the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 May 2026
Atmospheric methane growth rates, 2018-2024, driven mainly by emission changes but atmospheric photochemistry is important
Abstract. Global atmospheric methane growth rate has increased dramatically since 2007, peaking in 2021, yet the relative roles of emissions and atmospheric oxidation remain uncertain. Accurate attribution is essential for climate mitigation because emissions-driven and chemistry-driven changes imply fundamentally different policy responses. Here we investigate global methane production and loss from 2018 to 2024 using an ensemble Kalman filter coupled to the GEOS‑Chem model. We employ two methane inversion configurations: assuming climatological monthly OH distributions or jointly optimizing zonal mean OH with methane. The joint inversion reveals substantial interannual variability in OH, including an ~18% decline in 2020 followed by a recovery in subsequent years. Accounting for this variability reduces the inferred 2019–2020 emission increase by ~63% (14±6 Tg/yr versus 37±5.5 Tg/yr with fixed OH), demonstrating that changes in OH strongly influences source attribution. The total increased methane loss, 2024 minus 2019, is about 31±6 Tg/yr, including a temperature‑driven increase of the OH+methane reaction rate that represents about 40% of the total sink increase. Most of the sink variations originate from the tropics, where the largest shifts in emission occur. Despite year-to-year variations, emissions remain the primary driver of changes in the methane growth rate change, except in 2020. Both inversions identify significant emission reductions over tropical South America in 2023–2024, likely linked to regional drought. Broadly, our results underscore the necessity of jointly estimating methane sources and sinks to interpret recent atmospheric methane trends.
- Preprint
(2245 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 03 Aug 2026)
- RC1: 'Comment on egusphere-2026-2662', Anonymous Referee #1, 25 Jun 2026 reply
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 81 | 51 | 3 | 135 | 4 | 2 |
- HTML: 81
- PDF: 51
- XML: 3
- Total: 135
- BibTeX: 4
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors explore the relative importance of the CH4 sources and sinks in recent changes in the methane growth rate, by performing several different inversions of GOSAT CH4. In particular, their results emphasize the necessity of including both the sources and sinks in such an inversion to get accurate results as well as the large impacts of increasing temperatures on the CH4 loss to OH. While this is an interesting paper, the authors sometimes fail to justify their assumptions, leaving questions about the accuracy of the results, as outlined below. Once these issues are addressed, the paper will be suitable for publication in ACP.
Line 88: Where does the value of 1.2% come from? I see you used the same value in your 2023 paper, but that also has no justification.
Line 97: Sentence beginning with “Using” is poorly worded.
Line 110: Need a citation for the MERRA2 dataset. E.g. Gelaro, R., McCarty, W., Suarez, M. J., Todling, R., Molod, A., Takacs, L., et al. (2017). The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2). J. Clim., Volume 30, (Iss 13), 5419-5454.
Line 116: Are these annually varying OH fields or are they static for a given month? From later in the text, it seems the latter, but that should be explicitly stated here.
Line 125: Your reference is missing.
Line 131: Where does the 40% number come from for your uncertainty? How sensitive are your results to this value?
Line 134: How does this choice of adjusting the OH fields by latitude bands affect your results? Figure 3, for example, highlights the non-physical impacts of this methodology, which results in extreme discontinuities. If for example, you had adjusted the OH fields in 15 degree latitude bands, would you have gotten similar results? It is somewhat difficult to trust any quantitative conclusions without at least some discussion of the impacts of this scaling approach.
Line 142: You say here that 2018 is your spin up year, but you include 2018 in many of your figures and analysis. Sometimes, however, you omit discussion of 2018 from the analysis. Since you consider this to be spin up, why are you including it in the paper?
Line 160: This is where the inclusion of 2018 becomes confusing. According to your figure, OH in 2018 was even lower than in 2020. If you consider the results for 2018 to be meaningful, which including them in the figure would imply, then your statement here is incorrect.
Line 173: Your conclusions here seems entirely dependent on your choice of the size of the latitude bands for your OH scaling.
Line 214: I think you need to include more details about the GOSAT data in Section 2, particularly as they relate to XCO2. Are you using the same XCH4:XCO2 retrievals as in your earlier work? The discussion here would imply that, but there’s no discussion of this in the methodology.
Line 218: Are the biases compared to TCCON invariant in time? In that case I could see how they wouldn’t affect the interannual variability. If that’s true, you should state that. Otherwise, what evidence do you have that they don’t affect the interannual variations and how do you define “significantly”? That’s a very imprecise modifier.
Line 230: “whereas” is misspelled.