Multi-Sensor Observations Reveal Large-Amplitude Nonlinear Internal Waves in the Kara Gates, Arctic Ocean

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Igor et al. used multiple observation methods to reveal the large-amplitude nonlinear internal waves in the Kara gates. Basically, the reaserch is valuable. The manuscript is well written and give convincing and practical results. I have only some minor concerns listed below: 

 

1, How did the authors choose these 5 stations? Are they representative of the Kara Gates? Or are these locations more prone to the formation of NLIW? Some explanations can be given in the "Materials and Methods" section.

 

2, Has the observation data undergone quality control？

 

3, The manuscript only show the comparison between theoretical and observed phase speed for 3915 station. How is the situation for other stations. Besides, I think some discussions can be conducted about why the shallow water KdV model and that for strongly nonlinear waves agree best with the observation.

 

Comments on the Quality of English Language

None

Author Response

Dear Reviewer!

We are very thankful for the positive evaluation of our work and provided comments. In the revised paper, we have attempted to address your comments. Below are our replies to each of the comments made. Yet, we had a difficulty to fully reply to the last comment, and would be thankful if you can share your ideas on the question raised (re. why the shallow water KdV model and that for strongly nonlinear waves agree best with the observation).

Below are replies are given in italic.

1. How did the authors choose these 5 stations? Are they representative of the Kara Gates? Or are these locations more prone to the formation of NLIW? Some explanations can be given in the "Materials and Methods" section.

These stations were selected prior to the beginning of field works based on analysis of historical (e.g., Kozlov et al., 2015; Morozov et al., 2017 – these papers are mentioned in the text) and most current (e.g., Sentinel-2 image made 4 days prior to arriving to the KG, Fig. 5) satellite data, as well as prior in situ measurements and model results available for the region (Morozov et al., 2008; 2017 - also cited in the text). In these data, IW trains spreading along the strait were nearly always present and characterized by strongest radar contrast in SAR images (e.g., Fig. 9 in Morozov et al., 2017). Though the IW field in the strait could be rather complex (having multiple generation sites and propagation directions) due to complex and still not well resolved bottom topography, the waves propagating along the strait seem to be the most energetic due to much higher flow speed of the along-strait currents. In this sense, we believe that measurement locations were right in order to capture along-strait propagating IWs. We have added these explanations in brief to the revised paper.

2. Has the observation data undergone quality control？

Yes, surely. We did an intercomparison of water temperature records made by thermoprofilers (the key instruments used for the IW analysis) between each other (TPArctic vs Starmon Mini StarOddi) and with CTD profiles obtained from Sea-Bird Electronics 19 plus and AML BaseX profilers. The quality of the data was quite good, without any large biases from the precision values provided by the manufacturers. The data of navigational X-band radar JMA-9122-6XA was also calibrated prior to mapping.

3. The manuscript only shows the comparison between theoretical and observed phase speed for 3915 station. How is the situation for other stations. Besides, I think some discussions can be conducted about why the shallow water KdV model and that for strongly nonlinear waves agree best with the observation.

Yes, this is correct. We did so for several reasons: i) this is the most interesting station showing high-amplitude NLIWs, ii) it has a clear 2-layer structure so that a simplified 2-layer model could be applied, iii) it was interesting to apply a commonly used KdV-type model to check its applicability for the case of large-amplitude NLIWs, iv) adding the same analysis also for station 3913 (where 2-layer model could be also used) would extend the text even more (other reviewers strongly advised us to shorten the paper), but what is more important, only 2 sources of vertical profiling were available at 3913 (TPArctic and AML profiler), hence, not allowing to infer the true propagation speed and direction of IWs.

Regarding the second point, some previous works confirmed that the KdV model is a simple and attractive option working well in many realistic cases (Ostrovsky & Stepanyants, 2005). The results of scale analysis presented indicate that the relevant dimensionless parameters in our study area satisfy the conditions required by the shallow water KdV theory. Since the KdV model is an asymptotic model, it can provide a description of the wave field at some distance from the initial perturbation, i.e., at least at two or three wavelengths. In our case, high-amplitude NLIWs were observed at about 10 km from the presumed generation site, which clearly satisfies the above condition. However, we can’t suggest why the KdV model and that for strongly nonlinear waves agree best with the observations. We would appreciate much if the reviewer would suggest any relevant thoughts on the subject.

