A number of numerical experiments using different assumptions on the inital odd nitrogen content and the efficiencies and durations of heterogeneous reactions is selected based on the simulations of the observed trace gas concentrations during the Airborne Antarctic Ozone Experiment (AAOE) in 1987 (Rodriguez et al. this issue). The model is used to calculate the seasonal behavior of O3 in 1987 and its possible interannual variations as the chlorine content of the atmosphere increases from 1 to 6 ppbv. The seasonal behavior in 1987 is compared with the total ozone mapping spectrometer (TOMS) and AAOE data. In all the cases considered, the catalytic cycle associated with the formation and photolysis of Cl2O2 could account for more than half of the photochemical removal of O3 within the Antarctic vortex through mid-September. The reaction of BrO with ClO, which accounts for 15--20% of the removal in the same period, tends to play a more important role toward the end of September, when the concentration of ClO is expected to decrease. The detailed behavior of the calculated O3 depends on the assumptions for the spatial and temporal extent of denitrification, the initial concentrations of odd nitrogen species sequestered in the vortex in July, and the efficiences and durations of the heterogeneous reactions in months that follow.

Additional observations of ClO, odd nitrogen species, and O3 similar to those obtained in the campaign covering the period in July and October will be useful in providing additional constraints to narrow down the possibilities. There is no simple relationship between the increase in chlorine level and the interannual decrease in Antarctic O3. It is difficult to explain both the 40% decrease in October mean O3 within the hole since 1979 and the much slower decrease prior to the mid-1970s without involing effects from interannual variations of heterogeneous activities and initial odd nitrogen concentrations in July. Such effects must be isolated before the observed trends in the Antarctic O3 hole in the past 8 years can be used to predict future behavior of O3 based solely on projected increase in the chlorine content of the atmosphere. ¿ American Geophysical Union 1989