Analysing different groups of remote sensing vegetation indices to study rangeland vegetation types

Satellite images are currently used as a fast and low-cost tool for rangeland assessment. In this study, 10 vegetation indices from different groups were calculated using spectral bands. The correlation of these vegetation indices with canopy cover was then measured for three vegetation types in Semirom region-Isfahan. The percentage of canopy cover and other earth surface components such as litter, rock, gravel, stone and bare ground were determined using step-point method. Ten random points were selected in each vegetation type and four 150 meters transects in two perpendicular direction were established from each point, therefore 6000 points were studied per rangeland type. Ten different groups of vegetation indices including slope-based, distance-based and plant–water sensitive indices extracted from IRS-P6 satellite data (AWiFS sensor) were evaluated against total ground cover. The results indicated that vegetation indices have higher relationships within vegetation types based on their characteristics but the correlaion for each vegetation index was not consistent among different types. Considering all photosynthetic vegetation cover, the correlation between these factors and vegetation indices were much higher than degraded sites due to greater reflectance from bare soil and less photosynthetic area. In general, the results indicate that each vegetation index is appropriate for mapping a specific vegetation type and we should consider this issue in mapping and monitoring vegetation cover.