The manuscript firstly explores the different impact of the two commonly used O3 damage parametrizations which is an interesting comparison with relevant conclusions for the community. The authors additionally use measurements of O3 and meteorology to evaluate the model prediction which, however, could be more taken into account. In general, I feel more explanation and interpretation in the result section can imrove the manuscript, though it is overall well written and understandable. Please find my minor comments below:

•  l. 285: The terms 'warmings' and 'coolings' are not clear. This would more refer to model changes or even climate change experiments
•  l. 288 "[...] but it shows a high correlation (R=0.96)"
•  l. 292 For which model was it also reported ? Is it model-specific?
•  l. 296 mention the reason for the overestimation of O3 (counteract the overstimatio of wind speed?)
•  l. 298 "reports" of "overestimated" (the model overestimates)
•  l. 316 f: But the O3 damage not only depend on O3 concentration, right? How do you come to the conclusion that S2007 is more reasonable here?
• l. 333: 5.5% is this an average over the model region?
• l. 344/345 Please explain the reason for the different changes by the two schemes
•  You can be more concrete here.
•  L. 366/367 Why is the L2013 O3 inhibition constant over day?
•  l. 388/389: The reffering of the different values is not clear. Perhaps, there is a bug with one unit or the brackets.
• l. 423 ff. please split the sentence in two or shorten it
• l. 433-435: To my knownledge that shouldn't be the case? Didn't the other models consider leaf turnover?
•  l. 446 f: Be consistent with the O3 unit.
•  l. 464/465: I would rephrase to "However, this scheme show no significant different change for sunlit and shaded leaves"

This paper use the established methods of chemistry-meteorology-ecosystem modeling to simulate ozone damage on plants over China, and the associated impacts on surface energy balance, carbon sink, meteorology and air quality. The manuscript is well-organized. Compared to earlier papers in this topic, the authors focus on comparing several established methods of calculating ozone damage (S2007 vs L2013), which is an important and new contribution. Minor revision is recommended to address several linguistic and conceptual problems:

L48: Rewrite as “…adverse effects on ecosystem functions, global warming and O₃ pollution through…”

L60: rewrite as “…growth, suppressing ecosystem carbon uptake.”

L104: “surface energy balance”

L107: “but” -> “and”

L311: what is “instant O₃ concentration”?

L 310 – 313: Clearer explanation is required. L2013 (Table 2) has a lot of PFTs with 0 slopes. That means when stomatal O₃ flux is above 0.8 nmol m^-2 s^-1, the response of photosynthesis and stomatal conductance remain constant. I believe this causes the same phenomenon described in L 366 – 367, especially during ozone season. A few CUO and PFT plots could help explain/verify this.

L316 – 318: There is no direct observation suggesting plants in southwest receive less ozone damage. This is not a valid conclusion and not necessary for the paper. Remove this statement or provide more direct evidence. On the other hand it is fair to point out L2013 lacks distinction between sunlit and shaded leaves since direct evidence were given by the authors.

L 343 – 346: Like I explained above: for a lot of PFTs L2013 has constant response after stomatal O₃ flux is higher than a threshold, while S2007 depends on instantaneous stomatal O₃  flux. It’s more appropriate to highlight the difference in model structure/assumptions that leads to different result between S2007 and L2013 than judge which scheme is better without comparing with direct empirical evidence (e.g. plant trait and EC measurements). 

L 393: This paper suggests that O₃ damage increase isoprene emission because of increased leaf temperature, which is in line with previous studies (Sadiq et al., 2017). However, isoprene production is coupled to photosynthesis. There are empirical evidence, that high O₃ exposure actually reduces isoprene emission when O₃ exposure is prolonged enough to suppress photosynthesis (Bellucci et al., 2023). As an empirical parameterization, MEGAN does not include this effect. While this does not completely invalidate the O₃ feedback result, this possible artifact in isoprene emission and its potential impact on the result have to be discussed thoroughly.

Reference:

Bellucci, M., Locato, V., Sharkey, T. D., De Gara, L., and Loreto, F.: Isoprene emission by plants in polluted environments, Journal of Plant Interactions, 18, 2266463, https://doi.org/10.1080/17429145.2023.2266463, 2023.

Sadiq, M., Tai, A. P. K., Lombardozzi, D., and Val Martin, M.: Effects of ozone-vegetation coupling on surface ozone air quality via biogeochemical and meteorological feedbacks, Atmospheric Chemistry and Physics, 17, 3055–3066, https://doi.org/10.5194/acp-17-3055-2017, 2017.

The authors examined the meteorological and air quality feedback of O₃ damage to vegetation by coupling WRF-Chem with two O₃ damage schemes. This reviewer has a few questions.

First, S2007 seems to calculate instantaneous (for WRF-Chem’s model integration time steps or hourly) values of the undamaged fraction F, whereas L2013 calculates the ozone damage ratio for the entire growing season. So, was one constant L2013-calculated, plant-specific, O₃ damage ratio applied throughout the whole simulation period, whereas S2007-calculated O₃ damage ratios were time-dependent, when the schemes were coupled with WRF-Chem?

Second, the way the manuscript was written did not show the distinction between sunlit and sunshade in S2007- and L2013-calculated O₃ damage ratios, which leads to the question how the ratios were applied to NOAH-MP. This leads to the next question. Why were L2013-calculated sunlit and sunshade O₃ damage values for both photosynthesis and stomatal conductance were almost the same, whereas S2007-calculated ones showed such a contrast?  

Third, isn’t Eq. 5 supposed to be the integration of Eq. 4 according to its definition?