The manuscript, “A model for the Artic mixed layer circulation under a summertime lead: Implications on the near-surface temperature maximum formation” by Alvarez develops a simplified 2D model of a summertime Arctic lead using SHEBA data to examine the sensitivity of the near surface temperature maximum to changes in solar radiation, winds, and ice motion.

Global climate models have biased salinity and temperature distributions in the upper Arctic Ocean. It has been suggested this could be associated with resolution or parameterizations of vertical mixing (Rosenblum et al. 2020). Future changes in the upper ocean are also very uncertain as the changing sea ice state may modulate ocean circulation. The manuscript touches on an important issue of simulating processes controlling the structure of the upper Arctic Ocean but requires a clearer discussion of the novel conclusions that separate this study from previous studies of the summertime NSTM and to be put in a broader context in terms of the Arctic as a whole and other more complex models. I have listed these concerns below. My recommendation is major revisions.

1. Previous studies, such as Richter-Menge et al. (2001), Steele et al. (2011), and Gallaher et al. (2017) have recognized the importance of solar radiation, winds, and sea ice motion in the formation and persistence of the NSTM layer beyond summer. This is mentioned in the discussion, but the results seem to confirm previous findings, not add very much new. Could you expand on the novel results and the benefits of this model within the hierarchy of the models from the other studies?
2. While there was good discussion about the caveats and limitations of this modeling study, I had a question about a couple more. For example, the model developed here does not include important aspects such as large-scale ocean circulations and only has a very simple representation of freshwater input, which have both been highlighted in the previous studies as important for the NSTM. The model was also developed based on a very short time period from SHEBA and the Arctic has changed considerably since then (Dewey et al., 2018). Does the author expect the conclusions would change with a more recent case? Although it was located in different region of the Arctic, what about with MOSAIC?
3. There are instances throughout the manuscript that require some proofreading. For example, in the paragraph beginning on Line 194, the first and second sentences are missing “are” and “is”, respectively and in the last sentence computing should be changed to computing. Another example is in the use of minimum and maximum throughout the manuscript. However, I felt it was well constructed, as I appreciated that the methods section describing the model was easy to follow and the approaches taken are appropriate.

Dewey, S., Morison, J., Kwok, R., Dickinson, S., Morison, D., & Andersen, R. (2018). Arctic ice‐ocean coupling and gyre equilibration observed with remote sensing. Geophysical Research Letters, 45(3), 1499-1508.

Rosenblum, E., Fajber, R., Stroeve, J. C., Gille, S. T., Tremblay, L. B., & Carmack, E. C. (2021). Surface salinity under transitioning ice cover in the Canada Basin: Climate model biases linked to vertical distribution of fresh water. Geophysical Research Letters, 48(21), e2021GL094739.

This is very interesting work, but I think it's missing a few kew elements to make it significant - I am not convinced that it demonstrate the importance of both the calm and windy periods, as is claimed in the abstract and summary. 

My main question is about the run setup, where there is a calm period (5 days) before the wind is turned on. Why is that? Once the wind is on, even for the smallest wind speed, a 3 m/s wind with a 2% wind factor, would result in a displacement of 2 km per day for the ice. These add up to reasonably large distances over 5 days (particularly for the larger wind experiments), so the periodic nature of the experiment gets a little strange. The ocean field (for the wind forcing cases) is the results of local (calm) initial conditions and a lot of leads that occasionally input more solar radiation in the ocean. Does it make sense to compare only temperature profiles, if we are not sure if the total heat input has been the same? 

Maybe I’m missing something… Is the grid (Fig 3) changing as the “Ice sheets are assumed to move westward with the wind”?

The results (and modeling setup) make sense to me for the calm period, but it gets confusing when discussing the wind simulation. In particular, the impact of the initial conditions for the 3 m/s case doesn't make sense to me, since this should be the same in all cases - the lead that created the initial warm surface water (Fig 4b) is 10 km away in the 3 m/s case (Fig. 10a), and 20 km away in Fig 7b. Why would only the 3 m/s case show asymmetry (L300)?

It might make more sense to 'close the leads' at the beginning of the wind period, to decouple these two processes (adding more heat or freshwater at the surface), and adding mixing. Ultimately I'm not entirely sure what the simulations demonstrate as they are setup now. 

Some of the big questions that this study could really help answer are:

- Do we need calm conditions (ice not moving and lead being open for a long time) to get enough heat in the surface layer to get an NSTM?

- Do we need (large?) wind events to mix this water to depths comparable to what we observe?

- how quickly does the 'patchiness' of the formation (leads) disappear? If one was to increase the separation between leads, would that make the NSTM more patchy?

I believe that some of the results show here help with these questions, but I would like to see a better justification of the setup. 

The summary states that "sequence of calm and windy periods in the leads results in a final thermal structure characterized by a spatially distributed NSTM layer" - I'm not sure that this was convincingly demonstrated. Do you *not* get is without the calm periods? A few different scenarios show various levels of patchiness, but the periodic nature of the simulations (when ice is advected) makes it a bit confusing... 

L7: I understand that the focus of the paper is on early summer time, but it might be worth discussing a bit how it connects to winter and seasonality. In particular, while the second sentence of the abstract is probably correct, the heat exchanges are even stronger in the winter. I would probably delete this sentence.

L229: It is a little bit confusing how how is the ‘steady wind period’ (L226). From Fig 8, it seems that the wind is turned on at the beginning of “day 5” (the sixth day?)

L232: Is Fig. 7 a one-day average, or a snapshot?  What does “the circulation on the fifth day is considered” mean? A time average? The next sentence seems to indicate that it is a snapshot in time, but that wasn’t clear. Is the circulation (streamlines) calculated from a day-long average?

L255: ‘Reference profiles’ and “Initial profiles” are the same and without lateral variability. They should be called the same (it is a bit confusing here what the gray lines were, and if they changed from panels to panels). 

L277: It might be useful to add what day “the end of the simulation period” corresponds to (day 10?), to facilitate interpreting Fig 8 in the context of Fig 7 (and others). 

L 295: The salinity and velocity fields are ’considered’ since temperature is effectively a tracer. Rephrase: “Discussion of the results focuses on the impact on the thermal field.”

L300: Specify that the local temperature is warmer, and shallower in this scenario. Is the profile

L304: The ‘control run’ has also a depth of 28m, and temperatures seems similar (Fig 7b and Fig 9). This should be captured in the text. Ultimately I think that is the main point here - doubling the FW meltrate doesn’t impact the NSTM. Why? It seems that the sensitivity is mostly on using smaller melt rates… More melt doesn’t change anything (already isolating the temperature maximum), but less has a large impact… Is the depth-integrated heat content the same?

The last sentence of the abstract states that 'ice drift is key in the development of the NSTM layer'. Closing the lead would do the same, no?