Analysis of the Quality of Collection 4 and 5 Vegetation Index Time Series from Modis

Globally acquired data from both MODIS instruments are suitable for science quality time series, because the unique concept of pixel-level quality information of each MODIS land product allows a detailed analysis of the data usability. MODIS datasets are regularly updated and reprocessed to meet present science requirements. This study compares time series of present collection 4 and currently released collection 5 data for the vegetation index product (MOD13). Considerable adjustments were made including changes in cloud and aerosol detection, compositing, and a redesign of the quality information layer. A software package for time series generation of MODIS data (TiSeG) was adjusted to collection 5 products. The quality requirements of collection 5 data are stricter and collection 5 flags are more sensitive to atmospheric disturbances. The newly introduced reliability dataset is important for accurate cloud detection. Compared to the NDVI, the EVI time series has a higher dynamic range for vegetated units and a better inherent temporal consistency also for lower quality composites.