Estimate of satellite-derived cloud optical thickness and effective radius...

Kato, S., L. Hinkelman, and A. Cheng (2006), Estimate of satellite-derived cloud optical thickness and effective radius errors and their effect on computed domain-averaged irradiances, J. Geophys. Res., 111, D17201, doi:10.1029/2005JD006668.
Abstract: 

The process of retrieving cloud optical thickness and effective radius from radiances measured by satellite instruments is simulated to determine the error in both the retrieved properties and in the irradiances computed with them. The radiances at 0.64 mm and 3.7 mm are computed for three cloud fields (stratus, stratocumulus, and cumulus) generated by large eddy simulation models. When overcast pixels are assumed and the horizontal flux is neglected in the retrieval process, the error in the domain-averaged retrieved optical thickness from nadir is 1% to -32% (1% to -27%) and the error in the retrieved effective radius is 0% to 67% (0% to 63%) for the solar zenith angle of 30° (50°). Using the radiance averaged over a 1 km size pixel also introduces error in the optical thickness because of the nonlinear relation between the reflected radiance and optical thickness. Both optical thickness and effective radius errors increase with increasing horizontal inhomogeneity. When the 0.64 mm albedo is computed with the independent column approximation using retrieved properties from nadir (oblique) view for a solar zenith angle of 50°, the error is -0.3% to 14% (-5% to -30%) relative to the albedo from 3-D radiative transfer computations with the true cloud properties. The albedo error occurs even though the radiance at one angle is forced to agree because a plane parallel cloud with a single value of optical thickness and effective radius cannot consistently match the radiance angular distribution. In addition, the error in the retrieved cloud properties contributes to the albedo error. When albedos computed with cloud properties derived from nadir and oblique views are averaged, the albedo error can partially cancel. The absolute error in the narrowband 0.64 mm (3.7 mm) albedo averaged over a 1° x 1° domain is less than 1.5% (0.6%), 5.0% (4.1%), and 7.1% (11%) in order of increasing inhomogeneity, when albedos computed with cloud properties derived from viewing zenith angles between 0° and 60° are averaged and when the solar zenith angle is between 10° and 50°. When the solar zenith angle is 70°, the error increases to up to +24% (+37%) for all three scenes.

PDF of Publication: 
Download from publisher's website.
Mission: 
CERES