07/25/2011
Dissipating Instabilities in Climate Models
Summary
Climate model simulations are based on solving equations of motions in the atmosphere and ocean. To make the solutions feasible, the equations must be simplified, retaining only the most important terms. However, the solutions to these simpler equations can become unstable, resulting in unrealistic simulations. To address this issue, damping or dissipation is added to smooth out the unrealistic behaviors. Models either implement this dissipation directly, or the numerical methods used to solve the equations include smoothing effects. All climate models need smoothing, but there has not been a systematic evaluation of how various models achieve this. DOE-funded researchers have investigated and compared the dissipation processes used in the fluid dynamics component of climate models, providing a comprehensive survey of the diffusion, filters, and fixers in dynamical schemes of over 20 general circulation models. They focused on dissipation used in the Community Atmosphere Model (CAM), part of the DOE-supported Community Earth System Model (CESM) at the National Center for Atmospheric Research. Using idealized test cases, the investigators isolated causes and effects of individual dissipation mechanisms, demonstrating that the choice of the dissipation processes directly impacts the accuracy and stability of the simulations. Dissipation even has the potential to alter the large-scale circulation pattern and thereby the outcome of climate simulations. The survey reveals the important role that the stabilizing methods have on atmospheric dynamics and offers practical guidance in choosing adequate subgrid-scale mixing schemes.
References
Jablonowski, C., and D. L. Williamson. 2011. “The Pros and Cons of Diffusion, Filters, and Fixers in Atmospheric General Circulation Models,” published in Lauritzen, P. H., C. Jablonowski, M. A. Taylor, R. D. Nair (Eds.), Numerical Techniques for Global Atmospheric Models, Lecture Notes in Computational Science and Engineering, Springer, Vol. 80, 381–493. DOI: 10.1007/978-3-642-11640-7_13.