Mapping of soil degradation in semi-arid environments in the ouarzazate basin in the south of the central High Atlas, Morocco, using sentinel 2A data

Abstract

This study was conducted to assess the state of soil degradation in a semi-arid environment in the southern part of the Central High Atlas of Morocco (CHA), in the vicinity of the El Mansour Eddahbi Dam in Ouarzazate. This area is subject to soil erosion and degradation issues because of its geological features, and also because of climate change and human activities. The objective of this study was to use a remote sensing space (Sentinel 2 A) to map and evaluate the current status of the land degradation risks. For this purpose, a visit to the study area was carried out to collect the coordinates of the sites of the different types of soil degradation. These were overlaid on the satellite image to obtain a more accurate extraction of the spectral reflectance from eroded, non-eroded, and moderately eroded soil. This visit also provided a detailed view of the state of soil degradation in the region. Subsequently, two supervised classifications were made by the spectral angle mapper (SAM) and support vector machine (SVM) methods, using the Sentinel 2 A satellite imagery, to give high spatial and spectral resolution and effective monitoring of the soil. During the investigation, the mapping and modelling models were validated based on the values of the Kappa index (86% for SAM and 75% for SVM) and the convolution matrix of the two types of classifications, as well as the results of the calculation and the analysis of a set of spectral indexes and new components such as normalized difference vegetation Index (NDVI), normalized difference water index (NDWI), the color index (CI), the brightness index (BI) and two main components (PC1 and PC4). The SAM classification revealed that 18.44% of the soils are weakly degraded, 36.45% are moderately degraded and 27.74% are highly degraded. However, the SVM classification showed that about 36.16% of the soils are weakly degraded, 17.85% are moderately degraded, and 29.35% are strongly degraded. The highly degraded lands are found more in the northern part of the dam, which is intersected by a network of affluents originating from the southern slopes of the Central High Atlas (CHA). These affluents facilitate the phenomenon of water erosion. Geologically, the soils in the northern part of the dam correspond to recent formations (alluvium and silt) of the river valleys, and abrupt lateral slopes. Most of these soils also concern facies related to the Meso-Cenozoic cover, whereas in the southern part of the dam the soils are less degraded and all correspond to very resistant lithological facies (magmatic and metamorphic formations of the Precambrian basement (granites, eruptive rocks, and schists)). We conclude that the short and easy approach used in this research can be a powerful decision-making tool for protecting soils in semi-arid environments.