The mechanism of ozone stress on plant physiology, especially on photosynthesis, is well realized. However, the ozone stress scheme in land surface models is at the very beginning stage, introducing large uncertainties in GPP estimation. The authors compiled over 4000 paired data from peer-reviewed literature and proposed new response functions to reproduce statistically significant relationships between the relative reduction of physiological variables and accumulated phytotoxic O3 dose (POD). I believe this work potentially pushes the development of the ozone stress scheme in land surface models one big step further for the following reasons: 1) The new data samples are more comprehensive including almost all plant functional types of CLM5, and the new scheme is more reasonable than the previous schemes. The previous schemes only found skill response functions for crops, or estimated the photosynthetic response function based on yield/biomass response observations with the implicit assumption that all the processes and variables from photosynthesis to biomass were correctly modeled which is obviously not true; 2) the ozone-stress-induced global GPP reduction is predicted to be a more reasonable and moderate number of 13%. The original one is 28%, which almost indicates the original model without ozone stress is completely wrong in global GPP estimation, even the number more or less matches the observation-based dataset (e.g., FLUXCOM). While this work is publishable, I have a few minor comments that could enhance the quality of the manuscript, particularly regarding how the empirical relationships are derived from 4000 paired data. Please see my detailed comments below.

Figures 3 and 4: I acknowledge that these data are sourced from diverse literature and encompass a variety of environmental conditions and locations. Consequently, expecting a clear linear relationship might be unrealistic. Nonetheless, given the big unexplained variation, there may be potential to improve the fitting. I wonder if incorporating environmental factors or applying region-specific fitting functions could better capture the photosynthetic response. I recommend that the authors discuss potential strategies to account for the remaining variation in the data.

Line 63-67: The sentence is too long. Consider separating it into two sentences.

Line 87: “physical” I would say “biophysical”.

Line 115-116: Better put the version number of CLM and CESM and mention whether the current implementation is on the same or different versions of CLM and CESM. So, the reader will know whether or not L15 is comparable with this ozone stress scheme.

Line 166-167: “only data categorized as high and medium confidence defined by Lombardozzi et al., (2013)” Need generally mention the confidence level is defined based on what standard in Lombardozzi et al., (2013).

Line 169-171: “if the data are previously or more completely reported in another article” do you mean the data is repeatedly reported?

Line 290-291:”2000Clm50Sp” and “2000Clm45Sp” Better use a simple description rather than the CESM configuration abbreviation, which will be more friendly for those who don’t use CESM.

Line 306: 1.9^o should be 1.875^o .

Line 311: Missed one atmospheric forcing “Downward longwave radiation”

Line 313-314: “have no interannual variability” mislead. MODIS data, of course, have interannual variability. Maybe you just want to say “you use a prescribed climatology of vegetation distribution and structure, which is based on present-day MODIS satellite observation”

Line 319: “28.9655/47.9982” I'm not sure if we really need such high precision.

Line 340-341: “there is no need to use a function from one vegetation type for another” reads obscure. “no need” assumes readers know the original scheme derives the function from observation of another vegetation type. But I guess most of them don’t unless reading further. I suggest a direct introduction like “Each vegetation type owns its individual function based on observation.”

Line 481-486: I am wondering how are these modeled relative responses compared to the literature? More discussion will be helpful.

Line 631: “decouple”. How the results show the decouple? Maybe need more specific discussion. Also, I was expecting a photosynthetic rate versus stomata conductance scatter plot to show the decoupling.