the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Aug 2026
Characterising the long-term effects of stratospheric aerosol injection geoengineering on the stratospheric ozone layer with mid-infrared limb emission sounding instruments
Abstract. Stratospheric aerosol injection (SAI) has been proposed as a potential method to counteract anthropogenic greenhouse gas–driven global warming but it may perturb the stratospheric ozone layer. Here, we use existing SAI scenario simulations as pseudo-reality (PR) input to assess how a future mid-infrared limb-emission sounding observing system would characterise its modelled ozone response. Our PR scenarios were generated using CESM2(WACCM6) simulations following the SSP5-34-OS overshoot pathway, with and without SAI. Pseudo-observations (PO) of a future instrument, modelled around the satellite mission concept CAIRT (the Changing-Atmosphere Infra-Red Tomography explorer), were generated using mission performance simulators, providing full error propagation and spatial smoothing characteristics. Our results demonstrate that a CAIRT-like mission can monitor and quantitatively characterise global, regional and seasonal ozone impacts associated with this SAI scenario. The parent PR fields used here contain a modelled SAI-induced ozone response, including: a) a pronounced additional depletion of total column ozone in the southern hemispheric high latitudes, with austral springtime reductions exceeding 20 DU between 2033–2062, consistent with enhanced heterogeneous halogen activation on sulphate aerosols, b) a delay in Antarctic ozone recovery, and c) a moderate ozone increase in winter and spring at northern hemispheric mid and high latitudes, associated with altered transport and weakened subtropical jets. All these different decadal impacts of SAI interventions are observable with CAIRT PO and are fully distinguishable from a baseline non-SAI scenario. These findings highlight the importance of advanced satellite observations, which are not available nowadays, to monitor and evaluate these impacts.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Measurement Techniques.
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
(2134 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-4654', Anonymous Referee #1, 08 Sep 2026
-
RC2: 'Comment on egusphere-2026-4654', Anonymous Referee #2, 04 Oct 2026
This goal of this paper is to demonstrate that a future mid-infrared limb-emission sounding instrument, like CAIRT, has the precision/resolution to determine differences in stratospheric ozone which may occur due to a future geoengineering scenario between 2035 and 2060. The “data” used for this comparison comes from CESM2/WACCAM model projections under scenarios of business as usual and a geoengineering injection rate of 10 Tg SO2/yr. The differences in ozone between these runs is on the order of plus/minus a few percent to over 10% depending on altitude, latitude, and hemisphere, with generally ozone loss in the southern hemisphere. In addition to the direct model output of ozone, pseudo reality (PR) in both scenarios, the authors also calculate pseudo observations (PO), i.e. the observations a mid-infrared sounding instrument would make, were it available through the two scenario periods. If the differences, geoengineering to natural (del PR), are greater than del PR/PO, the differences between PR and PO, then the limb sounding instrument would be capable of detecting del(PR) proving the authors thesis that a CAIRT like instrument is needed to assess future possible del(PRs).
Fig. 1, the key figure in this paper, shows that the 1 sigma uncertainty, in the stratosphere, ranges from 0.05-0.15 ppmv for a limb sounding instrument. For a typical stratospheric ozone profile this corresponds to uncertainties of 0.5 – 3%. While the upper end of this difference approaches del(PR), the following figures in the paper demonstrate, over and over again, that del PR/PO is less than 1% in both scenarios, proving the authors point from the beginning. Never the less, the authors proceed to describe the differences globally, regionally, and seasonally. Every section proceeds the same way showing that del PR/PO is much less than del PR, and then list the differences in the text of the paper. This makes for very dry reading. If the value of these differences are important, in all the different data slices, they should be put into a table where they could be seen at a glance. Meanwhile the interesting fluctuations and differences in the two PR scenarios are not mentioned at all.
It is not clear why this paper should be published as the point is obvious from the beginning and CAIRT was not selected. Should the editor choose to proceed anyway, then the paper should be significantly shortened and homogenized.
All the endless text quoting numbers should be reduced to tables. All the figures should be homogenized as much as possible. Don’t show uncertainty in mixing ratio and then the latitude altitude ozone field in molecular concentration. Use one or the other units for both. Better would be to show uncertainty in percent, and the same for del(PR) throughout. Notice how little difference there is in del(PR/PO) throughout for both scenarios, see Figs. 4, 5, 7-10. At least in Figs. 7-10 the authors add percent difference. In Fig. 4 del(PR/PO) is so small the difference scale has to be amplified.
Here are a few additional comments/questions.
321 Isn’t it Fig.5 showing the latitude-time evolution?
246 isn’t 24 DU closer to 10% than 5%?
Fig. 8 What is causing the rather large annual to bi-annual oscillations in both the baseline and geoengineering cases? Why aren’t these oscillations correlated more closely between baseline and geoengineering? Why are they completely absent in the October means, particularly after 2050? Similar oscillations also occur in Fig. 9 where the anti-correlations are even more pronounced.
424 … hardly discernable…? That is not the case. There are easily differences of 10 DU or more discernable. It is just that with 12 month averaging, and unexplained differences in the short term variations between the baseline and geoengineering, that the average differences aren’t as large as in the Antarctic. Still the positive differences are 50% and the negative differences 25-50% of the Antarctic differences. These differences are borne out by Fig. 10 where the influence of the annual averaging is diminished.
