In this study, the authors used an up-to-date regional climate-chemistry-ecology model to quantify the effect of anthropogenic emission, meteorological factors, and carbon dioxide variations on O3 variation across China and highlighted the importance of considering CO2 variations. I suggest this article be published with some modifications to improve the clarity of some details and ambiguous presentation. My comments are listed below.

1. The main innovation is the emphasis on the role of CO2, but this only occupies a small part of this study. Other effects such as meteorology and emissions have been extensively discussed in previous studies and the authors need to elaborate more on the significance of this study.
2. Line 98-100: Why did you choose ERA-Interim data to evaluate meteorological variables simulation instead of using observations?
3. Line 134-139: Can you give more descriptions of the model improvement?
4. Line 162: “i,m=2008” should be the subscript.
5. Section 3.1: Why did you only compare simulations and observations in 2018? Did the model perform well in other years? I suggest you evaluate simulated surface meteorological variables because they significantly affect surface air pollutants. I also recommend you assess the spatial distribution of surface O3 and CO2. For example, you can additionally evaluate model performance in key regions like NCP, YRD, and PRD apart from the whole domain.
6. Line 243: This expression is ambiguous. Meteorological factors are favorable for ozone formation in summer.
7. Line 260-270: There are some contradictions in this part. You attributed the decrease in ozone concentration to increased cloud fraction, decreased SWF, increased precipitation, and enhanced wind speed. But how the warmer surface and higher PBL can accompany these conditions?
8. Figure 5: Please modify the value range of the color bar.
9. Section 3.4: How did you quantify the contributions of isoprene and precipitation to ozone concentration?
10. Figure 7: Please modify the value range of the color bar.
11. Line 342-343: What did you mean by “due to the slight increase in air O3 concentration”?
12. Section 3.5: I suggest adding a figure or table showing emission trends of main air pollutants and precursors to support the explanation.

Major comments:

This is an interesting study looking at the effects of emission changes, meteorological variability, and -for the first time- also impacts of CO2 (via radiative forcing, but also through isoprene emission).

The manuscript needs to be improved on several aspects. 

1) model descriptions, boundary conditions and a host of other assumptions made need to be improved and completed, to be able to understand what has been done in the study.Units need to be included in captions of tables and figures.

2) The use of a single year (2008) as a reference year for pregovernance 2009-2013 and post-governance (2014-2018) periods is problematic.Often what we are looking at is some specific feature of 2008 that is showing up as difference.

 A more classic attribution approach using e.g. constant 2008 emissions with variable meteo; or constant year with varying emissions would be more appropriate. Alternatively, I could suggest using the period 2008-2013 as reference

period (pre-governance) and just show the changes  compared to this period. Make sure that the units are correct and comparable.    

3) I suggest re-ordering of sections to describe effect of emissions; effects of meteorology; and CO2 effect in the order of importance. A 4 section could be added describing the addiviity

4) The authors should pay attention to the language- which is often confusing and inaccurate. I have indicated a number of places where this is relevant.

5) Figure/tables should be numbered in  order of appearance.

6) The  new aspect of this paper is the inclusion of effects of CO2- through affecting the radiative balance, but also through isoprene emissions. However, it is not clear to me how exactly CO2 has been included in the calculations; what about boundary conditions? 

Although the model bias is quite large (1 year offset), it is probably more important that the trends are realisticly included and therefore the changes are reflected in the results. One other aspect is that two things are coming together: changes in (regional) radiative forcing, and isoprene emissions, but there is no discussion at all of the magnitude of these

effect in the model

7) The authors mention that one of the improvements is an improved photolysis scheme. One of the post governance effects would be a gradual clean of aerosol (precursor) emissions, and consequently less aerosol scattering and diffusive radiation.

Could the authors elaborate (also in the manuscript) to what extent this included in the manuscript and how it affects the results.

All taken together I suggest major revisions before this manuscript can be accepted for ACP.

Detailed comments.