Reviewer 2 Report

Comments and Suggestions for Authors

This paper is a useful contribution to improved understandings of internal waves at high latitudes. The measurement program was well designed, and it provided sufficient information to provide some analyses of the dynamics of internal waves propagating into the Kara Sea as generated in Kara Gate.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The English Language content in the paper is generally good, but there are several grammatical errors. I include some of these in my detailed comments.

Author Response

Dear Reviewer!

We are very thankful for the positive evaluation of our work and very thorough review providing detailed comments and suggestions to improve the paper. This was really helpful, thank you for your time and efforts! In the revised version, we have attempted to address all your comments, and improve the quality of the paper as much as possible, including the shortening of the relevant parts of the paper. Below are our detailed replies to each of the comments made (given in italic).

Replies:

Section 2. Materials and Methods. Provided a thorough and detailed overview of the measurement program. The description of the UAV observations provided in lines 120-122 should be expanded to describe the camera that was used and how the imagery from the camera was interpreted.

Thank you for this notion. We have expanded the text. Now it reads:

“In order to record surface manifestations of NLIWs commonly observed in the form of series of coherent curved elongated slicks or rip bands, and correct the position of the ves-sel, photo and video recording was made using a DJI Mavic 2 Pro UAV from heights of 100 to 500 m at all stations where visibility conditions were favorable. The DJI Mavic 2 Pro was equipped with the Hasselblad L1D-20c visible band camera with image size of 5472×3648 covering an area of about 800×500 m when looking in nadir from 500 m height with spatial resolution of about 0.3-0.5 m. Identification of NLIWs’ signatures was based on visual inspection of UAV records made at stations.“

We didn’t add more details on UAV data processing because we didn’t extract any quantitative NLIW properties from the UAV data.

Section 3. Results. 3.1 Background conditions in Kara Gates.

The wind conditions described the wind speed but did not mention the wind directions (lines 165-166); please describe the prevailing wind direction.

Done. We have updated the text and the figure 2. Now the text reads:

“During the field works in the KG wind conditions were characterized by low to moderate winds of varying direction with mean velocity of about 5 m/s during the first three stations, rising gradually to about 7 m/s toward the end of observations (Figure 2). Southerly and southwesterly winds prevailed during the first two stations periodically changing to northeasterly for the rest of the record.”

The surface currents presented for Kara Gates (lines 190-193 and Figure 3c) were derived from thermal infrared satellite imagery measurements on Aug. 9, 3 days prior to the ship-based in-situ measurements (lines 185-186). The applicability of the Aug. 9 satellite-imagery derived surface currents to those prevailing over three days later during ship-based measurements should be discussed, in terms of how similar the forcing conditions were for surface currents (e.g. surface winds and ocean upper layer water properties) during these two different measurement times.

Thank you for this important notion! We have checked the wind conditions during Aug 9-11 using data of ASCAT scatterometers (https://manati.star.nesdis.noaa.gov/datasets/ASCATData.php). On August 9, the westerly and northwesterly winds of 5-8 m/s were prevailing. On Aug 10-11, the wind direction changed to southwesterly and the mean wind speed decreased to 2-5 m/s. Starting from Aug 12, it raised again to 4-7 m/s. So, in general, the near-surface winds on Aug 12 were about 20-30% weaker compared to Aug 9, but the wind direction (S and SW) was more favorable for the surface current to be directed aligned with the background current from the Barents to the Kara Sea. Some of this information is now included to the paper, while ASCAT winds on 9-12 August are shown in the Supplementary Information.

The presentation of the tidal conditions in lines 209-215 and Figure 4 should be supported by a literature reference to the Arc5km2018 tidal model.

Thank you for this notion! We missed it here because it was already introduced in Data and Methods. We have added the reference now.

This statement, that the total horizontal currents in the strait may be of the order of 1 m/s, is speculative and it is not supported by physical reasoning other than the tidal currents which are much lower than 1 m/s. Please provide a justification or rationale for this statement.

