The cloud-climate feedback problem remains a major source of uncertainty in modeling climate change. The proposed work addresses this problem by using MODIS observations of marine stratocumulus to document linkages between cloud properties such as pixel-scale and regional scale fractional cloud cover, cloud optical depth, layer altitude, droplet radius, cloud liquid water path, column droplet number concentration, etc., anticipated on the basis of various cloud models. A retrieval scheme that accounts for the fractional cloud cover within the 1-km MODIS fields of view will be used to characterize these relationships and to document artifacts in the MOD06 cloud products that lead to erroneous linkages. The artifacts result from the assumption that cloudy 1-km pixels are overcast when, in fact, many are only partly cloud covered. Among the artifacts are the large droplets (radii > 20 microns) found in the MOD06 product. For marine stratocumulus, many of the large droplets are from partly covered pixels coupled with the use of 2.1-micron reflectances to retrieve droplet radii. While the Collection 5 processing of the MOD06 product included screening checks to remove partly cloudy pixels, the deciding check was the use of the 250-m MODIS cloud flag to identify 1-km pixels that were overcast. Unfortunately, many of the 250-m pixels are identified as overcast when, in fact, they too are only partly cloudy. These findings are to be confirmed through the analysis of a representative sample of high spatial resolution ASTER scenes of marine stratocumulus. In addition, 3-D radiative transfer simulations will be undertaken to determine the extent to which holes in otherwise extensive layers of stratus could give rise to the large droplets retrieved using the 2.1-micron reflectances. Likewise, the overcast pixel assumption in the MOD06 processing leads to an erroneous linkage between cloud layer temperature and cloud optical depth. While the observed linkage would be consistent with expectations based on, for example, adiabatic parcel models, the linkage stems from pixels being partly cloudy. Through comparisons of the results of the partly cloudy pixel retrievals with the MOD06 products, artifacts arising from the overcast pixel assumption will be documented and linkages among the various properties of marine stratocumulus will be established. Such linkages should prove helpful in the development and assessment of models for marine stratocumulus. As are the ongoing comparisons of the partly cloudy pixel retrievals with the MOD06 products, the proposed work will be done in collaboration with Steve Platnick and Mike King (NASA GSFC), and the linkages among the cloud properties and 3-D simulations of the effects of holes in stratus layers will be undertaken through an ongoing collaboration with Andy Ackerman (NASA GISS).
