The impact of orographic variance on boundary layer clouds and its parameterization for climate models

A successful simulation of time-averaged low-level cloudiness and associated radiative forcing is required if climate models are to produce realistic representation of the earth's energy and hydrological budgets. To achieve success, radiative and turbulent processes responsible for the formation, maintenance, and dissipation of planetary boundary layer (PBL) clouds must be appropriately formulated. The orographic variance and associated thermal circulations induce inhomogeneities in the cloud field that can significantly alter the overall behavior of the PBL-ground system. This effect has been ignored in climate models.; This work investigates the macroscopic impact that orographic variance has on PBL clouds and proposes a parameterization of this effect. The strategy is based on performing a large systematic set of experiments with a cloud resolving model (CRM) in which the diurnally varying interactions between PBL clouds and the topography are explicitly simulated for a variety of configurations. The analyses of the results strongly suggest that the orographic effect on PBL clouds is parameterizable. A parameterization is then formulated and further implemented and tested in the UCLA atmospheric general circulation model (AGCM).; It is found that including variable topography in the CRM reduces the bi-modality that the PBL behavior exhibits otherwise. Parameterizability of the effect of orographic variance on PBL clouds arises from a quasi-equilibrium between cloud-radiative, turbulent and mesoscale processes that possesses an inherent degree of inhomogeneity in the cloud field. A statistical parameterization of the liquid water path (LWP) spatial distribution was then formulated as a function of the mean PBL state. The LWP distribution is characterized by an exponentially decaying function that describes the patchy clouds, and a bell-shaped function that describes the solid cloud decks. These results are supported by previous observational work. The standard deviation of the orography variance impacts both cloud fraction and the relative area cover of the patchy and solid clouds, mainly by reducing the incidence of the latter. The characteristic horizontal scale of the orography, on the other hand, has little influence on PBL cloudiness.; The inclusion of the parameterization in the AGCM also reduces the bi-modality of the PBL behavior. In particular, it prevents the PBL from falsely falling into a cold and cloudy regime and thus results in a much more realistic simulation of low-level cloudiness.