[Previous] | [Session 53] | [Next]
T.E. Dowling, A.P. Showman (U. Louisville)
A long-term goal in the study of planetary atmospheres is to build a single general circulation model (GCM) that accurately simulates all the major atmospheres in the solar system, both gas-giant and terrestrial. Such a model will be indispensable for comparative planetology studies if it can achieve state-of-the-art accuracy while handling the different bottom boundary conditions of the two cases. In regions stable to convection (entropy increasing with height) that are away from rigid boundaries, the isentropic vertical coordinate has been shown to be the most accurate because it eliminates vertical differencing in the absence of heating, one of the major sources of error in GCMs. However, the isentropic coordinate encounters problems near the lower boundary in both the terrestrial and gas-giant cases. In the terrestrial case, the difficulty is the steep intersection of coordinate surfaces with the ground, which leads to poor accuracy near the surface. In the gas-giant case, the difficulty is a reduction in resolution in the weakly stratified lower troposphere, which compromises the simulation of multi-layered clouds. And in both cases, the coordinate fails in convecting regions because it becomes non-monotonic. The most accurate coordinate near a solid surface is a terrain-following coordinate, for example the sigma coordinate that is formed by dividing the pressure by the surface pressure. Konor and Arakawa (1997, Mon.\ Wea.\ Rev.\ 125, 1649-73) have recently devised a generalized vertical coordinate that blends the sigma coordinate seamlessly into the isentropic coordinate and captures the benefits of both, including being well defined in convective regions, which allows much deeper extension into a gas giant. The idea is technically challenging because the pressure and entropy are now both dependent variables everywhere, while a combination of the two must self-consistently serve as the independent vertical variable. We describe our progress adapting this scheme to the EPIC model Version 4.0, which can be used to simulate all the major atmospheres in the solar system.
The author(s) of this abstract have provided an email address for comments about the abstract: dowling@louisville.edu