Subceiling probing of clouds using pairs of AVHRR images

Accurate determination of cloud parameters and, particularly, of the cloud-top temperature and cloud emissivity are considered. The recognized ISCCP algorithm relies on relationships between bispectral data taken at visible (VIS, /spl sim/0.65 /spl mu/m) and infrared (IR, /spl sim/11.5 /spl mu/m) wavelengths. Granted that such relating of the VIS extinction coefficient to the IR absorption coefficient is admissible, headway in developing a technique based upon thermal IR-channels is of vital significance, because it promises the retrieval of cloud parameters both day and night while avoiding some problems inherent in the VIS-IR approach. It is accepted that the emissivity of opaque clouds approaches unity thus permitting for infrared methods to retrieve accurate cloud-top temperature. Even that, current IR methods based solely on attenuation effect of clouds are incapable to measure radiative characteristics of opaque clouds. To cope with this problem, it was proposed recently to use submillimeter wave frequencies at which most clouds are transparent enough. Although no specific microwave cloud algorithms have been developed, it was pointed out that subceiling probing of radiometrically opaque nimbus clouds (rain cells) remains the challenge for microwave satellite sensing. On the other hand, many "multichannel" infrared cloud algorithms developed for semitransparent cirrus cases are based on a doubtful assumption that the cloud emissivity is independent of wavelength. In virtue of this, the authors intended to investigate this problem with a new dual-path (DP) approach, as the slant path (proportional to the secant of the satellite zenith angle) is different instead of wavelength thus obviating the obstacle of emissivity differences between wavelengths. Examination of AVHRR images from the triplet of nearly simultaneous NOAA. Polar orbiting satellites (namely: NOAA-9, -10, -12) revealed significant variances in the infrared radiation emitting by opaque clouds at different angles. To account for the effect of cloud temperatures changes with depth, causing the essential part of these angular differences, the current cloud radiative model is revised following the principles of radiative transfer. Based on the model analysis, the authors investigate in this paper the potential of DP method for retrieving parameters of opaque clouds.<>