the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Sep 2026
Climatology and trends in stratospheric mean age of air from correlations of multiple trace gases measured by ACE-FTS
Abstract. We develop and apply a novel method for estimating stratospheric mean age of air (AoA) from trace gases observed by the Atmospheric Chemistry Experiment – Fourier Transform Spectrometer (ACE-FTS). The method combines SF6 with five additional long-lived tracers, N2O, CH4, CFC-11, CFC-12, and HCFC-22, and is evaluated for ACE-FTS dataset versions 3.6 and 5.2. A proof of concept using the Chemical Lagrangian Model of the Stratosphere (CLaMS) shows that the multi-tracer approach reduces the uncertainty of zonal-mean AoA by about 50 % relative to conventional methods based only on SF6 and yields AoA estimates at individual profiles with an average uncertainty of about 0.3 years. Applied to ACE-FTS data, the multi-tracer method produces smoother AoA fields and strongly suppresses noise in sparsely sampled regions compared with the conventional convolution method. Comparison with in situ AoA estimates from SF6 and CO2 generally favors the multi-tracer product, although some discrepancies remain. The method provides a precise and spatially resolved satellite-based AoA product and a promising basis for future studies of long-term changes in stratospheric circulation. Comparison of ERA5 AoA with the new satellite product indicates a slow bias in the reanalysis. The long-term AoA trend pattern over 2004–2021 shows increasing age in the Northern Hemisphere stratosphere between about 18–28 km and decreasing age in the tropics and Southern Hemisphere subtropics, largely consistent with ERA5. In particular, AoA increases in the Northern relative to the Southern Hemisphere.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(14117 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 04 Nov 2026)
- RC1: 'Comment on egusphere-2026-4647', Anonymous Referee #1, 28 Sep 2026 reply
Data sets
Monthly zonal-mean time series of stratospheric mean age of air (AoA) derived from ACE-FTS versions 3.6 and 5.2 using SF6 and multiple long-lived trace gases Florian Voet https://doi.org/10.5281/zenodo.21342696
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 242 | 135 | 58 | 435 | 57 | 52 |
- HTML: 242
- PDF: 135
- XML: 58
- Total: 435
- BibTeX: 57
- EndNote: 52
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The study evaluates a new method of estimating mean age of aire from ACE-FTS measurements of five additional long-lived tracers, in addition to SF6, which has been used before. A proof of concept is carried out with the lagrangian chemistry model CLaMS to test the multi-tracer method against the SF6-only and the clock-tracer derived mean age. The results reveal that the multi-tracer method is more reliable and produces smoother fields. The application to two versions of ACE-FTS data confirms these results and the estimated trends over 2004-2021 reveal a smoother pattern with positive trends in the NH and negative in the SH tropics and subtropics, qualitatively consistent with ERA5 CLaMS results. I find the article a very important advance in the field, and the results are highly valuable; the article is extremely clearly written and the results well presented. I do suggest a few more analyses (not implying much difficulty, I believe) to strengthen the relevance of the paper and exploit the implications of the produced AoA observational estimate.
- What happens if the trend calculation period is extended to 2024 (or the end of the available dataset) with v5.2? Based on Fig. 9 it seems that the agreement with the reanalysis trends would be decisely worse. But also it seems that trends become more negative also in the NH (at least in the 25N latitude band). Given the relevance of these trends and the large impact of variability, this 3-year extension could help improve understanding and aid the discussion on the observational long-term trends, particularly in light of the recent results by Ray et al. (2026), which reveal negative mean age trends from the mid-1990s to the 2020s based on in situ measurements. It would be very useful to see the trend pattern for the extended period.
- Related to the previous question on the trend period, several studies have related the dipolar pattern to a southward shift in the tropical pipe- perhaps related to the QBO (e.g. Strahan et al. 2020) – which is particularly pronounced over the earlier part of the period (until 2017 or so). Is that the case in the new multi-tracer ACE-FTS estimate? This could be easily tested changing the start and end dates for the trend calculation.
- Hatching of the region where AoA estimates fall below 1.5 years. First a small note that this is first done in Fig. 1 but the explanation of why it is done does not arrive until lines 358-358, maybe could be moved earlier. Then, it is unclear if we should then discard the results in theis region, for example for the trends. I find interesting that the ACE-FTS estimates show negative trends in the tropical lower stratosphere (between the tropical tropopause up to 20 km), but this is hatched out due to a possible influence of extratropical tropospheric air. While it is indeed possible to have isentropic mixing from the extratropical stratosphere into this region, it is already in the overworld (above the lowermost stratosphere), and thus unlikely to be influenced by the extratropical troposphere. Rather, the trends in this region are most likely dominated by changes in tropical upwelling. Thus, hatching this region can be misleading and removing useful information, in my opinion. One possible solution could be to, instead of hatching points with AoA beow 1.5 years, to hatch the lowermost stratosphere. An even better option would be to use CLaMS to determine the region most influenced by parcels not crossing the tropical tropopause, perhaps based on previous work by the group.
- Could the trends in the other CLaMS estimates of mean age be shown? This would address two important points: 1) to what extent the AoA trends from the clock tracer (which represent the ‘real’ trends) remain consistent in the tracer-derived estimates? and 2) are these ‘real’ AoA trends best captured by the multi-tracer method than by the SF6-only method? This may imply extending the proof-of-concept simulations to longer time period, so I leave it as a suggestion, but it would be definitely a strong addition to the paper.