the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Sep 2026
What controls tropospheric carbon monoxide in the remote Southern Hemisphere?
Abstract. The remote extratropical Southern Hemisphere (SH) has the cleanest atmosphere on Earth, but is subject to rapid changes regarding the emissions landscape and a warming climate, which strongly impact local atmospheric composition. Carbon monoxide (CO) is a product of methane (CH4) oxidation and a modulator of the tropospheric oxidizing capacity. In this study we investigate if the rapid increase in tropospheric CH4 levels of ∼7 % between 2008 and 2022 has driven a positive response in CO in the remote SH, based on multi-platform observations and a global chemistry transport model. We find inconsistencies in observations from aircraft, surface and satellite. CSIRO surface flask observations, AGAGE in-situ measurements (2008–2022) and airborne observations from the HIPPO (2009–2011) and ATom (2016–2018) campaigns show CO increases of 5–10 %. CO columns inferred from IASI (2014–2023) show more moderate increases of 3–4 % (statistically insignificant). NOAA surface flask observations and CO columns from MOPITT (2008–2022) do not show changes over time. GEOS-Chem simulations are consistent with the observed positive trend, exhibiting a significant CO increase of ∼7 % throughout the southern extratropical troposphere (2008–2022). These CO increases are partly driven by enhanced production from rising CH4 (∼50 %) and further affected by increases in biogenic non-methane volatile organic compounds (NMVOC) (∼45 %) and the direct effects of rising temperatures (∼5 %). We further explore how simulated CO levels in the remote SH could be impacted by future enhancements in primary and precursor emissions, changes in the oxidizing capacity, and temperature.
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
(3866 KB) - Metadata XML
-
Supplement
(2813 KB) - BibTeX
- EndNote
Status: open (until 04 Nov 2026)
-
CC1: 'Comment on egusphere-2026-5476', Domenico Taraborrelli, 16 Sep 2026
reply
-
CC2: 'Reply on CC1', Domenico Taraborrelli, 16 Sep 2026
reply
Reference
Bergamaschi, P., R. Hein, C. A. M. Brenninkmeijer, and P. J. Crutzen (2000), Inverse modeling of the global CO cycle: 2. Inversion of 13C/12C and 18O/16O isotope ratios, J. Geophys. Res., 105(D2), 1929–1945, https://doi.org/10.1029/1999JD900819
Citation: https://doi.org/10.5194/egusphere-2026-5476-CC2
-
CC2: 'Reply on CC1', Domenico Taraborrelli, 16 Sep 2026
reply
-
RC1: 'Comment on egusphere-2026-5476', Anonymous Referee #1, 05 Oct 2026
reply
review of manuscript egusphere-2026-5476
Title: What controls tropospheric carbon monoxide in the remote Southern Hemisphere?
Author(s): Clara M. Nussbaumer et al.
General comments:
This manuscript investigates carbon monoxide (CO) observations in the remote extratropical Southern Hemisphere, a region where atmospheric observations remain sparse. The authors compare multiple observational datasets with simulations from the GEOS-Chem model and identify discrepancies among measurement platforms, which they largely attribute to calibration and measurement uncertainties associated with the very low CO mole fractions characteristic of these pristine environments. The study highlights the importance of maintaining reliable, consistent, and long-term observations to robustly assess trends in atmospheric composition in the remote Southern Hemisphere.
The analysis primarily focuses on long-term trends in annual mean CO and their potential drivers. In my view, this represents a missed opportunity. Greater attention to higher-frequency variability, interannual variability, and spatial differences among the observing sites, together with their consistency with GEOS-Chem simulations, could strengthen the conclusions and provide additional insights into the processes controlling CO in the region. Some of these analyses are already presented in the Supplement and could be moved to the main text to increase their visibility and scientific impact.
Overall, the manuscript addresses an important topic that is well within the scope of Atmospheric Chemistry and Physics. The study is carefully conducted and provides valuable insight into the challenges of quantifying long-term CO changes in the remote Southern Hemisphere. I recommend publication after minor revisions.
