In this paper we present a new technique for the estimation of ozone loss in the stratospheric polar vortex based on the assimilation of Earth Observing System Microwave Limb Sounder (EOS MLS) and Solar Backscatter Ultraviolet radiometer (SBUV/2) ozone observations in the Met Office data assimilation system. We focus on the northern winter of 2004/05, which was exceptionally cold in the Arctic stratosphere, with associated large ozone depletion due to heterogeneous chemistry. Our ozone loss estimate, which was calculated for the 1 February to 10 March 2005 period, peaks at 450 K (approximately 17–18 km), and is 0.6 ppmv at that isentropic level (our loss estimate for the vortex core only was somewhat higher (1.0 ppmv) and indicates uncertainties related to mixing at the vortex edge). This value is similar to or smaller than results from other studies, which estimate ozone loss in this period to be in the 0.6–1.2 ppmv range. When combined with results from other studies that estimate ozone loss occurring outside our assimilation period, we obtain an estimate of 0.8–1.2 ppmv for ozone loss from early January to early March 2005.
We find a second maximum in ozone loss for the 1 February–10 March period near 650 K (approximately 25 km) of around 0.4 ppmv. This is a lower figure than found in other studies, but ozone loss is actually much stronger at this level outside the vortex in a low-ozone pocket in the Aleutian anticyclone, likely due to the NOx catalytic cycle. Our results show that the ozone data assimilation method we have used to estimate ozone loss is very promising, and can lead to potentially more accurate ozone-loss estimates than other methods. © Royal Meteorological Society and Crown Copyright, 2008.