Dynamics of the Antarctic ozone hole dissipation revealed by ground-based and satellite observations

In order to examine dynamics of the Antarctic ozone hole dissipation processes in detail, an intensive observation was performed using ozonesondes at Syowa Station (39.6^oE, 69.0^oS) in late June 2003 through early January 2004 when the ozone hole was developed to the third largest in the past. Eighty seven vertical profiles of ozone were successfully obtained. Observed ozone increased earlier at higher altitudes in the dissipation phase of the ozone hole, which marked contrast with the developing phase when it decreased almost simultaneously in the ozone layer in the height region of 14-22km until late September. The earlier ozone recovery at higher altitudes started already in late August at the latest. The decent rate of the ozone mixing ratio level of 1.0 ppmv around 20km is estimated at about 1.1 ± 0.2km per month in a time period of late September through late October before the polar vortex breaking. A dominant process of the ozone recovery before the polar vortex breaking is considered to be downward transport by the diabatic circulation as discussed by previous studies, but this is not the only process.

Using data from satellite onboard Improved Limb Atmospheric Spectrometer-II (ILAS-II) which are distributed uniformly in the zonal direction, we examined the decent rate of ozone as a function of longitude in the polar vortex in the same time period, and found its significant longitudinal dependence. An analysis of dynamical structure of the polar vortex using ECMWF operational data shows that the longitudinal dependence is consistent with the decent rate of isentropes modified by a wavenumber 1 "quasi"-stationary planetary wave. Note that the decent rate is different from the vertical wind component of the planetary wave. An important fact is, however, that the longitudinal dependence remains even if the decent rate is estimated relative to isentropes. In other words, the ozone mixing ratio and its increase are not uniform in the polar vortex even on an isentropic surface. From a backward trajectory analysis, it is shown that the air parcels observed in the sector around the longitude of 180^o with large ozone mixing ratio were mostly transported from the edge region of the polar vortex, while the trajectories of air parcels in the sector around 0^o with small mixing ratio remain in the central part of the polar vortex. This fact means that the lateral transport/mixing is important even in the time period when the polar vortex is stable.

The mixing ratio of N₂O (one of a long-lived species) obtained by ILAS-II observations is also analyzed to examine the diabatic circulation in the polar vortex. An interesting result is that the decent rate of N₂O mixing ratio of 30 ppbv around 20 km is only half of that of ozone, suggesting that the ozone is recovered faster than expected by diabatic transport. This is consistent with the importance of lateral transport of ozone-rich air as suggested by the trajectory analysis. Another possible process to explain faster ozone recovery is in-situ chemical ozone production with solar radiation in spring. Quantitative estimation is needed using chemical transport models for future works.