Specific comments
Lines 9-10:
"NOAA surface flask observations and CO columns from MOPITT (2008-2022) do not show changes over time." I presume that these datasets do exhibit variability over time. It would be more precise to state that they do not show a significant long-term trend.Figure 1:
Consider replacing "generation" and "removal" with the more commonly used terms "sources" and "sinks". It would also be helpful to indicate the relative contributions (percentages) of the different sources and sinks. Figure 1 in https://acp.copernicus.org/articles/25/1105/2025/ provides a useful example.Lines 19-33:
While changes in anthropogenic emissions and future projections are discussed, the role and the future evolution of biomass-burning emissions is not addressed. Please add a brief discussion of expected future developments and associated uncertainties.Lines 40-75:
This section largely contains textbook knowledge and could be shortened substantially.Line 88:
Consider adding "cleaner combustion processes", as improvements in combustion technology have increased combustion efficiency and promoted more complete oxidation of fuel carbon to CO2, thereby reducing CO emissions.Lines 98-99:
It may be worth noting that the pause in atmospheric CH4 growth lasted only about eight years. CH4 concentrations were increasing before 2000 and resumed their increase after approximately 2007.Lines 120-121:
Please refer the reader to Table S2. It would also be helpful to include station names in Table S2.Figure 2:
Increase the symbol size for improved readability.Lines 126 ff.:
It is unclear which observational products were used as input. Were individual flask measurements ("event data" in WDCGG terminology) used, or monthly aggregated data? For the continuous records, please specify which temporal aggregation was applied.Lines 126-127:
"We only include measurements that are flagged as representative of the background in the analysis." Please clarify whether this refers to existing WDCGG background flags or whether the authors applied their own background-filtering procedure.Line 129:
Please clarify the deseasonalization procedure. For example: "Time series were deseasonalized by subtracting the contribution of the fitted harmonic functions." Is this correct?Lines 136-137:
What is the approximate magnitude of the calibration-scale differences discussed here?Line 153:
Consider rephrasing to: "...we additionally use CO spot-sampling observations from the Programmable Flask Package (PFP)...", to emphasize that these are discrete flask measurements rather than continuous observations, like the AeroLaser data. Could the PFP data also be used to quantify and potentially correct calibration-scale differences?Lines 166 – 167:
"We … use … PFP [data]… to ensure that … can be directly compared …". What does that mean in quantitative terms? Does that mean that you correct for scale differences?Lines 183-184:
"However, these datasets do not show ..." Please specify which datasets are being referred to.Line 193:
Please clarify whether the MOPITT product used here represents total-column CO or tropospheric-column CO.Line 195:
It is not clear why NO2 data are introduced here. Please explain their relevance.Line 217:
Spell out the abbreviation at first mention: "...nitrogen monoxide (NO)...".Line 259:
What is the cause of the enhanced isoprene emissions in 2020?Line 264:
Given the relatively short period used for the trend analysis, did the authors consider using the full 2000-2022 record and applying a higher-order fit rather than a purely linear trend?Figure 3 and accompanying text:
No discussion is provided for the years 2000-2008 or for their exclusion from the trend analysis. How do the authors explain the elevated measured and modelled CO levels during the period of stagnant atmospheric CH4 growth?Lines 282-283:
Is the GEOS-Chem output available only at monthly resolution, or was it aggregated after the model simulation? Furthermore, how were the HIPPO and ATom observations compared with the model output (e.g. Fig. 4b)? Please explain how differences in temporal sampling and resolution between observations and simulations were accounted for.Lines 285-286:
Please cite https://doi.org/10.5194/amt-11-5565-2018.Line 299:
Add "Figure" before "4b".Lines 365 ff.:
A comparison of the observed and simulated seasonal cycles may provide additional insight and could strengthen the analysis.Figure 7 and Lines 378 ff.:
Why not present seasonally resolved data within the same figure?Lines 384-385:
These values appear to be derived from GEOS-Chem simulations. Please make this explicit and compare them with the values shown in Figure 1 and with typical literature estimates.Figure 8:
Consider revising the caption to: "Time series of CH4 mole fractions and biogenic NMVOC emissions derived from GEOS-Chem simulations..."Line 424:
"...while weakening the loss of CO via OH." Please consider referring to Fig. 11b here. The role of OH is discussed again in Lines 469 ff., including evidence for increasing OH concentrations. Either move part of that discussion to this section or provide a forward reference.Figure 10:
Please clarify whether the figure shows values for the entire Southern Hemisphere or only the Southern Hemisphere extratropics.Data availability:
Please provide URLs and/or DOIs for all datasets used in the study.Line 560:
Please clarify what is meant by "supporting data".Citation: https://doi.org/10.5194/egusphere-2026-5476-RC1 -
RC2: 'Comment on egusphere-2026-5476', Anonymous Referee #2, 06 Oct 2026
reply
The authors clearly put a large amount of work into this study, though the paper is unfocused and the new scientific insight is unclear. Given the amount of work that needs to be done to revise the paper, I'm rejecting the manuscript so that the authors can take the necessary amount of time to carefully address my major concerns and specific comments below.
Main concerns
Title: The manuscript’s title is generic and doesn’t actually reflect the paper’s main findings. For instance, it has been known for decades what controls tropospheric CO in the SH. I suggest that the authors refer to ACP’s guidelines for authors (https://www.atmospheric-chemistry-and-physics.net/policies/guidelines_for_authors.html) for examples of appropriate titles.
