This paper presents an improved (version 2) catalog of global NOx point source emissions from TROPOMI NO₂ data. Point source rates are estimated on the basis of flux divergence applied to TROPOMI gradients. The paper includes detailed discussion of errors and of improvements relative to version 1, as well as evaluation with power plant data in Germany and the US.  The dataset is a very useful compilation, representing a significant advance over version 1, and the error analysis is clearly presented. The Abstract and Conclusions summarize the paper well. The presentation is very good.  I recommend publication after consideration by the authors of the following comments.

1. Figure 2: I found the presentation of AMF, NO2/NOx, and lifetime as scaling factors not helpful at all. I’m sure that these scaling factors are of interest to the authors but they are not to the reader. I recommend showing the actual quantities in Figure 2, and the scaling factors can be given as statistics in the text.
2. Section 3.3: Inference of the NO2/NOx ratio is crude and could be a significant source of error needing to be mentioned.. J  should depend on surface reflectivity and it’s a bit embarrassing that this is not recognized since it is key to the NO2 retrieval. The ozone climatology likely does not apply to the plume where ozone would be titrated at least in the near-field (and 15 km is close to that). I see no support for the authors’ claim that PSS is a reasonable assumption on the 15-km scale. Plume chemistry can be very weird. It would be good to discuss the literature for direct measurements of the NO2/NOx ratio in plumes.
3. Section 3.8: I don’t understand the iterative nature of the procedure, and this is worth explaining better because it’s brought up in a number of places including the abstract. As I (maybe incorrectly) see it, the procedure applies successively lower thresholds to the flux divergence thresholds. It’s not really iterative except that previously identified plumes are removed from the dataset. I may not have it right, which is the point. Clarify.
4. Section 3.9.2: do we understand the chemistry behind this NOx lifetime of 2 hours at low latitudes, 4-6 hours at higher latitudes? My recollection is that the mechanism for fast NOx oxidation in power plant plumes remains a bit of a mystery considering that ozone titration would be expected. It would be good to review some of that literature. Surely the lifetime parameterization should be a function of season as well as latitude.
5. Equation (10): I don’t understand the double integral. Shouldn’t it be a contour integral?
6. Section 3.11.1: I don’t understand the ‘statistical error’ terminology in that section. Are you referring instead to variability, such as standard deviation?
7. Section 3.11.3: a plume height of 300 m vs. 500 m is not enough to characterize the uncertainty in the AMF. In early afternoon when TROPOMI observes, vertical mixing up to the PBL depth (typically 2 km) can take place within 1 hour. That may give you an AMF error of more like 10%.
8. Line 421: Appendix ??
9. Table 1: the last column is key to understanding the factor of 3-4 increase relative to version v1, so I would give it a more helpful title and I would identify the principal contributors to the factor of 3-4 increase in the text.

General comments:

This study provides an updated version of global NOx point source emission inventory using TROPOMI NO2 data. The key idea of the approach is calculating NOx advection to detect point source, and integrating it in a defined area to obtain emissions. Compared with the previous versions, the main improvement includes AMF correction, new spatial derivative calculation, topographic correction, new spatial integration of emission estimate, and lifetime correction. The manuscript is well written.  I have a few specific comments.

Specific comments:

1. Line 113-114: What is the criteria used for removing cloudy pixels?
2. Section 3.2: Could you show more details of AMF correction? This study assume a plume at 500 m above ground; so what is the shape of final NO2 profile used for calculate new AMF?
3. Line 120: What is “ab initio”?
4. Line 421: “IN Appendix ??”. Please correct.
5. Line 616: “: American Meteorological ……”.Delete “:”.

Nice job with this new version of the algorithm. All comments are minor.

One overarching comment is in the lack of literature on the realistic nature of both the effective plume heights and the wind speed used. I am glad you performed a sensitivity analysis in Section 3.11 showing the effects of wind speeds on emission estimates, but I do think some literature supporting this analysis is warranted. I am not requesting additional analyses, but instead a longer discussion, including additional literature, on both any potential biases in the ERA5 wind speed/direction and how realistic a 500-m effective plume height is. For example is there a global bias or any regional variation in a wind bias? How is performance of the wind product near coastlines? Are there any references - using in situ aircraft observations - demonstrating that a 500-m plume height is approximately correct mean value?

Other minor comments:

Line 39. “far higher and thus more realistic” —> “larger and more realistic”. The word “higher” could refer to height in the atmosphere.

Line 63. Mention that the 500-m wind is used, and that a sensitivity analysis is performed with the 300-m wind.

Line 74. I see in Line 154 that you provide a range of NOx/NO2 ratios that were used range of NOx/NO2 ratios that were used, but maybe include here also? And quickly discuss (in either section) where ratios may be higher and lower?

Line 86. Remove “Basically”

Line 105. In the US, the CAMPD is used most often (https://campd.epa.gov/). I am assuming eGRID and CAMPD are identical. Can you confirm? And if so, can you add a sentence in the text mentioning this?

Line 121. Are there any references - using in situ aircraft observations - demonstrating that a 500-m plume height is approximately correct mean value? It’d be great for you to check out the literature and see if you can find anything.

Line 138. This is great, thank you for adding! But can you add a bit more detail? I’m not fully following how the new AMF is calculated. This section would benefit greatly from an illustrative figure showing a standard a priori profile vs an a priori profile with the “excess” at 500-m.

Line 178. When you say “this procedure” are you referring to de Foy and Schauer? I think yes, but can you make this clearer?

Line 220. Local maxima during the May 2018 - Nov 2021 average? Or daily average? I can imagine the latter is much harder.

Line 255. Define “area” source. Does this include mobile/vehicle emissions? Or something else?

Line 291. The NO2 lifetime is also function of the atmospheric composition (both total NO2 and VOCs) as well. I could imagine two locations at the same latitude and with similar wind speeds having differing NO2 lifetimes based on ambient atmospheric composition (Figure 1; https://www.science.org/doi/10.1126/science.aax6832). It seems like you are not accounting for this in your lifetime derivation. Is that a correct assumption? If so, please state explicitly, and discuss in here or in the Discussion section that this would be an opportunity for further improvement of the methodology.

Line 417. In the Supplement, I noticed that some of the power plants (the ones in Canada and the US at least) didn’t match exactly. Instead of saying “First match only”, perhaps say “Potential match”

Line 421. I think you meant to say in the “Supplemental Material”

Line 425. Tab 2? I think you meant to say “ For five of the point sources listed in Table 2…”.

Line 458. Seems like more than 41 points are in Figure 13. I could be wrong though. Please double check