Rangeland canopy cover estimation using Landsat OLI data and vegetation indices in Sabalan rangelands, Iran

Vegetation canopy cover is an important biophysical indicator for monitoring rangeland condition and rangeland management. The main purpose of this study is to assess the capability of Landsat 8 (OLI) image-derived vegetation indices (VIs) to estimate the vegetation canopy cover (CC) in Sablan rangelands. Field CC measurement was performed using the transect-quadrats method in 12 selected sites on the study area according to three vegetation life-forms canopy cover including, Forbs (FCC), Grasses (GCC), and Shrubs (SHCC) along with a combination of three life-forms as Total canopy cover (TCC). The estimated CCs were correlated and regressed with Landsat OLI reflectance values and vegetation indices. Results revealed that all the slope-based, distance-based, and orthogonal transformation VIs had the highest significant correlations with Grass canopy cover data (mean of R-2 = 0.70) at the 95% confidence level. The weakest and insignificant correlations were derived for FCC followed by SHCC. A number of VIs (with mean R-2 = 0.6) significantly correlated with TCC. The results of linear regression for GCC, TCC, and SHCC indicated that the reflectance values of spectral bands and vegetation indices were convincingly regressed with CC field data having R-2 values in the range of 0.72 to 0.83, while the FCC due to the third-order logarithmic regression models gained the weakest relationship with spectral data and also VIs (R-2 = 0.14 and 0.42, respectively). Results of validity test of the regression models declared that the models developed for SHCC are the best predictors followed by those made for TCC, GCC, and FCC, respectively. In this study, there was not observed an obvious difference between the performances of the three groups of VIs in predicting the CC of life-forms in the studied sites. It can be concluded that the studied VIs are more strong predictors of Grass canopy cover as the dominant life-form.