Quantifying Stochastic Forcing at Convective Scales

We are working to improve the prediction of convective weather systems and the effects of such systems on broader-scale forecasts. Specifically, we are developing a modeling tool, based on super-parameterization, that can be used to generate initial conditions for ensemble members that differ partially or even exclusively through stochastic perturbations of the parameterized cloud systems. This is beneficial because of the potential importance of convectively excited perturbations for generating spread among the ensemble members. Preliminary results will be presented.