| Cumulus convection is a major source of small-scale variability in the atmosphere, and a particular problem for the numerical models used for weather and climate prediction since the variability is strong on scales close to the model resolution. A stochastic representation of convection is typically required, but this requires knowledge of how the convection is coupled to other processes in the atmosphere. A crucial distinctionis whether the convective environment is sufficiently slowly varying for the convection to reach a statistical equilibrium. This presentation will review properties of equilibrium and non-equilibrium convection, and discuss consequences for stochastic parameterisation, ensemble prediction and data assimilation. |
|