A new framework for cumulus parametrisation informed by ARM data

Description

The basic principle underpinning the parametrisation of tropical convection in global weather and climate models is that there exist discernible interactions between the resolved model scale and the parametrised cumulus scale. Furthermore, there must be at least some predictive power in the larger scales for the statistical behaviour on small scales for us to be able to formally close the parametrised equations. The poster discusses both past and future work with ARM data in the tropics to develop a new framework for cumulus parametrisation based on the idea of separating the prediction of cloud area from that of velocity. Were show how observations at the tropical ARM site can be used to develop conceptual ideas for such a framework as well as serve to quantify key relationships between large and small scales. Examples are the identification of four key types of convection from observations, the quantification of the deterministic and stochastic components in the relationship between the large scale and cloud area fraction as well as the use of observations to derive the statistical properties of cloud scale vertical velocity. The implementation of the first stage of the new framework in the climate model of the Max-Planck-Institute for Meteorology shows great promise in improving the tropical variability without any major deterioration of the mean state. The next steps for the extension of the framework to allow for the co-existence of multiple cloud types and convective organisation will be highlighted. In doing so future ASR-ARM data gathering and analysis needs will be highlighted.