Ferdinand et al. (2025) Spatio-temporal variability of flooded areas in the Ouémé floodplain (Benin, West Africa) from 2015 to 2023

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-11-21
• Authors: Noémie Ferdinand, Alexis Chaigneau, Alexei Kouraev, Yves Morel, Olaègbè Victor Okpeitcha, Sylvain Biancamaria, Sylvain Ferrant
• DOI: 10.1016/j.ejrh.2025.102965

Research Groups

• Laboratoire d’Etudes en G´eophysique et Oc´eanographie Spatial (LEGOS), Universit´e de Toulouse, CNES, CNRS, IRD, UPS, Toulouse, France
• Space and Earth Sciences, University of Science and Technology of Hanoi (USTH), Vietnam Academy of Science and Technology (VAST), Hanoi, Viet Nam
• PRODATA SARL, Cotonou, Benin
• Institut de Recherches Halieutiques et Oc´eanologiques du B´enin (IRHOB), Cotonou, Benin
• International Chair in Mathematical Physics and Applications (ICMPA–UNESCO Chair), University of Abomey-Calavi, Cotonou, Benin
• Centre d’Etudes Spatiales de la Biosph`ere (CESBIO), Universit´e de Toulouse, CNES/CNRS/INRAE/IRD/UT3, Toulouse, France
• Laboratoire d’Oc´eanographie Physique et Spatiale (LOPS), University of Brest, CNRS/IRD/ Ifremer/IUEM, Brest, France

Short Summary

This study assessed the spatio-temporal variability of flooded areas in the Ou´em´e floodplain (Benin) from 2015 to 2023 using remote sensing and in-situ data, revealing a significant upward trend in flood extent driven by cumulative rainfall and river water surface elevation thresholds.

Objective

• To assess the spatio-temporal variability of flooded areas in the Ou´em´e floodplain (Benin) from 2015 to 2023, and to understand the hydrological mechanisms driving these floods using satellite remote sensing, in-situ measurements, and rainfall data.

Study Configuration

• Spatial Scale: Ou´em´e floodplain, southern Benin, West Africa (approximately 1500 km², between 6.33°N - 6.50°N and 2.33°E - 2.58°E).
• Temporal Scale: 2015 to 2023 (9 years).

Methodology and Data

• Models used: "Minimum valley" thresholding method for water detection, Normalized Burn Ratio (NBR) mask for burnt soil correction, Forest and Building Removal Digital Elevation Model (FABDEM) for permanent water body identification.
• Data sources:
  • Satellite:
    • Sentinel-1A C-band Synthetic Aperture Radar (SAR) (VV polarization, 10 m spatial resolution, 12-day revisit time) for flood detection.
    • Jason-2, Jason-3, Sentinel-6, Sentinel-3B altimetry data for Water Surface Elevation (WSE) at virtual stations (VSOu´em´e, VSSˆo).
    • Landsat-8 multispectral imagery for NBR mask.
    • NOAA VNG Flood V1.0 (VNGF) product from Suomi-NPP/VIIRS optical data (375 m spatial resolution) for comparison.
    • CHIRPS (Climate Hazards Group InfraRed Precipitation with Station) for rainfall data (0.05° spatial resolution).
    • Copernicus GLO-30 Digital Elevation Model (DEM) for FABDEM.
  • Observation/In-situ:
    • Pressure sensor (HOBO-U20L-01) at Ladji (Nokou´e lagoon) for in-situ water level data (since February 2018, every 20 minutes).
    • Field campaign (December 2024) for burnt soil validation.
    • AMMA-CATCH stations for in-situ rainfall data (for CHIRPS validation).

Main Results

• A clear seasonal cycle of flooding was observed from August to November, peaking in September.
• Flood extent varied significantly, from an average of 20 km² during the dry season to 160 km² during flood peaks.
• The Sˆo region was identified as the most flood-prone area, with flood frequencies reaching approximately 50% in September.
• A significant upward trend in flood extent was detected over the 2015–2023 period, with an average annual increase of about 6 km².
• Flood dynamics are primarily controlled by cumulative January–September rainfall, with a critical threshold of 1070 mm beyond which flooding expands rapidly.
• A rise in Ou´em´e river WSE above 6 m (from its low-water level, corresponding to an absolute water level of 12 m) induces significant overflows into the floodplain.
• In the Sˆo river, WSE variations exceeding 0.5 m (from its low-water level) trigger overflows due to low riverbanks.
• Temporary hydrological connections form during flood events, increasing connectivity across the basin.
• Sentinel-1 SAR data proved more reliable for consistent flood mapping in this cloud-prone region compared to the VNGF optical product.

Contributions

• Provides novel insights into the hydrological mechanisms driving floods in the Ou´em´e floodplain through a comprehensive remote sensing approach.
• Delivers the first detailed spatio-temporal assessment of flooded areas in the Ou´em´e floodplain from 2015 to 2023 using Sentinel-1A SAR data.
• Quantifies a significant upward trend in flood extent and establishes critical rainfall and water surface elevation thresholds for flood initiation and expansion.
• Highlights the crucial role of hydrological connectivity and low topography in flood propagation within the floodplain.
• Offers valuable metrics for the calibration and validation of hydrological models aimed at flood prediction.
• Lays the groundwork for developing early warning systems and vulnerability maps by integrating flood frequency data with socio-economic information for improved flood management and risk mitigation.

Funding

• "Lagune Nokou´e" TOSCA project (French National Center for Space Studies (CNES))
• “Space Climate Observatory” SCOast-DT program (French National Center for Space Studies (CNES))
• University of Toulouse PhD grant (for No´emie Ferdinand)
• French Space Agency (financial support to Alexis Chaigneau & Yves Morel)

Citation

@article{Ferdinand2025Spatiotemporal,
  author = {Ferdinand, Noémie and Chaigneau, Alexis and Kouraev, Alexei and Morel, Yves and Okpeitcha, Olaègbè Victor and Biancamaria, Sylvain and Ferrant, Sylvain},
  title = {Spatio-temporal variability of flooded areas in the Ouémé floodplain (Benin, West Africa) from 2015 to 2023},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102965},
  url = {https://doi.org/10.1016/j.ejrh.2025.102965}
}