Thank you for this important comment. Well, in fact, this is not a speculation. Thanks to your question, we have double checked the literature and found that Morozov et al. (Acta Oceanologica, 2003) also mentioned the possibility for the surface current speed to reach about 0.8 m/s. Below we give a citation from page 234 of Morozov et al. (2003), where they describe the current regime in the strait based on ADCP measurements made at 3 moorings. Their mooring # 2 was the closest to our ship track, but the measurement depth was 90 m, and the maximal current speed was equal to 0.35 m/s. Yet, more interesting data comes from the nearby mooring #3. We cite:

“The upper instrument at 65 m recorded the highest northeasterly currents with velocities exceeding 50 cm s–1. The mean velocity was equal to 26.5 cm s–1. The currents calculated from the density profile across the strait were close to 80 cm s–1 at the surface.”

So, though it was not obtained directly from ADCP data, they report a similar velocity value for surface currents as was obtained from our satellite data. We believe that such velocity values are perhaps anomalously high, but could still exist under certain (e.g., southerly) wind conditions superposed with the background NE current and enhanced spring tides. Lastly, the satellite-based velocity map in Fig. 3c seem rather realistic because i) the mean current speed over the entire strait is about 0.5 m/s, which agrees with earlier observations [Morozov et al., 2003; 2008], ii) the region of enhanced surface currents has an arc-shape very similar to the arc-shape of the central jet of the warm Barents Sea current in Fig. 3a, suggesting that the warm current could be topographically trapped and intensified over the underwater channel found in the strait. The latter is not seen in the standard bathymetry data, but well present in the navigational maps, and also has an arc-shape.

Part of these explanations is now added to the text, while we also corrected the value for the maximal current speeds in the KG in the Introduction.

Reference:

Morozov, E. G.; Parrilla-Barrera, G.; Velarde, M.G.; Scherbinin, A.D. The Straits of Gibraltar and Kara Gates: A comparison of internal tides. Oceanol. Acta 2003, 26, 231– 241. https://doi.org/10.1016/S0399-1784(03)00023-9

The surface signatures of NLIWs in the satellite data (lines 246-263 and Figure 5a) were obtained for August 8, 2021, four days prior to the ship-based in-situ measurements; why would these be applicable to surface conditions four days later?

This is a good question. First, we have processed a lot of SAR data for this region covering summer months of 2007, 2011, 2015, 2021, 2022, 2023 (which is a subject of another ongoing study, while some results are already published in Kozlov et al., [2015]; Morozov et al., [2017]) – most of them clearly show systems of NLIWs propagating southwest and northeast in the study region. Here we presume that NLIWs are constantly generated over the same topographic obstacles by background currents most intense in the along-strait direction (SW to NE), hence, generated waves should be also oriented primarily in this direction. Part of this explanation is now added to the text.

The SAR snapshots obtained on Aug. 12 and Aug. 13 (Figure 6) were interpreted to resolve “elongated slicks” (lines 268-269) indicative of surface circulation but they do not reveal signatures of NLIWs. How are these SAR observations relevant to this paper?

Thanks for the notion. We agree and have deleted Fig. 6.

Section 3.3.2. Vertical thermohaline measurements at stations. This sub-section provides the major findings in this paper, i.e., those derived from the ship-based in-situ measurements on Aug. 12-13, 2021.

In lines 247-248, it is stated that “Satellite data covering the study region were operationally communicated to the expedition team for planning the locations of stations” This is also mentioned in lines 99-100. It is not clear how the satellite data of surface conditions (shown in Figure 5 from Aug. 8?) were interpreted to provide information that supported the selection of the stations. Please explain this in the text.

We have partially answered this question above. The following explanation was added to the text: “We presume that NLIWs are constantly generated over the same topographic obstacles by background currents most intense in the along-strait direction (SW to NE). Hence, the generated waves should be i) oriented primarily in this direction and ii) found in close proximity to the surface signatures of NLIWs seen in Figure 5. The locations of stations were selected so that they would begin south of underwater hills and cross the NLIW trains seen in Figure 5 while the ship freely drifted NE.” We mean that receiving an actual satellite data with distinct manifestations of NLIWs helped us to finalize the exact locations where to start the drift.

