Abstract
A coupled chemistry–climate model is integrated for the period March 1979–January 2000 with sea surface temperatures and sea ice amounts specified from observations. Greenhouse gas concentrations and halogen loading are also taken from observations. The model contains a detailed stratospheric chemistry package that responds to and in turn influences the model temperature fields. The simulated temperature and ozone trends over the period 1980 to 2000 are broadly in agreement with observations. However, the Arctic ozone depletion is smaller than observed, and the impact of the Antarctic ozone hole lasts too long into summer. The coupling between the model ozone and temperature trends is demonstrated to occur in a similar way to that inferred from observations and could be important in generating the observed interannual temperature variability. It is concluded that further improvement in climate models is necessary before future trends in stratospheric ozone and temperature can be predicted with confidence.