The large-scale transport circulation in the stratosphere, the Brewer-Dobson circulation (BDC), strongly influences the distribution of radiatively important trace gases like ozone and needs to be properly represented in atmospheric models for reliable climate predictions. Its strength is best understood through observations of trace species, and the most common metric for the circulation strength is the mean transport time from the surface to a point in the stratosphere, the so-called mean Age-of-Air (AoA). Global observations of trace gases by satellite instruments are the most promising way to obtain consistent global estimates of mean AoA, however few satellite-based AoA products exist at present. Complementary to satellite data, in-situ observations form a data base for estimations of mean AoA. The AoA estimates from different data sources show considerable difference for a variety of reasons: sampling biases and retrieval uncertainties, the use of different trace gas species, and different methods to convert tracer mixing ratios to mean AoA. In addition to these inconsistencies between different observational estimates, global models also show a large inter-model spread in the simulation of mean AoA, indicating deficits in the simulation of transport processes in the stratosphere. Furthermore, long-term trends in AoA in global models are consistently negative, i.e. indicating a speed-up of the BDC, while the currently available tracer-based observational estimates indicate a slight increase in mean AoA in the northern mid-latitude mid-stratosphere. To conclude on the reasons for the deficits of global models to simulate transport processes and on the consistency of the long-term evolution of the BDC in models and observations, it is necessary to homogenize observationally-based mean AoA data products to the extent possible. Additional transport diagnostics will be necessary to provide information about the individual transport processes. The goals of this ISSI team are to
a) consolidate on the best possible method to calculate mean AoA from observations and apply this method to available trace-gas data,
b) identify which additional diagnostics can give information on the transport processes, and
c) establish how the tracer-based diagnostics can best be used to evaluate transport processes in models of different complexity.