LANDSAT SATELLITE IMAGES USED IN IDENTIFICATION OF LAND USE AND LAND COVER IN MOUNTAIN AREA

Landsat satellite images are valuable sources of information which help to obtain information on large surfaces. Using the Landsat satellite images to study the ecosystems, in change detection, in retrieved the land surface temperature and other studies involve identifying land use and land cover (LULC). In this paper were used two satellite images, Landsat 5 Thematic Mapper (TM) acquired on 1989 and Landsat 7 Enhanced Thematic Mapper Plus (ETM+) acquired on 2000 to identify LULC in the mountains area. Classification was done in ten classes using supervised classification method, maximum likelihood algorithm. LULC spectral signature were examined based on euclidian distances and the analysis in feature space. Overall accuracies for Landsat 5 TM was 73.68% with kappa statistics 0.66 and for Landsat 7 ETM+ 80.70% with kappa statistics 0.74. The results show that for LULC with similar spectral behaviour appear confusion in classification. Also, mountain forest can be classified only on forest formations or groups of formations forest and not on forest species.