Formalization for Subsequent Computer Processing of Kara Sea Coastline Data

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

Abstract: The abstract lacks key descriptions of the dataset, such as data source, spatial resolution, time series, and quantitative results regarding shoreline change. Furthermore, there is an overemphasis on the use of neural network models for predicting shoreline change, yet no mention is made of how these neural networks utilize the shoreline data for predictions in the methodology or discussion sections. I recommend adding a brief explanation of how the neural network is used for predicting shoreline changes based on the data.

Introduction: The literature review on AI-based shoreline detection is insufficient. It would benefit from additional references and descriptions of AI methodologies for shoreline detection. For example, Mapping intertidal topographic changes in a highly turbid estuary using dense Sentinel-2 time series with deep learning provides useful insights that could be referenced. Since this manuscript focuses on data, the introduction should also include an overview of shoreline detection methods using different satellite sensors, platforms, and resolutions, as well as their limitations. A relevant reference here could be "UAV Photogrammetry in Intertidal Mudflats: Accuracy, Efficiency, and Potential for Integration with Satellite Imagery."

Study Area: I suggest adding more information about the tidal conditions and trends in sea-level rise for the study area. This would provide a clearer context for understanding shoreline changes in the region.

Estimation of the Coastal Retreat: It is unclear what shoreline definition is used in this study—whether it refers to the mean high water line or the instantaneous water edge. I recommend clarifying this in the manuscript. Additionally, please include the tidal conditions corresponding to the image acquisition times in Table 1. Comparing shorelines at different tidal levels does not provide an accurate rate of shoreline retreat, as tidal fluctuations can significantly affect the position of the shoreline.

Results and Discussion: The authors introduce a method for quantifying shoreline change that differs from the DSAS approach. I strongly recommend that the authors compare the shoreline retreat detected by their proposed method with results from the DSAS method. This comparison will highlight the effectiveness of the proposed method in detecting real shoreline retreat and allow for a discussion of the differences between the two approaches. This could also be valuable for future studies that aim to adopt the new method proposed by the authors。

Author Response

Dear Reviewer,

Thank you for your review of our article and for your comments and suggestions.

Comments 1 [Abstract: The abstract lacks key descriptions of the dataset, such as data source, spatial resolution, time series, and quantitative results regarding shoreline change. Furthermore, there is an overemphasis on the use of neural network models for predicting shoreline change, yet no mention is made of how these neural networks utilize the shoreline data for predictions in the methodology or discussion sections. I recommend adding a brief explanation of how the neural network is used for predicting shoreline changes based on the data.]  

Response 1: The abstract has been rewritten to more clearly and comprehensively reflect the study's objectives, methods and plans. 

Comments 2 [ Introduction: The literature review on AI-based shoreline detection is insufficient. It would benefit from additional references and descriptions of AI methodologies for shoreline detection. For example, Mapping intertidal topographic changes in a highly turbid estuary using dense Sentinel-2 time series with deep learning provides useful insights that could be referenced. Since this manuscript focuses on data, the introduction should also include an overview of shoreline detection methods using different satellite sensors, platforms, and resolutions, as well as their limitations. A relevant reference here could be "UAV Photogrammetry in Intertidal Mudflats: Accuracy, Efficiency, and Potential for Integration with Satellite Imagery."]

Response 2: references to articles have been added. In our study, we determined the position of the bluff top and assessed its change. Remote sensing data and geodesic survey were used for that.

Comments 3 [Study Area: I suggest adding more information about the tidal conditions and trends in sea-level rise for the study area. This would provide a clearer context for understanding shoreline changes in the region.]

Response 3: The height of tides has been added. In our study, we determined the position of the bluff top which is composed of unlithified ice-rich sediments. The coastal retreat is associated with sea activity (after thawing) and positive air temperatures, which affect the thawing of frozen coasts.

Comments 4 [Estimation of the Coastal Retreat: It is unclear what shoreline definition is used in this study—whether it refers to the mean high water line or the instantaneous water edge. I recommend clarifying this in the manuscript. Additionally, please include the tidal conditions corresponding to the image acquisition times in Table 1. Comparing shorelines at different tidal levels does not provide an accurate rate of shoreline retreat, as tidal fluctuations can significantly affect the position of the shoreline.] 

