21. Contribution of Hyperspectral Images (Hyperion) and Spectroscopy for Mapping Soil’s Surface and Materials (Case of the Watershed of Oued Beni Zalten, the New Matmata, Tunisia)

Degradation in arid zone is reflected on the ground by the modification of the soil surface components. Several studies have investigated the mapping of the soil surface and components in drylands. A few studies are the outcome of ground-based observations, while others widely use multispectral remote sensing and calculating indices that are limited to samples of electromagnetic spectrum. This paper proposes an innovative method that is applied to geomorphology and environmental research. It seeks to develop a new methodology for mapping soil surface and components, based on hyperspectral remote sensing with a wide range of electromagnetic spectrum. Our case of study is Beni Zalten watershed, which is situated in the natural region of Matmata (Southeast of Tunisia). The region is characterised by limited natural resources, complex physical conditions and irrational human use. This led to the fragility of the environment and the proliferation of water-erosion problems. Therefore, mapping the soil surface and components of Beni Zalten watershed is highly useful to evaluate soil erosion sensitivity and eventual study on erodibility. The methodological approach of our research is essentially based on a Hyperion satellite image dating from December 2009, spectrophotometry measurements and spectroscopy data. This approach depends on a series of pre-treatments (radiometric and atmospheric corrections and reflectance passage), adding to a series of treatments (spectral similarity method: the spectral angle mapper SAM and a method to extract endmembers from the image: Spectral HourglassWizard). This paper is organised around three main parts: First, we will present the physical and human context. Then, we will describe the data and methodology. Finally, we will present our results and discuss the hyperspectral remote sensing and spectroscopy contribution to the cartography of soil surface and components in an arid zone.