A new cheap ozone tracer experiment was designed which allowed extensive testing of model tracer transport near the extratropital tropopause. This innovative model design allowed many more sensitivity experiments to be performed than would be tie case when running with full ozone Chemistry. It also allowed clean comparison of the tracer transport characteristics of different models since they all could be run with the same silified chemistry. scheme. Results were validated against, the MP large dataset of ozonesonde, and MOZAIC measurements (mord than 6001 ozone profiles and measurements oh 20001 aircraft flights were used) . This is the first time such a comprehensive model validation of this type, which included analyzing different latitudes and changes over the annual cycle has been undertaken.
-In general all models agreed with the observations within approximately a factor of two. There was, however, a distinct difference between the performance of models run from analyzed winds, and those from global circulation model (GCM) winds. The models run from GCM **** did not disturb the seasonal 'ozone accumulation in the lowermost stratosphere. However the models run from analyzed winds overestimated ozone above this accumulation especially during the spring. These errors are probably largely due to errors in the vertical velocities, in the GCMs and analyzed winds., Close to the tropopause all models, underestimate Ozone and its variability, except during winter. The strongest underestimation is found during summer. A sensitivity study with, the TM3 model demonstrated that a second-order moment advection scheme is required to model ozone in the low-cost stratosphere. The TOMCAT model, was\used to investigate the_ effect of different top boundary conditions and different horizontal grid resolutions on the ozone distibution.
It showed that top boundary constraints with observed ozone data down to 50 hta are recruited to simulate 6zon-, in the