Physical-based validation of unified parameterization based on satellite observation of cloud and precipitation properties

The vertical distribution of clouds and cloud properties in the atmosphere is a signature of complex interaction between multiple physical processes. These processes include boundary layer turbulence, convective transport, large-scale condensation, microphysical processes and radiation. In this work we describe a novel methodology for developing and validating unified parameterizations for global climate models. The crux of the new methodology is to infer possible problems and biases in parameterization by comparing observed and modeled cloud properties. To this end we created a multi-year long database, which combines collocated cloud and precipitation observations from CloudSat, CALIPSO and MODIS data collocated with meterological fields from the (MERRA) Modern-Era Retrospective analysis for Research and Applications. Our work focuses on the tropical and subtropical marine atmosphere and the transitions between cloud regimes. The cloud cases were simulated with our unified stochastic multi-plume parameterization (Suselj et al., 2012) in the framework of single-column-model inititalized and forced by the MERRA meterolology. The model results were compared to the observations as a function of thermodynamic and circulation regimes. In this presentation we describe how this methodology helped improve and validate description of physical processes in the parameterization.