Pope et al. use a combination of satellite measurements, surface observation, and modeling to investigate the severity and the drivers of ozone pollution during the heatwave over Europe in 2018. They find that meteorological processes were mostly responsible for the increased surface ozone, with emissions changes playing a smaller role. Increases in ozone at 500 hPa was associated with increased ozone transport from the stratosphere and the troposphere above the North Atlantic.

Heatwaves are getting more frequent and severe with climate change and it is important to understand its effect on air quality. This paper addresses this issue but has significant shortcomings, as described below.

General comments:

1. The study uses "novel" satellite observations of tropospheric ozone from the GOME-2 and IASI instruments, but the manuscript has few details about the retrievals. While the GOME-2 retrieval have been described in Miles et al., 2015, a few more details about it in the manuscript would be useful. I could not find details of the IASI ozone retrievals in the cited manuscripts (Pope et al., 2021, Palmer et al., 2022). It is important to discuss the vertical sensitivity of the retrievals (averaging kernel profiles), its the vertical resolution in the troposphere, how the retrieval compares to independent (ozonesonde) observations, etc. 

2. The model used for investigating the ozone pollution episodes in 2018 does not seem well-suited for this application. In particular, it fails to capture the observed increase in ozone in 2018 compared to 2017, suggesting that the sensitivity of the model to emission or meteorological charges is too low. Also, a 2.8°x2.8° horizontal resolution seems too coarse for regional modeling.

3. The back trajectory analysis to determine changes in ozone transport is confusing and unconvincing. The back trajectories are weighted by the modeled ozone to determine the amount of advected ozone. However, the frequency with which air arriving at a receptor site passes a particular location does not seem to be accounted for. Also, the choice of Paris and Berlin as receptor sites seems arbitrary. How does the back trajectory model handle convective transport from the boundary layer to the free troposphere? I wonder if it would be much easier just to derive the net influx of ozone into Europe from the 3D chemical transport model using the ozone concentrations and the wind fields.

4. Could the authors include in a discussion of what the results of this work imply for our understanding of ozone air pollution during heat waves? The results, as presented currently, seem to apply only to the particular episode in the summer of 2018, which limits the scientific significance of the work. Do any of the results apply more broadly? 

Specific comments:

1. The introduction lacks a discussion of previous work on this topic. 

2. Line 100: "cost function less than 200" - This needs some context. 

3. Line 104:  How were the 83 sites selected out of the hundreds of sites in the EMEP network?

4. Line 129: Please clarify that the fixed emission run includes fixed biogenic VOC emissions.

5. How do biogenic and lightning NOx emissions differ between the two years?

6. Line 211: Please verify whether GOME-2 actually has peak sensitivity in the lower troposphere. Miles et al. 2015, which describes the GOME-2 retrieval, states that it peaks at 500 hPa. It would be useful to include the averaging kernel profiles for GOME-2 and IASI retrievals in the supplement.

7. Line 242: Do you see any spatial consistency between the charges observed at the surface and from the satellite data?

8. Line 248: Again, I do not think that the tendency of the model to underestimate ozone charges between 2018 and 2017 can be brushed aside.

9. Section 3.4 can be considerably shortened. Much of text just repeats what is shown in the figures. Better to emphasize the main takeaways instead.

10. Line 455: The influence of stratospheric ozone intrusions is shown to be higher in 2018 than in 2017. Is this a coincidence or are more stratospheric intrusions linked to meteorological processes associated with heat waves? Also, this does not seem to be supported by the back trajectory analysis, which do not show any trajectories originating from the stratosphere.

11. Figure 5: What do the dotted lines show?

12. Figure 5: Wouldn't MDA8 ozone be a better metric for this?

References:

Miles, G.M., Siddans, R., Kerridge, B. J., Latter, B. G., and Richards, N. A. D.: Tropospheric ozone and ozone profiles retrieved from GOME‐2 and their validation, Atmospheric Measurement Techniques, 8, 385–398, doi:10.5194/amt-8-385-2015, 2015.

Palmer, P., I., Marvin, M., R., Siddans, R., et al.: Nocturnal survival of isoprene linked to formation of upper tropospheric organic aerosol, Science, 375 (6580), 562-566, doi:10.1126/science.abg4506, 2022. 

