This talk is presented as a Zoom Webinar and requires a passcode. For anyone outside the department; If you would like to attend, email [email protected]

Stochastic and machine learning assisted models of  population dynamics of convective clouds and their applications to parameterization

The expansion of computational resources and numerical methodologies has allowed global weather forecasting and experimental climate models to run at horizontal grid spacing of 10s of kilometers or less. But this opportunity comes with the need to re-examine the representation of sub-grid convection processes such as the assumed “quasi‐equilibrium” (QE) balance between large‐scale (resolved) forcing and convection. In that spirit a stochastic framework for modeling the population dynamics of convective and stratiform clouds is proposed. In this framework transition functions are used to represent the evolutions of the number of convective cells of a specific size and their cloud‐base mass fluxes under given large‐scale forcing. In this presentation a hierarchy of models from probabilistic and physical reasoning based simple models to machine learning assisted approach to simulating the interactions between convective and stratiform clouds are discussed. Ongoing work on the application of such models as parameterizations of convection in a regional model will be discussed.