performance when models were forced with biased satellite precipitation (e.g., IMERG-LR GPM).
• Product Consistency: Soil moisture datasets derived from Sentinel-1 (TUWIEN-RT1 and SMAP-Sentinel) provided better consistency in streamflow simulations compared to coarser products like GLEAM or SMAP-SMP.
• Resolution: Downscaling precipitation to 1 km did not significantly improve lumped model performance for most products, except for IMERG-LR GPM in specific catchments.

Contributions

• Provides a comprehensive evaluation of high-resolution (1 km) Earth Observation (EO) products for mesoscale hydrological applications in the Mediterranean.
• Introduces a robust multi-model, multi-data Bayesian framework to attribute uncertainties between forcing data, model structure, and parameters.
• Demonstrates the added value of Sentinel-1 soil moisture in constraining conceptual hydrological models, particularly in data-scarce regions where ground-based rainfall gauges are absent.

Funding

• European Space Agency (ESA) through the 4DMED-hydrology project (contract n. 4000136272/21/I-EF).
• Italian National Research Council (CNR) senior research grant (n. IRPI 001 2023 PG).

Citation

@article{Sivelle2025Investigating,
  author = {Sivelle, Vianney and Massari, Christian and Tramblay, Yves and Filippucci, Paolo and Dakhlaoui, Hamouda and Quintana‐Seguí, Pere and Clavera-Gispert, Roger and Boutaghane, Hamouda and Boulmaiz, Tayeb and Quast, Raphael and Vreugdenhill, Mariette},
  title = {Investigating hydrological modeling uncertainties in the Mediterranean region by combining precipitation and soil moisture products},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.103015},
  url = {https://doi.org/10.1016/j.ejrh.2025.103015}
}

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