Rangelands vegetation cover mapping using IRS-LissIII data image processing (Case study: Sorkh Abad Basin, Mazandaran)

This research was conducted to produce the map of vegetation cover percentage of Sorkh Abad rangelands with an area of 21717 hectares by using IRS-Liss III data. Images of IRS-LissIII were geo-referenced by digital topographic maps with a scale of 1:25000. Due to mountainous condition, for obtaining the high accuracy at the geometric correction process, elevation corrections were done on the images. In this study, in addition to main bands, vegetation indices, ratios and principle component analysis also were used.For field data collection, systematic random sampling method was employed. In each vegetation type, one key area was selected. In each key area, four vertical and horizontal transects with 100m length were established. Then along the each transect vegetation cover percentage was estimated in ten plots of 1m2. In the next step, correlation between vegetation indices, ratios and the field data was calculated. Then the indices that had stronger correlation with field data were used for classification with higher accuracy. Among the vegetation indices, RVI (0/74%) and NDVI (0/704%) had shown the highest correlation. Based on the obtained results, we may generally conclude that IRS-LissIII images have a good capability for cover percentage mapping in the study area.