This manuscript explores the sensitivity of ozone production in a future climate across three possible emissions scenarios.  The paper is extremely well written, the figures and tables are clear and self-explanatory and the conclusions are sound and supported by the evidence.  I have a few minor comments, mainly focused on the need to cite previous work, and to present the findings in the context of the new conclusions by IPCC AR6.

The paper would benefit from some discussion that places these results in the broad context of the recent findings of IPCC AR6.  I realize that AR6 was not published when this paper was submitted, so AR6 could not be referenced. But now that AR6 is publicly available, a comparison is warranted, especially in terms of Chapter 6, Section 6.5.1, “Effect of climate change on ozone”.  The broad message from this new analysis is that ozone in the mid-21^st century will be lower across the USA than it is today, under all three scenarios.  Presumably, the primary cause is the decrease in regional US emissions of NOx. However, part of the explanation could also be due to a shorter ozone lifetime in a warmer, more humid future.  How much of this decrease is due to emissions changes and how much is due to climate change?  Along these same lines, what is the impact of future heat waves, which are expected to be more intense in the future?  Stronger heat waves will lead to episodic surface ozone pollution events.  Will these future pollution episodes be more intense than present day events?  Is the average decline in ozone masking a few extreme high ozone events in the future?

In the Introduction (lines 36-41) several of the papers cited in terms of describing recent ozone trends are out of date.  For example, the data analysis in Lefohn et al. (2008) stops in 2005 so the paper does not report the strong decreases of ozone in the eastern USA that occurred after 2004. The paper by Akimoto et al. (2003) does not even report observations from the 21^st Century. Ohara et al., 2007 is also out of date.  Current papers are:

Simon, H, Reff, A, Wells, B, Xing, J and Frank, N (2015), Ozone Trends Across the United States over a Period of Decreasing NOx and VOC Emissions. Environ. Sci. Technol 49: 186–195. DOI: https://doi.

org/10.1021/es504514z

Strode, S. A., J. M. Rodriguez, J. A. Logan, O. R. Cooper, J. C. Witte, L. N. Lamsal, M. Damon, B. Van Aartsen, S. D. Steenrod, and S. E. Strahan (2015), Trends and variability in surface ozone over the United States, J. Geophys. Res. Atmos., 120, 9020–9042, doi:10.1002/2014JD022784

Lu, X., Zhang, L., Wang, X., Gao, M., Li, K., Zhang, Y., Yue, X. and Zhang, Y., 2020. Rapid increases in warm-season surface ozone and resulting health impact in China since 2013. Environmental Science & Technology Letters, 7(4), pp.240-247.

Chang, K-L, I. Petropavlovskikh, O. R. Cooper, M. G. Schultz and T. Wang (2017), Regional trend analysis of surface ozone observations from monitoring networks in eastern North America, Europe and East Asia, Elem Sci Anth., 5:50, DOI: http://doi.org/10.1525/elementa.243

Tarasick, D. W., I. E. Galbally, O. R. Cooper, M. G. Schultz, G. Ancellet,  T. Leblanc, T. J. Wallington, J. Ziemke, X. Liu, M. Steinbacher, J. Staehelin, C. Vigouroux, J. W. Hannigan, O. García, G. Foret, P. Zanis, E. Weatherhead, I. Petropavlovskikh, H. Worden, M. Osman, J. Liu, K.-L. Chang, A. Gaudel, M. Lin, M. Granados-Muñoz, A. M. Thompson, S. J. Oltmans, J. Cuesta, G. Dufour, V. Thouret, B. Hassler, T. Trickl and J. L. Neu (2019), Tropospheric Ozone Assessment Report: Tropospheric ozone from 1877 to 2016, observed levels, trends and uncertainties. Elem Sci Anth, 7(1), DOI: http://doi.org/10.1525/elementa.376

Gaudel, A., O. R. Cooper, K.-L. Chang, I. Bourgeois, J. R. Ziemke, S. A. Strode, L. D. Oman, P. Sellitto, P. Nédélec, R. Blot, V. Thouret, C. Granier (2020), Aircraft observations since the 1990s reveal increases of tropospheric ozone at multiple locations across the Northern Hemisphere. Sci. Adv. 6, eaba8272, DOI: 10.1126/sciadv.aba8272

Table 2 indicates that methane will be about 1364 ppb in 2050 under the SSP3-7.0-lowCH4, which is far lower than the present-day value of about 1890 ppb (https://gml.noaa.gov/ccgg/trends_ch4/). This would require major reductions in CH4.  To provide context for the reader, can you let us know when methane was last at such a low level in the atmosphere?  I’m guessing that it would be sometime around the 1960s based on Figures 2.4 and 2.5 of IPCC AR6.  According to Figure 2.5 of IPCC AR6, methane was approximately 1500 ppb in the 1970s, based on in situ observations.  According to Figure 2.4 of IPCC AR6, methane was about 1000 ppb in the early 20^th century (perhaps around 1940?).

