Evaluating of the most suitable vegetation indices of estimating of canopy cover and above-ground phytomass in arid rangelands during different growth periods

One of the major applications of remote sensing in environmental resources management is change detection and quantitative assessment of green vegetation. This research assesses the vegetation indices (VIs) derived from Landsat 8 images for modeling canopy cover (CC) and above-ground phytomass (AGP) in Marjan rangelands, Boroujen. CC was measured (using double sampling method) and AGP was also estimated (using grid quadrat method) in 4 sampling periods during growing season in spring till summer using 95 quadrats that were laid out along a 10-km transect in line with 19 sampling points, 3 each contains 5 centroid quadrats with 4-m distance from central quadrat (Total 380 quadrats between May to September 2014). Vegetation indices VIs calculated with outcomes of FLAASH atmospheric correction method for four Landsat-8 image sets obtained between May to September. Ground measurement of plant GCC and AGP between May to September 2014 was regressed against vegetation indices VIs.
Results of statistical analysis showed that ARVI, SARVI and EVI showed the highest correlation with CC (R2= 0.81) and with AGP (R2= 0.60, 0.61, 0.61 respectively).Even though, the correlation between CC and AGP with vegetation indicates was high, but CC shows the highest relationship with VIs in comparison to AGP. It can be conclude that arid rangelands vegetation can be accurately estimated with derived vegetation indices from Landsat-8 images, especially those concerned with atmospheric corrections, i.e., ARVI, SARVI and EVI.