The use of remote sensing & geographical information systems to identify vegetation: The case of dhofar governorate (Oman)

Extracting and identifying vegetation from satellite images is a major topic in Remote Sensing and GIS. Several techniques have been developed by scientists worldwide. This study aims to evaluate the approaches to the extraction of vegetation from satellite images and to examine each approach within the GIS environment. Three different approaches were used: The first is Normalized Difference Vegetation Index (NDVI), the second is Supervised Classification and the third is Unsupervised Classification. These three methods were examined using two Landsat scenes for Dhofar Governorate which lies in the Southern part of Oman. The two Landsat scenes were acquired on 4^th May 2001. The preliminary results show that there are variations in the total areas of vegetation between the three approaches. For example, the total vegetation area using NDVI approach is just above 900 Km², but it increases to more than 1700 Km² when using upervised classification, and then it increases again to more than 3300 Km² when using the unsupervised classification approach.