the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 01 Jul 2025
CO2 variability and seasonal cycle in the UTLS: Insights from EMAC model and AirCore observational data
Abstract. The complex distribution of CO2 in the upper troposphere and lower stratosphere (UTLS) results from the interplay of different processes and mechanisms. However, in such difficult-to-access regions of the atmosphere our understanding of the CO2 variability remains limited. Using vertical trace gas profiles derived from measurements with the balloon-based AirCore technique for validation, we investigate the UTLS and stratospheric CO2 distribution simulated with the global chemistry-climate model EMAC. By simulating an artificial, deseasonalised CO2 tracer, we disentangle the CO2 seasonal signal from long-term trend and transport contribution. This approach allows us to study the CO2 seasonal cycle in a unique way in remote areas and on a global scale. Our results show that the tropospheric CO2 seasonal cycle propagates upwards into the lowermost stratosphere and is most modulated in the extra-tropics between 300 hPa and 100 hPa, characterised by a 50 % amplitude dampening and a 4-month phase shift in the northern hemisphere mid-latitudes. During this propagation the seasonal cycle shape is also tilted, which is associated with the transport barrier related to the strength of the subtropical jet. In the stratosphere, we identified both, a vertical and a horizontal ‘tape recorder’ of the CO2 seasonal cycle. Originating in the tropical tropopause region this imprint is linked to the upwelling and the shallow branch of the Brewer-Dobson-circulation. As the CO2 seasonal signal carries information about transport processes on different timescales, the newly introduced tracer is a very useful diagnostic tool and would also be a suitable metric for model intercomparisons.
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
(2382 KB) - Metadata XML
-
Supplement
(1160 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-2648', Anonymous Referee #1, 29 Jul 2025
-
RC2: 'Comment on egusphere-2025-2648', Anonymous Referee #2, 09 Aug 2025
Review: CO2 variability and seasonal cycle in the UTLS: Insights from EMAC model and AirCore observational data
# Summary
The present work by Johannes Degen et al. studies the distribution and seasonal cycle of CO2 in the upper troposphere-lower stratosphere (UTLS) based on observations as well as model results in the time range of 2000 to 2024. Observational data is obtained via in-situ AirCore measurements; the global distribution of CO2 is simulated using the Atmospheric Chemistry model EMAC using CO2 tracers, from which the seasonal signal is derived. Overall, observed and modeled CO2 mixing ratios agree very well with each other. The CO2 seasonal cycle exhibits a dampening and time lags with altitude, as well as a tilt, which is related to the subtropical transport barrier. Both a vertical and a horizontal (latitudinal) tape recorder are presented, revealing information on the seasonality of the BDC tropical upwelling and shallow branch.# General comments
## Relevance and Overall Quality
CO2 is of high relevance to climate studies not only due to its impact on radiative balance, but also as an almost passive tracer, revealing information on large-scale transport through its distribution and via Age of Air. The UTLS is of special importance due to the strong radiative response to changes in its GHG composition. This study provides a unique comparison of highly resolved in-situ measurements of vertical CO2 profiles in the UTLS with global Chemistry Climate Model results, as well as an innovative approach to disentangling the seasonal cycle from trends and other variability modes using model tracers. The presented findings are closely related to the seasonal variability of the BDC and the subtropical jet, rendering this study extremely valuable for understanding global circulation and the composition of the UTLS. Overall, I find the methods, results and discussion in this work outstanding; the manuscript is also well-written and features excellent graphical representations of the data.## Strengths
- The comparison between model and observational data reveals valuable information on detailed vertical profiles, as well as the global distribution of CO2. Both complement each other well, and it is interesting to see how closely they match.
- Thorough comparison between different boundary conditions
- Excellent, clear and original representation of data in plots
- Comprehensive analysis of CO2 seasonal cycle depending on latitude and altitude. Interesting to see a horizontal as well as a vertical tape recorder.
- Results are thoroughly discussed and linked to dynamical processes
- The deseasonalized CO2 tracer is a novel approach to disentangling the seasonal CO2 cycle from long-term trends and interannual variability
- AirCore observations provide composition profiles that are more highly resolved than satellite measurements and reach farther into the stratosphere than aircraft measurements## Weaknesses
- Some parts of the discussion could be clarified further, see detailed comments below. Especially during the discussion of figures, it would be helpful to refer to individual subplots more often.
