Yonaba et al. (2025) Hydrological evaluation of top-down and bottom-up rainfall products in West Africa: Model performance, parameter range and uncertainty propagation

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-11-24
• Authors: Roland Yonaba, Tazen Fowé, Gyskel Ngonga, Axel Belemtougri, Arsène Kiema, Idi Souley Tangam, Harouna Karambiri
• DOI: 10.1016/j.ejrh.2025.102957

Research Groups

• Laboratoire Eaux, Hydro-Systèmes et Agriculture (LEHSA), Institut International d’Ingénierie de l’Eau et de l′Environnement (2iE), Ouagadougou, Burkina Faso
• Département Gestion des Ressources Naturelles et Systèmes de Productions (GRN/SP), Institut de l’Environnement et de Recherches Agricoles (INERA-BF), Ouagadougou, Burkina Faso
• Département Climat-Eau-Sol (AGRHYMET), Centre Climatique Régional pour l′Afrique de l′Ouest et du Sahel (CCR-AOS), Niamey, Niger

Short Summary

This study evaluated the hydrological performance of four top-down and three bottom-up satellite rainfall products in three West African Sahelian river basins using the SWAT model. It found that model skill varied across basins, with some gridded products outperforming gauge observations, and demonstrated that carefully selected rainfall products can significantly enhance hydrological modeling and water resource planning in the region.

Objective

• To evaluate the hydrological performance of top-down and bottom-up rainfall products across multiple Sahelian watersheds.
• To investigate how rainfall products affect the calibration ranges and identifiability of hydrological model parameters.
• To quantify how uncertainty propagates from rainfall inputs into runoff simulations.

Study Configuration

• Spatial Scale: Three major river basins in Burkina Faso: Mouhoun River Basin (approximately 100,500 km²), Nakanbé River Basin (approximately 20,865 km²), and Sirba River Basin (approximately 38,750 km²). The study region is the West African Sahel.
• Temporal Scale: The main study period for analysis was 2007–2024, with a warm-up period of 2005–2006.

Methodology and Data

• Models used:
  • Soil and Water Assessment Tool (SWAT)
  • SWAT Calibration and Uncertainty Program (SWAT-CUP 2012) using the Sequential Uncertainty Fitting Procedure (SUFI-2)
• Data sources:
  • Rainfall products:
    • Top-down Satellite Precipitation Products (TD-SPPs): ARC2, RFE2, CHIRPS v2, TAMSAT v3.
    • Bottom-up Rainfall Products (BU-RPs) derived from SM2RAIN: SM2RAIN-ASCAT, SM2RAIN-CCI, SM2RAIN-GPM.
  • Observation data:
    • Daily rainfall, minimum temperature, maximum temperature, and potential evapotranspiration records (1981–2024) from the National Meteorology Agency in Burkina Faso (ANAM-BF).
    • Daily discharge observations from the General Directorate of Water Resources (DGRE) at basin outlets (MRB: 1998–2024, NRB: 1998–2024, SRB: 1998–2017).
  • Static inputs for SWAT:
    • Elevation data: FABDEM Digital Elevation model (30 m spatial resolution).
    • Land use and land cover: National land use map of Burkina Faso (2012).
    • Soil characteristics: Harmonized World Soil database v2.0 (30 m spatial resolution).

Main Results

• Hydrological model skill (KGE) varied across basins for the gauge-forced baseline (SIM_OBS): 0.83/0.78 (Mouhoun), 0.70/0.52 (Nakanbé), and 0.78/0.83 (Sirba) for calibration/validation, respectively.
• Several gridded products matched or outperformed SIM_OBS: CHIRPS (KGE 0.90/0.80) and RFE2 (KGE 0.81/0.85) in the Mouhoun, and CCI (KGE 0.83/0.72) and CHIRPS (KGE 0.80/0.75) in the Sirba.
• Accurate rainfall products (e.g., CHIRPS, RFE2) resulted in narrow and consistent posterior ranges for runoff and routing parameters (e.g., CN2, SOLK, CHN2).
• Uncertain rainfall inputs (e.g., some BU-RPs) led to compensatory broadening of groundwater controls (e.g., lower SOLK, higher CN2, elevated GWQMN and RCHRGDP) to sustain baseflow and damp peaks.
• Error propagation analysis showed that rainfall uncertainty could be either attenuated (R < 1, especially for some BU-RPs in Mouhoun and Nakanbé) or amplified (R > 1, generally for TD-SPPs and in Sirba).
• Variance decomposition revealed that model parameter uncertainty accounted for most of the discharge spread in all basins, with rainfall product differences being secondary and intermittent.

Contributions

• Provided a comprehensive hydrological evaluation of top-down and bottom-up satellite rainfall products in West African Sahelian basins, extending beyond typical meteorological validation.
• Demonstrated that the quality and structure of rainfall inputs directly influence the identifiability and stability of hydrological model parameters, with more accurate inputs yielding tighter, more physically realistic parameter ranges.
• Quantified the non-linear propagation of rainfall errors through the hydrological modeling chain, highlighting that catchment filtering and calibration choices significantly condition error transmission.
• Offered guidance for improving rainfall product selection, calibration strategies, and uncertainty treatment in hydrological applications, particularly relevant for water management in data-scarce semi-arid regions.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Citation

@article{Yonaba2025Hydrological,
  author = {Yonaba, Roland and Fowé, Tazen and Ngonga, Gyskel and Belemtougri, Axel and Kiema, Arsène and Tangam, Idi Souley and Karambiri, Harouna},
  title = {Hydrological evaluation of top-down and bottom-up rainfall products in West Africa: Model performance, parameter range and uncertainty propagation},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102957},
  url = {https://doi.org/10.1016/j.ejrh.2025.102957}
}