Response 4:  We did not compare shorelines at different tidal levels, we investigated bluff top changes.  

Comments 5 [ Results and Discussion: The authors introduce a method for quantifying shoreline change that differs from the DSAS approach. I strongly recommend that the authors compare the shoreline retreat detected by their proposed method with results from the DSAS method. This comparison will highlight the effectiveness of the proposed method in detecting real shoreline retreat and allow for a discussion of the differences between the two approaches. This could also be valuable for future studies that aim to adopt the new method proposed by the authors]

Response 5: done.  

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

This paper presents an interesting investigation. The authors developed a new methodological framework. There were analysed 3 study area, with appropriate input data quality for the application of presented methodology and future machine learning. That is the only proper approach, unfortunately today many studies don’t have that approach. The author used adequate geostatistical approach fir analyse of shoreline changes. Time series of data were adequately used (shoreline changes, meteorological data).

The paper needs some minor changes. I would like that paper more highlights used and developed methodology than future application of that data.

Figure 2 – which is location of it?

Figure 3 - the calculation of coastal retreat is adequately explained, it is necessary to better explain the connection between c and d panels of the image, they are not really connected and clearly explained.

 

Figure 7 (and 4). I would like to see the map with presented coastal transects, to see the part around prof 160 which did not change and others.  

Author Response

Dear Reviewer,

Thank you for your review of our article and for your comments and suggestions. We made the necessary revisions to the manuscript to all your recommendations. Specifically, the conclusion has been significantly shortened to avoid repeating the content of the main text. We believe these changes have enhanced the clarity and structure of the paper.

best wishes,

Daria

Round 2

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

The authors have addressed all the issues well. I think it can be published in the current form.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

I thank the authors of the paper for the reflected changes

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper describes data preparation for erosion studies over 3 Arctic coastal sections. The novelty incorporated in the subject is minimal, while the latter is presented vaguely, incompletely, and with loose scientific rigour. This leaves the reader with a number of questions and concerns about the credibility of main aspects of the paper including its conclusions. The most critical of those issues are briefly stated in the following.

1. The meaning of averaged values over advancing time is distorted and the relevant conclusions unsound. In lines 219-220 when discussing Fig. 5 it is stated for the coastal retreat that "...we see an upward trend in the past" and also "...observing a slowdown" for larger time periods. Additionally, this slowdown with time of the retreat is ascribed to "a decrease of wave impact" with no explanation, when the general trend for most coastlines is the opposite, due to the current climate change. These conclusions are not right if you consider what a moving average represents, as well as its dependency on the selected starting point of time. In fact the graphs presented in the paper do not support the said conclusions. Think of the car sensor that gives you the fuel consumption through time. If the value shown increases with time, even at a lower rate, this does not mean that you consume less fuel/km but the opposite!

2. It is further suggested to use the long-term retreat values in prediction models, l. 308, based on the assumption that these are more stable! However, on top of the fact that these values may not actually be more stable (see #1), such suggestion ignores the high possibility that older values would rather be unfit for predictions into the future, given the climatic variability that controls coastal erosion.

3. The variograms presented in Fig. 7 are not discussed, but it is decided in lines 263-264 that "...entire sections of the coastline represent a single ensemble..". The meaning of this statement and its justification are not given, while the relevant graphs (with an undefined 'dispersion' term on the vertical axis) do not give a clue leading to such a conclusion.

4. The result from Fig. 6 (the second of the two figures 6), claimed by the authors, is that coastal retreat changes systematically, but what this exactly means or how it is derived remains unanswered.

5. Along the same lines, Fig. 4 reveals an "inverse correlation phenomenon (sic)" that cannot be observed by the reader. This 'phenomenon' is not described or discussed in scientific/technical terms.

6. Also, Fig. 8 is not discussed in a comprehensive fashion, but rather arbitrarily it is said that normal or log-normal distributions may fit the results without providing the corresponding representative lines of those distributions against the results or any comparison with other model distributions.

7. The Abstract mentions interesting methods and tools that are absent from the full text (in which chapter 4 is missing).

There are additional review comments not included here, less critical than the above though.

Comments on the Quality of English Language

English is OK in general, with a few language slips dispersed in the manuscript.