Assessment of NDVI and SAVI Vegetation Indices Potential to Monitor Grazing Impact in a Rangeland Ecosystem

Abstract

The vegetation changes of semi-desert and desert landscapes are temporally and spatially heterogeneous. Vegetation indices derived from remotely sensed data have long been proposed as a source of information for predicting green biomass levels. This research aimed to assess vegetation status via remote sensing techniques using Normalized Difference Vegetation Index (NDVI) and Soil Adjusted Vegetation Index (SAVI) in semi-desert and desert environments. The feasibility of applying such techniques is tested for assessing grazing impact in Forish district of Uzbekistan. Time-series of ecological indicators of land degradation that are collected synoptically from local to global spatial scales can be derived from the 43-year and continuing Landsat satellite archive. Consequently, this study was conducted in the Forish rangeland ecosystem using a time-series of standardized Landsat imagery for the period 2010 to 2015. Two common vegetation indices derived from Landsat Surface Reflectance (LSR) imagery. Five images, one image for each year were used for the analysis. The research had the objectives to characterize and map the historical trends of a remotely-sensed index of vegetation response, a correlate of vegetation cover or phytomass, and field geobotanical monitoring results and define the appropriate values for four degradation levels (weak, average, strong and very strong). Those values were applied for 2010-2015 years period to define degradation level in order to monitor grazing impact in the study area.