ALBERT

Description: Simple geometric considerations suggest that in the same way as broken cloud cover appears practically total near a ground observer's horizon, the anisotropy of the longwave radiation emitted to space from an area of broken cloud should be substantially enhanced. Using a simple model consisting of parallel bands of “black” clouds of rectangular cross section, imbedded in an otherwise horizontally homogeneous vertically stratified atmosphere, we evaluate the longwave radiance emergent in different directions, averaged over the cloud pattern, and find this expectation to be confirmed in a wide range of cases. At large zenith angles, the emergent radiance can exhibit considerable dependence on azimuth, relative to the direction of the band structure. Although such azimuthal dependence practically disappears for many 3-dimensional broken cloud fields (e.g., hexagonal cells), the enhanced zenith angle dependence remains. This can be a source of bias in cloud extractions or Earth radiation budget determinations based on geostationary satellite data, since zenith angle is directly related to geographical location of the point under observation. With our model cloud arrays, and considering radiative transfer in the atmosphere, we have evaluated the bias in the broad-band longwave (3–100 μm) radiant exitance values derived, assuming horizontal homogeneity, from either broad-band or narrow-band (10.5–12.5 μm) radiance data. We find that the bias for near-nadir observations can be as large as 20% (40 W m−2) in extreme cases involving bands of opaque high-level cloud over warm ground. It may remain substantial in situations which are encountered in nature. We examine the dependence of the broken cloud effects on the geometrical parameters of the field. We discuss how such effects might be detected and how they can influence determinations of Earth radiation budget components and of cloud cover. We note that “brokenness” will be a factor in cloud-radiation feedback.