Abstract: Again, I recommend the authors refer to ACP’s guidelines for abstract composition. I’m not sure what the novel findings of the paper are. For instance, “We find inconsistencies in observations from aircraft, surface and satellite.” leaves me wondering why they are inconsistent. Is it simply because the various datasets represent different layers of the atmosphere (i.e., not an apples-to-apples comparison)? Would one expect them to give the same trend given the shortish lifetime of CO? The abstract also left me wondering if changes in the oxidizing capacity had any influence on the modeled CO trend.
Scientific gap that is addressed by this manuscript: Line 30: This statement is not compelling. First, is the remote SH such an important place to study? Why? What are the implications of future increases in CO in this region? Second, there have been quite a few attribution studies over the last three decades that have looked at in situ production and transport of CO to the SH. I recommend that you take a thorough look at the literature and summarize the findings of previous work. That will set the stage for what new insight that your work brings – what your work is adding to the body of scientific knowledge. The paragraph starting on Line 94 is similarly vague on what scientific gap in understanding that your work is addressing.
Specific comments
Lines 38-72: This part of the manuscript reads like a textbook. The information is well known and could easily be condensed to one paragraph without the need to add reactions explicitly.
Line 39-42 & Line 77: Depending on your definition, anthropogenic CO emissions are composed of fossil fuel burning and biomass burning (i.e., intentionally set fires) or just fossil fuel burning. I wonder if you are mixing your definitions. Biomass burning emissions increased since the 1980s in the SH and they make up a larger percentage of total anthropogenic emissions (fossil fuel + biomass burning) than in the NH.
Paragraph beginning on Line 76: I recommend that you focus this discussion on the SH or remote SH, especially since you’ve already said that transport between the hemispheres is minimal for directly emitted CO. What is known concerning CO trends over the last 50 y or so and what explanations for these trends have been presented in the literature?
Section 3.1: I recommend that you provide a roadmap at the beginning of the section, saying what the important findings are up front. “In this section, we show that various observations indicate …”. That is, tell the reader what your important conclusions are and then lay out how you came to those conclusions in the section.
Section 3.1: I would like details on how you compare the simulated concentrations to the observations. For instance, how many observations are there per month/year at the NOAA and CSIRO sites and is the model output sampled similarly? On Line 282, you casually say that you don’t filter the model output as you only have a monthly averages. How does this affect your comparison? I assume that you similarly use monthly averages for the aircraft observation comparison.
Line 338: MOPITT and IASI measure two different vertical extents of the troposphere. I would like to see a discussion on how this affects your discussion. SH CO is not well mixed horizontally, as your plots indicate, and it is not necessarily well mixed vertically either. Does your model simulation provide any insight?
Line 360: You could merge Section 3.1 and 3.2 instead into one coherent discussion.
Figure 6: Why are you only calculating trends for 2008-2022? It would be helpful to add vertical cross-sections to see how trends vary with altitude.
Figure 7: This may be more informative if you compared the contributions in the remote and non-remote SH. This raises the question of how do you define the remote atmosphere (e.g., CO < X ppbv)?
Section 3.2: Much of discussion seems to be rambling and I’m having trouble tying it all together. I recommend that you provide a roadmap at the beginning of the section, saying what the important findings are up front. “In this section, we show that our model simulations indicate …”
Figure 11: Why aren’t you showing this for the remote SH? Also, it may be interesting to see spatial maps.
Line 527: As mentioned above, you are comparing apples to oranges as the observations represent different atmospheric quantities. Given CO’s short lifetime, it would be compelling to see if the observations along with the model simulation provide any insight into trends throughout the remote SH troposphere.
I recommend that the authors refer to the ACP guidelines for help composing the conclusions.
Citation: https://doi.org/10.5194/egusphere-2026-5476-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 471 | 208 | 83 | 762 | 80 | 66 | 73 |
- HTML: 471
- PDF: 208
- XML: 83
- Total: 762
- Supplement: 80
- BibTeX: 66
- EndNote: 73
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors analyse and quantify the contribution of methane and BVOC oxidation to CO in the Southern Hemisphere. The key intermediate is HCHO whose sinks are said to be reaction with OH and photolysis. However, there are both evidence and understanding of other processes like reaction with HO2 (in the tropical UTLS) and cloud processing converting HCHO to formic acid. Are these processes accounted in the GEOS-Chem simulation analysed here? Is HCHO scavenged based only on physical solubility or also on its hydration and reaction with OH(aq)? The cloud processing of HCHO is consistent with the study by Bergamaschi et al. (2000) pointing to a likely CO-yield from methane oxidation being lower than 0.88 and potentially as low as 0.71. What is the simulated CO-yield for this model? Atmospheric models usually have a near-unity yield but there are exceptions. I would appreciate if the authors discussed this aspect in relation to the simulated 7% increase in CO during 2018-2022 in the southern extratropical troposphere.