The figure shows comparisons of SF6 profiles simulated with the Eulerian chemistry transport model SILAM (driven with the ERA-Interim reanalysis for 1980-2018) and MIPAS observations of SF6. The model includes the chemical sink of SF6 in the mesosphere, and the effect of gravitational separation; sensitivity runs for testing the appropriate diffusivity of the model in the altitude range of interest have been performed. The blue symbols with error bars are seasonal means of MIPAS SF6 observations for 4 seasons and 6 latitude bands of the year 2007. The error bars indicate the systematic uncertainties of MIPAS observations that do not cancel out by averaging. The solid coloured lines (pink, green, blue) are results from the model for the same year with different diffusivities interpolated to the locations of MIPAS measurements. The dashed coloured line show the same, however for full zonal averages from the SILAM model output on all model gridpoints. The black vertical error bars indicate the vertical resolution of the MIPAS SF6 measurements, derived from the averaging kernels.
The mesospheric sink of SF6 has a major impact on the mixing ratios above 20 km. The depletion impact is especially strong in the wintertime polar areas due to the descent within a polar vortex. A set of sensitivity tests showed that molecular diffusion and gravitational separation of SF6 are responsible for up to a few percent of further reduction in SF6 mixing ratios in the upper stratosphere. A good agreement of the simulated SF6 distribution with the MIPAS observations up to the altitudes of 30-35 km was shown. The SILAM model simulations were able to reproduce the apparent SF6 AoA derived from the MIPAS observations. This highlights the role of fast mesospheric destruction of SF6 due to the electron attachment mechanism. The mesospheric sink has severe implications on the AoA derived from the SF6. The apparent over-aging introduced by the sink is large and variable in space and time. Moreover, the over-aging due to the sink increases as the atmospheric burden of SF6 grows. All this makes SF6 unsuitable to infer AoA above 20km.
For a fully-passive SF6 tracer, the variable rate of emissions (i.e. the deviation from linear increase) causes deviations from the ”ideal age”, and these deviations can be compensated to some extent. However, correcting the deviations due to the mesospheric sink of SF6 is hardly possible. These deviations appear as long-term trends in the apparent AoA. These trends differ from the trends in the “ideal-age AoA”, and have no direct correspondence to the actual trends in the atmospheric circulation.
Procedures used to derive the AoA from observations of various tracers in the atmosphere are inevitably based on assumptions and idealisations that have limited and often unknown area of applicability. The resulting uncertainties in the AoA are large enough to preclude the use of apparent AoA and its trends for evaluation of changes in atmospheric circulation or for validation of atmospheric models. Observations of the tracers themselves, however, have well quantified uncertainties, so direct comparisons of simulated tracers to the observed ones are a very promising means for the atmospheric model evaluation. AoA in turn is a convenient means for model inter comparison if a protocol of the AoA derivation is well specified. A method to overcome the problems related to AoA has been developed by von Clarmann and Grabowski (2016).
For more information see: https://doi.org/10.5194/acp-20-5837-2020