General

The paper points to the importance of aqueous iron chemistry in cloud droplets for the removal of anthropogenic SO₂ in the lower atmosphere in a chemistry climate model. There are other CCMs with similar aqueous chemistry and slightly different equations which are mostly cited in the introduction. 

It is, however,  not clear with which boundary conditions the sensitivity studies concerning Fe³⁺ and the pH are performed, also the definition of 'improved case' is fuzzy and has to be improved.   Nevertheless the study is interesting and worth to be published after revision.

Specific

A statement that reactions in liquid aerosol are not included but only reactions in the gas phase and in cloud droplets, is at the beginning of section 2.2. It might be good to have that also earlier, even in the abstract.

Section 2.1: What are the boundary conditions for aerosol? Is dust included or not? Is it related to the iron in the droplets? The provided links are not unique here since there are plenty of scenarios. More details please.

Table 1: Are the chemical equations complete? I wonder about missing HCHO and CH₃CHO and sometimes different products compared to other models. Is HONO in the text and HNO₂ in the table the same? If yes please decide for one notation.

Section 2.2, after line 186 and section 2.3: An initial setting of [Fe³⁺] cannot explain how the different scenarios and the spatial distributions are constrained. More information please. Do you mean every time step at cloud formation, if meteorological conditions are favorable for the selected cloud types?

Section 3: In the introduction reactions on mineral dust are mentioned, but not later in connection with the mismatch between model and observations in China, please improve. It might be also useful to include an extra line in Fig 4c with high Fe (e.g. from dust in droplets). Or is this automatically included in China?

Table 2, Figures 4, 6 and 7: The different units are very distracting. Better decide for one common unit and provide the (approximate) conversion factors in the caption of Table 2 or the text.

Section 4, line 344: I suppose these studies are performed with and without the corresponding equations of Table 1. Please mention that clearly if that is the case or explain how you proceeded.

Section 5.1, line 370 (and earlier): it not quite clear how [Fe³⁺] at different values is superimposed to the variable values at different locations. It should be better explained already in Section 2 what the "improved case" is with this respect.

Line 376: added to what (and where)?

Section 5.2: Please more details how the pH is entered into the simulations. Is it only via the directly dependent equations or for example via constraining of [H⁺] using Eqns. 33 to 58 of Table 1 in the droplets? Feedbacks with sulfate and nitrate as indicated in section 2.3? Please improve text here.

What is the difference between Fig. S7 (former Fig. 7 in first draft) and Fig. S10?  You may skip one because there appears to be almost no difference (except a slightly different caption and differences in Africa for JJA and SON for pH 6). Or explain better in text (section 5 and 2) and the captions. 

Technical corrections

Please provide maxima and minima or the range of the color bars in the figures (also in supplement).

Include blanks after ';' in citations.

Line 356, 393: Don't use 'trend' here, wrong word.

In the references often pagenumbers are missing. The doi is provided with 'DOI', 'doi' or nothing before, please use consistent notation. 

General comment

This manuscript concerns with the global impact of aqueous-phase chemistry on the simulation of atmospheric SO2 with the CESM2 model. They apply a framework for integrating this chemistry that is similar to the one in use in other models like GEOS-Chem. This approach does not foresee the online calculation of pH. Additionally, not having soluble Fe emissions the authors go on with sensitivity simulations with plausible, at least regionally, of pH and [Fe3+]  applied globally. Then, an attempt of separating the effects of different parts of the mechanism on SO2 is done but limited by the choice of the reaction categories to xclude at once (more to it below). In general, the manuscript is written well and with a good and clear structure. The topic and the results are interesting. However, I have identified a few points to be addressed/clarified.

Comments

In Table 1b most of the references for the reaction kinetics are given to three collections/mechanisms published earlier. Please cite alongside the primary literature for each reaction and not just the modelling studies that collected sets of reactions. Moreover, it is not explained/justified why the authors blends these (secondary) sources with others (primary) to obtain their own chemical mechanism in CESM. Although section 2.2 contains the reaction tables that well done (apart from the secondary references), in the text little is written for describing in words the salient features of the mechanism used here. I understand the authors wrote a comprehensive introduction on the SO2-relevant known chemistry. Howeover, the latest and most reliable data is not necessarely reflected in their mechanism. For instance, the oxidation of S(IV) by NO2 (also mentioned in the text). However, in Table 1b one finds only reaction 199 for HSO3- and nothing for SO4--. Is the latter neglected because only important at high pH values that are seldom represented in model simulations? Furthermore, although Cheng et al., 2016 is cited in the introduction, the low rate constant by Lee and Schwarzt 1982 (by the way the reference is incomplete, please add at least the url https://www.osti.gov/biblio/6567096) is used. This is a 2nd-order reaction rate constant while reaction 199 is obviously 3rd-order. The former is about one order of magnitude larger as reported by a few later studies with the last one being Spindler et al. 2000, DOI: 10.1016/S1352-2310(03)00209-7 . These issues I have make me doubt about the statement on line 18 in the abstract about the minor role of N-chemistry for SO2 oxidation.   

It is remarkable that the mechanism the authors put together completely lacks the chemistry of methanol, methyl hydroperoxide and formaldehyde. The latter, in its hydrated form, combines with HSO3- to produce HMS which reacts quickly with OH to yield SO5-. Recent work on the importance of such chemistry are DOI: 10.1029/2020JD032706 and DOI: 0.5194/acp-21-457-2021 . Why only reactions of organic acids with sulfur are considered? Why has this chemistry been neglected? It is well known and acknowledged as a significant source of O2-and thus OH. I think this chemistry needs to be considered especially in a study about the importance of SO2 aqueous-phase sinks.   

Table 1a, line 163: here again the kinetic data seems to come from the modelling study by Liang and Jacobson (1999). However, the primary references for the kinetic paramaters are missing. Please cite the primary sources for the data.

Line 164, footnote b: why using "C" for a rate constant and "k1" and "k2" for the pre-exponential factors and T-dependency? This is confusing. Why then "k2" is multiplied by 500? "k1" is obviously k(298K) but "k2" is not a rate constant. By the way in Table 1a the units of "k1" and "k2" are not given.

line 343: this separation of the effects by HOx-chemistry and Fe-chemistry is artificious as much of the fuel (HO2 and O2-) for the Fe-chemistry is produced by the HOx-chemistry. By shutting down the latter, the former is also strongly depressed. Thus, one can not really see separate effects. It makes sense to always keep the HOx-chemistry (which also has sulfur-sulfur reactions!) and switching off the other parts of the mechanism. Otherwise the sulfur-sulfur chemistry without involving HOx should be separated from the "HOx-chemistry". Anyway, the mentioned "parts" of the mechanism do not interact linearly and their respective effects are not additive. More appropriate for the goal of the section would be a budget for the loss of S(IV) in the aqueous-phase. This would be a good addition to the manuscript.  

lines 405-406: the little change is likely affected by the authors using a low rate constant for the S(IV) oxidation by NO2 (see my previous comment)

Minor comments

line 215: citing Xu et al. 2019 and Yi et al. 2017 for justifying the one-year model spinup is not necessary and very suspicious of authors self-citation. Please remove these citations that anyway have thematically very little to do with the topic of the manuscript.

line 236: in Figure 2 mixing ratios are shown and not concentrations. Obviously they are not same. This mistake is done in many other places. It needs to be corrected.

line 53: Poschl -> Pöschl