Dynamical Environments of Organised Deep Convection in Climate Models

A longstanding limitation of all climate models is their inability to produce the correct distribution of precipitation, with light rain being too frequent and heavy precipitation being too rare. In the tropics, such intense rainfall is often the result of organized deep convection such as mesoscale convective systems. These features manifest as an extensive cover of deep convective and thick stratiform anvil clouds. While similar cloud signatures exist in climate models, it is unknown if they reflect the fidelity of convection in models.

Model convection is represented through parametrization schemes, in which the statistical effects of convection are quantified through relationships with the resolved large-scale variables. In this study, we analyze this relationship in models to uncover the origins of clouds that in observations are signs of organized deep convection. To do so, we use a cloud regime approach to identify organized deep convection in observations. The cloud regimes are derived using data from the International Satellite Cloud Climatology Project. The regimes have been shown to be good indicators of the state of tropical convection, including the degree of organization. Cloud regimes can also be defined in models, thus allowing us to examine the relationship between the organized deep convective regime and the large-scale dynamical environment in both observations and models.

The models are generally able to produce the cloud regime associated with organized deep convection in the real world. However, while the cloud structure and frequency of occurrence are comparable, the geographical distributions differ significantly from observations. Furthermore, vertical motions associated with this regime in many models are too strongly ascending. This implies that organized deep convective clouds in the models originate more from resolved dynamics of the large-scale environment and less from the self-organization of convection.