Winter polar stratospheric nitrous oxide (N2O) measurements made during two NASA polar aircraft field campaigns provide a unique opportunity to evaluate the performance of the 1¿ latitude resolution version of the Geophysical Fluid Dynamics Laboratory's ''SKYHI'' general circulation model. This high-resolution model has been integrated 20 months, producing one Antarctic and two Arctic winters. Power spectra of the dynamically controlled tracer N2O are used as a diagnostic of wave activity. Comparison of the spectra of SKYHI and the observations shows that the SKYHI Arctic winter lower stratosphere is dynamically active enough to generate realistic mesoscale tracer variability but that the SKYHI Antarctic has deficient variability at scales of 220--3000 km. Low-pass filtering is applied to a new type of analysis that attempts to discriminate between different sources of atmospheric variability, to the extent that different sources are characterized by different timescales. The goal is to diagnose mesoscale sources of tracer variability in the model and in the observations and then to assess whether SKYHI generates variability for the right physical reasons.

This analysis shows that variability from ''slow'' processes such as planetary wave breaking dominates and is generated in realistic amounts in the SKYHI Arctic winters. The SKYHI Antarctic vortex shows insufficient ''debris'' from planetary wave breaking at scales below 700 km. The balance between diabatic descent inside the vortex and wave breaking in the ''surf zone'' generates N2O gradients at the vortex edge in the model and the real atmosphere. Because the diabatic circulation is driven by wave activity, the strength of model wave activity diagnosed by the spectral analysis and the mean N2O gradients can be used to evaluate SKYHI's diabatic circulation and net tracer transport. In the Arctic, SKYHI temperatures, spectral results, and realistic N2O gradients at the vortex edge suggest a reasonable diabatic meridional circulation and transport. Antarctic spectral results, low vortex temperatures, and flatter N2O gradients at the edge all support the conclusion that the diabatic circulation and wave activity in the model southern hemisphere is too weak. ¿ American Geophysical Union 1994