It would help to briefly mention the significance of the term “regional rivalry” when describing the future scenarios.  This term is mentioned twice in the paper, so it must have some importance, but I really have no idea what it means.

When using the TOAR data products, the following data link should be cited, in addition to the peer-reviewed publication by Schultz et al. (2017):

Schultz, M. G, et al. (2017): Tropospheric Ozone Assessment Report, links to Global surface ozone datasets. PANGAEA, https://doi.org/10.1594/PANGAEA.876108

Line 321

globe should be global

Figure 6 and elsewhere

When reporting trace gas values in units of ppb, one cannot use the term concentration, which is mass per volume. The expression mixing ratio must be used.

Liu et al. examine the changes in tropospheric ozone between present day conditions and several future scenarios in model runs from UKESM1 as contributed to the AerChemMIP model intercomparison project. They show that the future tropospheric ozone burden and surface ozone mixing ratio are sensitive to the changes in the emissions of short-lived ozone precursors (NOx and NMVOC) and to changes in the assumed surface mixing ratio of methane. In the high-emissions SSP3-7.0 scenario the tropospheric ozone burden and the surface mixing ratio of ozone increases, while these decrease in variants of this scenario in which the emissions of ozone precursors or the methane mixing ratio decrease.

Some high-emission regions show different trends to the trend in global surface ozone mixing ratio, which the authors relate to the modelled NOx-sensitivity of the ozone production regime in those areas, for example declining NOx emissions are associated with ozone surface mixing ratio increases in winter in all future scenarios in Europe and North America, but decreases in summer, consistent with NOx-saturated (or VOC-limited) conditions in these regions in present-day winter and NOx-limited conditions in present-day summer.

So far, these results are not particularly novel. Where Liu et al. attempt to bring some novelty to the analysis is with a determination of the modelled local ozone production regime in individual model grid cells based on the modelled ratio of NOx and VOC mixing ratios (the authors call this “concentration”, but they have clearly used molar mixing ratios in their calculations). Such an analysis is potentially interesting, as it would enable determination of the ozone production regime from a single model run rather than from a comparison of two runs. The authors do indeed do exactly this, and their results do make sense (eg. areas with high NOx emissions are generally NOx-saturated, and these can change to NOX-limited as the NOx emissions are reduced). Unfortunately, the authors do not give enough information about their method to enable a proper understanding of it. This is itself unfortunate, since the novelty of the paper depends strongly on this analysis.

Firstly, for such an important piece of analysis, the authors give no mention of any previous work that has performed it. This reviewer is not aware of any previous published attempt to quantify the ozone production regime in global model grid cells. If the authors are also similarly unaware of such work, they should say so! Alternatively, they should discuss their approach in the context of any previous work. And while there may not be a lot of literature on the simulated chemical regime in global model grid cells, there is certainly a mature literature on the general topic of ozone production regimes.

The authors define the ozone production regime based on the relative abundance of NOx and VOC, citing Sillman (1995) as a source for this. But Sillman (1995) discusses this regime based on the ratio of nitric acid and peroxides, based on the products of the NOx-dependant dominant loss pathway for radical species; when NOx is high, the dominant loss pathway is NO2 + OH -> HNO3, while when NOx is low, the dominant loss pathway is HO2 + HO2 -> H2O2 + O2. The relative abundance of HNO3 and H2O2 from radical termination reactions has in fact been used already in regional modelling studies as an indicator of the ozone production regime for the purpose of ozone source attribution (Dunker et al. 2002, Kwok et al. 2015). Did the authors consider a similar approach?

Kleinman (1994) showed that these regimes of radical loss are equivalent to the ozone production regimes and can also be characterised by the ratio of the NOx source to the radical source. The addition of VOC to a NOx-saturated system results in the production of carbonyl compounds (especially HCHO) which act as a source of radicals by their photolysis, hence the equivalence between “NOx-saturated” and “VOC-limited” chemical conditions.

So, the ozone production regime depends on more than just the local ratio of NOx and VOC. At any given ratio between NOx and VOC, the chemical system could be NOx-saturated or NOx-limited depending on several other factors: the intensity of solar radiation; the background abundance of ozone itself; and the OH reactivity of the VOC present. For example, the same ratio of NOx and VOC could lead to NOx-limited conditions in summer and NOx-saturated conditions in winter. Similarly, the same ratio of NOx and VOC could be NOx-limited if the VOC are highly reactive (eg. biogenic isoprene) but NOx-saturated if the VOC are relatively unreactive (eg. most anthropogenic VOC). By defining the ozone production regime in terms of the relative abundance of NOx and VOC, the authors miss all this complexity.

