Séminaire

The age of stratospheric air is calculated over 22 years of the ERA-Interim reanalysis using an off-line Lagrangian transport model and heating rates.

At low and mid-latitudes, the mean age of air is in good agreement with observed ages from aircraft ights, high altitude balloons and satellite observations of CO2 and SF6. The mid-latitude age spectrum in the lower stratosphere exhibits a long tail with a peak at 0.5 y, which is maximum at the end of the winter, and a secondary flat maximum between 4 and 5 y due to the combination of fast and slow branches of the Brewer-Dobson circulation and the reinforced barrier eect of the jet. At higher altitudes, the age spectrum exhibits the footprint of the annual modulation of the deep Brewer-Dobson circulation.

The variability of the mean age is analysed through a decomposition in terms of annual cycle, QBO, ENSO and trend. The annual modulation is the dominating signal in the lower stratosphere and in the tropical pipe with amplitude up to one year. The phase of the oscillation is opposite in both hemisphere beyond 20C and is also reversed below and above 25 km with maximun arising in mid-March in the northern hemisphere and in mid-September in the southern hemisphere. The tropical pipe signal is in phase with the lower southern stratosphere and the mid northern stratosphere. The maximum amplitude of the QBO modulation is of about 0.5 y and is mostly concentrated within the tropics between 25 and 35 km. It lags the QBO wind at 30 hPa by about 8 months. The ENSO signal is small and limited to the lower northen stratosphere.

The trend is signicant and negative, of the order of -0.3 to -0.5 y dec^{-1}, within the lower stratosphere in the southern hemisphere and under 40N in the northern hemisphere below 25 km. It is positive (of the order of 0.3 y dec^{-1}) in the mid stratosphere but there is no region of consistent signicance. This suggests that the shallow and deep Brewer-Dobson circulations may evolve in opposite directions. It is however difficult to estimate a reliable long-term trend from only 22 years of data. For instance, a positive trend is found in the lower stratosphere if only the second half of the period is considered in agreement with MIPAS SF6 data excepted in the northern polar region and in high altitude.

Finally, it is found that the long lasting influence of the Pinatubo eruption can be seen on the age-of-air from June 1991 until the end of 1993 and can bias the statistics including this period. In our analysis, this eruption shifts the trend towards negative values by about 0.2 to 0.3 y dec^{-1}.