This work estimates the global methane emission for the year 2019 at 2^ox2.5^o spatial resolution based on inverse modeling technique by utilizing the TROPOMI and GOSAT satellite observations. It validates TROPOMI and GOSAT observation against TCCON methane column measurements, using the GEOS-Chem CTM and also provides the sensitivities parameters such as averaging kernel and degrees of freedom for signal which quantify the number of independent pieces of information on the distribution of methane emissions. The paper provides the discussion on estimated methane emission at different major methane source regions around the world using GOSAT, TROPOMI, and joint data-set. I recommend its acceptance after the minor revision with following specific/minor comments:

specific comments:

1). There are techniques such a 4-dimentional variational data assimilation (4D-var) and local ensemble Kalman filter (LETKF) that also provides grid-based flux estimation of methane by assimilating satellite observations. How close the inversion technique used in this study is to those techniques? I suggest author to add a paragraph that discuss the limitations of the high-resolution Bayesian inversion technique using satellite observations. 

2). The methane emission has been estimated using annual mean methane concentration data of GOSAT and TROPOMI. How would the seasonal variability of methane concentration affect such emission estimate? Is it possible to extend such inverse modeling set-up to estimate the methane emission at monthly scale?

Minor comments:

Line 367, We conducted ……. grid cell.

How significant it is to apply equal ratios to all sectors in the grid cell. How better this is in comparison to isotopic fractionation method.

Line 371, Study claimed that 2ox2.5o grid makes sectoral attribution more accurate, but we don’t know true sectorial contribution.

Line 385, In China,……………..Plain. Please cite the figure number here.

Line 399, the analysis is performed for 2019, why did the author cite 2014 report.

In Figure 3, large difference between GOSAT and TROPOMI can be seen during DJF at northward of 30oN. How does it affect inversion estimate?

Figure 7, Over India the estimate looks pretty close from all the methods, over Brazil, the joint inversion estimate is close to GOSAT, almost same story for Europe, but over CONUS the joint inversion is very high compared to both the inversion. How would you explain the joint inversion behavior over CONUS?

The authors performed inversion analyses of atmospheric CH₄ using column average data from TROPOMI and GOSAT. Since the TROPOMI product is at the initial stage, the retrieved CH₄ data may have substantial uncertainties. Considering those uncertainties, this study elucidated characteristics of the TROPOMI data specifically for inverse results by comparing with the GOSAT data, which are at a mature stage.  The inversions and related statistical calculations they performed revealed significant discrepancies between TROPOMI and GOSAT in terms of error characteristics and information content, and much better quality of the GOSAT data at this moment. Detail information of the TROPOMI CH₄ data obtained in this study is surely useful for improving the data quality of the TROPOMI data in a near future. The manuscript is well structured and most descriptions are clear to me. I think this paper can be published for Atmospheric Chemistry and Physics after minor revisions suggested below.

L43: “Climate change action” is better than “Climate action”, isn’t it?

L45: Here, “GOSAT” first appears in the main text. Therefore, it should be written as “Greenhouse Gases Observing Satellite (GOSAT)”

L53-54: Like “TROPOMI (GOSAT) observes light intensity at 2305-2385 (1630-1700) nm wavelength”

L64: “inversion of a chemical transport model” => “inversion with a chemical transport model”

L89-90: “one year of data” => “one year data”

L105: TCCON first appears here. It should be “Total Carbon Column Observing Network (TCCON)”. Furthermore, it is better to make a brief explanation of TCCON so that people who are not familiar in this field can understand the purpose of the use of TCCON.

L141: Why was the 4x5 degree used only here (Fig. 3)? The other analyses seem to be done with the 2x2.5 degree.

L201-202: I think this kind of pulse calculations assumes that the model is linear. Does the GEOS-Chem satisfies the model linearity?

L203: Only one-year-long inversion fluxes would have some errors attributed from the initial mole fraction field, which was optimized just by the globally uniform factor, especially for the earlier period. I think some discussion about that error is needed.

L256: How is the “35” derived?

L266: The term “averaging kernel matrix” is familiar in the satellite retrieval field, but not the case for flux inversions. Although, it is theoretically correct, it might be better to put some note to avoid confusion.

L292: The ratios of the posterior/prior non-wetland emissions in Fig. 5 show values close-to-zero or over 2 in many places (e.g., over Europe and Africa in the joint case). Does this mean that the inversion zeroed/doubled non-wetland emissions? If that is the case, are the resulted emissions reasonable?

L314: It is hard to see the effect of TROPOMI. It would be better to show sensitivity differences from those of the GOSAT inversion case in the right bottom panel of Fig. 5.

L324: Is this the same for GOSAT?

L347: Why does the GOSAT inversion show negative biases against the GOSAT observations almost everywhere? Is that contributed by the strong prior constraint to wetland emissions?

L371-372: Can you estimate magnitude of the potential errors due to the strong constraint on wetland emissions?

L386: Specify the version of EDGAR.

L392: Not clear what is “inconsistent” with the TROPOMI inversion.

L402: What is the “base” inversion?

L424: The number of 34 Tg a^-1 seems not consistent with Fig. 7, which looks like around 40 Tg a^-1. Furthermore, I am not sure why the joint inversion increased the emissions estimate, as each inversion showed downward estimates from the prior estimate.