Towards an operational atmospheric correction for AVHRR land surface products

The German Remote Sensing Data Center (DFD) of the German Aerospace Center (DLR) has been operating a ground segment for high resolution picture transmission (HRPT) data acquisition, archiving, and distribution since the early 1980s. Since 1994 daily normalized difference vegetation index (NDVI) synthesis maps for all of Europe have been derived from Advanced Very High Resolution Radiometer (AVHRR) measurements in near real-time with a horizontal resolution of about 1 km. In 1998 the processing chain has been expanded to European land surface temperatures (LST). Due to the presence and high variability of atmospheric ozone, water vapour and aerosols in time and space, the atmosphere significantly impacts the solar channels and results in erroneous NDVI data and noisy time series. A robust operational near real-time atmospheric correction scheme is presented. Until ENVISAT data is available, total ozone data derived from ERS-2 GOME and water vapour from NOAA-TOVS are considered for correction. An aerosol retrieval scheme based on GOME and ATSR data is currently being developed. For the determination of LST from the AVHRR brightness temperature measurements also an explicit correction of the atmospheric influence, mainly caused by water vapour absorption, is necessary. Therefore a simple split window technique is applied instead of a single channel method in terms of operational processing. The results have been validated over different sites in Germany, Egypt and Jordan