A study on the quality of the vegetation index obtainded from MODIS data

A Vegetation Index (VI) is easily accessible and is associated with a wide spatial and temporal availability. However, are these datasets really useful? One of the main difficulties related to their use, such as MODIS time-series, is the potentially high number of pixels with low radiometric quality. In this work, a Quality Assurance analysis was performed using four MODIS VI scenes from 2011, based on the flags Pixel Reliability and Quality Detailed. The considered dataset covers two different areas of Portugal and two distinct species. Two VI products were analyzed: Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI). The EVI presented generally better results than the NDVI. Additionally, both NDVI and EVI indexes presented a seasonal dependence with better quality in the dry season, and a geographical/land use influence supported by better quality observed in the south area. In conclusion, quality indexes should be considered in VI applications.