Citation: https://doi.org/10.5194/egusphere-2026-4654-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 201 | 96 | 68 | 365 | 78 | 70 |
- HTML: 201
- PDF: 96
- XML: 68
- Total: 365
- BibTeX: 78
- EndNote: 70
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This paper tests whether satellite mission concept CAIRT could actually detect the ozone changes due to stratospheric aerosol injection (SAI), if such a mission ever flew during a deployment. CAIRT is a mid-infrared limb-sounding mission concept that ESA considered for its Earth Explorer 11 program but ultimately (and sadly) didn't select. The authors use a set of synthetic "pseudo-observations" that include the retrieval noise, smoothing, and bias a real CAIRT-like instrument would introduce, including the extra effect of enhanced stratospheric aerosol loading on the retrievals themselves, and ask if those pseudo-observations preserve the model's simulated SAI-induced ozone signal, or if the instrument limitations would wash out the signal. Overall the paper's central result looks well-supported by the quantitative comparisons throughout, and the paper is perfect for this journal.
I have some mostly minor comments that need to be addressed before recommending publications.
Major comments
1) I would frame the discussion a bit better in the broader context of literature around time of emergence of a climate signal from natural variability in the context of both climate change and ozone recovery. The authors do this in places already by discussing linear trends and calculating roughly how many years would be required to detect a signal, but they use a period in which the signal itself is already strong. Of much more interest could be to ask whether, if CAIRT was launched contextually with the SAI deployment, the mission would be able to detect the initial signal, and how long would that take.
Contextually to this comment, I want to note that while the paper is very strong and scientifically robust, it is a bit of a shame the authors selected to use a rather silly and unfeasible SAI scenario, especially when many multi-model results exist (i.e. both GeoMIP simulations and the new CCMI runs with SAI are out there, and have the same kind of ozone information this paper uses, see Jörimann et al. (2026) and Bednarz et al. (2026)), that have the added perks of having 1) more serious, believable scenarios compared to the Overshoot one, which would fit better with the “realistic SAI scenarios” the authors mention in line 159 and 2) multiple models which would enable the authors to better talk about detection/emergence within the context of different simulated ozone responses. Even if the authors still only looked at CESM2-WACCM6 here, they would have better enabled follow-up studies where someone else could have looked at this same question in other models within standardized scenarios.
2) One potential discrepancy: the abstract's point (c) claims "a moderate ozone increase in winter and spring at northern hemispheric mid and high latitudes." But Table 1 shows (0–60°N) increasing while (60–82°N) actually decreases slightly in both the annual mean (−0.74 DU) and in March specifically (−0.73 DU). There's no NH high-latitude winter analysis shown anywhere I could find to support the "winter" part of the claim. As written, the abstract overstates/misattributes the high-latitude finding, and it should probably be split into "increase at NH mid-latitudes" versus "weak, non-robust decrease at NH high latitudes," or supported with an added winter-specific analysis. Related to this, the abstract's causal mechanism — "associated with altered transport and weakened subtropical jets" should be better discussed and referenced (there are plenty of works out there discussing this) in the paper.
3) The authors define the TOC in line 320-321 as the integral from 10 to 52 km – I found this a bit weird, as that is a partial, stratosphere-only column, not the conventional total-atmosphere column. But the term "Total Ozone Column" is already used in Figure 2 and its caption (p. 9), well before this definition appears, without stating the integration bounds there. Since typical satellite/ground-based "total column ozone" (for instance, from OMI) products integrate the full atmosphere including the troposphere, this could confuse readers comparing the paper's DU values to standard products. I suggest moving the definition earlier, and justifying it a bit better in the context of other products.
Minor comments
Abstract, line 20-21: "to assess how a future mid-infrared limb-emission sounding observing system would characterise its modelled ozone response": "its" is ambiguous (does it relate to the system's response? the SAI scenario's response?). Maybe "...would characterise the modelled ozone response."
Line 122-124: "which is investigated in details in this study" should be "in detail"
Line 155: "in the hypothesis of a series of such instruments proving consistent observations", "proving" should almost certainly be "providing." "In the hypothesis of" also reads weird; "under the assumption that a series of such instruments provides..." would be clearer.
Figure 2 caption: "Global maps of the total ozone column averaged derived from model over the period 2033-2062" is a bit garbled. Suggest: "Global maps of the model-derived total ozone column, averaged over 2033–2062, for..."
Line 552-553 (Conclusions): "the recent selection of NASA's STRIVE mission demonstrate that..." should be "demonstrates."
The reference list includes at least one entry (Budyko, 1977) that I could not find cited anywhere in the text.
Figure 4 uses molecules/m³ (nO₃, ×10¹⁸ molec/m³), but the text discussing the same quantity using mol m⁻³. Clearly the authors use mol as a shorthand for molecules, not moles, but just be clear about it.
References
Bednarz, E. M., Butler, A. H., Haywood, J. M., Henry, M., Jones, A., Kravitz, B., Lee, W. R., MacMartin, D. G., Maycock, A. C., Sekiya, T., Watanabe, S., and Visioni, D.: Stratospheric ozone projections under sulfur-based stratospheric aerosol injection: Insights from the multi-model G6-1.5K-SAI experiment, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-310, 2026.
Jörimann, A., Sukhodolov, T., Tilmes, S., Plummer, D., Watanabe, S., Akiyoshi, H., Chiodo, G., Visioni, D., Vattioni, S., Rozanov, E., Bednarz, E. M., Josse, B., Yamashita, Y., and Peter, T.: Middle atmosphere chemical and dynamical effects in the CCMI-2022 stratospheric aerosol injection scenario, Atmos. Chem. Phys., 26, 11207–11234, https://doi.org/10.5194/acp-26-11207-2026, 2026.