Evaluation of Desertification in the Middle Moulouya Basin (North-East Morocco) Using Sentinel-2 Images and Spectral Index Techniques

摘要

This article focuses on the quantitative assessment of desertification in the Middle Moulouya basin located in the North-East of Morocco. Indeed, this study aimed to map the degree of desertification at the level of the basin in 2018 using the Sentinel-2 image. To do this, we have adopted a methodology based on several stages. First, we extracted the spectral indices, in particular, the NDVI, the albedo, the TGSI and the MSAVI. Then, different combinations of these indices were the subject of a linear regression analysis (NDVI-albedo, MSAVI-albedo, albedo-TGSI and TGSI-MSAVI) to use the best correlated combinations to construct the feature space. The results obtained showed that the NDVI-albedo and MSAVI-albedo combinations are the best correlated with respective correlation coefficients of r = - 0.73 and r = - 0.76, respectively. As a result, they were used to propose the desertification degree index (DDI) by exploiting the NDVI-albedo and MSAVI-albedo feature spaces. Finally, a desertification map was generated for the entire basin. It has five degrees of desertification (extreme, severe, moderate, low, non-desertification). According to our results, the situation of desertification in the basin is alarming. Indeed, 86.86% of the study area is located in the moderate to extreme desertification class. While only 12.25% and 0.89% fall in the low and no desertification categories, respectively. The MSAVI-albedo model gave a high overall accuracy of 93.75%, so it is perfectly effective for the quantitative analysis and monitoring of desertification at the level of the basin studied.