Combining representations of entrainment and organization to better model atmospheric variability

Entrainment and organization are two key aspects of convective phenomena that have been difficult to represent in climate models. Convective parameterization closure has tended to be based on a pseudo-adiabatic approximation to determine strength and vertical extent of convection. This approach includes no consideration of liquid/ice processes or entrainment effects and is thought to be at least partly responsible for poor GCM performance when simulating climate variability features such as the Madden-Julian Oscillation and the diurnal cycle. Organization is ubiquitous to convective processes and tends to follow a common pattern of shallow to mid-level to deep in extent, occurring in features spanning many horizontal scales.

We have developed a convection scheme that employs representations of entrainment and organization aimed at reproducing the convective information flow seen in observations. Essentially the scheme allows for high sensitivity to environmental moisture during low-organization scenarios and low sensitivity during high organization scenarios. The convection-scheme configuration will be described briefly and the results from a simple 4-level dynamical beta-plane model will illustrate its benefits. The inclusion of a greater sense of information flow into the formulation of entrainment and organization and their co-relationship results in improved performance, such as realistic wave activity, even in this simple modeling framework.