Radiative effects of inhomogeneous clouds

Clouds are a vital part of the climate system, with important effects related to solar and longwave radiative transfer, latent heating, and precipitation. Inhomogeneous clouds can have quite complicated radiative effects, making them difficult to parameterize effectively. Adequate knowledge of the characteristics and radiative effects of these clouds is essential to ensure their accurate portrayal in large-scale models. This work was directed toward improving this knowledge for small tropical cumulus and single-layer summertime arctic stratus clouds.; A variety of data sets were analyzed to explore the characteristics and radiative effects of small tropical cumulus clouds. The size distributions and fractal dimensions of these clouds exhibit scaling behavior, with a break which may represent the maximum size of individual convective elements. Cumulus fields are spatially clustered at all scales, rather than random or regular. Smaller clouds grow upward more vigorously than larger clouds. Simulations show that cumulus clouds can be difficult to remove from satellite remote sensing data and that even small cumulus cloud fractions can lead to significant errors in retrieved sea surface temperature. As expected, larger cloud fraction and higher solar zenith angle produce greater solar albedo. Cumulus fields can exert significant shortwave forcing at the surface. Only some of the larger, deeper cumulus clouds are found to precipitate.; Next, 150 high-resolution cumulus cloud fields were retrieved from MODIS Airborne Simulator images, and a Monte Carlo radiative transfer model simulated the solar fluxes through these fields using four radiative transfer methods: full 3D, independent pixel approximation (IPA), tilted IPA, and plane-parallel approximation (PPA). The retrievals and the simulations strove for maximum realism in their operations and input parameters. The average 3D radiative effects of these cumulus cloud fields are little more than 1 W/m2 for reflected fluxes and 0.3 W/m2 for absorbed fluxes. This follows from their small cloud fraction and small (though still significant) total effect on reflected fluxes. The IPA, TIPA, and PPA errors correlate with some of the parameters of the cloud fields, including the statistics of the optical depth distribution, so it may be possible to predict and correct these errors when these approximations are used.; Finally, data from FIRE.ACE and SHEBA were used with an explicit radiative transfer model to explore the effects of inhomogeneous cloud and surface properties on solar and longwave radiative transfer in the summertime Arctic. Three representative cases were selected from three flights in May and July 1998, with detailed specification of cloud, surface, atmospheric, and aerosol properties. Results show that the surface inhomogeneity affects the fluxes at small scales but not in the domain average. The IPA and PPA can produce accurate domain-average fluxes for low arctic stratus clouds, although the IPA can create large local errors. Measured and modeled fluxes are reasonably well matched, within the bounds of uncertainty in the model inputs, of which surface albedo has the largest impact. The fluxes are also sensitive to cloud liquid water path, cloud droplet effective radius, and aerosol amount.; This work suggests that two cloud types which might be expected to have significant 3D radiative effects—small tropical cumulus and singer-layer summertime arctic stratus—have only modest 3D effects on large-scale radiative transfer.