A PV View of Zonal Flow Vacillation

Abstract

It is suggested here that the zonal flow vacillation of the Southern Hemisphere is caused by a mutual interaction between the barotropic shear of the zonal flow and the evolution of baroclinic eddies during the later stages of their lifecycle. An index of the zonal wind anomaly difference between 40° and 60°S is defined for a 13-yr record of European Centre for Medium-Range Weather Forecasts (ECMWF) analyses. Periods are selected during which the zonal flow is at the extremes of its characteristic vacillation, which are distinguished by a broad westerly jet with minor maxima near 30° and 60°S and a narrower jet peaked near 40°S. The narrower 40°S jet has stronger cyclonic shear across the baroclinic zone. These changes are accompanied by shifts in poleward eddy momentum flux on the 500-hPa surface and potential vorticity fluxes on the theta surfaces that lie in the midtroposphere, which are both consistent with the maintenance of the zonal wind anomalies against friction. Series of synoptic maps of potential vorticity (PV) on isentropic surfaces suggest a very different evolution of the lifecycles of baroclinic waves in the two extreme composites. In the broad jet configuration, the vortex (region of high PV) is confined to polar latitudes, and cyclones that develop are relatively quickly distorted by anticyclonic zonal wind curvature. In the narrow midlatitude jet case, the vortex is expanded and cyclones develop within the region of cyclonic zonal wind curvature within the vortex. These cyclones tend to have a more isotropic shape and maintain their identity longer. In the later stages of their development they tend to roll downstream in the cyclonic curvature of the zonal flow and contribute to the maintenance of the expanded vortex structure. The different evolution of cyclones during the later stages of their lifecycles reinforces the zonal flow anomalies in such a way as to provide the mechanism for the vacillation. The different cyclone lifecycles found in the observations here correspond to differences found in idealized numerical experiments when similar differences in initial zonal wind structure are imposed.