The radiative balance of the earth drives the global circulation, with latent heat effects responsible on average for some 20% of the energy transport. It is important to assess the role of clouds in this balance---not only from latent heat considerations, the detail of which is dependent on topics of this article, but also from the view point of such clouds in absorbing and scattering solar and thermal radiation. Changes in CCN give changes in cloud drop size distribution; it may be hypothesized that changes in CCN give changes in cirrus crystal size distribution as discussed earlier. Such effects may be anthropogenic---aircraft exhausts, urban or industrial effluent; they may also be of volcanic origin. In general radiative effects of water clouds are subject to Mie computation, subject to a known spatial distribution and size distribution of the droplets. Real and imaginary refractive index of water is known, and these calculations can be performed (Chýlek et al, 1992). Indeed, calculations demonstrate the extent of laser transmission and absorption at specific wavelengths, and demonstrate the ability to evaporate fog droplets under specific conditions (Young and Tomlinson, 1993).
Problems arise primarily from unknown absorption by soot and mineral particles, which under some conditions can be important. It is appreciated that many small ice crystals in cirrus can dominate the radiative properties, and valiant attempts are being made to compute these effects (Takono et al, 1992); the existence of such crystals is being well documented by aircraft penetration (Arnott et al, 1994). Not only is cirrus a cloud with complex structure but the size spectrum, habits and skeletal structure of cirrus crystals changes from time to time and place to place. Some generalities exist, but prediction of radiative budgets pose a problem. Laboratory studies can provide a significant contribution here, in terms of empirical measurements of the wavelength dependence of scattering and emissivity in well controlled crystal sizes shapes (Arnott et al 1995). This is almost impossible to do in the real atmosphere because of the variability; it is almost equally difficult to compute because complexities of shape are often of the order of the wavelength of interest. A possible solution is to measure the properties in the laboratory and determine the sensitivity to input ice distribution, assessed by a mesoscale model input on the scale of some hundreds of kilometers. A major challenge is to produce in the laboratory uniform particle size and shape. Analytical formulation of particle spectra are often figments of the modelling imagination. Observational and hence trajectory prediction are required to make sense of the detailed structure of the cirrus in terms of particle nucleation and growth. Significant advances in this area would seem within our grasp.