20 improved compared to a previous version?

23-24 Examples are not really for an abstract

25 CO2 includes only radiative effects (in which case it is somewhat like a meteorological). Or also 

ecophysiological? The effects rather suggest secondary rather than critical.

26-29 not sure if the inclusion of CO2 variations is so important.

32 'the ecosystem' is a bit general. Rather suggest vegetation growth.

33 specie=>species or compound. Influencing the earth's radiative balance.

40 performed=>issued; initialised.

46- clarify that the previous paragraph was about emission.

47 force=>anthropogenic radiative forcer of the climate system

57-62 Suggest to integrated these sentences in the next paragraph as the connection to CO2 is not immediately clear.

79 should not be negligible=> can be substantial.

88 sure that emissions have increasing- but here what counts for CO2 is how concentrations have been increasing?

98-101 ERA-interim is not really observed- it of course assimilates observations, and can be a useful approximation if observations are absent or difficult to retrieve.

Explain better the rationale of using ERA-interim. And why in contrast stations were used for ozone. (I guess they were not available to the ECMWF/CAMS global ozone assimilation).

119 what is an authentic atmosphere?

112-127 The model description is rather empty.

 A short description of the what the two models are doing would be useful (e.g. what chemistry and physics schemes in the RegCM-Chem), what boundary conditions were used? What are the critical weaknesses for 

this study- that will be addressed in more detail?

What are the characteristics of YIB?

129 Recommend to use clear version numbers rather than "previous version of xxx". 

131 I guess it would be approriate to give a short description of the radiative transfer scheme; and also of the photolysis scheme?

Are these two unified, or separate routines? Are they consistent?

143: not clear Are you including CO2 as a tracer in the regional model? If so, where do you get the boundary conditions from?

Figure 2: EAR=>ERA. I assume that the Boundary conditions were taken from ERA interim, but what about the RH? I notice also the much finer resolution of ERA-interim, but

it is not clear what the resolution of the reigonal model was. Both should be given.

197 and table 2: Explain what is MDA8? What is the evaluation period (summer 2018?). I notice that the difference of CO2 between the simulations and the OBS are really high- and could be much better if the boundary were taken into account

appropriately into account. Perhaps it is not a super-critical issue as long as the CO2 trends are appropriately taken into account- but this is to proven.

215 also high temperatures and humidities are conductive to ozone production.

218 Can these regions be graphically shown, together with some qualitative information? I have no clue what/where is: NCP, the YRD, the PRD, the SCB, and the FWP

217 The ozone seems to increase mostly in 2017 and 2018. While the 10 plots are useful, I could suggest to use the target analysis regions (tables), and show also average concentration change as a fucntion of year (lines).

232 If table 5 is discussed first, it should be placed before. Tables (and figures) appear in order of discussion. Also I guess these are reponses in ppb not ppb per year? Table is not clear about units. 

243 What is meant with: "The meteorological factors were generally unfavorable to O3 formation during the study period". I think they are very favourable. But I guess the authors intend to say that in general changes in meteo conditions, 

led to overall (small ) changes in O3. Of course using one reference year for meteorology is somewhat misleading (figure 4). Also Figure 6 shows that some of the changes are quite determined by the year 2008 (especially for precipitation).

245-253: which table are these results?

256 changes in meteo factors are unfavorable (not the meteo itself).

288 the distinction of preG and postG needs to be revised to tell variability from signal.

293 driver of climate *change*.

321 give units also in figure caption (hard to read).

325 units?

351 multiplied=>multiple

366 give units.

373 Suggest to have a seperate section that discusses Figure 11, and also gives an attribution. (is the red line fully explianable by the blue, green and black line)

377 this is not an uncertainty analysis, rather a discussion of why boundary conditions were kept the same in the sensitivity studies. (I guess CO2 was however not kept constant). This information should be given much earlier. 

A uncertainty analysis could try to explain how this would affect the results. And there are other factors as well that need to be discussed.