General comment:

My major concerns pertain to the clarity and novelty of the finding, rigour of the analysis and quality (and quantity) of illustrations.

As I understand it, the core message of the manuscript is that the seasonal cycle of speed (and shear) in the upper SAO is tightly linked to sea ice and has strengthened. The underlaying drivers are named (increased stratification in summer and increased ventilation during winter) but not shown in the manuscript (I understand that these are earlier findings by the same author(s)). Instead the focus appears to be on arguing that wind-driven near-inertial energy is distributed within the SML. This is in accordance with the over 50 year old concept of slab models, which the authors mention in passing but do not really incorporate in their analysis. Despite of the host of data presented, their argument remains largely qualitative and I find it difficult to decide how novel the insights actually are.

Specific comments:

The only indication for a systematic change in seasonal cycles of currents, shear and sea ice is in the qualitative comparison of the 6-year average 2004-2010 with the 10-year average 2013-2023. In the time series presented, there is no obvious systematic change.

The method introduced to separate wind-driven inertial currents from tides is not sufficiently explained, nor tested (but I also don't think it is necessary to attempt this separation in the first place).

There are far too many figures (over 100 panels in total) that are unfocused, hard to read and contain sometimes odd, unexplained choices. Most of them are hardly referred to and many do not clearly support the reasoning in the text.

These are my short notes:

-Figures 1,2,3,4,6 all show seasonal cycle of upper ocean currents and shear in relation to sea ice. It should be possible to make this point with 2, max. 3 figures.

-Figures 7-10 basically do the same for the 2014-2023 period.

-Figure 11 has no clear message; the increase of shear at the bottom of the SML is not really visible in most panels.

-Figure 13 shows much of the same information as figure 12, but apparently with an inexplicable 2018-2021 time average.

-Figure 14: what is the physical motivation for the averaging times in c? d is the same as the top right panel of figure 2 and e is for some reason only given for a depth range from 60m, even though a and b go up to about 45m, which would be a very interesting depth for buoyancy frequency.

The list of references is not complete.

Suggestions for revision:

- Focus the manuscript on the core message of changing seasonal cycles from 2004 to 2023 (e.g. by showing full time series and deriving useful metrics).

- The point that there is little difference between all the MB moorings can be made much more succinctly.  

- Remove many of the figures (or at least put them into supplementary), revise and focus the remaining figures to directly and clearly support the argument.

- Perhaps add quantitative estimates using slab-models or at least incorporate the theory (and related findings) properly into the introduction.

- Decisions should be clearly motivated and transparent (why (and how) to use a 2-day window to separate tides from wind-driven currents, why the split at 2004-2010/2013-2023, why show 2018-2021 averages (fig. 13), why not show buoyancy in the 45-60m range (fig. 14)).

Line-by-line:

23: Either expand on the "logistical challenges" or remove the sentence.

28: How can a cycle follow a coherence?

29: Can the absence of ice "drive" something?

54: Missing a "the".

54-56: I think there should be references.

56-58: Same here. 

60-61: No brackets necessary.

62-65: If this is to suggest a connection between inertial currents and Arctic Ocean circulation it should be explained here.

69-71: References should be provided for this statement.

94: Is this no problem for the magnetic compasses in the ADCPs?

95: Odd title for the section.

123: I don't understand how this method could separates tides from wind-driven near-inertial oscillations. Where does the 2-day window come from? How is it implemented? What is the sensitivity? The explanation is lacking and the figure S1 does not provide useful information. In the context of this story, I think the authors could just proceed with NIC that contain wind driven and tidally driven currents (i.e. just do a 10-14h bandpass).

153: Specify which moorings are considered to be on the shelf.

157: This is only (ever so slightly) visible for MB9.

159-161: Sea-ice also provides a friction barrier for tidal currents.

192: The error bars appear to be exceedingly small; are they correct?

203: Velocities in figure 2 mostly do not look surface intensified.

221: Connection instead of connectivity?

223: In the figure it is not possible to see if it is before, at the same time or after.

225: How is the significance calculated?

227: This does not appear to be the case for MB6.

229: The differences often appear to be negligibly small.

234: Figure 7: Figure title talks of M1, should be MB1? Why split at 2010?

242: Wind speed is not the only meaningful variable; what about the wind direction or wind spectra?

247: In line 244 data from 2004-2023 is mentioned; why is only 2013-2023 considered here?

254: Why is the data from 2004-2009 not shown?

286: No obvious consistent change is visible in figure 10.

314: This is hardly visible (if at all) in most panels of figure 11.

319: Should be Figure 13a-d.

322: Reference formatting.

320-339: Isn't this the premise of all slab-models since about 1970?

334-336: How are near-inertial wind signals "transformed" to non-inertial residual currents?

346: Why not show the full data set?

352: Where does the 2018-2021 average come from? 

362-363: This is very confusing. Where am I supposed to see stratification decrease in winter? In figure 14c, buoyancy frequency tends to be highest in winter. Neither can I see greater currents or late winter shear in figure 14d. 

365: a and b are never discussed, it is unclear how the averaging periods in c are defined; why is buoyancy only shown from 60m, when T and S go up to 45m? 

395: This is not visible in figures 9 and 10.

402: Where and how is this shown?

410-412: This is not shown.

421-423: Is this new?

This manuscript investigates how seasonal shifts in sea-ice concentration and vertical density stratification regulate current speed and shear across the Laptev and East Siberian seas.  The authors present an extensive mooring data set and deliver a largely qualitative interpretation of upper-ocean variability, highlighting the summertime prominence of near-inertial currents.  Although the analysis is descriptive rather than quantitative, the topic is important and the observations are valuable for understanding ongoing Atlantification of the Siberian Arctic.

Several points require attention.  Methodology needs greater transparency: lines 76–94 and 95–99 list instruments and programs, yet direct links or citations to the underlying mooring, sea-ice and wind datasets appear only later in the Data-Availability statement, which can be confusing.  The ERA5 winds should be referenced in full (e.g. Hersbach et al. 2023, DOI 10.24381/cds.adbb2d47) rather than cited only by a portal link.

The procedure described in lines 119–133 is understandable in outline, but additional detail is essential.  There are no citations to the earlier works of Pnyushkov & Polyakov (2012) or Baumann et al. (2020, 2022); one must therefore assume previous approaches relied on Fourier transforms, whereas the present study may be using a wavelet-style filter to isolate the 12–14 h signal band.  It is unclear what the declared “2-day window for NIC detection” actually means, given that NICs are already obtained through band-pass filtering.  The 2012 paper referenced does not discuss spectral or frequency analysis, only time- and space-normalised diagnostics, so perhaps the authors intended the sliding-window harmonic analysis introduced in the Arctic Tidal Current Atlas by Till M. Baumann.  The manuscript should explain why a 2-day window was selected, how window width influences the results, and how sensitive the analysis is to this choice.

A number of references cited in the main text do not appear in the bibliography, for example More and Polyakov 2025, as well as the methodological papers mentioned above.  Completing the reference list will help readers trace the provenance of data and techniques.

Finally, the presentation could be streamlined.  The manuscript contains many multi-panel figures, some of which repeat similar seasonal or depth–time information.  Reducing the total number, improving resolution, and tightening captions would strengthen the flow and keep focus on the principal findings.