Methane emissions are predominantly  responsible for record-breaking atmospheric  methane growth rates in 2020 and 2021

Feng, Liang; Palmer, Paul I.; Parker, Robert J.; Lunt, Mark F.; Bösch, Hartmut

doi:https://doi.org/10.5194/acp-23-4863-2023

Articles | Volume 23, issue 8

https://doi.org/10.5194/acp-23-4863-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-23-4863-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 23, issue 8

Research article

| Highlight paper

|

25 Apr 2023

Research article | Highlight paper |

| 25 Apr 2023

Methane emissions are predominantly responsible for record-breaking atmospheric methane growth rates in 2020 and 2021

Liang Feng, Paul I. Palmer, Robert J. Parker, Mark F. Lunt, and Hartmut Bösch

Download

Final revised paper (published on 25 Apr 2023)
Preprint (discussion started on 17 Jun 2022)

Interactive discussion

Status: closed

RC1:
'Reviewer comment on Feng et al', Anonymous Referee #1, 18 Jun 2022

See attached.

Citation: https://doi.org/10.5194/acp-2022-425-RC1
- AC1:
  'Reply on RC1', Paul Palmer, 24 Jun 2022
  
  The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-425/acp-2022-425-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/acp-2022-425-AC1
  - RC2: 'Reply on AC1', Anonymous Referee #1, 24 Jun 2022
    
    This reviewer appreciates the quick response from the authors. Although I would like to clarify one point in case it was not well articulated in my initial review.
    It is this reviewers understanding that previous work arguing for a role of chemistry claims that chemistry could explain some large fraction of the methane growth. My reading of this manuscript is that the central conclusion is that methane emissions are responsible for for the methane growth. The former claim that a process can explain a phenomena is justified through forward simulations. In this reviewer's opinion, arguing that a process is responsible for a phenomena requires a higher burden of proof. This reviewer was unconvinced that the numerical experiements presented in the inital manuscript meet the standard to justify the latter claim. This is why I suggested either reframing the conclusions to include approriate caveats or presenting additional numerical experiements to fully justify the central claims. Apologies for any confusion in my inital review.
    
    Citation: https://doi.org/10.5194/acp-2022-425-RC2
    
    AC2: 'Reply on RC2', Paul Palmer, 24 Jun 2022
    
    This reviewer is correct that studies proposing that OH played a large role in the atmospheric growth in 2020 used forward simulations. Point well taken. However, these studies were effectively blind to the methane emission hotspots that we can see from space, drawing attention to the disconnect between different modelling approaches used by different parts of the community. Also, these studies tended to make their claims a little more robustly than suggested by this reviewer.
    We sincerely thank the reviewer for keeping us honest and challenging us to provide additional evidence to support our claim. In our submitted response, we have suggested expanding our caveats and including additional experiments. Our preliminary follow-on experiments are consistent with our original claim, which we will detail in the revised manuscript.
    
    Citation: https://doi.org/10.5194/acp-2022-425-AC2
CC1: 'Comment on acp-2022-425', Janne Hakkarainen, 01 Jul 2022

I noticed that the second formula in Appendix B has an extra k in the denominator.

Citation: https://doi.org/10.5194/acp-2022-425-CC1
RC3: 'Comment on acp-2022-425', Anonymous Referee #2, 06 Aug 2022

Feng et al attempts to attribute the stunning increases in the methane growth rate from 2020-2021 using satellite observations of methane in conjunction with surface data. The subject is very timely and worth getting out into the community for further analysis and discussion. However, there are a number of issues that the authors need to address, which are incorporated into the annotated text of the paper. In particular, there are no uncertainty estimates of the fluxes, no real explanation of the Eastern African increase, and an unrealistic description of OH change. Spending more time on these elements will strengthen the paper and adding credibility to the results.

Citation: https://doi.org/10.5194/acp-2022-425-RC3
AC3: 'Integrated and updated author responses to reviewer comments of acp-2022-425', Paul Palmer, 20 Dec 2022

The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-425/acp-2022-425-AC3-supplement.pdf

Citation: https://doi.org/10.5194/acp-2022-425-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Paul Palmer on behalf of the Authors (20 Dec 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (21 Dec 2022) by Bryan N. Duncan

RR by Anonymous Referee #1 (27 Dec 2022)

Suggestions for revision or reasons for rejection

First, I think the manuscript is greatly improved and I commend the authors for both adding additional inversions and using simulations that more accurately represent previous work. However, in this reviewer’s opinion, the paper needs to be restructured. I am generally reluctant to suggest structural changes to a manuscript, but the current structure resulted in both confusion and misunderstandings by this reviewer about what the authors *actually* did and what they report. Frankly, this reviewer is still confused about which numbers are being reported in the abstract and what contributes to their uncertainties. This reviewer feels the manuscript still needs major revisions before being publishable.

