Ozone Forecasts of the Stratospheric Polar Vortex–Splitting Event in September 2002

The southern hemisphere major warming event in September 2002 has led to a break-up of the vortex in the middle and higher stratosphere and a corresponding splitting of the ozone hole. Daily 3D ozone forecasts, produced at KNMI with a tracer transport and assimilation model based on the ECMWF dynamical forecasts, provided an accurate prediction of this event a week prior to the actual break-up of the vortex. The ozone forecast model contains parametrizations for gas phase and heterogeneous chemistry. Initial states for the forecast are obtained from the assimilation of near-real time ozone data from the Global Ozone Monitoring Experiment (GOME) on ERS-2. In this paper we discuss the ozone forecasts and ozone analyses as produced before, during and after the event. These fields are compared with ground based Dobson, ozone sonde and TOMS observations. The total ozone comparisons show that the location of the vortex edge is generally well described by the 5 to 7 day forecasts in September and October. The GOME assimilation compared with TOMS shows a good correspondence concerning vortex location and ozone features, but also reflect clear dierences in the average ozone amount between the two retrieval schemes. The assimilation system produces realistic ozone profiles, apart from a systematic underestimation of ozone around 150 hPa inside the vortex in August-October.