- Some of the figures require minor polishing (see "Figures" section below)
- Grammar could be refined in some places, also decide between American and British English
# Specific comments## Text
- 267: Please provide a bit more detail on the statistical methods. Which correlation coefficient are used; e.g., Pearson's/Spearman's? Were the assumptions for computing correlation coefficients checked, e.g., normality? How is the Mean Absolute Deviation (MAD) computed here?
- 336-339: Did I understand the reasoning here correctly such that a weak tropopause leads to enhanced transport of CO2-rich air into the stratosphere, which is seen in AirCore observations, but not in the model, leading to the deviations between both datasets? That would seem plausible to me. If so, please clarify that in the text, since I found the connection between "weak tropopause" and "scale problem" not so obvious.
- 348: With "compared data", do you mean the AirCore observations, since they only cover specific regions? If yes, please clarify that in the text.
- 365: Is this related to the figure? If yes, it's better to start with, e.g., "Figure 4 shows...". Also, if this is part of the figure discussion, please use the same height coordinate as in the figure (between 8-35 km -> use pressure instead).
- 365: "hemispheric spring", which hemisphere is meant here? Or reword to "spring in each hemisphere". Also please refer to individual subfigures to support your point.
- 371: Could you briefly mention/discuss why this region shows enhanced seasonality?
- 384: Clearer: "contribution of tropical air in the extratropical LMS..." or "export of tropical air into the extratropical LMS..."
- 385-387: Please refer to individual subfigures for clarity. Also, which layer is meant here? Are you referring to the decrease of CO2 near the surface as seen in the 2019-08 plot? Is the main point here that, in summer, fast quasi-isentropic mixing of CO2-rich air into the extratropics counteracts the CO2 sink due to photosynthesis, or that a layer of low CO2 (biogenic) can be observed below the CO2-rich air (mixing) in the LMS? Please clarify.
- 389: Please elaborate on the hemispheric asymmetries with regards to jet strength. E.g., which hemisphere usually shows the stronger jet, and how does that influence the CO2 distribution?
- 390: Which characteristics of the shallow BDC branch (e.g., hemispheric differences and seasonal variability) can you see in your results?
- 399: ...the long-term trends and seasonal cycle of CO2 sources and sinks?
- 433-434: refer to panels, e.g., a) and f)
- 436: envelope of the curves in Figure 6a)?
- 442: Please specify in which pressure region the free troposphere lies
- 442-445 and 446-448: Please refer to individual panels; are these sentences discussing Fig. 6a?
- 455: What exactly does the stratospheric residual mean/ how can we interpret it? From Eq. 1, I understand that CO2_seas is the seasonal deviation from the deseasonalized "baseline"; so does that mean the seasonal signal of CO2 in the stratosphere is permanently negative? How exactly can we infer negative CO2 flux into the stratosphere from that? I do understand the reasoning in the following lines (455-463), but still, the meaning of the residual isn't clear to me.
- 463: By the "findings described above", do you mean the negative residual?
- 472: with decreasing pressure?
- 492: "15 km", please use altitude coordinates consistent with the figure (or add a km scale to the figure).
- 504: What do you mean by "features that are not so pronounced"?
- 513: "expected" because of the larger land coverage and therefore stronger sources and sinks in the NH? Also, it would help to explicitly describe the hemispheric differences in this sentence, i.e., stronger seasonal signal in the NH.
- 526: "...in the extratropical LMS throughout all months/seasons..."
- 532: Is this related to Fig. 8?
- 533: Homogeneous in what sense? From looking at Fig. 8, I can still see strong variations in the seasonal signal with both latitude and height.
- 544: "its distinctiveness" -> "the distinctiveness of the seasonal cycle"? Suggest rephrasing the sentence for clarity.
- 546: Suggest explicitly mentioning the hemispheric differences
- 563-566: A more detailed description of interannual variabilities would be interesting
- 568: "spring to autumn ...", "October": Please indicate that you are referring to the NH(?) and relate the observations to the corresponding subfigures.
- 569: Suggest to explicitly mention that the BDC is stronger in winter
- 571-574: I have difficulties following this part of the discussion: AMA should, as far as I know, accelerate upwelling and mixing into the stratosphere -- but in NH summer. Why is the ascent starting from January linked to AMA here? Also, in "the transport from the (sub)tropics is likely to be the main origin for this feature", which feature exactly is meant here?## Figures
3)
- Very nice plot clearly showing the agreement between observation and model data.
Subfigure 3e)
- Annotation of UTLS plot in light blue is hard to read; please choose a darker colour.