To start, this manuscript would greatly benefit from a Table that explains the various simulations including what was held constant, what was perturbed, etc. It could list both 3D inversions, forward simulations, and the Box Model. This should follow a general description of the methodology.

The material in the FIVE appendix sections should be moved into the main text. Much of this material is central to their conclusions and, in this reviewer’s opinion, should not be relegated to the appendix. The main text references an OH sensitivity run, but the manuscript now includes at least 3 different descriptions of OH: their 5% OH change over CO2 source regions, OH described in Appendix D, and OH inferred in Appendix E. Following this, it is not clear to this reviewer why the authors still include the 5% OH change over CO2 source regions. Both reviewers criticized this sensitivity run because it does not represent what previous work found. Yet this is currently the only OH sensitivity run that is described and discussed in detail in the main text.

Regarding the reported numbers and conclusions, it seems that the authors should be reporting results from their joint inversion of both OH and methane emissions. This seems like the numerical experiment that allows them to make a quantitative statement about the relative importance of sources and sinks, which is the central claim of the manuscript. However, from my reading of the manuscript, I do not think this is the numerical experiment that is used for the numbers in their abstract/conclusions, although I am not actually sure which experiment their numbers come from.

Regarding the title, I still think their title is too strong given the conclusions. From their inversions, it seems like both emissions and sinks are found to have changed and driven the rise. It sounds like 25-30% of the changes are attributed to sinks and 70-75% to sources. Yet someone reading the title and abstract would get a very different message. If so, that seems important and a more nuanced title is needed.

Regarding the box modeling, does the box modeling reflect the updated findings regarding OH? It seems that the text was not updated even though the authors state: “after considering the effects of methane sinks, we find that a one-box model calculation…” (Line 133).

It is still not clear how much the authors are perturbing global mean OH in their sensitivity runs. This would be important to state. It seems that it is still less than other work reports.

I have a number of questions regarding the OH inversion as I think this is the numerical experiment that actually supports their conclusions and claims:
- What do the spatial patterns of the OH inversion look like?
- What about the temporal pattern?
- The authors mention that results in Appendix D and E are consistent but provide no numbers, figures, or really anything to back up that claim. How was this assessed? What is the bar for "consistency"?
- The basis functions for OH differ from methane, does that matter?
- The authors mention that we do not have sufficient constraints for OH, yet they aim to solve for both longitudinal and latitudinal changes. Are those well constrained? Why not just solve for OH as a function of latitudinal bands?

Another paper was recently published in Nature that uses similar data and reaches similar conclusions (~50/50 sources and sinks; https://www.nature.com/articles/s41586-022-05447-w). This work should be mentioned as it is directly relevant. This paper also included both methane and OH in the state vector (as did Zhen et al., 2022). Therefore I stand by my earlier review that the bar for claiming emissions are responsible for the changes necessitates inverting for both methane and OH.

Hide

ED: Reconsider after major revisions (31 Dec 2022) by Bryan N. Duncan

AR by Paul Palmer on behalf of the Authors (15 Mar 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (15 Mar 2023) by Bryan N. Duncan

RR by Anonymous Referee #1 (17 Mar 2023)

ED: Publish subject to technical corrections (18 Mar 2023) by Bryan N. Duncan

AR by Paul Palmer on behalf of the Authors (29 Mar 2023) Author's response Manuscript

Post-review adjustments

AA: Author's adjustment | EA: Editor approval

AA by Paul Palmer on behalf of the Authors (19 Apr 2023) Author's adjustment Manuscript

EA: Adjustments approved (19 Apr 2023) by Bryan N. Duncan

Executive editor

After a period of near-zero growth of atmospheric methane, a major greenhouse gas, its growth rates have increased, with values in 2020 and 2021 exceeding all prior values since the beginning of systematic measurements in 1983. This recent acceleration, particularly during the covid-19 years 2020 and 2021, raises the question whether the methane increase was due to stronger sources or reduced sinks in the troposphere. A reduction of the tropospheric abundance of the hydroxyl radical (OH), the reaction of which is the main tropospheric methane sink, could be plausibly explained by global-scale reductions in nitrogen oxides due to pandemic-related industry shutdowns. Using an inversion scheme, Feng et al. demonstrate that such a reduction only accounts for about 34% in 2020 and 10% in 2021 of the observed methane rise. Instead, the authors attribute increased methane emissions to hydrological anomalies and microbial sources over the tropics, i.e., Eastern Africa and tropical South America, and temperate North America. The study demonstrates the importance of simultaneously accounting for changes in methane emissions and sinks for an improved quantitative understanding of the evolution of the methane concentrations to assess the role of greenhouse gases in climate change and tropospheric composition.