Evaluating the fraction of vegetation cover based on NDVI spatial scale correction model

Vegetation index (VI) is an important variable for retrieving the vegetation biophysical parameters. With different kinds of remote sensing data sets, it is easy to get the VI at different spatial and temporal resolutions. However, the main concern is whether the relationship existing at some scale between the VI and biophysical parameters is still applicable to other scales. This paper first presents a method to correct the spatial scaling effect of NDVI by mathematic analysis, and then analyses NDVI scale sensitivity with data from a spectral database. The result shows that the NDVI obtained by reflectance up-scaling is larger than the up-scaled NDVI from NDVI itself in most situations. The NDVI scaling effect is more significant when water exists in a pixel, and increases with the increase in the difference of the sum of visible reflectance and near-infrared (NIR) reflectance between the vegetation and soil. Finally, a method is proposed to estimate the fraction of vegetation cover (FVC) on the basis of the NDVI spatial scaling correction model. The method is accurate enough to assess the FVC taking into account the scaling effect