This manuscript summarizes experiments that compare dry deposition schemes available in the CMAQ model. The conclusion that careful diagnostic grid-scale and land use specific dry deposition variables should be adopted especially for model intercomparison purposes is an important message of this work.

In general, I think this manuscript presents a reasonable analysis with many valuable details that will be of interest to the relevant community. My major comment is really related to what I felt was a lack of focus in the earlier parts of the manuscript:

There is a lot of text in this paper (throughout the Methods and Section 3.1) that spends time explaining differences in their simulations with the output from model experiments in Appel et al. (2021). I was a bit confused by this. The main purpose of this manuscript is to compare the M3Dry and STAGE dry deposition schemes in an internal CMAQ implementation from 2016. To give this manuscript a more targeted focus, I would have suggested just to evaluate their model implementation directly without commenting so much on the indirect evaluation through the Appel et al. (2021) implementation. Likewise, I was a bit confused why there are so many tables and figures focused on evaluating 2010 model output, when none of this output is relevant to the dry deposition scheme discussion that uses only the 2016 model experiment (from Sections 3.2 onward).  A great deal of unnecessary (and seemingly unrelated) text could be avoided if the focus was simply on the 2016 model experiments to begin with. This focus would also help remove many tables and figures (e.g., avoiding at least Table 2b, 3b, and many panels throughout Figure 1a-2b) to tighten the content.

To illustrate my point, one of the sentences early in the conclusion section mentions that lateral boundary conditions are a driver of the main differences between the present simulations and those from Appel et al. (2021). Why is this relevant to the internal comparison of CMAQ dry deposition schemes? Following this one (and only) mention in the conclusion, the remaining two and half pages of summary discuss, appropriately, only on the 2016 dry deposition scheme comparison. I highly recommend a substantial revision of the earlier sections of this manuscript so that it focuses on the relevant 2016 simulations only, and perhaps demphasizes all the comparisons with Appel et al. (2021). In my opinion, all this extra material detracted from the motivation of the paper. Otherwise, perhaps I missed something important, and the authors could provide better motivation for why all this analysis was included. 

I really don’t have many other substantial comments. Sections 3.2 and 3.3 contain a great deal of valuable information about dry deposition modeling in CMAQ, with implications for other models as intercomparison projects, that I believe could be of interest to many readers of ACP.

Minor comments:

Line 171: First introduction of “VEGF” – please define.

Line 494-496: “…treats cells with more than 10% water as either all land or all water depending on whether the fractional water coverage is below or above 50%”. I had trouble following this sentence. Could it be reworded or broken up to help with its clarity?

The manuscript is a very detailed investigation of dry deposition within the CMAQ model, which places the runs for the AQMEII4 exercise in context with other configurations and inputs that have been used recently or are more typically used for CMAQ. The sensitivity studies provide a great deal of new or updated information about the impact on air quality concentrations and fluxes of land use definitions and treatment, the underlying details of the resistance parametrization, and inputs such as meteorology and boundary conditions. The results will be primarily of interest to atmospheric modelers, but also are a significant contribution to the larger project that is of broad interest to the air quality and deposition impacts communities. Given the length and level of granularity, the paper needs to be carefully organized, and the authors have done a good job of this. I recommend publication with very minor revisions.

Specific comments:

Section 2.2: the description of the set of model runs is missing the M3DRY_APPEL_EMIS_2016 one

Section 2.4: There’s significant discussion in the paper about results in southern Canada; what was the reason for not using data from Canadian sites (or Mexican, if available)?

Fig. 1: it’s a bit difficult to see the observations distinctly. Can you make them dashed, or points, to bring that line out a bit better?

Fig. 6: maybe choose a more contrasting color for the WRF lines

Fig. 8: are these plots linear scale? I can’t tell with only 2 values.

Fig. 9: I suggest you swap the M3DRY and STAGE order on the plots to better correspond with Fig. 7

L. 695-6: I would include deciduous broadleaf in the list of LUCs where NLCD > MODIS flux, it’s a significant difference (here and in the summary/abstract)

L. 732: pretty sure you mean NO2, not SO2, here

Section 4: It would be helpful to provide some context as to how significant are these deposition velocity (or flux) differences between configurations compared to differences that are seen between CMAQ and other models with different schemes entirely. Presumably this will be discussed in the larger AQMEII4 exercise, but there are studies in the literature that could be cited to give the reader a sense of the relative magnitude.

References: I expect you are aware that the Clifton “in prep” paper is now citable on ACPD