Calculation of the biophysical parameters of vegetation in an arid area of south-eastern Kazakhstan using the normalized difference moisture index (NDMI)

A comparative analysis was carried out on the accuracy of vegetation indices and NDMI (Normalized Difference Moisture Index) based on Landsat-8 data. The paper describes the peculiarities and the effectiveness of different spectral indices in recognizing sparse desert vegetation and calculating the basic biophysical parameters of vegetation. The theoretical and technical limitations and advantages of different approaches and the application of vegetation indices to different types of vegetation cover are discussed. The original narrow-band water content in the green biomass broadband index was calculated from Landsat-8 data. NDMI was comparatively tested with a number of vegetation indices, based on red and near-infrared bands of satellite imagery. Pearson’s correlation coefficients were considered, calculated for three basic vegetation biophysical parameters and spectral indices. The transformed NDMI demonstrates a higher correlation with all the basic biophysical variables of vegetation (grass cover, biomass, and productivity) compared to NIR-RED-based vegetation indices for the intrazonal vegetation of the desert and semi-desert territory of Kazakhstan. NDMI appears to be a promising approach in studies based on the remote detection of non-homogenous vegetation cover in arid areas.