SO<sub>2</sub> emissions derived from TROPOMI observations over India using a flux-divergence method with variable lifetimes

Chen, Yutao; van der A, Ronald J.; Ding, Jieying; Eskes, Henk; Williams, Jason E.; Theys, Nicolas; Tsikerdekis, Athanasios; Levelt, Pieternel F.

doi:https://doi.org/10.5194/acp-25-1851-2025

Articles | Volume 25, issue 3

https://doi.org/10.5194/acp-25-1851-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-25-1851-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 3

Research article

|

11 Feb 2025

Research article |

| 11 Feb 2025

SO₂ emissions derived from TROPOMI observations over India using a flux-divergence method with variable lifetimes

Yutao Chen, Ronald J. van der A, Jieying Ding, Henk Eskes, Jason E. Williams, Nicolas Theys, Athanasios Tsikerdekis, and Pieternel F. Levelt

Download

Final revised paper (published on 11 Feb 2025)
Supplement to the final revised paper
Preprint (discussion started on 22 Apr 2024)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-1094', Chris McLinden, 17 May 2024

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1094/egusphere-2024-1094-RC1-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2024-1094-RC1
- AC1: 'Reply on RC1 and RC2', Yutao Chen, 29 Jul 2024
  
  See response in supplement.
  
  Citation: https://doi.org/10.5194/egusphere-2024-1094-AC1
RC2:
'Comment on egusphere-2024-1094', Anonymous Referee #2, 20 May 2024

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1094/egusphere-2024-1094-RC2-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2024-1094-RC2
- AC2: 'Reply on RC1 and RC2', Yutao Chen, 29 Jul 2024
  
  See response in supplement
  
  Citation: https://doi.org/10.5194/egusphere-2024-1094-AC2

Peer review completion

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

AR by Yutao Chen on behalf of the Authors (29 Jul 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (12 Aug 2024) by Jianzhong Ma

RR by Anonymous Referee #3 (02 Oct 2024)

Suggestions for revision or reasons for rejection

The study "SO2 emissions and lifetimes derived from TROPOMI observations over India using a flux-divergence method" by Chen et al.
presents emission estimates over India based on TROPOMI observations and a flux-divergence method.
The paper generally reads well and presentation quality is good.
However, I don't see the main conclusion to be sufficiently supported by the presented results, as detailed below.
While I don't think that all the open questions need to be solved within this study before publication,
the authors need to extend the discussion substantially and have to present their conclusions more cautious and less absolute.

Before publication, the following issues need to be resolved:

Major issues:

1. Title
SO2 emissions have been derived from TROPOMI observations.
However, the lifetimes have not! They are actually based on CAMS simulations, and the CAMS-based lifetimes are the basis of the conclusions that emissions in CAMS are too high.
Thus, the title is misleading, and "and lifetimes" should be skipped.

2. SO2 lifetime

The main finding of the study is that TROPOMI SO2 columns over India are lower than CAMS model simulations, and the conclusion of the study is that bottom-up emissions of SO2 are too high. However, the discrepancy could also have other reasons, in particular that the modeled SO2 lifetime is too high.

I see the following aspects that need further discussion:

(a) OH climatology: The SO2 lifetime is estimated based on CAMS OH climatology. However, this approach completely ignores the very special conditions within the power plant plumes: In particular due to the co-emitted NOx, plume-OH is probably highly variable and can substantially differ from the climatology!

(b) Clear sky: The authors claim that only reaction with OH is relevant, as only cloud free observations were considered. However, this is not completely true:
- A cloud filter was applied, but pixels with up to 30% cloud fraction are still included! (as I understand from Fig. 9; the cloud threshold has to be added to section 2.1)
- Even if a pixel is cloud free at TROPOMI overpass, the observed air mass might have been in contact with clouds over the last hours.
- Also heterogenous reactions with aerosols can play a role.

My strongest concern about the main conclusion is actually raised by Fig. 9: While the difference in SO2 columns is clear, this could indeed be due to input emissions, but also due to the loss processes in CAMS. It is hard to tell from the Figure alone, but it looks like CAMS (even if values would be halved) shows substantial outflow of SO2, e.g. over the oceans, that is not observed from TROPOMI. Thus I would conclude that the CAMS SO2 lifetime is definitely too high.

The authors need to
- include a discussion about plume chemistry (affected by strong local NOx emissions) on OH and how far it is appropriate to use the OH climatology here,
- be aware that "cloud free" does not really mean free of any clouds, and heterogeneous reactions on clouds and aerosols need to be discussed as well, which actually results in lower SO2 lifetimes,
- include a discussion about remaining uncertainties, and the possibility that the CAMS lifetime is actually too long, as indicated in Fig. 9.
With these open questions, I consider the given uncertainties of 35% (monthly) and 10% (annually) to be far too low.

3. Implementation of the derivative

The authors state that the analysis was performed on 0.4° grid, on 0.1° grid, and "TROPOMI measured pixels".
It is not clear to me what the latter actually means:
Is the analysis done on a regular grid with 5.5 km × 3.5 km resolution (as indicated in line 267)?
Or is the original "grid" of the TROPOMI measurements (along x across) used for calculating the derivative?
I would not understand the first option, and I see problems with the pixel size changing across the swath.
If the second option is meant, this should be explained and motivated in more detail, and a reference to de Foy and Schauer, 2022, who proposed this approach, has to be added.

The information content of the divergence term comes from the change of horizontal flux from one pixel to the next.
For this, a resolution of 0.4° is far too coarse.
I would expect that (a) calculating the divergence on the TROPOMI grid and (b) re-gridding and averaging it afterwards on high spatial resolution (0.05°) yields high-resolution information about the location of SO2 point sources, without the need for modifying the difference quotient.

Additional comments:

Line 103: "and its divergence": "its" refers to the sinke term, whereas the divergence is derived from the horizontal flux.
Line 123: Please specify the version of the COBRA SO2 product.
Line 334: This pixel size is valid for nadir. Towards the swath edges, pixels become significantly larger (even larger than 0.1°!)
Figure 9: Resolution is poor. Lat/lon numbers should not be bold.

Reference:
de Foy and Schauer, DOI 10.1088/1748-9326/ac48b4

Hide

ED: Reconsider after major revisions (05 Oct 2024) by Jianzhong Ma

AR by Yutao Chen on behalf of the Authors (15 Nov 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (17 Nov 2024) by Jianzhong Ma

RR by Anonymous Referee #3 (09 Dec 2024)

ED: Publish as is (10 Dec 2024) by Jianzhong Ma

AR by Yutao Chen on behalf of the Authors (18 Dec 2024) Manuscript

Download

Article (10384 KB)
Full-text XML

Short summary

There is a lack of local SO₂ top-down emission inventories in India. With the improvement in the divergence method and the derivation of SO₂ local lifetime, gridded SO₂ emissions over a large area can be estimated efficiently. This method can be applied to any region in the world to derive SO₂ emissions. Especially for regions with high latitudes, our methodology has the potential to significantly improve the top-down derivation of SO₂ emissions.

SO2 emissions derived from TROPOMI observations over India using a flux-divergence method with variable lifetimes

Download

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection

SO₂ emissions derived from TROPOMI observations over India using a flux-divergence method with variable lifetimes