A COMPARISON OF 2 BULK MICROPHYSICAL SCHEMES AND THEIR EFFECTS ON RADIATIVE-TRANSFER USING A SINGLE-COLUMN MODEL

Increasingly, numerical models in climate studies are using prognostic bulk microphysical schemes to predict grid-scale cloud cover and prop erties. These schemes provide information which can lead to improved c alculations of radiative transfer, and a better understanding of the i nteraction of radiation with cloud microphysics on the large scale. In this study a one-dimensional, hydrostatic column model with fixed ver tical velocities includes two different bulk microphysical schemes to investigate how the type of scheme influences the hydrometeor content in a cloud, and the effect of this on the radiative heating rates thro ugh the cloud. Two test cases are performed, one representing the stra tiform region of a tropical cloud cluster, the other a dissipating tro pical cirrus cloud, Each test is first performed using a microphysical scheme that carries only one variable for solid water (MS1), and then using a scheme that separates solid water into ice crystals, snow, an d graupel (MS2). Further sensitivity tests are made using MS2 to exami ne the effects of excluding graupel, and of allowing the ice crystals to fall. The influence of these modifications on the hydrometeor conte nts of the clouds and the corresponding radiative heating rates is con sidered. In a simulation of the stratiform region of a tropical cloud cluster, MS1 is shown to produce significantly larger hydrometeor cont ents than MS2. However, in the simulation of a dissipating cirrus clou d, the ice content predicted by MS2 remains much larger than the ice c ontent predicted by MS1, throughout a 24-hour integration. This is bec ause there is a non-precipitating ice category in MS2 which is very sl ow to convert to snow at low ice-water contents. The use of a non-prec ipitating ice variable is shown to have a major impact on both the sol ar and the infrared radiative heating rates at the cloud top, and in s ome cases to give unrealistic predictions of cloud ice contents.