An automatic system for AVHRR land surface product generation

Making products from the Advanced Very High Resolution Radiometer (AVHRR) on board the National Oceanic and Atmospheric Administration (NOAA) polar-orbiting satellites can be time consuming and an automated technique for image processing is required to generate long time series of AVHRR imagery. This paper aims to describe the development of a system for fully-automated AVHRR image processing, including radiometric calibration, precise geo-registration and generating land-surface products, such as vegetation indices, maximum value composites and cloud masks. Tests for crop monitoring purposes were carried out using High Resolution Picture Transmission (HRPT) images between October 2003 and April 2004. The region used to evaluate the system was the State of Parana, one of the primary soybean producers in Brazil. Results have shown that for severely cloud-filtered images, the system was effective in generating geometrically precise image products, with geolocation errors less than a pixel. The developed system can be operated with no human intervention and can be used as an important tool for NOAA-AVHRR image users especially those who need to use long time series.