Depletion of column ozone in the Arctic during the winters of 1993-94 and 1994-95

The total ozone reduction in the Arctic during the winters of 1993/94 acid 1994/95 has been evaluated using the ground-based total ozone measurements of five SAOZ spectrometers distributed in the Arctic and from number density profiles of a balloon-borne version of the instrument. The ozone change resulting from transport has been removed using a 3D Chemistry Transport Model (CTM) run without chemistry. A cumulative total ozone depletion at the end of winter in March of 18% ± 4% in 1994 and of 32% ± 4% in 1995 was observed within the polar vortex, and of 15% ± 4% in both years outside the vortex. This evaluation is not sensitive to the vertical transport in the model. The periods, locations and altitudes at which ozone loss occurred were tightly connected to temperatures lower than NAT condensation temperature. The maximum loss was observed at 50 hPa in 1994 and lower, 60-80 hPa, in 1995. Half of the depletion in 1994 and three quarters in 1995 occurred during the early winter, showing that a late final warming is not a prerequisite for large ozone destruction in the northern hemisphere. The timing, the geographical location and the altitude of the ozone losses are well captured by the 3D CTM photochemical model using current chemistry, but its amplitude at low sun during the early winter, is underestimated. The model simulations also capture the early season reductions observed outside the vortex. This suggests that the losses occurred in situ in the early winter, when low temperatures are frequent, and not later in March, when ozone is most reduced inside the vortex, which would be the case if leakage from the vortex was the cause of the depletion.