Recently Achatz and Branstator (1998) have shown that the internal variability of a perpetual January GCM can be well reproduced by reduced (EOF-based) models with empirically determined linear parameterizations of the impact of unresolved physics and spatial scales. Given that already 30 degrees of freedom where sufficient for a faithful simulation of essential features such models have a much increased transparence and handiness in comparison to current GCMs, thus possibly enabling deeper insight into atmospheric dynamics. For reasonable applicability of semi-empirical reduced models to climate studies, however, the impact of the atmosphere's coupling with the ocean has to be taken into account. With this in mind a reduced coupled ocean/atmosphere model is currently being developed along lines followed also in the atmosphere-only context. The empirical work rests on data from a long run of a complex coupled AGCM/OGCM (ECHAM3/LSG). The model design has the following central ingredients: (1) Baroclinic (i.e. threedimensional) EOFs as basis of state space. They are being calculated using a total energy metric so that the various prognostic fields are weighted in a dynamically meaningful manner. (2) Energy conserving nonlinear terms calculated from the primitive equations. (3) Empirical linear parameterizations of the impact of unresolved physics and spatial scales, as well as of the coupling between the two subsystems. The reduced models' climatology is compared with that of the complex coupled AGCM/OGCM, with special emphasis on the dependence on the number of basis patterns treated explicitely.
Close window or click on previous window to return to the Conference Program.
12th Conference on Atmospheric and Oceanic Fluid Dynamics