Abstract
Throughout most of the shortwave spectrum, atmospheric gases do not absorb the abundant amount of incoming solar radiation. The shortwave-scattering albedo of clouds is very large. The combination of large amounts of incoming solar radiation, low gaseous absorptivity, and large cloud-scattering albedo enables clouds at one level of the atmosphere to affect the shortwave radiative transfer at all other atmospheric levels. Absorption by atmospheric gases is much stronger in the longwave. This localizes the effects of clouds in the longwave. Since longwave absorption is weakest in the window region (8–12 μm), cloud effects there will have the greatest chance of propagating to other levels of the atmosphere. In partially overcast conditions, individual cloud geometry and optical properties are important factors. Longwave calculations of most GCMs ignore individual cloud geometry. For liquid water clouds, the optical properties of clouds are also ignored.
Previous work in the window region by Takara and Ellingson considered opaque clouds with no absorption or emission by atmospheric gases. Under those conditions, the effect of cloud scattering was comparable to cloud geometry. In this work, the comparison of longwave scattering and geometric effects in the window region is improved by including partially transparent clouds and adding absorption and emission by atmospheric gases. The results show that for optically thick water clouds, it is sufficient to model the geometry; scattering can be neglected. The window region errors are less than 5 W m−2 for fluxes and 0.05 K day−1 for heating rates. The flat-plate approximation worked for ice clouds; the window region flux errors are less than 3 W m−2 with heating rate errors less than 0.05 K day−1.
Corresponding author address: Ezra Takara, Department of Meteorology, University of Maryland at College Park, College Park, MD 20742.
Email: ezra@atmos.umd.edu