Pope, R. J., Kerridge, B. J., Siddans, R., et al.: Large enhancements in southern hemisphere satellite-observed trace gases due to the 2019/2020 Australian wildfires, Journal of Geophysical Research: Atmospheres, 1–13, doi:10.1029/2021jd034892, 2021.  

Review of “Investigation of the summer 2018 European ozone air pollution episodes using novel satellite data and modelling”

This study explores the reasons for the observed ozone enhancement in Europe during the 2018 heat wave compared to the previous year. Satellite data, surface measurements and model simulations were used to investigate summer time ozone during the European heat wave of 2018. Based on model simulations, meteorology was found to play a large role in the ozone enhancement in 2018.

Major comments

The model horizontal resolution is fairly coarse and it's likely to average out any large regional emissions of ozone precursors. As well as the model representation of ozone precursors emissions, the coarse model resolution is likely to impact on ozone formation and ozone deposition; these impacts will be largest near the surface. Can the authors justify the choice of model resolution for this study, particularly when looking at surface ozone and comparing to EMEP surface station data? Similarly, is the vertical resolution in this model appropriate for this study? The authors should give an idea of the vertical resolution, e.g. number of levels in the boundary layer and mid latitude free troposphere, in section 2.1.

The authors should consider differences in biomass burning emissions of ozone precursors between 2018 and 2017.

A lot of results, including model evaluation and comparison with observations, are currently in the supplementary material. I believe a lot of the analysis and discussion in supplementary material should be transferred to the main body of the paper.

Figure S7 shows that TOMCAT fails to reproduce surface ozone’s seasonal cycle compared to observations and also the extent of the ozone enhancement in 2018 compared to 2017, possibly due to the model’s coarse resolution. This is also supported by comparison of figure S8-S11. Are the authors justified in using TOMCAT to disentangle different processes leading to the ozone enhancement in 2018?

Minor Comments

The introduction should be extended to include a broader discussion on the many factors affecting Summer time European ozone, including intercontinental transport and the impact of interannual variability in biomass burning and BVOC emissions, see for example work by Honrath et al. 2004 (Honrath et al., Regional and hemispheric impacts of anthropogenic and biomass burning emissions on summertime CO and O3 in the North Atlantic lower free troposphere, J. Geophys. Res., 109, D24310, doi:10.1029/2004JD005147)

p3, l 103: the authors should say a bit more about the EMEP measurements and add appropriate references.

p3, l 113-114; “year-specific anthropogenic emissions from MacCity”. Historic emissions for this dataset end in 2010 but the model simulations are for 2017 and 2018. Could the authors explain how these emissions were extended to cover 2017-2018? If scenarios emissions are used, what pathway do they follow, and can we say that the emissions are year-specific if they are using a scenario?

p4, l 135: can the author explain more about the O3 tracer. Does it have a fixed lifetime, does it decay once in the troposphere? Can they cite previous work using this tracer with TOMCAT or other models?

Figure 1: the caption in Fig 1 contains the sentence “…retrieved from MetOp-A IASI, MHS and AMSU by the IMS scheme.” This should be expanded, appropriate references added and moved to section 2.1. ‘LHS’ and ‘RHS’: please replace with ‘left’ and ‘right’ (same for Fig 4, 6, 7, 8) or explain somewhere what the acronym means.

Figure 2: the green polygons make it hard to understand what’s significant (it’s not clear if it’s inside/outside the contour lines); please use shading, hatching or crossing to make this figure clearer.

Figure S7 d: why is data not shown for Oct-Dec?

Minor corrections

p3, l 88: ozone is repeated twice

p3, l 105: years is repeated twice

p4, l 138: Pope 2020 should be Pope et al. 2020

p4, l 141: verses should be versus

p8, l 337: change “centre panels” to “middle column”

p11, l455: “Intrusion of stratospheric O3 into the mid-troposphere has a moderate influence on the observed/modelled O3 enhancements with contributions of up to 15.0-40.0%”. A contribution of 15-40% is a large contribution, not moderate.