Assimilation of VGT and AVHRR NDVI data to build long series dataset

This paper describes the assimilation methods and results of SPOT-VGT and NOAA/AVHRR NDVI data in order to build a time series dataset for many applications such as monitoring crops. In our process, we have five years of VGT data and ten years of AVHRR data, we select five-year series of these two datasets during the same period. We first build the AVHRR dataset from raw data through standardized methods, retrieve the environmental variables such as NDVI and temperature. Then we use the overlap datasets to create the VGT and AVHRR relationship. This relationship was used to extent the VGT dataset back to 1991. The relation coefficient and variety are also studied to analyze spatial diversities of the two datasets. Three levels are studied including pixel, stratum and province levels.