Gustafson Jr., William I., Po-Lun Ma, Balwinder
Singh, Jerome D. Fast, and Philip
J. Rasch; Atmospheric Sciences and Global Change Division, Pacific Northwest
National Laboratory
No matter the scale
simulated, all non-idealized atmospheric models must account for physical
processes such as turbulence, radiation, and clouds. However, modeling at
different scales requires different approaches. A large eddy simulation (LES)
handles clouds differently than a global climate model, and a mesoscale model
lies between these two extremes. At one end of this spectrum, one can view bulk
cloud processes as practically resolved, but at the other end they are entirely
parameterized. Even so, micro-scale processes, such as condensation, must be
parameterized at all practical grid spacings. The continuum of atmospheric
models requires careful consideration of the scales involved and the best way
to handle each process for the given scale. Today, WRF contains physics modules
designed for the full model spectrum, from the eddy-scale to the global-scale.
What are the differences between the approaches used in the schemes? How do
cloud schemes designed for global resolutions behave at finer scales? This
presentation will give an update on implementation of the Community Atmospheric
Model v5 physics suite in WRF, and compare this physics suite against a
traditional mesoscale physics suite to better understand how the schemes behave
at mesoscale resolutions. The Separate Physics and Dynamics Setup (SPADS),
which allows one to target parameterization behaviors at different scales
through the use of independent dynamics and physics resolutions, will be
employed to understand the scaling behavior of the two physics setups.