-

The objective of this study was to assess the application of vegetation indices derived from Land sat Thematic Map per imagery for estimating rangeland vegetation canopy cover percentage of Jahan-Nama protected area. Stratified random sampling was designed and 370 sample plots on 37 sample units were allocated in the field, using Global Positioning System (GPS). Land sat Thematic Map per image, dated 1998, was geometrically corrected with RMS error of less than 0.98-pixel size. Corresponding mean values of 9-pixel elements belonging to sample units for all created indices were extracted. Corresponding field measurements were regressed against TM data (as an independent variable) and 69 vegetation index values created by different combinations of TM bands (as dependent varibables). Then, statistical models were developed. Results show that if canopy cover percentage without its arrangement in classes are entered in the regression analysis, land sat data will be unable to detect their changes. The results also show that land sat data are capable of detecting high canopy cover percentage classes when field data are categorized. Regression analyses show promising results in estimating canopy cover percentage of rangeland vegetation. The index, (TM5-TM3)/(TM5+TM3), shows high correlation (0.99) with the first canopy cover class (71.9-88.1%), (TM3-TM7)/(TM3+TM7) and (TM6)/(TM3+TM5) indicate high correlation with the second class (58.6-71.9%). Having more than 0.85 regression coefficient, (TM6)/(TM7-TM2) demonstrates high correlation with the third canopy class with 45.2-58.6 percent canopy coverage. For the last class, no index could be introduced to estimate coverage
percentage as there was no significant correlation between canopy cover and all indices involved.