In Lines 278-280, signatures of NLIW apparently were observed at four stations, but not at station 3917. Was there no NLIW activity present at the last of the five stations, or was the NLIW activity smaller and less evident?

The measurements at station #3917 showed a pronounced 12-m high solitary depression in the middle of the record. First, we presumed that this solitary pattern could be associated with one of the surface current convergence zones widely present in the strait and seen in Sentinel-1 SAR images (earlier Figure 6, now deleted). However, the true origin of this feature remained unclear. Moreover, we also had to pull out TPs at this station due to repositioning of the ship, making the quality of the entire record somewhat questionable. Hence, we decided not to show this record in the paper. We have added this subsection to the supplementary material.

In Lines 292-293, the interpretation of Figure 7a is that stratification and thermocline was confined to depths of 25-40 m. The CTD sampling was limited to depths less than 45 m, so how is it known that stratification and the thermocline were not present at depths greater than 45 m?

We made full depth CTD profiles at all stations (see Supplementary material), but showed them only within depth limits of TP measurements for better intercomparison. If the Reviewer finds it necessary, we can plot all of them in one figure, but this would result in further extension of the text and paper length. We have added the information that CTD profiles were full-depth profiles to the Data and Methods section.

The word “period” is used through this sub-section (e.g. lines 322-323, lines 427-428) but most of the NLIW events seen in the temperature profiles were not clearly periodic in nature. Is “period” the right word here? The described NLIWs appear to be a set of discrete events of vertical displacement of isotherms.

Well, this is correct, but the term “period” is clear to everyone as something meaning the length in time of particular (even discrete) oscillation. Surely, we can use some other word, but, we guess, this would cause even more questions from the readers than the use of “period”. So, if changing, would “time length” suite for this purpose? In this case, we would better keep the initial wording.

The surface slicks described in lines 328-330 are not clearly seen in the visual imagery of the UAV in Figure 9, especially for the western portion of the “slicks”.

We have added the contrast to the figure and made it larger to make slick boundaries more visible. It seems to become much clearer now. Please also note that you may deal with a compressed version of the figure made after the paper submission to the system. The original figure (to be used in final version) should have a better quality and visibility.

References to Figure 9a,b in lines 346-353 appear to be incorrect. The references should be to Figure 10.

Thank you for noting this typo, we have corrected it. In the revised paper Fig 10 is now Fig 9 (as we deleted former Fig 6).

“From the latter, it follows that the surface manifestation spread in the NE direction at a very low speed of ~0.05 m/s.” (lines 359-360): What is the uncertainty in these speed values? The very small difference may be highly uncertain.

Yes, indeed, the given value is close to the uncertainty threshold, hence we simply corrected it so: “From the latter, it follows that the surface manifestation is also nearly stationary.”

Line 389: “one can see several oscillations of 10-15 cph…” What is this statement based on, given that Figure 11b does not show the buoyancy frequency?

Yes, this is correct, we have deleted mentioning the cph in the text.

Lines 396-398: This result emphasizes that the use of thermal profiling in the strait is quite justified, since the vertical water stratification significantly depends on temperature in this Arctic region.” This statement need to be reworded. It suggests that the change in density (determined from CTD data) is due only to changes in temperature (from separate thermal profiling). In fact, there may have been similar changes in salinity (not measured) which reinforced the temperature-driven changes in density.

Thank you for this notion, we have reworded it using the suggested phrase in the following way: “The temperature and salinity variations were likely closely coupled in this Arctic region which allowed the use of temperature as an indicator of the vertical density structure.”

Line 413 “…followed by about 5-m thick pycnocline…” The pycnocline appears to be about 10-m thick (15 to 25 m) based on the salinity profile, rather than 5-m.

Thank you, we agree, this is corrected now.

Line 470: “the difference” The difference in what? Thickness?

Yes, we missed the word ‘thickness’, thank you!

