The monthly mean difference fields were computed for the six years of the simulations. A t-test was applied to the monthly mean difference to judge if the changes between the integrations were significant. The significant changes in temperature were for the most part confined to the levels above 200 hPa. In the zonal mean the patterns of differences were largely consistent with regions of the ozone variations. Deeper tropospheric penetration of temperature differences occurred in October near the South Pole in the region of the "ozone hole". The significant temperature changes at the lowest model level ( ~992 hPa) were confined to very small areas.
The 200 hPa zonal wind differences demonstrated that the stationary wave structure was evidently altered by the ozone differences. Although the ozone specifications were zonally symmetric, the zonal wind differences were zonally asymmetric at 200 hPa.
The differences in the two simulations were generally small below 100 hPa. However, the differences are not negligible and it would be make model intercomparison problematic if all the participants in AMIP do not use the same ozone data. This problem is most serious for models using the older, pre-ozone hole data sets. (pdf file)