General comment

• This manuscript, “Critical Load Exceedances for North America and Europe using an Ensemble of Models and an Investigation of Causes for Environmental Impact Estimate Variability: An AQMEII4 Study” presented by Maker et al., summarized their excellent work in the model intercomparison study of AQMEII4. The authors reported a comprehensive investigation regarding the model variability and suggested potential research directions for future studies. I believe that this kind of study is crucial to progressing our knowledge of the model themselves and their applications. Most of my comments are minor specific and technical comments for better reading and presentation quality; however, I would like to make one major request regarding the current manuscript.

Major comment

• As described in Section 2.0, model simulations over Europe were carried out in 2009 and 2010, which is why there is a large difference in the meteorological field. From this manuscript, I can follow the discussion in the U.S., which targeted a significant SO2 emission reduction between 2010 and 2016. However, from the result over Europe such as shown in Figs. 3 and 5 and Table 4, the difference between the years 2009 and 2010 was small. This discussion could be presented in other companion papers in AQMEII4 project, but one possible conclusion from this manuscript is that there is little impact on the estimates of critical loads despite the significant impact from the meteorological field? Actually, I did not fully figure out what was the most variated parameters (temperature, wind field, precipitation, etc.) between these years. For more information derived from this study, please consider including this point.

Specific comments

• L37 and L40: In the abstract, “New” is repeated, and this wording will be ambiguous. It will be better to use another specific term.
• L157-160: As written in L473-475, it is better to mention that this time is the year 2021 status explicitly.
• L177: For modelers, “process analysis” will be associated with the model-embedded process analysis tool (https://www.cmascenter.org/cmaq/science_documentation/pdf/ch16.pdf). Could you change the wording here?
• L198: In the latter part of the discussion, we can follow the wording “reduced ensemble”, but this term is suddenly used here without any introduction. Please rewrite or define this wording here.
• L416: What is “pant” here? Is this a typo of “plant”?
• L413-415: “A second model…” is described in L415, so “A first model” can be explicitly stated in L413?
• L423: For wide readers, “PRISM” should be shortly introduced.
• L476-477: Because the following sentences started from the U.S. and then stated Europe, it is better to change the position here (i.e., the years 2010 and 2016 for North America, and 2009 and 2010 for the European region).
• L549: This section introduced participating models. We can follow the result section (L708) from these descriptions, but I think it would be helpful for readers to summarize in a table which model was performed for which or both domains (Northern America and Europe).
• L961 (Fig. 14), L1397 (Fig. 22), L1422 (Fig. 23), L1490 (Fig. 24), L1667 (Fig. 27), and L1740 (Fig. 30 ): The gray grid indicated the negative value from the color bars; however, I think this is just out of the calculated domain.
• L1575 (Fig. 26): Because the color scale is based on blue-red bars, it is tough to distinguish blue and red lines, which indicate the predominant land use category. Please revise this figure.

Technical corrections

• L98: Please correct “A a”.
• L119: “simple mass balance (SMB) model” will be better.
• L224: No need “)”.
• L236, L239, and L240: Please use subscript for “PM2.5” and “PM10”.
• L335 (Table 1): Maybe there is no need for parenthesis in “source” description.
• L342, L339, L371, L393, and L408: For the consistent expression of the subsection name in L429 and L447, it could be used “:” like, “1.2.1 Canada: Aquatic Ecosystem Data”.
• L363 and L364: The charge for each ion should be presented.

Through inter-comparison among models and comparison with monitoring results, the performance of CTM models for modeling S and N deposition in US and Europe was systematically evaluated. The critical load exceedance was further estimated based on the ensemble deposition simulation. It is of great importance that the future improvement of the models was suggested.

However, the paper is not a research report and is limited in length. Although critical loads were needed to calculate the exceedance by S and N deposition, the calculation method of critical load need not be detailed described in the main text, because only existing critical load databases were used in this study. Instead, relevant literature (or supplement) can be referred. There were too many figures and tables on the comparisons of modeling results in the paper. I suggest to make further summary and move most of them to the supplement.

In addition, the focus of this paper is on the uncertainty of critical load exceedance (CLE). Recognizing the uncertainty of deposition modeling, especially the underestimation of wet deposition, the degree of underestimation of CLE should be shown in the paper.

Some detailed comments are as follows:

Line 122-123: Deleting “, denitrification, nitrogen immobilization in the rooting zone, run-off volume, and a critical value of the non-sodium base cation to aluminum ion ratio”

Line 224: Delating “)”

Line 236: For the whole text, 10 in PM10 and 2.5 in PM2.5 should be in subscript.

Line 272: Moving detailed introduction on critical load to the supplement or delete.

Line 329, Figure 1: Adding explanation of the dashed lines.

Line 681-700: The text can be shorter with Table 3.

Line 892, Table 4: Can the emissions of major pollutants such as SO2, NOx and NH3 in each year be added? Also S and N deposition?

Line 1067: Changing “workT” to “work.”

Line 1103: Adding full point in the end.

Line 1170, Table 8: Moving this like tables to the supplement.

Line 1429: Missing “;” before ca.

Line 1438: Mission “.” After “AQMEII4”.

Line 1623: Deleting “HNO3 summer than”.

Line 1779: The conclusion can be shorter.

Line 1819: Here the underestimation of CLE caused by the bias of deposition modeling is of interest.

Line 1855: Same comment as above.

Line 1902: Same comment as above.

Line 1931: Same comment as above.

Line 2636: The reference is repeated.