Radiometric validation of satellite vegetation indices using flux tower measurements

Satellite vegetation indices (VIs) serve as an important means to monitor and characterize seasonal changes of terrestrial vegetation. It is, therefore, critical to ensure quality of such VI products. One method of assessing VI product quality is cross-comparison with in situ flux tower measurements. Two methods of deriving the normalized difference vegetation index (NDVI) from flux tower data have been used, both using photosynthetically active radiation and global solar radiation data; one partitions the global solar radiation into visible and near-infrared wavebands, whereas the other does not. In this study, we evaluated the partitioning impacts by comparing these two methods and using them to validate daily NDVI products from Visible Infrared Imaging Radiometer Suite (VIIRS). In general, tower NDVI values from the non-partitioning method were slightly higher than those from the partitioning method, and produced slightly higher R² values with VIIRS NDVI. We conclude that the non-partitioning method was a preferred approach to validate satellite NDVI due to its simplicity and higher correlations.