- Spell out "Mean Absolute Deviation" in the figure caption and/or mention again that this is a metric for determining the deviation/similarity between modeled and observed data (not everyone reads the Methods chapter ;)).
- A legend and additional description in the caption would help to interpret the box plots: Which data range do the coloured boxes cover, which errors are included in the error bars (only standard deviation or including systematic errors?) and what do the open circles mean (outliers?)? Do the vertical lines in the coloured boxes represent the medians?4)
- Please add letters a)...f) to each panel.
- Since these are quite many plots: is there a specific reason why you chose to show individual months instead of seasonal averages, which might be better suited to show seasonal differences?
- In the text discussing the figure, you refer to "8-35 km", while the figure uses a pressure scale. Suggest to add a geometrical height scale to the figure, or change the discussion accordingly.
- It would help adding theta annotations to the contours in every plot
- I also recommend annotating selected wind contours with values, or at least stating the lower threshold in the figure caption. Please also specify in the caption and/or legend whether you considered zonal or horizontal wind speeds.
- Suggest rewording the caption: "...of EMAC CO2 tracer (CO2_MBL_pbl)", "...potential temperature surfaces indicating the UTLS."5)
- Red and orange might be difficult to discern for colourblind readers
- Caption: "...these pressure levels..."
- Do the symbols represent individual AirCore flights or averages thereof?
- Otherwise, trends and seasonal cycle are very well represented here6)
- Excellent representation of phase shifts and dampening of seasonal cycle
- Red/green colour contrast might be difficult to read for some users, suggest checking figure with a colourblind simulator or switching to explicitly colourblind-friendly palettes
- Can observations be added to this plot, or would that over-clutter it?8)
- Add subfigure letters a)...f)
- Consistency: Why are odd months selected here, while even months are shown in Figure 4?
- Please give the time range of the climatology in the caption
# Technical corrections
- "WMO tropopause" spells without a dash; please correct throughout the manuscript (also in the figures).
- Throughout the manuscript, "Figure" needs to be spelled out at the beginning of a sentence.
- 181/182, 202, 204, 232 etc.: Reference instead of inline link
- 211: corresponds
- 233: Spelling: "analysed" is British English, "summarized" American English
- 321: Suggest using "beneath" instead of "below" to clarify that the considered region is situated below 350 hPa in the sense of altitude, not pressure
- 324: Again, characteriZed spelled with z, use either AE or BE
- 329: no comma after "... vegetation period"
- 335: no comma after "... cases"
- 340: No return behind "... UTLS well."
- 341: Despite the fact that...
- 345: Suggest rewording for clarity: These vertical shifts also appear for other simulated species...
- 352: remove comma in "opportunities for analysis, that are"
- 381: remove comma after "CO2"
- 414, 415: Suggest writing values with uncertainties like this: (2.44 +/- 0.16) ppm/yr
- 435: climatological
- 498: simultaneous
- 500: remove comma after "both"
- 507: The "again" in "definitely more complex again" reads a little awkward here; suggest moving it to the beginning of the sentence: "Again, the short-term variability ..."
- 511: Suggest rewording: "As can be seen from the panels..."
- 519: "Hadley cell" is spelled without dash
- 517, 522-524 and 541: The parts in brackets disturb the flow; suggest formulating as actual clauses
- 546: for different latitudes
Data sets
NOAA AirCore Atmospheric Sampling System Profiles Bianca Baier et al. https://doi.org/10.15138/6AV0-MY81
Goethe University Frankfurt AirCore profiles Johannes Degen et al. https://doi.org/10.5281/zenodo.15274043
EMAC global CO2 data and high resolved EMAC data along AirCore Flights (S4D) J. Moritz Menken https://doi.org/10.5281/zenodo.15583480
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 512 | 24 | 10 | 546 | 21 | 14 | 22 |
- HTML: 512
- PDF: 24
- XML: 10
- Total: 546
- Supplement: 21
- BibTeX: 14
- EndNote: 22
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This paper compiles about 260 AirCore vertical trace gas profiles of CO2 and combines them with EMAC model results to investigate the distribution and variability of CO2 in and around the UTLS. After validating some features of ECAM using AirCore data, the paper focuses on patterns in the ECAM CO2 seasonal cycle using a deseasonalised CO2 tracer (CO2_seas), which is a novel feature newly implemented in ECAM. CO2_seas isolates the CO2 seasonal signal from the transport contribution and long-term trend, allowing the authors to address how the seasonality of CO2 in the free troposphere propagates into the Lowermost Stratosphere.