Line 472: “close to half-difference between thicknesses of the layers…” The upper layer is 15-20 m thick while the lower layer is about 100 m thick. The half-difference is 80 m vs. 50 m stated here?

Well, if we consider the upper layer is 15 m thick and the lower layer is 110 m thick (from 30 m down to 140 m), the half-difference would be (110-15)/2 = 47.5 m or simply ~50 m. Correct?

Lines 517-526: In Table 2, the values of C are considerably higher than those of co. The wording in lines 518-519 suggests that C is the theoretical phase speed, but this is not consistent with the statement that “theoretical values are somewhat lower than the observed phase speed”. Please clarify.

Perhaps the reason of this is an incorrect table caption. The corrected caption is “Observed () and theoretical values of linear () and NLIW () phase speeds at station 3915”. As you see, C values are higher than those of c0 because the latter (c0) are theoretical phase speeds of linear waves, while C – the same for NLIWs (calculated using c0). The observed speed is a constant single value of 0.86 m/s (Cobs). As you see most theoretical C values (starting from third column and further to the right) are smaller than 0.86 m/s apart of one predicted by KdV model (=0.87-0.98 m/s). So, there is full consistency in the statement “theoretical values are somewhat lower than the observed phase speed”. The problem is (as seems to us) that you thought that c0 is the observed value, but it is not, c0 is just the theoretical phase speed of linear waves used to calculate C (phase speed of NLIWs). We show c0 values in the Table 2 because we further use them for the Froude number calculations below.

Given the comment above, and the very extended set of computations presented in lines 462-526, it would seem that this portion of the paper did not directly and explicitly lead to any conclusions which provide insights into the characteristics of the NLIW beyond what was derived from the in-situ observations prior to line 462. This section of the paper does not appear to be used in Section 5. Conclusions. This section of the paper should be reduced in size or simply not included in the paper.

We have explained, that the comment above was raised by not very clear Table caption, but the text was consistent with the results. Yet, we have improved it now. We have also shortened this part of paper where it was possible, trying to keep the clarity of the approach and calculations. We also believe that it is a very useful part of the work (see some reasoning below), so we’d like to keep it. Other reviewer, instead, asked us to give some more details for this part (more references, etc.). Moreover, this section is actually used in the Conclusion (prelast paragraph, Lines 660-662 in the initial paper version).

Though this part doesn’t give new results about the NLIW properties in the study region, it importantly shows that a simplified and easy-to-use two-layer approach is working to predict the phase speed even of large-amplitude NLIWs in the Arctic Ocean. To our knowledge, there are no similar works dealing with Arctic NLIWs (that have some certain peculiar features of generation) in the same manner. Many theoretical papers attempt to establish strict formal boundaries of various models of weakly/fully/strongly nonlinear waves. But here we go from the practical side – we observe high-amplitude waves and check whether a simple model can adequately describe their phase speeds. The obtained result that the use of simple theory is a working option at least for this case could be important for some other applications, e.g., when deriving NLIW phase speeds directly from sequential satellite observations (one of our ongoing tasks) and checking their theoretical phase speeds from available hydrography.

Line 541-542: “As a coarse estimate, we presume that the latter doesn’t exceed 2% (using analogy to Nansen–Ekman ice-drift law [67])” This is a very rough estimate. Is there any data available from ship handling studies or analyses to substantiate this?

Yes, we completely agree, this is a very rough estimate. We talked with experienced colleagues trying to find a better solution, as well as looked for literature that would describe more precisely a wind-induced ship drift component – it didn’t help much. But what is important here, even if we take 3% or more, supercritical conditions (Fr>1) would be still fulfilled and the result wouldn’t change. If the Reviewer knows a better solution, we would appreciate to use it here.

Section 4. Discussion.

This section is generally informative, especially Figure 15. However, the section should be reduced in length and it should avoid speculative statements that cannot be tested with the results derived in this paper.

Thank you! We have deleted all irrelevant statements.

Lines 600-615: As these two paragraphs are largely speculative, they should be reduced to a shorter single paragraph

Done.

Section 5. Conclusions

Much of this section is repetitive and it should be reduced in length to concentrate on the major findings of the study.

