Contributions of multispectral images to the study of land cover in wet depressions of eastern Tunisia

• Main Author: Walid Chouari
• Format: Article
• Language: English
• Published: Elsevier 2021-12-01
• Series: Egyptian Journal of Remote Sensing and Space Sciences
• Subjects: Eastern Tunisia; Wet depressions; Classification methods; Moisture index
• Online Access: http://www.sciencedirect.com/science/article/pii/S1110982320303380

Eastern Tunisia is dotted with varied wetlands such as lagoons, sabkhas, garaas… These wetlands, at the interface between land and water, are rich, diverse and dynamic environments. Remote sensing shows potential and efficiency in the detection and characterization of these environments, although these humid depressions are difficult to define. The implementation of new optical satellite sensors characterized by High and Very High Spatial Resolution (HSR and VHSR) and by a high temporal repetitiveness, completes the field data and allows to envisage a delineation of the wet depression and a detailed detection of its land use. The purpose of this article is to determine whether the VHSR image, with its very rich spatial resolution, offers an added value for the detection of these diversified depressions, in comparison with the HSR image with richer spectral resolution. The results of the analysis show that the Ikonos image at VHSR compensates for the spectral richness of Sentinel-2 images at HSR. The “object-oriented” classification method, widely used today in image processing for various applications, appears to be more suitable than the “pixel-based” method in VHSR image classification. The Normalized Difference Water Index (NDWI) and the Transformed Soil Adjusted Vegetation Index (TSAVI) have multiple interests in wetland detection, although moisture masks show some inefficiency in improving the quality of classifications.