Review of “Source contribution to ozone pollution during June 2021 in Arizona: Insights from WRF-Chem tagged O3 and CO” by Guo et al.

This is a study on the source attribution of surface and tropospheric O3 and CO in Arizona during one month (June 2021) when fire emissions impacted the area. The authors use WRF-Chem tagged simulations as well as WRF-Chem sensitivity simulations (with and without fire) to determine the budget of O3 for the region, including the fire contribution. They also investigate the meteorological, diurnal, and chemical conditions that influence O3 production for two case studies in detail. While the study covers only a small spatial and temporal region/period, this kind of detailed case study is very interesting as there remains a high level of uncertainty and variability in O3 production during fire events. My comments below are mainly related to some missing information and discussion that, if added, would make the results clearer and put into better context.  

Line by line comments:

Lines 60-64: The wording could be improved since it’s not the vehicles or industrial buildings that are combusting. Better to say fossil fuel combustion by vehicles, industry, and power plants. Also add “natural” in “as well as the natural biogenic emissions” since the start of the sentence says anthropogenic activities.

Lines 75-86: I realize this isn’t supposed to be an exhaustive list, but you could add GEOS-Chem, which has had O3 tagging options for quite some time. Described originally in Wang et al (1998) and updates in Zhang et al (2008) and used in several studies, including Whaley et al (2015), for example.

Wang, Y., and D. J. Jacob (1998), Anthropogenic forcing on tropospheric ozone and OH since preindustrial times, J. Geophys. Res., 103, 31,123–31,135, doi:10.1029/1998JD100004.

Zhang, L., et al. (2008), Transpacific transport of ozone pollution and the effect of recent Asian emission increases on air quality in North America: An integrated analysis using satellite, aircraft, ozonesonde, and surface observations, Atmos. Chem. Phys., 8, 6117–6136, doi:10.5194/acp-8-6117-2008.

Whaley, C. H., K. Strong, D. B. A. Jones, T. W. Walker, Z. Jiang, D. K. Henze, M. A. Cooke, C. A. McLinden, R. L. Mittermeier, M. Pommier, et al. (2015), Toronto area ozone: Long-term measurements and modeled sources of poor air quality events, J. Geophys. Res. Atmos., 120, 11,368–11,390, doi:10.1002/2014JD022984.

Section 2: can you include information about what kinds of vegetation are in this region that burned in the wildfires?

Figure 1a: Can an outline of the city of Pheonix be marked on the left panel? The caption implies that this whole area is considered Phoenix.

Section 2.2: Does the WRF-Chem model include the radiative effects of smoke on the air temperature and the photochemical production of O3? E.g., when smoke is heavy and shading the surface, does O3 production and surface temperature go down? If so, please say explicitly, as not all models do this. And if not, then this is a caveat to all of your O3 results that come in Section 3.2, as it’s an important missing process.

Figure 2: how come you don’t include TROPOMI O3 and CO as well?

Figure 4: Similarly, why is TROPOMI O3 not shown in Figure 4?

Fig 4 vs Fig S2: These two figures should either (a) have the same colour scales per species, so that we can more easily see the difference between the two case studies, or (b) have colour scales that highlight the spatial distribution of the enhancements. The HCHO row does both (a&b), the CO row does the latter (just b), but the NO2 row seems to do neither. Can you please adjust the colour bars of the NO2 row in these figures to at least do one or the other? 

Fig 5 & Fig S3: Ok, now I see that the colour scales above were set to match those for these model figures which follow. If that’s the priority, then I suggest that you re-jig these figures so that the TROPOMI and WRFChem tropospheric columns appear right on top of each other (e.g. one row for TROPOMI HCHO, and next row for modelled HCHO). You could potentially also add a model-minus-satellite row as well to better see the differences..

Line 351: I believe you mean Figures S3-S4 here.

Lines 390-395: This background information seems out of place. You could have included this when you first mention CO in the paper. Some of it (e.g., that CO is emitted when there is incomplete combustion) you can presume the reader already knows.

Line 402-404: Is the background O3 really an “absence of local sources”? Or are local natural fluxes (e.g., isoprene, soil NOx, lightning NOx, stratospheric O3 decent) included to contribute to this high background O3 during the heatwave? I think you probably mean absence of local anthropogenic sources... and maybe you can clarify whether biogenic and natural sources are included in panel (b). Line 404 says that the background O3 is due only to “regional and global influences on a monthly basis”, which the reader could interpret as only long-range anthropogenic sources.

Line 410-411: A few additional words can help clarify this sentence: “Mexico's anthropogenic contributions *to O3* (Figure 8f) have a larger impact than *they do for* CO (Figure 7f)…”

Figure 8: I notice here and in other O3 figures, there is never a negative contribution to O3 concentrations, which could occur if, for example, the emission source caused O3 titration or a transition on the O3 formation chemical regime, or if the fire emissions caused shading that reduced photochemical production of O3.  Do those circumstances really never happen in the model in your study? Or have they just been averaged out in the June mean? Perhaps the authors could include in the discussion the fact that there is no reduction in modelled O3 from any source at any part of the time series and why that may be (missing process in the model?).

Line 431: “Here” should instead be “In Figure 9”.         

Line 439-440: I’m not sure I see this in Figure 9. June 17^th only seems remarkable in that it’s the day when there is the least good match between the model and measured O3 in the time series, and so I don’t think it should be emphasized. Also, the background O3 contribution, which may or may not include local natural sources (see comment above) is the dominant O3 source throughout this time series. The red, which represents the Arizona anthro contribution is also significant throughout, and in particular on June 15^th, so I’m not sure why Mexico is emphasized in the text.

Lines 460-469: As per the comment above about what fire processes the model is including: Does it include the shading effects of the smoke on the O3 production? If not, then please include that in the discussion here as a caveat. If it were, the O3 increase with fires may not be so high (and may even decrease O3).

Figure 11: Can you please make the colour scales the same between panels (a) and (b). Similarly, make the colour bars the same for panels (c) and (d), as the purpose of the figure is to compare the O3 diff technique with the O3 tag technique.

Lines 532-535: could you please add here what causes the negative values in Figure 13c? While I was expecting/looking for negative values to appear for O3, I don’t know why negative values would appear for the fire influence on CO.

Lines 591-593: Don’t you mean “…the influx of VOCs from the fires can shift the chemical regime from VOC-limited to NOₓ-limited, altering…”?

Figure 16c and f: could you add in the caption which chemical regime red represents a move towards and which chemical regime blue is a move toward?

Guo et al. used WRF-Chem model to identify the source of ozone in summer of Arizona. Overall, the method is robust and the results are reliable. However, the novelty of this study is not fully revealed. Some conclusions seem to be well known. I suggest the authors should stress the major findings and the novelty of this study. The detailed comments are as follows:

1. The authors only introduced the importance of source attribution techniques, while this study lacks of the introduction of the novelty in Arizona compared with previous studies in California and many other regions.
2. Section 3.1 The authors should simply explain the reasons for the selection of this period.
3. Line 359: FNR often shows large uncertainties. Why threshold could you use to distinguish the VOC- or NOx-limited regions.
4. The authors used H2O2/HNO3 to identify the oxidizing capacity of the atmosphere and the relative contributions of different chemical pathways to O₃ Please examine the predictive accuracy of H2O2 and HNO3 firstly. Besides, the threshold also shows large uncertainties. Please explain the detailed reasons.
5. The limitations of this study should be added in the conclusion.