Done, we have shortened it substantially, but please note that this is not an abstract and allows to put things in a bit broader context.

Lines 644-646: How were the sea surface state conditions obtained from the satellite data and how was this information used to inform a sampling strategy for NLIWs? (See previous comments on this topic).

This was explained above, but for simplicity we have deleted this statement.

Lines 647-648: Salinity likely had an important role in determining density, as is the usually the case in Arctic waters. However, the temperature and salinity variations were likely closely coupled which allows the use of temperature as an indicator of the vertical density structure.

This is corrected now.

Lines 664-665: If the ship-based Acoustic Doppler Current Profiler had been working, these measurements would have provided simultaneous earth-referenced currents that coincide with the thermal profiler and CTD data sets used to characterize the NLIWs. It is not clear that fixed point mooring velocity measurements would have been required.

Yes, we agree and have deleted this remark.

Reference section. This section includes 71 papers and reports, which is a very large number for a non-review paper. This list of references should be reduced to the essential papers/reports by providing one reference to a statement instead of 2 or more references as is seen several times in the paper.

We have significantly shortened the Reference section down to 50 references, keeping only the most relevant to the analysis.

Detailed Comments

Line 16 clarify wording of “…from four stations evidence …” to “from four stations providing evidence…”

We have corrected it to “Analysis of the field data during an 18-hour period from four stations provides evidence that…”.

Line 19: clarify wording from “Triangle-shaped measurements of three thermal profilers…” to “Triangle-shaped measurements using three thermal profilers…”

Thank you, corrected!

Line 57: The acronym IW (Internal Waves?) should be defined here on first usage.

Corrected, we have defined it earlier in Line 55.

Line 134: Reference [34] is a key reference to presenting technical information on the most important in-situ data source, the “thermal profiler (TP) TPArctic”. This reference is incorrect because the link provided is no longer valid. An alternate reference to the paper was found: http://ecological-safety.ru/en/repository/issues/2023/01/11/

Yes, we see the problem with doi number and will contact the journal to solve the issue. Instead, we have added the link you have provided. Thank you!

Line 256: The “white rectangle” in Figure 5a is not readily apparent.

We changed the wording to “white box”. But the box in the center of Fig 5a is well seen, isn’t it? It is well present in our paper version.

Line 290: Improve wording: “at the beginning of station” change to “at the beginning of sampling”

Corrected.

Line 327: Missing word: At this station, a UAV survey was made…”

Corrected.

 

Line 378: Change “raised” to “increased”

Corrected.

Line 475: The meaning of the wording “simple use” is not clear.

We have deleted this phrase.

Line 573: “Kara Gates” is misspelled.

Thank you, corrected.

Line 617-618: Improve the clarity of the wording: “could be still enough energetic” to “still be energetic enough”

This paragraph is already deleted to shorten the section.

Line 621: poor wording, change: “Our observations evidence…” to “Our observations provide evidence …”

Thank you, corrected accordingly.

Line 629: “Reword: “might also be forming”

Thank you, corrected.

Thank you again for very detailed comments!

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Manuscript entitled “ Multi-Sensor Observations Reveal Large-Amplitude Nonlinear Internal Waves in the Kara Gates, Arctic Ocean" shows that the large amplitude nonlinear internal waves in the Arctic Ocean Karamen can be observed by using multiple sensors. In general, the analysis of this paper is meaningful and promising. It may be suitable for publication in Remote Sensing after some detailed revisions. The comments are as follows:

 

1.      Equations in general seems to come from nowhere. Are there other works that use these formulae? If yes, please cite them. If no, please explain better the formulae.

2.      Part of the subtitles in the third section are not numbered, which makes the content of the manuscript confusing. It is suggested to revise and perfect the small paper to solve this problem.

3.      The data and text in Table 2 are not aligned in the center. It is suggested to modify this part to make the manuscript more professional.

4.      The author marked the position of the ship in Figure 6b, but there may be other factors interfering with the location of the mark is inaccurate, and it is suggested that another means be used to verify the feasibility of the existence of the ship.

