Intercomparison of methods of coupling between convection and large-scale circulation

Monday, 15 December 2014: 5:21 PM
Chimene Laure Daleu1, Steve J Woolnough2, Robert Plant2, Adam H Sobel3, David J Raymond4, Sharon L Sessions4, Shuguang Wang5 and Gilles Bellon6, (1)University of Reading, Reading, RG6, United Kingdom, (2)University of Reading, Reading, United Kingdom, (3)Columbia University, Department of Applied Physics and Applied Mathematics, New York, NY, United States, (4)New Mexico Tech, Socorro, NM, United States, (5)Columbia University of New York, Palisades, NY, United States, (6)University of Auckland, Auckland, New Zealand
In the past decade, a set of methods of parameterizing large-scale dynamics in limited area models have been developed based on our physical understanding of the tropical atmosphere. One of the methods, the weak temperature gradient (WTG) is to derive the large-scale circulation from the assumption that its action counteracts the local buoyancy anomaly and thus, maintains the horizontal variations of potential temperature close to zero in thedeep tropics. Another method, the damped-gravity wave (DGW) is to derive the large-scale circulation directly from the momentum equation. Other methods of parameterizing large-scale dynamics have been developed but this study considers only the WTG and DGW methods.

These methods of parameterizing large-scale dynamics offer a computationally cheap way to study a range of problems in which two-way interactions between tropical convection and large-scale dynamics are essential. They have been applied in both Cloud-Resolving Models (CRMs) and Single Column Models (SCMs). From such studies, much insight has been learned on what controls large-scale variation of tropical deep convection. However, both agreement and discrepancies between different studies are seen in the published results.

The GASS-WTG international intercomparison project was initiated in early 2014 to compare the WTG and DGW methods with a consistent implementation in a number of CRMs and SCMs, and to compare the behaviour of CRMs and SCMs under consistent parameterizations of the large-scale dynamics. Here, we discuss some of the main differences that have emerged between the models and methods. For instance, the WTG simulations produce a wider range of behaviour either comparing models against each other or against their own radiative convective equilibrium state. In addition, they are more sensitive to the choice of the initial states than the DGW simulations. In contrast to the DGW simulations, two distinct equilibria corresponding to either a persistent, precipitating state or a non-precipitating state are shown to exist in most models that applied the WTG methods. Furthermore, some models show a strong SST dependence in their behaviour, and finally some models produced
increased precipitation from the dry initial states.