In this paper we study the baroclinic equilibration of zonally varying flow, using a simple q-g two-layer model. The model is relaxed linearly to some zonally asymmetric equilibrium profile. The forcing is strong enough that there remains at equilibrium significant longitudinal contrasts in baroclinicity over the channel. Our aim is conceptual, to understand the role played by finite propagation time scales on the wave-mean flow interaction.

Baroclinic adjustment theories (for zonally symmetric flow) assume that because the dynamical time scales are much faster than diabatic time scales, the time-mean flow should be close to neutral. However, in the asymmetric problem there is an additional time scale, corresponding to the time it takes a wavepacket to propagate through a length scale characteristic of the zonally varying baroclinicity. If this propagation time scale is much longer than the dynamical time scale, one would expect the eddies to equilibrate locally over the more unstable regions, and to have vanishing amplitude over the less unstable ones. On the other hand, when the propagation time scale is comparable to the dynamical time scale the eddies do not have time to equilibrate locally, and are more sensitive to the zonal mean properties.

To understand these two limits, we have designed a series of experiments in which the propagation time scale is artificially changed by adding a uniform component to the zonal wind. We discuss the dynamics of the different regimes observed and their possible relevance for the Northern Hemisphere winter storm track.