5.      This presented work is technically correct. However, the scientific findings are relatively weak. Therefore, it is recommended to highlight or add some scientific findings.

6.      In general, the author does not fully summarize the innovation of the manuscript in the conclusion, and it is suggested to further highlight this content.

Author Response

Dear Reviewer!

We are very thankful for the positive evaluation of our work and provided comments. In the revised paper, we have attempted to address all your comments. Below are our replies to each of the comments made. Our replies are given in italic.

Replies to the Reviewer comments:

 Equations in general seems to come from nowhere. Are there other works that use these formulae? If yes, please cite them. If no, please explain better the formulae.

Sorry, but we didn’t understand this comment as it fully contradicts the content of the paper. Each (!) equation in the text is supported by the exact reference to the paper, where it was introduced or explained before. In particular, Ref. [42] explains dimensionless parameters (Line 492 in the updated paper version), Ref. [35] describes eqs. (1,2,5-8), Ref. [45] introduces eq. (3), Ref. [43] – eq. (4), Ref. [40] – eq. (9). Moreover, each parameter in every single equation is explained in the text.

2. Part of the subtitles in the third section are not numbered, which makes the content of the manuscript confusing. It is suggested to revise and perfect the small paper to solve this problem.

Thank you for this notion, we have added numbering to the sub-subsections.

3. The data and text in Table 2 are not aligned in the center. It is suggested to modify this part to make the manuscript more professional.

Thank you, we have aligned numbers in Table 2 to the center. But please also note that most small technical issues like the one you noted will be adjusted by the editorial team at later stages.

4. The author marked the position of the ship in Figure 6b, but there may be other factors interfering with the location of the mark is inaccurate, and it is suggested that another means be used to verify the feasibility of the existence of the ship.

Thank you for this notion. According to the suggestion of another reviewer, Figure 6 was deleted from the paper, as it doesn’t add much to the analysis.

5. This presented work is technically correct. However, the scientific findings are relatively weak. Therefore, it is recommended to highlight or add some scientific findings.

Well, we have attempted to highlight the findings as mush as possible. To address it, and following the advice of other Reviewers, we have shortened the Conclusions and made them more focused. Main scientific findings were precisely formulated in the i) title, ii) abstract and iii) Discussion and Conclusions.
Here are they:

• Our record shows much larger internal waves than were observed or predicted before in the Kara Gates
• The observed NLIWs were of the same scale as narrow sill crests passed by the ship during a free drift
• An intense along-strait flow composed of the mean current from the Barents Sea superposed with pronounced barotropic tide results in strong baroclinic response and NLIW generation in the Kara Gates
• Our observations provide evidence that the NLIWs generated in the KG are among the largest ever documented in the Arctic Ocean
• The amplitudes of the observed IWs ranged from 4 to 40 m, revealing the existence of record-high NLIWs in this region
• The total IW energy per unit crest length for the largest waves was estimated to be equal to 1.0-1.8 MJ/m
• The most intense NLIWs were recorded during the ebb tide slackening and reversal when the background flow was predominantly supercritical
• Theoretical predictions of NLIW phase speed made within a two-layer approach showed that the shallow water KdV model and that for strongly nonlinear waves agreed best with observations, suggesting that the former is applicable to predict the phase speed of large-amplitude NLIWs in the Arctic Ocean.

We find these findings strong and focused enough compared to the current state of the NLIW research in the Arctic Ocean, and we can’t elaborate anything beyond what was really obtained.

6. In general, the author does not fully summarize the innovation of the manuscript in the conclusion, and it is suggested to further highlight this content.

Please see the answer above. Also, we have added in the Conclusions, that the waves we have observed are record-high NLIWs in the study region and among the highest ever observed in the Arctic Ocean. We have also described the reasons of their generation, their vertical and kinematic properties. Given that such information is very limited in the Arctic Ocean up to now, we believe the paper is rather self-contained. Yet, we would be also thankful if the Reviewer can suggest what other aspects of the results obtained could be highlighted more. Please also note, that we were recommended by other Reviewers to describe everything as it is and to avoid any unnecessary speculations.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed all my comments carefully, i have no more comments.