Cloudiness associated with extratropical cyclones is currently poorly represented in GCMs due to incorrect and insufficient representation of subgrid-scale processes, leading to erroneous cloud-climate feedbacks. This project will develop an understanding of the relationship between mesoscale variability and large-scale synoptic forcing so that mesoscale variability and its impacts on clouds and water vapor may be parameterized in a GCM. Observations from the ARM SGP/CART site will be used to characterize mesoscale distributions of cloud cover, temperature, and moisture as a function of large-scale forcing. Observations from the SGP March 2000 IOP will be combined with output from the Regional Atmospheric Modeling System (RAMS) forced by observed boundary conditions to provide a detailed set of cloud and dynamical information. An ensemble of runs will be carried out to investigate what processes are responsible for producing mesoscale circulations and cloud structures and how much the large-scale synoptic forcing determines specific mesoscale features. Output from a high-resolution run with enhanced microphysics will be made available to the ARM community for March 2000 IOP cases. The results of the observational analyses and RAMS simulations will be combined with theory for frontal rainbands to develop a parameterization representing the mesoscale distribution of synoptically generated cloudiness. The resulting parameterization will be tested and implemented in the GFDL Flexible Modeling System and the NCAR Community Atmosphere Model.