Analysis of vegetation index NDVI anisotropy to improve the accuracy of the GOES-R green vegetation fraction product

Abstract

Green Vegetation Fraction (GVF) is the fraction of area within the instrument footprint occupied by green vegetation. Information on GVF is needed to estimate the surface energy balance in numerical weather prediction (NWP) and climate models. For the Geostationary Operational Environmental Satellite-R Series (GOES-R) Advanced Baseline Imager (ABI) algorithm development, a normalized difference vegetation index (NDVI) based linear mixture algorithm has been chosen to convert NDVI into GVF. The GVF algorithm has been developed and tested using a proxy dataset from the Spinning Enhanced Visible and Infrared Imager (SEVIRI) sensor onboard the European Meteosat Second Generation (MSG) geostationary satellite. Studies of SEVIRI data have shown that NDVI strongly depends upon the viewing and illumination geometry of observations, especially over dense vegetation. If not corrected, this angular anisotropy of NDVI causes substantial spurious diurnal variations in the derived GVF. An empirical kernel-driven model to correct NDVI for angular anisotropy has been developed and implemented in the GVF algorithm. Its kernel weights for the GVF algorithm were also determined empirically from the SEVIRI clear-sky data. The preliminary validation estimates show that the model's performance is good.