Abstract
This paper provides a Digital Vegetation Model (DVM) which can be used for ecoclimatic analysis in Africa. Principal Component Analysis (PCA) of the 5-year monthly mean NOAA/AVHRR satellite data for the Visible (Channel 1), the Near Infra-Red (Channel 2) and the Thermal Band (Channel 4) was performed. Using the first component for each channel, we could retain and enhance the spectral properties of each land cover type. Accordingly, densely to sparsely vegetated areas all over the continent were classified into 7 main land cover categories. The seasonal properties and the area of each vegetation type were explored. The 3-dimensional relationship between vegetation and topography was shown using the DVM and a digital elevation model. The DVM is shown to be consistent with, and to upgrade, other available African vegetation maps produced mostly from one-year NDVI databases.
