Above-ground biomass estimation for Quercus rotundifolia using vegetation indices derived from high spatial resolution satellite images

Abstract

The estimation of vegetation parameters, such as above-ground biomass, with high accuracy using remote sensing data, represents a promising approach. The present study develops models to estimate and map above-ground biomass of Mediterranean Quercus rotundifolia stands using one QuickBird satellite image in pan-sharpened mode, with four multispectral bands (blue, green, red and near infrared) and a spatial resolution of 0.70 m. The satellite image was orthorectified, geometrically and radiometrically corrected. Object-oriented classification methods and multi-resolution segmentation were used to derive a vegetation mask per forest species. Data from forest inventory (24 plots) and vegetation indices (NDVI, EVI, SR and SAVI) derived from high spatial resolution satellite images were used for an area of 133 km2, in southern Portugal. The statistical analysis included correlation, variance analysis and linear regression. The linear regression models fitted included the arithmetic mean and the median values of the vegetation indices per inventory plot as explanatory variables. The overall results of the fitted models show a trend of better performance for those with the median value of the vegetation index as the explanatory variable. The best fitted model (R2 = 75.3) was associated with the Simple Ratio (SR) median value as an explanatory variable. A Quercus rotundifolia above-ground biomass map was produced.