Another issue that the authors should address is the potential dependence of their ozone sensitivity metric on model resolution. The lifetime of NOx is relatively short, and the error of instantaneously diluting concentrated NOx emissions (characteristic of NOx-saturated regions) into relatively large grid cells has been well studied (eg. Wild et al., 2006).

This reviewer recognises that it may be challenging to diagnose the chemical regime in all grid cells of a global model from the relative strengths of the sources of NOx and radicals, and that perhaps the mixing ratios of HNO3 and H2O2 may not have been saved in the model output before this analysis. But the authors should at lease discuss their approach and its limitations in the context of the previous literature on the topic. A good place for this discussion would be around the top of page 14, and an even better place for it would be collected in a new Subsection 2.3.

Along with this discussion, the authors should give more information about exactly what they did. Clearly the method has some use (as shown by the analysis in Sections 5 and 6), but for others to reproduce this work, more information is required. For example, which VOC are included in the quantification of the total VOC? Is it all VOC including intermediate oxidation products? Are there criteria for selecting which VOC to include? Is the OH reactivity of each individual VOC accounted for? Is methane included in the total VOC? If not, why not? What about CO? Which grid cells were used in the determination of the threshold ratio? Just the surface? Just the boundary layer? All the tropospheric grid cells?

I hope that the authors see that a lot more information is needed for the interested reader to understand what was done, how it relates to earlier work, how it could be reproduced, and what the potential limitations might be.

General comment

The authors should avoid using the term “concentration” when they mean “mixing ratio” (multiple places throughout the text and figures).

Minor comments

Page 2, line 25: “positive radiative forcing on climate forcing” is better as “positive radiative forcing”.

Page 2, line 36: There has been a decrease in extreme ozone events, but the overall trend in ozone exposure is less clear, and recent work indicates a modest rise in ozone-related mortality (Sicard et al., 2021).

Page 2, lines 45 and 50: see the major comment above.

Page 4, line 101: Lin et al. (2020) showed that vegetation can also be a smaller sink for ozone in warmer climates due to reduced deposition. Is this effect also included in the study?

Page 9, line 160 and page 16, line 256, and in the abstract: Do increases in BVOC (mostly isoprene) really offset decreases an anthropogenic and biomass burning VOC? Butler et al., (2018) showed that biogenic VOC produce more ozone than anthropogenic VOC over these regions, so wouldn’t an equal increase in BVOC at the expense of other VOC lead to an increase in ozone?

Page 14, line 232: Please provide more information about how the “dominant” source region was identified in each part of the NOx-VOC space.

Page 16, line 258 and page 18, line 267: How can changes in methane cause large changes in ozone itself, but not in ozone sensitivity to NOx and VOC? Methane is after all another VOC. This needs more discission.

Page 19, line 283, and in the abstract: Further emission controls on VOC and CH4 are unlikely to lead to reductions in winter ozone in these regions, as local photochemical production is extremely slow during winter. Reduced NOx emissions in winter act to reduce the titration of background ozone, so once the titration effect has been removed, what remains is the background ozone.

References

Butler, T., Lupascu, A., Coates, J., and Zhu, S.: TOAST 1.0: Tropospheric Ozone Attribution of Sources with Tagging for CESM 1.2.2, Geosci. Model Dev., 11, 2825–2840, https://doi.org/10.5194/gmd-11-2825-2018, 2018.

Dunker, A., Yarwood, G., Ortmann, J., and Wilson, G.: Comparison of source apportionment and source sensitivity of ozone in a three-dimensional air quality model, Environmental Science and Technology, 36, 2953–2964, https://doi.org/10.1021/es011418f, 2002.

Kleinman, L. I.: Low and high NOx tropospheric photochemistry, J. Geophys. Res., 99, 16831–16838, 1994.

Kwok, R. H. F., Baker, K. R., Napelenok, S. L., and Tonnesen, G. S.: Photochemical grid model implementation and application of VOC, NOx, and O3 source apportionment, Geoscientific Model Development, 8, 99–114, https://doi.org/10.5194/gmd-8-99-2015, 2015.

Lin, M., Horowitz, L.W., Xie, Y. et al. Vegetation feedbacks during drought exacerbate ozone air pollution extremes in Europe. Nat. Clim. Chang. 10, 444–451 (2020). https://doi.org/10.1038/s41558-020-0743-y

Sicard, P., Agathokleous, E., De Marco, A. et al. Urban population exposure to air pollution in Europe over the last decades. Environ Sci Eur 33, 28 (2021). https://doi.org/10.1186/s12302-020-00450-2