Please find the full comment in the Supplement pdf file.

Tu et al. present an analysis of TROPOMI methane observations over the coal-rich Shanxi province of China. They use their wind-assigned anomaly method to quantify regional methane emissions for three clusters of Shanxi coal mines. They compare their estimates with three bottom-up emission inventories, EDGAR v7.0, CAMS-GLOB-ANT, and a measurement-based coal mine methane inventory by Qin et al. (2023). They find good agreement with the Qin et al. estimates but much lower emissions (~factor of 2-3) than reported in the EDGAR and CAMS inventories.

The paper is interesting and a good fit for ACP, but in my view substantial changes are needed before it can be published. I see two major weaknesses. First, the methods need to be explained in much more detail, not merely by pointing to previous publications. I found it very difficult to follow the discussion of results because the paper does not adequately explain the wind-assigned anomaly method and its interpretation. Second, the uncertainty analysis appears to be incomplete. The authors report uncertainties <4% (<2% in two of three cases) on their regional methane emission estimates inferred from TROPOMI. These values are unrealistically low; regional emission errors reported elsewhere in the literature are routinely in the range ~20%-30%. I am therefore left with the impression that the authors have overlooked important sources of error, for example having to do with background subtraction and wind speed.

Specific comments

• L. 18-19: The reported uncertainties (<4%) are unrealistically small. Uncertainty in regional emissions derived from TROPOMI tend to be in the 20%-30% range (or more). There must be other, larger sources of error besides what is reported.
• Shen et al. (2022) used TROPOMI to estimate methane emissions for ~20 US oil and gas basins and reported mean errors of 30% based on an elaborate uncertainty analysis (https://acp.copernicus.org/articles/22/11203/2022/). Error bars for emissions from individual countries estimated by Shen et al. (2023) are of similar magnitude (https://www.nature.com/articles/s41467-023-40671-6). TROPOMI analyses by Cusworth et al. (2022), Chen et al. (2023), and many others found similar results.
• L. 20: Which bottom-up inventory?
• L. 21-22: That may be, but it’s not entirely clear given the unrealistically small uncertainties reported for the TROPOMI emission estimates.
• L. 23: How do the estimates help to develop climate mitigation strategies?
• L. 31: Would it not be more accurate to say that China is “the leading emitter”, rather than just “one of” them?
• L. 33: China did not sign the 2021 Global Methane Pledge, so for clarity it would be best to use another word besides “pledge” here.
• L. 33-35: It would be useful to include a reference for the Glasgow Agreement. Perhaps something like this 2021 US State Department press release: https://www.state.gov/u-s-china-joint-glasgow-declaration-on-enhancing-climate-action-in-the-2020s/
• L. 68: It’s unclear what “solar radiation […] radiated from the Earth” means.
• L. 88: What wind speed is used? The speed at 10-m? 50-m? Something else?
• L. 95: Regions of China or of Shanxi?
• Figure 1: Suggest increasing font size for legibility.
• L. 104-105: What are those estimates by Qin et al. (2023) based on? A brief description of the dataset would be valuable.
• L. 105-106: I do not understand the sentence beginning “Near 30 small coal mines…”
• L. 110: Qin et al. (2023) used eddy covariance measurements to construct their facility-scale inventory. Would it be appropriate to describe their work as a measurement-based inventory of coal mine emissions?
• Subsection 3.2: Suggest moving this subsection to section 2 (data and methods), including description of the Qin et al. (2023) dataset.
• Figure 4 (left): A log y-scale would be helpful here.
• Section 3.3 and Fig. 5: Significantly more explanation is needed on how the wind-assigned anomalies are calculated and how the wind direction segmentation is performed and what these things mean. The methods section on the wind assigned anomaly method only describes the cone plume model. Section 3.3 is very difficult to follow, and readers shouldn’t need to read the authors’ previous papers to understand what is going on; the paper should be readable on its own. It’s unclear to me what the middle panel of Fig. 5 is showing. How are the estimated emissions distributed between the different coal mines within a cluster? How are the emissions calculated from the TROPOMI wind-assigned anomalies? Are all the mines scaled up/down together following the spatial distribution of the underlying inventory?
• How is the TROPOMI methane background subtracted? Background subtraction tends to be a major source of error in regional emission estimation.
• L. 231: What “calculated average” value is being reported here? It’s unclear what the ppb values refer to.
• The uncertainty analysis varying the plume model, wind product, and inventory is a good start, but not sufficient. What is the sensitivity to background subtraction scheme? What about wind speed levels (e.g., using the 10 m wind rather than 100 m)?
• How are the current error values calculated? Do they represent 1-sigma errors? Reporting the uncertainty as the range of estimates from a broader estimation ensemble might be clearer.
• L. 246-250: These sentences are contradictory. Should the first sentence only refer to the “bottom-up” inventory (which, again, does not seem to me to be a “bottom-up” inventory – rather a measurement-based inventory).

Typos

• 14: “process” → “progress” ?
• 27: “emission” → “emissions”
• 51: “emissions” → “emission estimates”
• 53: “from satellite” → “from satellites”
• 70: “unprecedented high spatial resolution…” → “unprecedented combination of high spatial resolution…”
• 6: “bottum” → “bottom”
• 243: “achived” → “achieved” ?
• 252: “boarded” → “bordered” ?

References

Chen, Z., Jacob, D. J., Gautam, R., Omara, M., Stavins, R. N., Stowe, R. C., Nesser, H., Sulprizio, M. P., Lorente, A., Varon, D. J., Lu, X., Shen, L., Qu, Z., Pendergrass, D. C., and Hancock, S.: Satellite quantification of methane emissions and oil–gas methane intensities from individual countries in the Middle East and North Africa: implications for climate action, Atmos. Chem. Phys., 23, 5945–5967, https://doi.org/10.5194/acp-23-5945-2023, 2023.

Cusworth, D. H., Thorpe, A. K., Ayasse, A. K., Stepp, D., Heckler, J., Asner, G. P., Miller, C. E., Chapman, J. W., Eastwood, M. L., Green, R. O., Hmiel, B., Lyon, D., and Duren, R. M.: Strong methane point sources contribute a disproportionate fraction of total emissions across multiple basins in the U.S., Earth ArXiv, https://doi.org/10.31223/X53P88, 2022. 

Shen, L., Gautam, R., Omara, M., Zavala-Araiza, D., Maasakkers, J. D., Scarpelli, T. R., Lorente, A., Lyon, D., Sheng, J., Varon, D. J., Nesser, H., Qu, Z., Lu, X., Sulprizio, M. P., Hamburg, S. P., and Jacob, D. J.: Satellite quantification of oil and natural gas methane emissions in the US and Canada including contributions from individual basins, Atmos. Chem. Phys., 22, 11203–11215, https://doi.org/10.5194/acp-22-11203-2022, 2022.

Shen, L., Jacob, D.J., Gautam, R. et al. National quantifications of methane emissions from fuel exploitation using high resolution inversions of satellite observations. Nat Commun 14, 4948 (2023). https://doi.org/10.1038/s41467-023-40671-6