Review on Satellites reveal a 28% drop in Ukraine’s Nitrogen oxides emissions during the Russia-Ukraine war in 2022

by Yu Mao, Weimin Ju, Hengmao Wang , Liangyun Liu, Haikun Wang, Shuzhuang Feng, Mengwei Jia, and Fei Jiang

The manuscripts is a detailed application description of the method introduced in a previous publication Mao et al 2024 (Mao, Y., Wang, H., Jiang, F., Feng, S., Jia, M., and Ju, W.: Anthropogenic NOx emissions of China, the U.S. and Europe from 2019 to 2022 inferred from TROPOMI observations, Environ. Res. Lett., https://doi.org/10.1088/1748-9326/ad3cf9, 2024).

It shows in detail where the war in the Ukraine destroyed most of the ecomic and sozial (and human) live. The results are based on the comparison of modeled columns based on prior emissions, which are than adapted to satellite observations. The study focuses on the years 2019 (a pre-covid baseline) and the year 2022. The pandemic is included as small side note. The impact of the pandemic Ukraine’s NO_x emissions was negligible for compared to the Russian invasion.

Mayor comments

According to (https://www.temis.nl/airpollution/no2col/tropomi_no2_data_versions.php, March 2025) version 2.4 or higher of the TROPOMI NO₂ data is reommended. There was a major version change in July 2022, I am not sure the version 2.3.1 (page 4 line 124) is appropriate for the presented study focusing on 2022. I recommend to double check the results using latest version of TROPOMI NO₂ data. In this context I ask for your apologies that I did not realize in the pre-review.

Minor comments

The war has continued for 3 too long years. How did the emissions change in the second year?

page 5 line 138: The inversion is performed on a Monthly basis (due to the observation gaps of the satellite). Please elaborate in more detail what this means. Does this include that the wind fields are averaged over a month. Or where the daily observations for TROPOMI and the meteorological data used and the emission assumed to be  constant for the one month.

P6/7 L188: The assumption if the dominant sector is really constant over time is not justified – Also the authors themselves state that it might not be correct. However I am afraid that large parts of the conclusions are based on this assumptions. Or is there any other way to distribute the NOx emissions among the different sources?

P13 L339: The prior emissions are stated to be overestimated by 80%. This I s quite large has this been confirmed by similar studies?

Figure 2: The Crimea peninsular in the South of Ukraine has been occupied since 2014. What causes the NOx reduction there?

Technical comments

p3 l 93: the sentence beginning with Meteorological data can be split into two, remove the word “while”.

p5 l141: “global” instead of “g lobal”

In this manuscript Mao et al. evaluate the change in nitrogen oxide emissions in Ukraine due to the Russia-Ukraine war in 2022 using inversion methods. Overall, the discussed merit is interesting an deserves investigation. Before I can recommend publication in ACP, the authors need to address some fundamental aspects. 

Major comments:

1. I would suggest to rethink your title. Your methodology relies on inversion techniques and uses TROPOMI data as an input. The phrase "Satellites reveal" is thus misleading. In addition the 28% is associated with uncertainties and I would thus remove it from the title.

2. I would expect a significant contribution of direct and indirect war related NOx emission (e.g., infrastructure fires). In your study, however, you seem to not include any war related emissions. Keeping in mind the fast changing nature of this war, what uncertainties does this introduce to your methodology and your results? To what degree do you think did war related emissions compensate the reductions reported in your study?

3. Please provide further details on the TROPOMI retrievals. The VCDs are obtained from a polar orbit meaning that the same time of day is observed. How does this impact your methodology, especially considering a shift in activities to the night?

4. To improve data quality, you perform the inversion of the anthropogenic NOx emissions on a monthly scale. Since you frequently highlight the fast changing nature of the war, it sounds that using monthly averaged data is not an optimal choice. To what degree does this influence the predicted results?

5. I was surprised to see such a coarse model resolution being used when focusing on such a small region. At the same time, TROPOMI provides data at a km scale. How can you justify such a coarse resolution knowing that other inverse modelling infrastructures provide resolutions at the km scale? How does this affect your results?

Minor comments:

- Line 13: Replace "economic production" with e.g. "society".

- Line 18: Your abstract only mentions decreases in NOx emissions, even though you document increased emissions in urban areas in West Ukraine. This can be misleading and should be mentioned in the abstract.

- The introduction would greatly benefit from a figure which shows yearly average NOx VCD from TROPOPMI for the base year as well as 2022.

- Line 196: Please elaborate on the factors 1.2 and 0.7.

- Line 244: How do you account for the population migration in your emission datasets?

- Line 271: What "drivers" are you referring to? How does military activity related to transport compensate these changes?

- Line 417: Please elaborate on what policies you are referring to.

- Fig. 1 and 2: Please fix the inconsistencies in the x-axis labels.