Simultaneous, in situ observations of ClO, BrO, O3, N2O, pressure, and temperature are used to examine the kinetics of ozone destruction within the Antarctic polar vortex. The high azimuthal symmetry of the vortex is used to simplify the problem such that the time rate of change of ozone on isentropic surfaces between 450 and 360 K (18.3- to 13.5-km geometric altitude) can be calculated based on proposed catalytic cycles involving ClO, BrO, HO2 and O(3P). The availability of simultaneous, spatially resolved data on N2O, pressure, and temperature is used to reduce scatter in observed O3; this clarifies the quantitative analysis of ozone loss rates on isentropic surfaces. The kinetics of ozone destruction is thus cast in an N2O-potential temperature coordinate system to suppress atmospheric variability. We conclude, based upon observed radical concentrations and the most recent laboratory kinetics data, that the chlorine dimer mechanism (Molina and Molina, 1987) rate limited by ClO+ClO+M→ClOOCl+M, referred to herein as mechanism I, constitutes the largest single contribution to the integrated rate of ozone destruction within the vortex on isentropic surfaces between altitudes of 14 and 18.3 km.

We find that approximately 40% of the observed ozone loss rate in this altitude interval results from this mechanism, using the recently established pressure and temperature dependence for the rate of dimer formation reported by Sander et al. (1989). The coupled bromine-chlorine catalytic cycle (McElroy et al., 1986) rate limited by ClO+BrO→Cl+Br+O2 (herein mechanism II) constitutes approximately 20% of the ozone loss rate integrated over the course of the mission, August 23 to September 22, 1987. Mechanism III, suggested by Solomon et al. (1986), which is rate limited by ClO+HO2→HOCl+O2 (in close competition with OH+O3→HO2+ O2), constitutes 4% of the observed ozone loss rate based on observed ClO, but calculated HO2 concentrations. We find that the time-honored catalytic coupled ClO+O→Cl+O2 followed by Cl+O3→ClO+O2, identified by Molina and Rowland (1974) as the key reaction linking chlorofluorocarbon release to middle- and low-latitude stratospheric ozone depletion, consitutes 3% of the observed loss in the Antarctic vortex. The sum of all four mechanisms thus constitutes 65% of the observed O3 loss with an experimental uncertainty of ¿30%. An array of other halogen-oxygen radical catalytic cycles are considered. The manifold of possible catalytic cycles is large, particularly when the temperature and pressure regime characteristic of the Antarctic lower stratosphere is considered. Quantitative analysis is impeded primarily by the absence of laboratory data on key reaction rates. For example, a potentially critical alternative channel, designated herein as mechanism V, is the termolecular process ClO+O3+M→ClO⊙3+M, followed by photolysis, for which no kinetic data exist. Explicit note is also taken, in the quantitative conclusions, of the ozone flux divergence within the chemically perturbed region reported by Hartmann (1989). This constitutes an additional 20¿10% requirement on the rate of chemical removal. ¿ American Geophysical Union 1989