Some of the artifacts observed in SMOS brightness temperature images are originated in an inconsistency between correlation and antenna temperature calibration errors, which produces a multiplicative bias that scales with the scene mean antenna temperature. Its most important effect, although not the only one, is a land-to-sea contamination near coastal areas, which cannot be easily mitigated by post-processing correction techniques. A simplified end-to-end simulator allows introducing separate errors in both antenna temperature and correlation, thus providing an important insight on this problem and helping understanding its origin. Results from simulations compare very well with actual results from measured data.

Peer Reviewed