This is an original and generally well-written study (although it could perhaps be streamlined) that describes a number of interesting features pertaining to how the CO2 seasonal cycle propagates into the stratosphere. I have two major comments, one of which is practical and calls for specific, generally minor revisions to make the paper easier to follow. The other is more of an overall conceptual criticism, which doesn’t necessarily need to be addressed and indeed may not be possible to address.
1) Practical. The text points out many detailed features of the figures, referencing the pressure level in hPa where they occur. Yet the figures have a sparse Y axis that is labeled only at 10^3, 10^2, and 10^1 hPa. This makes it challenging for the reader to locate the feature being described. I would suggest including more Y axis labels on the right as well as guiding lines or contours that delineate key relevant features like the STJ, the LMS, and the tropopause. This comment pertains in particular to Figures 6-8 -see also my specific comments.
2) Conceptual. The use of AirCore data is somewhat limited and the paper is based mainly on ECAM model output, particularly the tracer CO2_seas. This approach is justified in Figures 3 and 5, in which ECAM is shown to simulate well the observed CO2 profiles at selected latitudes (Fig 3) and the AirCore seasonal cycles at different pressure levels (Fig 5). The abstract states that CO2_seas “is a very useful diagnostic tool” but it is not clear if and how CO2_seas can be derived from AirCore observations. The authors only address this issue in the very last paragraph of the conclusions, where they admit that, “such an approach is challenging.”
Specific Comments
Line 16. Please spell out EMAC (assuming ACP has a policy of no undefined acronyms in the Abstract).
Section 2. What is the vertical resolution of the AirCore profiles? (Later sections describing EMAC emphasize its “coarse resolution” of 90 vertical levels.)
Line 196-198 and 222-228. Please state more clearly whether the seasonal cycle of CO2 is prescribed in the standard configuration or calculated prognostically based on couple land and ocean carbon cycle modules. (Many readers will not be familiar with the details of the CMIP6 protocols.) Is the prognostic CO2 seasonal cycle from the coupled land/ocean/atmosphere model being “nudged” to the prescribed observed seasonal cycle?
Table 1 last row, last column, It would be better to describe CO2_seas as “CO2_MBL_pbl minus
CO2_deseas” in the table rather than the more vague “based on CO2_MBL_pbl and CO2_deseas”?
Line 264 change “It is calculated” to “The weighted average was calculated”
Line 266 What are “The two analysed species”?
Line 312. What is meant by “on top of it”?
Line 366. These reversed gradients are not obvious in Figure 4. Could they be shown better with vertical profile line graphs?
Line 382. Similarly, this feature is not obvious in Figure 4 and perhaps could be shown in a line graph. Also, please define the approximate altitude range of the LMS.
Line 405. By “information” do you mean “AirCore information” ?
Figure 6a. The lines on the right Y axis are helpful. But why not actually label them? There would be room if the width of 6a is reduced slightly. Also, could the same labeled lines be added to Fig 7a?
Line 446. What exactly is meant by the “strongest modulation”?
Line 448. Should 20 km be expressed in hPa, since everything else is.
Line 454. A “residual influence of -0.2 ppm seems to remain”. Is this simply CO2_seas as defined in Equation 1? Or has there been further processing of the model output? Please explain more clearly how the curves in Figure 6 are normalized/detrended to create a “climatology.”
Figure 6a and 7. Perhaps a black line showing the position of the tropopause would be useful, especially since the Y axis label has only 3 tick marks.
Figure 7c. This figure is confusing. If it is not illustrating an essential point, please consider deleting.
Line 511. “As can be seen”
Line 511. Probably better not to begin the paragraph referencing a supplementary figure that most readers won’t see.
Figure 8. Can you draw in the STJ (as done in Fig 4) for March and July to illustrate the points described in Lines 520-525? It is not obvious that the gradient is stronger in March.
Line 541. Is 5hPa even shown on Figure 8? If not, maybe delete this sentence.
Figure 10. I am not an expert on “tape recorder” effects, but it seems like it might be a stretch to call Figure 10 a “horizontal tape recorder.” Is there a precedent for this in the literature with H2O or other trace gases? Can the authors really be sure of what causes the hemispherical symmetry at and above 127 hPa? For the tropical signal to mix equally into both hemispheres seems at odds with the Brewer Dobson Circulation, which upwells in the tropics and descends into the winter hemisphere.
Lines 632-634. I don’t follow this argument. Is it important enough to be in the conclusions? In general, the conclusions should probably be trimmed to focus on the most key points.