This paper investigates the sources of atmospheric oxidized mercury at a mountain top site combining ambient measurements with a chemical transport model. This paper demonstrates that chemical transport model underestimates oxidized mercury concentrations, yet current sensitivity modeling experiments are insufficient to pinpoint the causes. The measurements seem to be appropriate and state-of-the-art. I recommend that this study should include several additional sensitivity experiments, such as examining iodine oxidation effects and deposition rate variations. These experiments could help unravel the reasons behind the observed underestimation of oxidized mercury levels. Unfortunately, the current analysis lacks this critical component. Thus I suggest major revisions of the current manuscript.

General Comments:

This study uses an atmospheric mercury chemistry model (GEOS-Chem) to analyze a measurement campaign conducted at a mountaintop site in the Western US. The authors identify that GEOS-Chem significantly underestimates the amount of oxidized mercury measured during this campaign. This highlights an important knowledge gap in Hg models; however, it is an issue that has been identified in many previous publications about GEOS-Chem (e.g., Gustin et al., 2023; Shah et al., 2021; Shah et al., 2016) and the 3 sensitivity simulations conducted here are not sufficient to bring us closer to resolving this issue. I believe that more sensitivity simulations and analysis would be required to yield more actionable results, and I suggest some ideas below. As it currently stands, I believe the article is more suitable as a measurement report due to the limited scope.

Main Issues:

1) The sensitivity simulations aim to test the impact of different oxidation rates on the comparison with the observed results. However, there is no simultaneous adjustment of the photoreduction rate or total Hg emissions, which leads to a highly unlikely Hg0 budget for SA1 (Figure 3). Because there is too little total Hg in the atmosphere, it is not surprising that for SA1 there is too little oxidized Hg. The authors admit this issue at several points, but it would strengthen the paper to have a scenario where the model is retuned (using higher emissions or photoreduction) so that a high oxidation scenario can be fairly compared to the results of the Base scenarios. I would also consider doing simulations where oxidation rates are only adjusted locally rather than globally, as the authors state that this episode is likely due to a in-situ oxidation process (P7L158).

2) I was intrigued by the discussion in Section 3.3 about using HgII:Hg0 as a potential metric for evaluating HgII deposition processes. There are definitely uncertainties related to deposition in the models in general and the parameters assigned to deposition of different HgII species, and the paper would benefit from a more detailed discussion of these uncertainties. I think there is a clear opportunity here to conduct additional sensitivity simulations adjusting the HgII deposition rates to constrain the potential bias in HgII deposition rates, using the slope derived from observations. Without this evidence it is difficult to know how biased the HgII deposition processes are in the model, or if the HgII:Hg0 slope depends on other factors.

3) In Section 3.4, the authors use the Hg0 temporal variability as an indicator of oxidation (P14L247). However, Hg0 variability will also be affected by issues in emissions, deposition, and transport. Therefore, I believe it is too simplistic to assert that the net oxidation is four times too low in the simulations, since even the simulation with higher oxidation (SA1) does not seem to improve the simulated amplitude of Hg0 during the episode (Figure 3). 

4) Building on the previous point: I think the change in magnitude of Hg0 can also be strongly impacted by dynamics; i.e., the specific dynamics of a mountaintop site that will be difficult to resolve in a global model. Even the high resolution simulations (~30 km x 30 km grid boxes) might not necessarily capture the dynamics above such a complicated topography. Is the model able to reproduce the variations in other atmospheric compounds measured at this site? Such a comparison would be helpful to identify whether this is a Hg-specific issue or a dynamics issue in general. 

Technical/Minor Comments:

P2L35 - Recent mercury literature (Zhou et al., 2021; Sonke et al., 2023) suggests that dry Hg0 deposition is a substantial portion of the overall atmospheric Hg sink. 

P4L94 - This term is unclear to me: expanded uncertainty

P7L147 - specify whether HgII here includes particulate species or only gas phase species

P13L232 - Slopes greater than -1  -> it’s not clear if you mean greater in absolute magnitude or more positive (i.e., -1<slope<0).

P16L272 - should be Hg(OH)2

P16L275 - A major reason for the predominance of HgCl2 is its stability to photoreduction compared to other Hg(II) species, see Saiz-Lopez et al., doi: 10.1038/s41467-018-07075-3 (2018)

P16L276 - Speciation discussion - there has been no molecular determination that these measured Hg compounds actually contain N- and S- species; it is just inferred based on similarity of thermal desorption profiles and a few Hg(iI) compounds that are available commercially (which don’t include Hg(OH)2 or HgBrOH that are predicted to be prevalent)

- Table 3 - not all parameters in this table are fully understandable (HgII:Hg0 (r^2) - shouldn’t this be a negative correlation?) nor are all discussed in the text (HgII:Hg0 intercept: what does this signify?)

- Figure 7 - add trend lines for simulations and observations, so that we can see visually how they compare