Mapping coastal marine ecosystems of the National Park of Banc d’Arguin (PNBA) in Mauritania using Sentinel-2 imagery

A novel approach using Sentinel-2 satellite imagery and machine learning successfully mapped coastal marine ecosystems in the National Park of Banc d'Arguin (PNBA) in Mauritania with high accuracy (>80% overall accuracy, kappa index 0.75). This method, combining supervised Support Vector Machine (SVM) classification with field data, provides an updated and detailed ecosystem database for this UNESCO-listed RAMSAR site, critical for biodiversity conservation and carbon sequestration. The study demonstrates that while Sentinel-2 offers robust classification capabilities, higher-resolution sensors like SPOT-6 can further refine mapping of small, patchy ecosystems. Notably, the use of freely accessible Sentinel-2 data and open-source geoinformatic tools ensures reproducibility, affordability, and scalability for local conservation efforts.

Pottier, A., Catry, T., Trégarot, E., Maréchal, J.-P., Fayad, V., David, G., Sidi Cheikh, M., & Failler, P. (2021). Mapping coastal marine ecosystems of the National Park of Banc d’Arguin (PNBA) in Mauritania using Sentinel-2 imagery. International Journal of Applied Earth Observation and Geoinformation, 102, 102419. https://doi.org/10.1016/j.jag.2021.102419