A conditional sampling of boundary layer plumes for the validation of mass flux parameterizations

Recently, more and more regional and climate models have a boundary layer parameterization based on a mass flux scheme that describe thermals from the surface layer to the top of the cloud layer. Nevertheless there is a lack of observations of coherent structures in order to validate such scheme especially in the transition from the subcloud layer to the cloud layer. Coherent structures in the dry convective boundary layer have been investigated for a while but, the construction of the conditional sampling is still a critical issue. Here we propose a conditional sampling based on the total water mixing ratio perturbation and applied on LES to characterize the organised structures in the cloud free and cloudy boundary layer. LES by resolving the coherent structures consist of an adapted tool to evaluate different conditional samplings. Especially, they provide information on the transition between the subcloud layer and the cloud layer where there is a lack of available observations and also have the advantage of providing a full three dimensional fields of different variables. This new conditional sampling enables to retrieve plume characteristics, entrainment and detrainment rates, variance and fluxes in Large Eddy Simulations and therefore extensively evaluate a mass-flux parameterization in the subcloud and cloud layers. The main advantage of such sampling is to be able to describe the transition from the subcloud to the cloud layer. First, this new sampling is evaluated against observations and more traditional samplings such as cloud and core samplings. Then, it is used to evaluate the mass flux parameterization of Rio and Hourdin (2008). It reveals some deficiencies of the thermal plume model leading to improvement of the vertical velocity equation and redefinition of entrainment and detrainment rates.