General comments

This paper describes model improvements when using a fully 3-dimensional hydrodynamic model within a regional climate model.  The authors describe the simulation improvements with the 3D model as compared with a 1D model, and explore the physical processes that lead to this improvement.  As the authors acknowledge, this coupling has been performed before and several other papers have highlighted the importance of improving deep lake representation.  The novelty in this paper is that they explore the physical reasons as to why these improvements occur.   The results in the latter half of the paper are interesting, and the authors do a thorough job explaining the physical processes underlying the model improvement.

My main comments are surrounding the paper organization, figure clarity, and being sure to accurately acknowledge prior work in this space.  These revisions are relatively minor, and I recommend publication with this minor changes.

Process description

• The process-level description that the authors are highlighting isn’t explained with equations until much later in the paper (5.2 for the heat transport, 5.3 for the vertical mixing). It would have helped if the authors had used a more traditional framework and described the important component models upfront (e.g., in Section 2.2) to make all the processes clear before getting to the results
• In the same way, there is one paragraph on the 1D model in Section 2.2 that is out of place. Given that the paper focuses on how the results are so different, more time to clarify the key difference of the model (in equations) would have set up the paper better.
• This would be helpful for later interpretation, e.g., Section 5.1 – line 523-4 states “Instead, only ice thermal dynamics are simulated as in the 1D lake model.” Something that describes this 1D process would be helpful for the reader.

Figure consolidation – Many of the figures have redundant information in them, e.g.,

• Figures 4/10/13 (vertical T profiles at the Spectacle Reef Site): Many duplicate panels in these three figures.  While I understand the intention to step through the different experiments, I often wanted to see these figures side by side.  I think these panels could be effectively combined to make one comprehensive figure.
• Figure 5/8/14 (spatial distribution of ice cover): Same as above – lots of redundant information on these figures, and it would help the reader to see some consolidation here.
• All figure panels could use some additional labeling on rows/columns, as they change from figure to figure (e.g., sometimes the different months are the rows, sometimes they are the columns). Also, many of the fonts and legends are *extremely* small and hard to read (e.g. Figures 9, 11, 12).  Some sublabels (e.g., labeling the panels a, b, c, etc.) would help to connect specific figures to the text.

References to prior work: The authors do acknowledge that some work has been done in this space before, but I don’t think that they have fully acknowledged all that has been done in the regional climate community, e.g.a few key ones that are missing include

1. 1. Leon et al. 2007 ELCOM in the Canadian regional model (CRCM)
   2. Turuncoglu et al. 2013 ROMS in the Regional climate model (RegCM)
   3. Bryan et al. 2015, showing the impacts of 1D lakes in RegCM

Minor editorial:

1. Line 35-36:  change to “…. this study identified the key processes influencing….”
2. Lines 153-160: This is a long list of possible obstacles followed by a long list of references. Could the authors parse this out more so there is more clarity in which studies were investigating specific components?
3. Line 179: LISSS, first use of acronym with no description.
4. Suggest to remove Table 1 as this information is all in the text
5. Figure 2 – why not include the 1D model on here as well for comparison?
6. Line 460 – can you show these observation locations on one of the spatial figures? Also, you note they are selected because of the highest ice coverage – what about observations on Lake Erie?  That usually tends to be the most ice covered.
7. Figure 7 – These line plots are very hard to read. Could it have the observations on one y axis, and the two model versions bias on a second y axis?
8. Line 574 – change “… GLSEA is not able to well capture….” to “…. GLSE cannot capture…”
9. Figure 11 – this figure is so hard to read, yet seems to be a very important one. Note that dT/dt is the black line in the caption.  It is very hard to see the difference between the black and the purple line, so perhaps dash one of them.
10. Lines 620-623: Can you comment more on the importance and implications of this issue?
11. Line 632: what specifically is meant by “sophisticated”?  At this point, most readers would want to know specifically which model is used (with a reference).
12. Figure 12 – there are six terms in equation 3, yet only five are shown in the plot. Is there a reason why the horizonal diffusion is not plotted?  If too small, then this should be included in the caption or the text.