A dual mass flux framework for boundary layer convection. Part I: Transport

Title
A dual mass flux framework for boundary layer convection. Part I: Transport
Report
Date Published
03/2007
Series/Collection
ECMWF-ARM Report Series
Document Number
2
Author
R. Neggers
M. Köhler
Anton Beljaars
Abstract The eddy diffusivity -mass flux (EDMF) approach for turbulent transport in well-mixed layers is extended into the modeling of shallow cumulus convection. Model complexity is enhanced to enable representation of conditionally unstable cloud layers that are flexibly coupled to the mixed layer. This significantly expands the range of applicability of EDMF, in principle including all major regimes of boundary layer convection and transitions between those. The treatment of subgrid transport and clouds is integrated by parameterizing both in terms of the same turbulent joint-distribution. This potentially skewed distribution is reconstructed using an ensemble of resolved updrafts, rising from the surface layer. Part I of this study concerns the formulation of this multiple updraft framework. A key new ingredient is the application of flexible area partitioning in the updraft ensemble, which is determined by the coupling between cumulus clouds and the sub-cloud mixed layer. This is achieved by defining and resolving two specific groups of updrafts; dry mixed-layer updrafts that never reach their lifting condensation level, and moist updrafts that condense and become positively buoyant cumulus clouds. This technique facilitates the representation of gradual transitions to and from shallow cumulus convection, and implicitly represents the impact of cloud base transition layer stability on cumulus transport. Other upgrades include i) flexible updraft entrainment rates, ii) stability feedbacks on the vertical structure of cloudy mass flux, and iii) the introduction of an entrainment efficiency closure for transport into the cumulus inversion. Impacts of these new components on boundary layer structure and equilibration are assessed.