Land use land cover diachronic change detection between 1996 and 2016 of region of gabes, Tunisia

Abstract

Remote Sensing Change Detection is designed to detect stand-replacing disturbances such as land cover, harvest and wildfire. Digital change detection essentially jnvolves the quantification of temporal phenomena from multi-date imagery. The main purpose of this study was to distinguish change in land cover within each land cover type (class), and to find the real changes on the land cover features between 1996 and 2016 in the region of Gabes, Tunisia. Two images were downloaded from Google Earth and then georeferenced and masked out the study area, which is imported since Kml File from Google Earth is too. The two images were also enhanced and then classified using the Maximum Likelihood algorithm. Five classes were identified: Water, settlements, vegetated area, bare soil and zone under development. The supervised classification was assessed with 94% for 1996 imagery and 95.5% for 2016. The result showed a decrease in bare soil class (from 115 to 94 km2) and an important increase in zone under development (from 96 to 834 km2). Regarding the water, the vegetated area and settlement classes, a slight increase was noted.