COST-action number: ES0905
Basic Concepts for Convection Parameterization in Weather Forecast and Climate Models
Objectives and Description
Current numerical models of the atmosphere, both for numerical weather prediction (NWP) and for climate projection, have difficulties in accounting for unresolved variability, most notably, the variability associated with convection. Convection is a crucial driver of the atmospheric general circulation and is a key process for the vertical distribution of energy. It is responsible for the bulk of global precipitation. Among other known deficiencies of climate projections, global forecast models produce too early an onset of afternoon convection and fail to represent the 20-60 day planetary-scale tropical oscillation.
Current operational parameterizations use various ad hoc assumptions that often lack robust physical basis. The most notorious example is a closure that is required in order to define the total strength of convection produced by a parameterization. One more example is the rate of entrainment and detrainment characterizing the lateral exchange of air between a convective plume (convective tower) and environment. In spite of their critical importance in defining the vertical extent of deep convection, these parameters are currently simply tuned often in physically unreasonable way. More fundamentally, current parameterizations are designed to represent particular elements of the model physics (e.g., deep convection, boundary layer turbulence, radiation), whereas coupling between different physical processes (modules), that is crucial in view of the overall model performance, is to a large extend missing.
An increasing resolution poses new problems and requires a new generation of physical parameterizations. This is a challenge that urgently needs a strong action. Therefore, a core group focusing on Theoretical Studies of the Convection Parameterization Problem, a critical missing element, is implemented as COST Activity ES0905.