the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Sep 2026
The Added Value of Amazon observations to estimate methane budget over South American river basins in a global inversion system
Abstract. Tropical South America is one of the world's largest natural methane (CH₄) emitting regions, yet regional methane budgets remain highly uncertain due to poorly constrained wetland emissions and sparse atmospheric observations. Here, we quantify methane emissions from the Amazon, Orinoco, and Paraná river basins for 2010–2019 using the Jena CarboScope atmospheric inverse modelling system. We investigate the variability in the methane budget through a series of sensitivity experiments. Our inversion estimates methane emissions of 4.44 ± 0.88 Tg CH₄ yr-1 for the Orinoco, 28.60 ± 5.15 Tg CH₄ yr-1 for the Amazon, and 30.10 ± 3.95 Tg CH₄ yr-1 for the Paraná basin, together accounting for ~59 % of the total estimated South American methane emissions (105.5 ± 11.10 Tg CH₄ yr-1). The inversion substantially modifies the seasonality and interannual variability of prior wetland emissions in all three basins. The dominant source of posterior variability differs among basins. The data-sparse Orinoco is most sensitive to the wetland prior, indicating strong dependence on prior information. Regional observations mainly constrain the Amazon, which accounts for 55 % of the variability in South American posterior methane emissions. The Paraná is largely insensitive to regional observations and is constrained primarily by the global network. These results show that uncertainty sources vary across South America. Improving wetland priors and expanding regional atmospheric measurements are essential for reducing uncertainty in regional methane budgets.
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
(6930 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 24 Nov 2026)
- RC1: 'Comment on egusphere-2026-5245', Anonymous Referee #1, 17 Oct 2026 reply
Data sets
Monthly Posterior Methane Fluxes over South America from the Jena CarboScope Global Atmospheric Inversion System, 2010–2019 Shrutika Wagh https://edmond.mpg.de/dataset.xhtml?persistentId=doi:10.17617/3.YH7RYS
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 280 | 103 | 28 | 411 | 35 | 30 |
- HTML: 280
- PDF: 103
- XML: 28
- Total: 411
- BibTeX: 35
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The study by Wagh et al., entitled “The Added Value of Amazon observations to estimate methane budget over South American river basins in a global inversion system” investigates the effect of observational data and different prior inventories onto the methane budgets in South America across three different river basins, including the Amazon, Parana, Orinoco basins. This is tested with a number of sensitivity studies by varying the prior and observational conditions of the inversion and analyzed by the posterior regional fluxes, their seasonal variability, and change in uncertainty contributions.
This manuscript is a valuable contribution to the research community, as it investigates the effect of prior and observational choices onto an inversion framework. I would recommend it for publication in ACP once the authors have addressed the comments and suggestions raised in this review.
General Comments:
The authors should motivate and explain their general perspective of their study more in the introduction and the conclusion. It is not quite clear how the structural sensitivity is meant to be investigated in their inverse system and what value their study has for the community and for the current methane budget.
Could the authors include more details regarding their transport model TM3? The horizontal and vertical resolution seems quite large for using it for in-situ airborne and ground measurements. How good is the transport model considering turbulence and convection? What is the observational coverage/footprint in the Orinoco and Parana from the Amazonian observations? The authors only included the effect of the observations onto the inversions but it is not quite clear if this is due to no coverage or dilution/mixing over long-distance travelling air masses/emissions or just well-agreeing emissions with the inventories.
Did the authors consider the effect of spatial resolution differences throughout the priors? Are the spatial resolutions correctly aggregated, so that these comparisons can be made? Particularly for heterogeneous sources such as wetlands, coarser grid cells can smooth emission hotspots, change the spatial covariance structure, alter the prior uncertainty assigned to the domain, and change how the inversion allocates emissions in response to atmospheric observations (e.g., Ma et al., 2021). Particularly for the comparison between the different wetland inventories (“wet1”, “wet2”) this might be crucial to discuss in more detail.
Please add numbering to the equations.
The authors should include a discussion regarding inter-annual variability on their results. Particularly during meteorological extreme events (El Nino/ la Nina) and the environmental changes on tropical ecosystems and if CarboScope is able to reproduce those. Or discuss those with appropriate literature references.
The authors include a short section (2.1.3) of methane sinks, but do not really include the effect of methane sinks into their sensitivity studies. Would it be possible to add another sensitivity experiment investigating the OH-, Cl-, soil sinks onto the inversion results? Or include more discussions regarding the importance of sinks in those regions based on literature.
This study would benefit in general from more literature references to support the findings and put them into a broader context (e.g., include literature from more recent studies, studies of other tropical regions with similar methods, from more heterogenous authors/groups).
Specific comments:
Line 36: Add sensitivity “studies”.
Line 46: To what do the authors refer to by uncertainty sources?
Line 51: Remove “Following a temporary pause in the early 2000s” and just start with “Atmospheric methane”.
Line 58: change “wetlands” to “wetland”.
Line 60: Change to “known to be the largest methane source”.
Line 66 + 68: correct to “approaches”
Line 95-96: What about more recent work on this? WetCHARTs was also used on more recent top-down inversion approaches investigating tropical wetland methane emissions (Ort et al., 2026, Parker et al., 2020, Balasus et al., 2026).
Line 137: What time period the ATTO observations cover? Is the full extent of the data used in this study published in Botia et al, 2020 and van Asperen et al., 2023? Otherwise, a time series of the ATTO data should be included in the Appendix.
