| The role of convection and clouds in the global climate system is important but not well-understood. Cloud feedbacks are the largest source of uncertainty in estimates of climate sensitivity, therefore a better understanding and representation of convection and cloud processes is of paramount importance for climate science. Global atmosphere models used in climate studies and weather prediction have resolutions that are too coarse to resolve convection explicitly. Instead, the bulk effect of convection on resolved-scale flow is represented through parameterization schemes. Traditional, deterministic parameterizations of convection are not functioning well, as their reliance on the quasi-equilibrium assumption is becoming more problematic with increasing model resolution. In the research proposed here, a stochastic convection parameterization scheme will be developed, making use of a recent mathematical innovation in which subgrid-scale processes are represented by stochastic processes fitted to data. The stochastic nature of the scheme allows to overcome the inherent limitations of deterministic schemes. We will use data from convection-resolving Large Eddy Simulations (LES) to fit the scheme, giving a systematic, data-driven approach to the formulation of a stochastic convection scheme. The scheme will be implemented in the climate model EC-Earth and evaluated for present climate conditions using observational datasets. We will assess the impact of the new scheme on modeled climate sensitivity through cloud feedbacks. |
