Optimal selection of satellite XCO<sub>2</sub> images for urban CO<sub>2</sub> emission monitoring

Danjou, Alexandre; Broquet, Grégoire; Schuh, Andrew; Bréon, François-Marie; Lauvaux, Thomas

doi:https://doi.org/10.5194/amt-18-533-2025

Articles | Volume 18, issue 2

https://doi.org/10.5194/amt-18-533-2025

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

https://doi.org/10.5194/amt-18-533-2025

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

Articles | Volume 18, issue 2

Research article

|

29 Jan 2025

Research article |

| 29 Jan 2025

Optimal selection of satellite XCO₂ images for urban CO₂ emission monitoring

Alexandre Danjou, Grégoire Broquet, Andrew Schuh, François-Marie Bréon, and Thomas Lauvaux

Download

Final revised paper (published on 29 Jan 2025)
Preprint (discussion started on 16 Oct 2023)

Interactive discussion

Status: closed

RC1:
'Review of Danjou et al.', Anonymous Referee #1, 21 Nov 2023

The comment was uploaded in the form of a supplement: https://amt.copernicus.org/preprints/amt-2023-199/amt-2023-199-RC1-supplement.pdf

Citation: https://doi.org/10.5194/amt-2023-199-RC1
- AC1: 'Reply on RC1', Alexandre Danjou, 12 Mar 2024
  
  The comment was uploaded in the form of a supplement: https://amt.copernicus.org/preprints/amt-2023-199/amt-2023-199-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/amt-2023-199-AC1
RC2:
'Comment on amt-2023-199', Anonymous Referee #2, 01 Dec 2023

The comment was uploaded in the form of a supplement: https://amt.copernicus.org/preprints/amt-2023-199/amt-2023-199-RC2-supplement.pdf

Citation: https://doi.org/10.5194/amt-2023-199-RC2
- AC2: 'Reply on RC2', Alexandre Danjou, 12 Mar 2024
  
  The comment was uploaded in the form of a supplement: https://amt.copernicus.org/preprints/amt-2023-199/amt-2023-199-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/amt-2023-199-AC2

Peer review completion

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

AR by Alexandre Danjou on behalf of the Authors (11 Apr 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (03 May 2024) by Andre Butz

RR by Anonymous Referee #1 (18 May 2024)

Suggestions for revision or reasons for rejection

Review of the revised manuscript “Optimal selection of satellite XCO2 images over cities for urban CO2 emission monitoring” by Danjou et al.

I would like to thank the authors for their revisions, which improved the descriptions and the understandability of the study. However, some of the authors’ responses to the reviews need more explanation and discussion in the manuscript which is why I still recommend major revisions before publication. The comments can be found below. The line numbers refer to that of the revised manuscript.

Re their responses on the reviewers’ comments

The authors asked in their responses where in the manuscript it is written that the method is not universally applicable to all cities and is still work in progress. The answer to this is e.g. in line 344: “However, despite the application of this criterion, the variability of the error distribution remains large across cities.” and in the conclusions line 559: “These significant remaining biases raise the question of the current reliability of the results obtained on a single given city.” So the comment by Reviewer 2 is still open: “However, [the manuscript] should also describe how the proposed method would be applied to real satellite measurements and critically assess, whether it’s applicable at all within the error budget.”

Re response by the authors about the interpolation methods used: There is no "classic" interpolation as mentioned by the authors in their response to the reviewers' comments. So which method is used? Please add this information somewhere around line 140.

Around Fig. 4 and its results: The threshold of 2.1 ktCO2/h is only based on the city emissions from the inventory. But there is no discussion whether the current satellites are able to measure emissions in this order of magnitude. Since the manuscript is based on OCO-3 and CO2M (e.g. random noise according to the precision requirement of 0.7 ppm of CO2M) there has to be a discussion somewhere in the manuscript which is the lower limit of emissions that is observable by the satellite. Otherwise, this threshold is fine in theory but not at all applicable in practice. For instance, typical parameters could be used to convert the emission of 2.1ktCO2/h to XCO2 enhancement using your equations as done in the literature, e.g. [1], [2]. If this is larger than the noise of the instruments (~0.7 ppm), then it is fine, if not, the emission corresponding to a 0.7 ppm enhancement has to be used as the emission threshold in the study.

[1] https://amt.copernicus.org/articles/4/1735/2011/amt-4-1735-2011.pdf
[2] https://www.sciencedirect.com/science/article/pii/S0959652623006832

Sect. 5.1: Since this analysis motivates the use of the machine learning algorithm described in Sect. 4.2, I still think that it would be better to move this between Sect. 4.1 and 4.2. In addition, the reader has to always move back and forth between 4.1 and 5.1 to understand all the details mentioned in both sections. I understand the authors’ argument that it is common practice to describe methodology first, but I think it is different when the methodology depends on previous results. And also from the point of view of the reader it would make the manuscript better understandable.

Fig. A1: I still think that because the description of this figure is in the main text, it should appear there (Sect. 2.3 or 2.4). Or at least there should be a text describing the results in Appendix A. Currently, it is just the figure without any description.

