UPSCALE INTEGRATION OF NORMALIZED DIFFERENCE VEGETATION INDEX - THE PROBLEM OF SPATIAL HETEROGENEITY

Spatial integration of radiometric parameters that describe the Earth's vegetation cover is an important issue important when studying global scale land-atmosphere interactions. The Normalized Difference Vegetation Index (NDVI) is currently used to characterize the land surface in terms of vegetation cycles and primary production. The NDVI is sensitive to fractional vegetation cover, canopy density, leaf architecture, and leaf physical state. A potential problem with NDVI, however, is that it does not depend linearly on radiance or reflectance. Since vegetation can be highly heterogenous spatially, and since the relationships between NDVI and vegetation parameters are established locally, using NDVI derived from coarse resolution sensors such as NOAA Advanced Very High Resolution Radiometer (AVHRR) may not be appropriate. The purpose of this paper, therefore, is to analyze the correspondence between NDVI calculated from average reflectances, M(NDVI), and NDVI integrated from individual NDVI's, I(NDVI) by simulating AVHRR data from high spatial resolution SPOT 1 Haute Resolution Visible (HRV) radiometer and Landsat Thematic Mapper (TM) data. For the considered sites, located in tropical West Africa and temperate France, and the scales analyzed, 300-1000 m, a strong correlation exists between the two types of index. The relationship is almost perfectly linear, with a slope depending slightly on the variability of the vegetation cover. Effecting the scale change using M(NDVI) instead of I(NDVI) does not introduce significant errors, especially when these errors are compared to those resulting from uncertainities in the relationships between NDVI and vegetation parameters, which are typically one order of magnitude higher. Other sites should be examined, however, and the variability within TM and HRV Radiometer pixels quantified in order to conclude definitely and generally about the adequacy of using M(NDVI) in parameterizations obtained at the local scale.