Desertification assessment using Sentinel-2A images, spectral and TCT indices in the Admer-Ezem watershed, upper Oum-Rbaa basin, Morocco

This paper focuses on a quantitative assessment of desertification in the Admer-Ezem basin, using Sentinel-2A satellite images and indices to map the degree of basin-wide desertification in 2023. The methodology is based on the extraction of several spectral indices, such as the normalized difference vegetation index (NDVI), albedo, topsoil grain size index (TGSI) and the modified soil adjusted vegetation index (MSAVI), as well as the tasseled cap transformation (TCT) indices: brightness (TCB), greenness (TCG) and wetness (TCW). Different combinations of these indices were then created and analyzed using linear regression to identify low correlation rates. Based on these results, three characteristic zones were selected: TGSI-Albedo (r = -0.95), TGSI-TCW (r = -0.74) and TCW-TCB (r = -0.73). The resulting map subdivides the region into five categories of desertification: extreme, severe, moderate, low and non-desertified. The results show that the forest is undergoing progressive desertification. In general, and according to the different models used, more than 50% of the area studied is subject to severe to extreme desertification, while less than 30% of the area is classified as no-desertified, or slightly affected. Thanks to the high accuracy of the models used (Area Under the Curve (AUC) from 79 to 83%), this approach proves to be almost optimal for the quantitative analysis and monitoring of desertification in the Moroccan Middle Atlas. Finally, this work can help planners identify the most suitable sites for future forest development.