Diurnal Cycle of Local Convection at South Great Plains from ARM Observations and Large Eddy Simulation

Global climate model has a well-known problem in simulating diurnal cycle of continental convective clouds and precipitation. In this study, we use ARM long-term observations to study the factors and mechanisms that controls the transition among locally-generated convective cloud types, such as forced and active shallow cumulus, and the shallow cumulus that do not stay shallow but transit to late-afternoon precipitating deep convective hot tower clouds. Particularly we focus on the impact on convective clouds from the temporal anomalies and the spatial heterogeneity of soil moisture and the associated meso-scale temperature, humidity and wind. In a parallel study using large eddy simulation, we focus on the effect on the length scale of surface flux heterogeneity and its coupling with the background wind speed. We find that when a non-dimensional parameter, F_hetero, is greater than 20, a well-defined non-precipitating shallow cumulus will transit to late-afternoon deep convection. Such non-dimensional parameter is further shown to distinguish the mesoscale transport of water vapor from turbulence transport in the associated meso-scale circulation development with increasing F_hetero. The goal of this work is to improve our process-level understanding of locally generated convections and thus to provide some insight on parameterization improvement and development.