The permeability of the Antarctic polar vortex is investigated in late austral winter and spring by comparing isentropic and three-dimensional (3-D) trajectories. Trajectory computations were performed with the help of the Royal Dutch Meteorological Institute (KNMI) trajectory model, using data from the European Centre for Medium-Range Weather Forecasts (ECMWF) from August to November 1998. Large numbers of air parcels were initially released inside and outside the polar vortex on the 350, 450, and 550 K isentropic surfaces. They were integrated 4 months forward in time in an isentropic mode, as well as in a 3-D mode that uses all three wind components from the ECMWF and takes into account diabatic heating and cooling effects. For the isentropic trajectory calculations, very little transport (0.37%/week) was found for August and September, while October and November gave somewhat higher transport rates (1.95%/week). The 3-D trajectory calculations for October gave much more exchange between the vortex and midlatitudes than the isentropic ones owing to a significant number of parcels that descended inside the vortex. Descent rates were calculated for 350 K (October), 450 K (August–October) and 550 K (October). Overall, the results show that 3-D trajectories will provide more accurate leakage rates than the isentropic ones. Also, despite the large-scale mixing in the polar vortex or in midlatitudes, little ozone-depleted air leaks from the ozone hole into the midlatitude stratosphere.
M Öllers, PFJ van Velthoven, H Kelder, L Kamp. A study of the leakage of the Antarctic polar vortex in late austral winter and spring using isentropic and 3-D trajectories
published, J. Geophys. Res., 2002, 107