Specific Comments

• L13: In the main text it is always 2.1, not 12.1 ktCO2/h.
• L79: Please add "about" before 2km x 2 km because it varies with the distance to nadir and the ratio will not be exactly a square.
• L88: It is not clear what "realistic" means here. Is realistic in the sense of the OLAM model applied? Is it the grid spacing which is similar to the expected satellite footprint size?
• L138: PBL not defined
• Sect. 2: There has to be a note somewhere that your 40-day simulation is free-running because then the forecast skill after 14 days is small and the meteorology is not realistic (but of course consistent within the model).
• L149: Why are the dates different from the previous manuscript? Are the authors sure that the simulation did not start at 01 August? That would also be consistent with 8 images per day for 31 cities.
• L154: "that expected for CO2M": This is only true if the city is in the center of the nadir swath of CO2M. Please clarify and rephrase.
• L169: "most emitting pixels": As pixels cannot emit themselves I think the authors mean the pixels with largest emission sources. Please clarify and rephrase.
• L286: "We thus obtain at most 4 subsets": It's not really clear where the number 4 is coming from. Is it because the decision tree has a depth of two and the whole set of 9920 images is split into two subsets in each step? A reference to Fig. 2 would be helpful here, I guess.
• Fig. 2: "pseudo image" should be replaced by "synthetic image" here as well for consistency.
• L315-317: This citation looks weird although I realize that this the recommended citation. Maybe this can be abbreviated in some way?
• L338: "the optimisation does not converge": Do the authors mean the minimization process described in Sect. 3? What is the criterion for convergence?
• L369: What is GP2?
• Fig. 4: replace pseudo-images by synthetic images
• L443: It is not clear what is meant by "accuracy" here. Is it the bias or the spread of the distributions shown in Fig. 5? Please clarify in the manuscript.
• Fig. 6: panel (a) still saying 11 degrees as the threshold.
• L515: I think these percentages are the fraction of cities that pass all the criterions. I don't agree with the statement that their emissions are "easier to quantify than cities on other continents" because this is not relative to other continents. Please rephrase.
• L518: 0.7 ppm is the precision requirement for CO2M, not accuracy of OCO-3. Please clarify and rephrase or provide references here.
• L528: This sentence still suggests that ERA-5 has only 37 vertical levels, which is not true. The authors should add here that they are using the ERA-5 product on constant pressure levels. By the way, it would be better to use the native resolution of the reanalysis with 137 vertical levels to ensure that the best resolution within the boundary layer is used, which is most relevant for this study.
• L585: It is confusing that 7 degrees are used here whereas in the main text, 12 degrees was found to be the threshold to be used. Is this an analysis that preceded the analysis of the main text? It would be helpful then if this is mentioned in the text.
• Fig. A1 is not referenced anywhere in the text. Please add a reference somewhere and put the figure at this place.
• Fig. B1: Why is the GP2 line in panel b different from Fig. 3? They should be the same. If I understand it correctly, this figure is the same as Fig. 3 in the main text. If not, please explain what is the difference.
• Fig. B1: The graphs inside each panel overlap with them quite significantly which makes them difficult to read. There must be another way to plot this, e.g. by adding another figure or decreasing the size of the small graphs and placing them better in each panel.
• Fig. B1: The y-axis of the small graphs does not have a label so that it is not clear what is illustrated there. It is also not described in the figure caption.
• Fig. B1: Panel c has an x-axis different from the other panels. Please explain and add the correct label for the x axis.
• Fig. B1: The caption does not describe what is actually shown in the figure. Please add this information or at least a reference to the similar figure in the main text.
• L603 and L606: Please remove "very" as this it is not clear what makes it "very" instead of just similar / different.
• L612: same comment as above: Why 7 degrees?

Technical corrections
• L120: typo: approximately
• L144: typo Generation
• L170: remove dot between "selection" and "is"
• L175: remove dash between New and York
• L224: have --> has
• L314: sur --> such
• L360: Supp.Mat. --> Appendix
• L535: Annex --> Appendix
• L544: enabless --> enables
• L579: add a "to" between "referred" and "as": referred to as GP2

Hide

ED: Reconsider after major revisions (22 May 2024) by Andre Butz

AR by Alexandre Danjou on behalf of the Authors (08 Oct 2024) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (26 Oct 2024) by Andre Butz

AR by Alexandre Danjou on behalf of the Authors (08 Nov 2024) Author's response Manuscript

Short summary

We study the capacity of XCO₂ spaceborne imagery to estimate urban CO₂ emissions with synthetic data. We define automatic and standard methods and objective criteria for image selection. The wind variability and urban emission budget guide the emission estimation error. Images with low wind variability and high urban emissions account for 47 % of images and give a bias in the emission estimation of −7 % and a spread of 56 %. Other images give a bias of −31 % and a spread of 99 %.

Optimal selection of satellite XCO2 images for urban CO2 emission monitoring

Download

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection

Optimal selection of satellite XCO₂ images for urban CO₂ emission monitoring