An approach for spaceborne aperture synthesis radiometer calibration

Abstract

This paper outlines an approach which was developed and investigated for the SMOS spaceborne imaging radiometer to be launched end of 2007. SMOS is an L-band Y-shaped aperture synthesis radiometer intended for global measurements of soil moisture and ocean salinity. Since the directly measured quantities are correlation values in the spatial frequency domain, a calibration procedure has to transform them to brightness temperatures, which are the typical radiometer output signals and specified as the SMOS level-1 data products. Furthermore, the imperfect system response has to be corrected, for which some on-board implemented hardware tools are available. However, those are only able to correct part of the errors and, typically, the residual imperfections are more expensive to eliminate. Finally the map of correlation values has to be converted to an image in the spatial domain in order to assign the determined geophysical data to their corresponding geographical location. The here proposed image reconstruction and calibration approach is based on the in-space measurements of well-known calibration targets like point sources in a low-power background as it could be provided by the Moon and the deep space. The use of multiple scenes, where the point-source is located on different positions within the usable field of view, allows the determination of a G matrix describing the whole imperfect system. Once determined, the matrix can be used to compute efficiently the corrected and calibrated brightness temperature map of the actually observed scene. Simulation results for various realistic conditions in order to assess the performance of the approach are shown and discussed.