General Comments

The manuscript describes the development of WRF-GC-Hg model, based on the already developed WRF-GC model and also shows a case study of the application of WRF-GC-Hg model in understanding the high Hg wet deposition in Southeast US. I do have several major comments:

• Does the paper focus on the study of the high Hg wet deposition in Southeast US only or also the development of the WRF-GC-Hg model? I would suggest including the development of the WRF-GC-Hg model as an important component of the paper. If so, the title should be revised as something: Development of WRF-GC-Hg v1.0 and its application in studying Hg wet deposition in Southeast US. This will make the paper stronger and more applicable. With that, the paper will need to be reorganized to include one part to focus on the new development and its evaluation and another part to focus on the study of the high Hg wet deposition in Southeast US. It will be also nice to extend the domain to the whole continental US for the WRF-GC-Hg evaluation part.
• The paper conducts two WRF-GC-Hg sensitivity simulations with a horizontal resolution of 50 km and 25 km respectively and compare the results to the GEOS-Chem simulation with a spatial resolution of 4º x 5º. As expected, the WRF-GC-Hg simulations with a finer spatial resolution will resolve more spatial signals. The comparison will be more meaningful to include the GEOS-Chem nested Hg simulation results, which are comparable to WRF-GC-Hg simulations.
• Some of the sentences are confusing. I would suggest improving the English language throughout the manuscript.

Specific comments:

1. Line 23-26, can you be more explicit here about the heights, different types of precipitation etc.?
2. Line 28: It is atmospheric Hg that can undergo long-range transport, it is not accurate to say that Hg goes through long-range transport here.
3. Line 46-48: I am not sure what it means here. What is the 80% of rainfall?
4. Line 108: The model is run as one domain with 50 km and 25 km, right? “The model horizonal resolution is set as ranged from 50 km to 25 km, with 50 vertical layers.” It sounds like the model is run as a nested domain. Please clarify here.
5. Line 120: what are the CMAP data? Are they merging observation and model data? Please provide more information here.
6. Line 131-132: I do not quite understand the sentence here.
7. In Table 1, please spell out the lw and sw.
8. Line 140: the average total precipitation increases to 4.63 mm/day and 4.33 mm/day, so is it 4.63 or 4.33?
9. Line 152: Are the eight AMNet sites shown in Figure 1? If so, can you use different symbols to differentiate them from the MDN sites?
10. Line 205: However, we find that this increase of resolution is finite. What do you mean here?
11. Line 254: whilst?
12. Line 156-158, 273-275: WRF-GC-Hg simulated Hg⁰ concentration is 1.61±0.20 ng m^-3, which does not quite agree with the GEOS-Chem and observation concentration. Do you know why WRF-GC-Hg simulated higher Hg⁰ concentration than both GEOS-Chem and the ground observation, even though WRF-GC-Hg simulated better Hg wet deposition? It is due to the atmospheric redox chemistry or something else?
13. Line 279: in this area, which area do you mean here?
14. Line 450-454: (Zhang et al., 2016a) and (Zhang et al., 2016b) are the same.
15. Figure 7, 9, 10 are hard to read.
16. In section 3.2-3.3: did you compare the model simulated precipitation vs precipitation measured at MDN sites?
17. This work only focuses on only one year (2013) study, what do you think of the interannual variability of the precipitation and Hg wet deposition?

This study presents a representative demonstration of GEOS-CHEM multi-phase, multi-species mercury atmospheric chemical transport algorithms in the WRF-GEOS-Chem model and addresses one extant scientific question through multi-scale regional simulations.  This novel tool is a major incremental advance in regional mercury modeling capabilities in scope for GMD and for scientific questions within the scope of EGU. This is also a first demonstration of the ease of transferring GEOS-CHEM algorithms to WRF-GC without the additional model development, porting, and QA/QC efforts usually required, another large incremental advance in regional atmospheric chemical transport modeling beyond the mercury application shown here. Thus, the title reflects the contents of the article as demonstration of model development. The overall presentation well-structured and clearly written, and ready for copy editing for fluency and precision. The original contributions are highlighted, and the number and quality of references are appropriate.

The manuscript nearly meets criteria for initial model demonstration, with a few important omissions. I recommend acceptance following minor revisions to evidence, context, methods and assumptions, and supplemental materials to support reproducibility:

1. Comparison of capabilities and performance to prior regional Hg modeling in the domain and elsewhere is a basic expectation, and notably absent. First, the authors must identify the differences and advantages of this model's atmospheric chemical transport mechanisms to other regional models that resolve Hg (e.g., CMAQ, CAMx, WRF-CHEM, STEM-Hg), cite them, and clearly communicate the value and novelty. A table comparing mechanism features would help. Performance comparison to regional scientific and regulatory modeling over the study domain is essential to communicating the applied value of this tool and would best be achieved by comparing to contemporary community model performance benchmarks (e.g., Emery et al., 2017) for the same season. Qualitative and quantitative summaries of the performance advantages of this regional model should appear in the abstract.
2. The multi-scale comparison presented requires additional evidence to support the conclusions and more information on data and assumptions. Fig. 3 must include results from all three scales. The article must describe the spatial and temporal surrogates and processes used to allocate emissions from each sector. The authors are strongly encouraged to present maps of total and sectoral emissions maps at each scale in the supplemental materials. In Fig. 4, panels 2 and 3 appear identical—this should be addressed to provide evidence of different resolutions, spatial patterns, and the differences in magnitude described in the manuscript. The discussion (206-213) should reflect the roles of the emissions inventory and downscaling as limiting factors in resolution for this study rather than an inherent process resolution issue below 50 km.
3. The arbitrary filter of 70% data availability for site inclusion should be justified, and other observational data QA/QC steps described.
4. The article and the GitHub and zenodo repositories include all files necessary and sufficient for reasonable reproducibility. Addition of a WPS pre-processor file for the domains and scripts for the analysis and figures presented would bring the study close to full reproducibility.

Reference:

Christopher Emery, Zhen Liu, Armistead G. Russell, M. Talat Odman, Greg Yarwood & Naresh Kumar (2017) Recommendations on statistics and benchmarks to assess photochemical model performance, Journal of the Air & Waste Management Association, 67:5, 582-598, DOI: 10.1080/10962247.2016.1265027