Key points are not available for this paper at this time.
A multilevel ensemble-averaged model has been developed to study the cloud-topped atmospheric boundary layer (ABL). Turbulence closure is formulated by using an equation for the turbulent kinetic energy and either a diagnostic formulation of the integral length scale or a parameterized version of the dissipation equation. The latter two options are compared. The model is used to study various combinations of physical processes in a cloud-topped ABL and their combined effect on the turbulent structure. The physical processes considered are an upward buoyancy flux at the surface, longwave radiative cooling near cloud top, shortwave radiative heating in the cloud, and wind shear near cloud top. We discuss a case with only a surface buoyancy flux (no radiation) and a case with only longwave radiative fluxes (no surface fluxes). The usual concept that the latter is the upsidedown version of the former is not confirmed by the model results. Furthermore, we apply the model to the datasets of Brost et al. and Nicholls. Tile pronounced differences in the observed turbulent structure of the ABL in these two cases (due to different combinations of physical processes) are well simulated by the model.
Duynkerke et al. (1987) studied this question.