Line 153: Why is the horizontal resolution of the model approximated?
Line 157: Please add a short explanation why the inversion was performed specifically for the chosen time period (2001-2020). Aren’t later observations and inventories available? And why is the main focusing period set to 2010-2019? What happened to the years 2003-2009? Are there large differences in the results by using different spin-up periods? Particularly as during 2001-2003 global methane emissions did not change a lot.
Line 162: Add the Bayesian inversion cost function as equation here and introduce the components, as they are also used in the subtitles of the method section.
Lines 171-173: Prior uncertainties differ quite a lot depending on the source. Are relative fractions of sources within a multi-source cell available?
Lines 194-195: But for all sources combined?
Line 197: Add a dot after the equation.
Table 1: Please explain more how you achieve those numbers. Maybe add the individual total fluxes for each source category here.
Line 226: Add “are” in front of “described” and change “described” to “shown” or “listed”.
Line 254: Consider adding here more “ATTO”-literature related to greenhouse gases (e.g., Sierra et al., 2025, Botía, Santiago, et al., 2020).
Figure 1: Correct “South America” in the Figures caption and “Global stations”, “Amazon station” in the Figure.
Line 288: Correct superscript of O(1D).
Line 295-296: Please correct this sentence.
Line 301: Add a comma after “study”.
Line 339: Change “to” to “in”.
Line 344: Add “can” in front of “constrain”.
Line 351-356: Explain why the threshold of 2500m was chosen to separate the vertical profiles.
Line 379: What about TROPOMI satellite measurements?
Table 3: Correct superscript of O(1D) (3x).
Line 436: Add a dot after the equation.
Line 450: Add a dot after the equation and use an equation format.
Figure 2: Could the color of the posterior bars be changed, so that the error bar becomes more visible? Or the color of the error bars (avoid black on black). And why is the Orinoco pie chart smaller than the others? Please correct or explain the reason in the caption.
Lines 467-481: I think the caption belongs to a previous version of this plot. Pie charts are displayed in (c) not in (a).
Line 495: There is a “.” Missing at the end of the sentence.
Figure 3: Is the same scale necessary for the uncertainty and the spread? Otherwise, I would suggest adapting the scale of b) to its values for better readability.
Line 537: SA = South America? Define when using first.
Line 548: “Here,”
Figure 4: Is this seasonality averaged for the whole time period? Please specify in the caption. Also define IAV, when first used.
Lines 609 – 622: Isn’t the change in removing individual stations an effect of the different seasonality’s the Amazon basin has, found in previous studies (Basso et al., 2021)? So the heterogenous distribution of wetlands? You may extent on this also in your discussion in Lines 877-888 as well.
Lines 673-675: By comparing the monthly anomalies for total posterior fluxes for the ST_all and REF cases, they do no quite look that similar to me. There are large differences in some years (2011, 2013, 2018, 2020), which might be to be pointed out.
Line 676-692: Why is Figure A7 a supplemental figure? Consider to include it in the main text here as this paragraph is only focusing on it.
Line 679: remove “s” in “shows”.
Line 752: Please add a reference here for the Taylor diagram.
Line 768: change “concentrations” into “mixing ratios”. You are referring to “ppb”!
Lines 786-788: On what grid scale level they did calculate emissions and which time periods are those studies addressing?
Lines 910-913: Is it still a bimodal seasonal pattern if the peaks basically merge into one through the change of years (peak from Nov-Feb)?
Line 924: “Basin” is written capitalized here. Try to make it consistent over the manuscript.
Line 935-939: This sentence is really long and may be separated into shorter sentences for better readability.
Line 944 – 947: Wetland-river continuum areas have been observed to be particularly underestimate in the Amazon in a recent study (Ort et al., 2026). Have you considered using the WetCHARTs inundation product for your inversion comparisons?
Figure A2: Make sure to adapt here the color same as Figure 2 and that all error bars are fully shown (some seem to be cut at the top).
Line 1053: The definition of SA should come earlier.
Literature:
Parker, Robert J., et al. "Exploring constraints on a wetland methane emission ensemble (WetCHARTs) using GOSAT observations." Biogeosciences 17.22 (2020): 5669-5691.
Balasus, Nicholas, et al. "2019–2024 trends in African livestock and wetland emissions as contributors to the global methane rise." Atmospheric Chemistry and Physics 26.7 (2026): 4601-4617.
Ort, Linda, et al. "Airborne observations reveal underestimated riverine methane emissions across the Amazon." Geophysical Research Letters 53.13 (2026): e2026GL122310.
Ma, S., Worden, J. R., Bloom, A. A., Zhang, Y., Poulter, B., Cusworth, D. H., et al. (2021). Satellite constraints on the latitudinal distribution and temperature sensitivity of wetland methane emissions. AGU Advances, 2, e2021AV000408. https://doi.org/10.1029/2021AV000408
Sierra, Carlos A., et al. "The flask monitoring program for high-precision atmospheric measurements of greenhouse gases, stable isotopes, and radiocarbon in the central Amazon region." Earth System Science Data 17.11 (2025): 5871-5884.
Botía, Santiago, et al. "Understanding nighttime methane signals at the Amazon Tall Tower Observatory (ATTO)." Atmospheric Chemistry and Physics 20.11 (2020): 6583-6606.