Direct Bayesian model reduction of smaller scale convective activity conditioned on large-scale dynamics

Abstract. We pursue a simplified stochastic representation of smaller scale convective activity conditioned on large-scale dynamics in the atmosphere. For identifying Bayesian model describing relation different scales we use probabilistic approach by Gerber and Horenko (2017) called Direct Model Reduction (DBMR). This is between categorical processes (discrete states), formulated via conditional probabilities. The available potential energy (CAPE) applied as flow variable combined with subgrid time series for vertical velocity. found CAPE up- downdraft day night. strategy part development process parametrizations models atmospheric representing effective influence unresolved motion flows. direct provides basis further research other possible drivers.