Dear authors,

The revised draft mostly addresses the concerns of the referees. The paper can be published after attention to the following comments.

Both referees requested more information on implications of the different model chains for the two cities. The authors have instead deemphasized this aspect of the analysis, stating a model comparison is beyond the work’s scope. While this is fine, I encourage the authors to state this directly in Sect 2.1, because, by describing the model studies for the two cities in the same paper, the models are being compared de facto. Secondly, both referees asked for a discussion of the uncertainties that result from the decision to use different model chains. The revised text only says: “We should note that the range of uncertainty in the results presented here is probably underestimated due to the choice of a single model chain for each case study.” Please elaborate. It is not obvious that or why this would amount to an underestimate in the uncertainty.

Referee 1: “Page 27059, Section 5.5: These are interesting results, but I feel like one paragraph isn’t sufficient to describe what’s going on. This could be expanded.” (This is the paragraph on photochemical regime indicator ratios.) This comment was never addressed in the author reply. Additionally, regarding the use of MD8hr averaging of indicator species, please clarify whether these are averaged over the same time window as MD8hr O3, or if each indicator species has a unique MD8hr time window. In a city, I wonder if the MD8hr NOy is driven by high urban nighttime NOx, if so, is this the right parameter to use? I recommend adding your reply to Referee 1 to the text, “the analysis was also performed implementing the daily averages but results remained the same.”

Referee 2: “Figure 3. To improve robustness of model evaluation, these plots would benefit from the addition of error bars that show the variability of the mean for the model and observations. Also, correlation coefficients of the model against observations should be reported somewhere.” It is not clear to me from your reply why error bars and correlation coefficients for period averages cannot be provided. Please add a little text to this effect.

Additional Editor’s comments:

Page 3, line 24-27: It is first stated that O3 and PM in Paris are largely affected by local emissions, while in Stockholm effects are mainly regional. This idea is referenced throughout the manuscript and used to interpret model results. However, evidence for this distinction is neither cited nor provided. Please include a reference and a brief summary of the justification.

Define “local scales” in the introduction.

Add a subtitle number to the heading ‘Urban-scale emission projections.’

Best,
Sally Pusede