University of Utah
Salt Lake City, UT 84112
Based on aircraft measurements of mid-latitude cirrus clouds, ice crystal size distribution and ice water content (IWC) are shown to be dependent on temperature (Figure 10.1). This dependence is also evident from the theoretical consideration of ice crystal growth. Using simple models of the diffusion and accretion growth of ice particles, Liou shows that the computed mean ice crystal size and IWC compare reasonably well with the measured mean values. The temperature dependence of ice crystal size and IWC has important climatic implications in that the temperature field perturbed by external radiative forcings, such as greenhouse warming, can alter the composition of ice crystal clouds. Through radiative transfer, ice microphysics can in turn affect the temperature field. Higher IWC would increase cloud solar albedo and infrared emissivity, while for a given IWC, larger crystals would reduce cloud albedo and emissivity. The competing effects produced by greenhouse temperature pertubations via ice microphysics and radiation interactions and feedbacks are assessed by a one-dimensional radiative-convective climate model that includes an advanced radiation parameterization program.
The radiation model includes the delta-four-stream approximation for radiative transfer in nonhomogeneous atmospheres, the correlated k-distribution method for non-gray gaseous absorption, and the scattering and absorption properties of hexagonal ice crystals. The broad band results computed from the delta-four- stream method are found to be about 5% accurate over the full range of zenith angles and optical depths. In comparison, delta-two-stream (or Eddington) calculations, while less computationally expensive, show errors of 20-30% for a number of conditions (Figure 10.2). The correlated k- distribution method is a new technique for computing transfer of radiation involving absorption lines and is superior to the conventional band model approach that is coupled with the scaling and two-parameter approximations. It is exact for a single line and periodic lines, and the absorption coefficients so derived can be directly incorporated in the multiple scattering model for non- homogeneous atmospheres.
The radiation program is driven by the ice water path, the product of IWC and cloud thickness, and mean effective ice crystal size, which are parameterized in terms of temperature based on measured cloud microphysics data. It is important to use the scattering and absorption properties of hexagonal ice crystals in the calculation of solar albedo for ice crystal clouds. Use of equivalent spheres will provide a significant underestimation of solar albedo, because spheres scatter more in forward directions as well as absorb more incident radiation than non-spherical ice particles.
The feedbacks and interactions involving ice crystal diffusion and accretion growth, radiative processes in terms of solar albedo and infrared emissivity, and equilibrium surface temperature are investigated by using a one-dimensional climate model with CO2 doubling as the climatic forcing. Increasing temperature raises the ice crystal size through diffusion and accretion growth which reduces cloud albedo and emissivity. The combined effect of the feedbacks is to raise the 2 x CO2 surface temperature increase from a control run of 2.4 ° to 3.0 ° C with a net positive temperature- emissivity feedback dominating a net negative temperature-albedo feedback (Figure 10.3). Uncertainty factors, such as the parameterization of ice microphysics as a function of temperature, the radiative properties of small ice crystals and various ice crystal shapes, and the effects of cloud horizontal nonhomogeneity on radiation, require more comprehensive analyses of cloud and radiation data derived from composite field experiments and model simulations.
