Pielke Sr., R.A., D. Stokowski, J.-W. Wang, T. Vukicevic, G. Leoncini, T. Matsui, C. Castro, D. Niyogi, C.M. Kishtawal, A. Biazar, K. Doty, R.T. McNider, U. Nair, and W.K. Tao, 2007: Satellite-based model parameterization of diabatic heating. EOS, February 20 2007, pp 96-97.
This new approach of using observations to construct the parameterizations of the physics in the models, with the exception of advection and the pressure gradient force, offers an opportunity for much more computationally efficient and also accurate simulations of weather and climate. This approach also recognizes (as Henk Tennekes is emphasizing on his Climate Science weblogs), that weather and climate models are engineering tools, not fundamental physics.
Up until the present, the parameterizations of the physics in these models have used the concept of constructing models within models. However, as can be easily shown (e.g. see page 198 in
Pielke, R.A., Sr., 2002: Mesoscale meteorological modeling. 2nd Edition, Academic Press, San Diego, CA, 676 pp.),
ALL of the existing parameterizations contain tunable coefficients in which the tuning usually occurs from observations from specially selected data which does not encompass the real world variations in these physical processes. An attempt to introduce even more computationally intensive models within the parent model (which has been called âsuperparameterizationâ?) is not only computationally very expensive but that framework also requires significant assumptions, such a two-dimensional cloud fields.
Since these parameterizations are not fundamental representations of the physics, however, all that is needed is that they accurately represent the real world response to a set of input forcings. It is not important that the parameterization retain the physics of the real world as long as it faithfully replicates the response in the real world to the input forcings.
We refer to this new parameterization approach as a unified parameterization, and also as a âLook-Up-Tableâ? or âLUTâ?, which emphasizes that this is an engineering code. The LUT will use remotely sensed and real world observations for its construction as described in our EOS paper.
This technique can be applied to both numerical weather prediction models (where the initial conditions provide the reason for much of its skill) and to climate models (where the results depend on accurately simulating the forcings and feedbacks within the climate system).