The paper by Li et al. with the title “Tropospheric bromine monoxide in Ny-Alesund: source analysis and impacts on the atmospheric chemistry” presented a comprehensive dataset, including BrO, ozone, aerosol, and mercury observations and modelled results. Bromine explosion is an old topic, but it still has many unknowns for the atmospheric research community. The paper has good in-depth new findings of both measurements and modelling work. I would support publishing this work if the authors could address my questions and concerns listed here.

Major comments:

Line 25-26: This is a misleading claim. Contacting time does not necessarily mean contributions to BEEs. The longer contact with multi-year sea ice only suggests the possibility, but can not be used as a conclusive reason. Multi-year sea ice has lower salinity has been viewed as less possible as a major bromine source. To support this argument, more analysis is needed. E.g., can the increase in BrO signals from GOME be linked to the multi-year ice area? How can we rule out that 23.8% contact with first-year ice is not the real major contributor to the bromine? Some analysis, such as extreme cases, might help. E.g., if you can find any case that is only in contact with multi-year sea ice.

Line 57-59: since your major finding is that multi-year sea ice plays important role than first-year sea ice, more descriptions of current understandings and previous findings on these two sources should be provided. What are the differences between first and multi-year sea ice in terms of bromine chemistry? If they are both considered as sources, are they considered on a similar or different level in the BEE contributions?

Line 118-124: I understand this is a complicated research topic and the background is rich. The author did a good job of providing some background. However, I still read that the introduction part is not well organized. E.g., the author talked about the general features of studies of BEE first, then discussed the local observations from Ny-Alesund. In this part, they go back to the other field findings again. 

Line 178: In March, I would expect the site still to have snow cover. Can the author explain why surface albedo is set to only 0.1? Even in the referenced paper, this 0.1 surface albedo is not specifically discussed. But, for satellite, typically 0.9 will be used for such conditions. Comments?

Line 219-220: p-TOMCAT is driven by ERA5, while HYSPLIT is driven by GDAS. These datasets have different vertical, horizontal, and even temporal resolutions (6 hr vs. 1 hr). Any comments on the impact of different meteorological inputs? Since p-TOMCAT is a transport model, is it possible to directly analyze sea-ice contact information from p-TOMCAT simulations?

Line 263-268: Isn’t such emission flux information part of the p-TOMCAT simulation results? If yes, what does it look like? If not, what are the main challenges here? Anyway, the point is that hybridizing results from a transport model with another trajectory model seems redundant and questionable (mainly, they are driven by different meteorological fields that have very different resolutions).

Figure 5: The general n-shape curve for the number of profiles in Fig 5a and 5c is reasonable. But why Fig 5e show double peaks?

Line 368-369: Does this “sea ice contacting time” include both multi-year and first-year sea ice? Please make this clear. Also, if yes, can you break down the sea ice contact time into multi-year and first-year sea ice?

Figures 6 and 7: as the author claimed in the abstract and conclusion that one key finding is that multi-year sea ice plays a key role, I would expect the panels to have “sea ice contact time,” meaning the contact with multi-year sea ice (not just all types of sea ice). Please make this clear in the captions and the related discussion part.

Line 430-436 and 492-495: My understanding is that the BEE in p-TOMCAT is mainly driven by this SSA production mechanism. But, here it seems the author also wants to emphasize the role of multi-year sea ice. What kind of role does the multi-year sea ice play in p-TOMCAT’s bromine simulation? I am not against the hypothesis, but is it possible to quantify the emissions from SSA and multi-year sea ice? I understand this could not be easy, but currently, the description is very tangled and confusing. Or maybe the author is suggesting blowing snow on multi-year sea ice is more productive than blowing snow on one-year sea ice? The correlation between modelled BrO and contact time with different sea ice types is very low, and their difference (0.17-0.29 vs. 0.03-0.23) is very small. This is not a good support for the argument in lines 494-495.

Technical issues:

Line 140: Some information, such as the distance between the Arctic Yellow River Station and Zeppelin Station, should be provided.

The paper ‚Tropospheric bromine monoxide in Ny-Ålesund: source analysis and impacts on atmospheric chemistry’ by Li et al. provides an extended study on the relationship between enhanced BrO levels, ozone depletion events and influencing meteorological factors. The core data set consists of time series (2017 – 2023) of BrO and aerosol amounts derived from MAX-DOAS measurements. In addition, many other data sets including in situ measurements, satellite observations, trajectory analyses and model results are used to explore the conditions which favor the formation and recycling of enhanced BrO levels. One important conclusion from the study is that sea ice is the dominant bromine source. The data set is unique, and many of the conclusions are important and relevant. Nevertheless, the study lacks some scientific rigor and clear conclusions.

Major comments

The overall conclusions of the paper are rather weak and vague, especially with respect to the number and extent of the used data sets. As an example, the last sentence of the abstract states: ‚These results provide long-term observational evidence linking BrO to sea-ice and SSA processes, advancing understanding of Arctic bromine activation and its implications for springtime ozone depletion.’

In essence, this statement says nothing. The authors should be more specific about their findings.

Important points which should be addressed are:

-what can be concluded from the study about the importance of initial release of reactive bromine compounds versus recycling? What is the role of aerosols in either of both processes?

-In Spitsbergen, relatively infrequent BEE occur. How representative are the results in general and for other locations?

-what is the impact of enhanced BrO on GEM? (the GEM data are used in the study, but no further conclusions are drawn). This should be explored and also discussed in the abstract and conclusions.

Minor comments

-clouds can have a strong influence on both satellite and ground-based observations. How are clouds treated in both analyses? Are the data filtered for cloud effects? Are cloud effects corrected?

-Fig. 3: how does MAX-DOAS AOD compare to AOD from AERONET? AERONET data at Ny Alesund are available for several years. Such a comparison is needed to strengthen the confidence in the MAX-DOAS results.

Also, a correlation analysis between AERONET AOD and BrO should be performed (like in Fig. 7).

-Fig. 3: also the BrO results from GOME and the in situ results for GEM should be included

-In Fig. 4 the BrO values for different O₃ mixing ratio bins are shown. But it would also be important to see the relationship between individual observations or daily averages. How do these correlations look like for the whole time series (2017-2023)?

-Section 3.3: in 2020 the BrO values were rather high. I suggest to add another year with relatively low values to better cover the different conditions. 2019 might be well suited because of the rather high number of measurements.

-Fig. 7: the correlation between BrO (and aerosols) with the ocean and sea ice contact times should be performed separately for the first and second half of the shown period. It seems from Fig. 6 that there is a high correlation between the occurrence of ocean contact and BrO events for the second half of the time period.

Is there an explanation for the different behaviors in both periods?

Maybe add correlation plots for the vertically integrated quantities (BrO and AOD versus ocean and sea ice contact times).

Fig. S5: a similar figure for the ocean contact time would be interesting

Technical corrections:

Fig. 6: add information on the time resolution of the different data sets

Fig. S1: the y-axes should be labeled ‚BrO partial VCD’

Fig. S5: add the altitude range of the BrO VMR to the figure caption