, program directors.
SUMMARYWe have used a hierarchy of numerical models for cirrus and stratus clouds and for radiative transfer to improve the reliability of general circulation models. Our detailed cloud microphysical model includes all of the physical processes believed to control the lifecycles of liquid and ice clouds in the troposphere. We have worked on specific GCM parameterizations for the radiative properties of cirrus clouds, making use of a mesoscale model as the test-bed for the parameterizations. We have also modeled cirrus cloud properties with a detailed cloud physics model to better understand how the radiatively important properties of cirrus are controlled by their environment. We have used another cloud microphysics model to investigate of the interactions between aerosols and clouds. This work is some of the first to follow the details of interactions between aerosols and cloud droplets and has shown some unespected relations between clouds and aerosols. We have also used line-by-line radiative transfer results verified with ARM data, to derive a new set of gas absorption data for the sort of rapid radiative transfer models used by GCMS. In total 9 people, two of whom are graduate students, have been working part or full time on this and other projects related to our work for DOE.
FINAL REPORTWe used a hierarchy of numerical models for cirrus and stratus clouds and for radiative transfer to improve the reliability of general circulation models. Our detailed cloud microphysical moclcl includes all of the physical processes believed to control the lifecycles of liquid and ice clouds in the troposphere. Below we outline work on GCM parameterizations, cirrus cloud studies, investigations of the interactions between aerosols and clouds, as well as studies of radiative properties of clouds.We have developed a consistent treatment of cirrus cloud microphysical and radiative processes for use in GCMs. The resea.rch-grade GCMs are faced with two related problems:(1) the bulk-water cloud paraiiieterization assumes size clistributions, particle shapes, and process rates t1ia.t are inappropria.te Tor cirrus: and (3) the radiation codes use size distributions and particle shapes that are decouplecl from, and inconsistent with, those used in the microphysics. We have stucliecl these problems using the PSU/NCAR MM4 mesoscale model which employs a bulk-water cloud scheme similar to those being incorporated in GCkIs. A s an example of the first problem, we found that the original cloucl schenie developed by Liii et (if. and Hobbs el nf., and widely used, assumed too many large cloud particles and unclerestimstes the total number and total surface area. As a result, the bulk fall velocity, accretioii, and other processes were overestimated. We have
DISTRIBUTION OF THIS DOCUMENT I S UNUMKED