Evolution of the Floe Size Distribution in Arctic Summer Based on High-Resolution Satellite Imagery

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Exploring the size distribution of sea ice floes is a meaningful work for global climate and physical processes in the polar oceans. The structure of the manuscript is good, and the research results are interesting to sea ice society, and can be used for practical applications by other researchers. However, there are still some printing errors in the manuscript. Although they do not reduce the practical significance of the manuscript, but they should be clarified and corrected accordingly. Additionally, I have some minor comments as below.

(1)  In Table 1, why did the author only provide the longitude and latitude of a point, rather than a longitude and latitude range? Although the size range of the study area is given in Table 1, I think it would be better to complete the longitude and latitude range.

(2)  Please pay attention to the aesthetics and resolution of the images. Figure 1 and Figure 2 have a problem of taking up a large amount of space, which affects the aesthetics. Additionally, there are some distortions in the images in Figure 2. I am not sure if this is caused by the compression function that comes with the software Word.

(3)  As exploring the impact of air temperature on the size distribution of ice floes at the end of the manuscript, the average temperature of image intervals was used as the monthly average temperature. Therefore, is the Monthly average temperature in Table 2 the average temperature of four months or the average temperature of image intervals? If it is the former, then this line of data should be considered for deletion in subsequent work. If it is the latter, I think an explanation is necessary.

(4)  In Table 4, I noticed that the results of D are between 0.83 and 1.32, which are similar to other research results in Table 4. However, there is a significant gap with Denton's research results, even though both studies are on FSD in the Canadian Basin. Why is this phenomenon occurring, and what does it mean to obtain D in different ways as described here?

(5)  In the introduction section, you used the word 'reveal' to describe the impact of temperature on the FSD parameter. However, combined with subsequent results, only the regularity of the change in FSD parameter with temperature and the relationship between temperature and FSD parameter were found. There was no study of the underlying physical processes. I think it would be better to change 'reveal' to 'find'. And in future study, I hope to continue to explore the physical processes behind the changes in FSD parameter.

Author Response

Comments1:In Table 1, why did the author only provide the longitude and latitude of a point, rather than a longitude and latitude range? Although the size range of the study area is given in Table 1, I think it would be better to complete the longitude and latitude range.

Response1:Revised accordingly, we revised the Table 1 to Table A in line 92 of the manuscript, and the latitude and longitude ranges were also added.

Table A. Temporal and spatial distributions of images and other information

Region	Time period	Longitude and Latitude	Location	Resolution (m)	Scope (m×m)	Ice age
CBIB_N	2014/06/09- 2014/09/09	82.01°N- 82.03°N 164.96°W- 165.11°W	Canadian basin	1.00	2627×2439	First-year ice
CBIB_S	2014/06/13- 2014/09/11	79.50°N- 79.57°N 149.98°W- 150.13°W	Canadian basin	1.02	3138×3893	First-year ice

Comments2:Please pay attention to the aesthetics and resolution of the images. Figure 1 and Figure 2 have a problem of taking up a large amount of space, which affects the aesthetics. Additionally, there are some distortions in the images in Figure 2. I am not sure if this is caused by the compression function that comes with the software Word.

Response2:Revised accordingly, we reuploaded Figure 1 and Figure 2, adjusted their size and resolution. The modified images are shown in attachment, and Lines 93 and 171 of the revised manuscript

Comments3:As exploring the impact of air temperature on the size distribution of ice floes at the end of the manuscript, the average temperature of image intervals was used as the monthly average temperature. Therefore, is the Monthly average temperature in Table 2 the average temperature of four months or the average temperature of image intervals? If it is the former, then this line of data should be considered for deletion in subsequent work. If it is the latter, I think an explanation is necessary.

Response3:Thanks for your comment. Monthly average temperature data given in Table 2 is the average temperature of image intervals. The average temperature of image intervals means that we use the average temperature of the 30 days between images as the average monthly temperature. For example, if there is an interval of 30 days between June 9 and July 9, we make the average temperature of these 30 days as the monthly average temperature for July. We gave the above explanation in Line 115 of the revised article and labeled Line 123 of the revised manuscript ‘Monthly average temperature means the average temperature of image intervals.’

Comments4: In Table 4, I noticed that the results of D are between 0.83 and 1.32, which are similar to other research results in Table 4. However, there is a significant gap with Denton's research results, even though both studies are on FSD in the Canadian Basin. Why is this phenomenon occurring, and what does it mean to obtain D in different ways as described here?

Response4: Thanks for your comment, the reason for this phenomenon is that the method used in this manuscript is different from the method used by Denton et al.[1]. The method used in this manuscript is the cumulative density function while Denton used the non-cumulative density function, which is the meaning of obtaining D in different way. Even though the methods used in the two studies are different, the trends observed for D are the same. Moreover, there is a large difference in the FSD parameters calculated using the non-cumulative and cumulative density function. For example, Stern et al.[2] studied FSD using non-cumulative density function in the Beaufort and Chukchi Seas in 2014, deriving D values between 1.9 and 2.8. While the D is between 1.0 and 1.5 according to Wang et al.[3], who take a cumulative density function. We added the above explanation to Line 332 of the revised manuscript.

Comments5: In the introduction section, you used the word 'reveal' to describe the impact of temperature on the FSD parameter. However, combined with subsequent results, only the regularity of the change in FSD parameter with temperature and the relationship between temperature and FSD parameter were found. There was no study of the underlying physical processes. I think it would be better to change 'reveal' to 'find'. And in future study, I hope to continue to explore the physical processes behind the changes in FSD parameter.

Response5:Revised accordingly. The ‘reveal’ has been replaced to 'find' in Line 71 of the revised manuscript. This paper mainly focuses on the influence of temperature on FSD, and we will also add the physical processes to the research of FSD in the future. We explained it in the last paragraph of the conclusion.

References

1. Denton, A.A.; Timmermans, M.-L. Characterizing the Sea-Ice Floe Size Distribution in the Canada Basin from High-Resolution Optical Satellite Imagery. The Cryosphere 2022, 16, 1563–1578, doi:10.5194/tc-16-1563-2022.
2. Stern, H.L.; Schweiger, A.J.; Stark, M.; Zhang, J.; Steele, M.; Hwang, B. Seasonal Evolution of the Sea-Ice Floe Size Distribution in the Beaufort and Chukchi Seas. Elementa: Science of the Anthropocene 2018, 6, 48, doi:10.1525/elementa.305.
3. Wang, Y.; Holt, B.; Erick Rogers, W.; Thomson, J.; Shen, H.H. Wind and Wave Influences on Sea Ice Floe Size and Leads in the B Eaufort and C Hukchi S Eas during the Summer‐fall Transition 2014. JGR Oceans 2016, 121, 1502–1525, doi:10.1002/2015JC011349.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Ice floe size is important in studies of polar physical processes and global climate change. This paper utilizes high-resolution satellite images to study the floe size distributions (FSD) in the Canadian Basin. Overall, the experiment was well designed and the results are convinced. Therefore, my opinion is to this manuscript is to accept after minor revision.

The comments and suggestions are as follows:

1)        what does ‘oversemented regions’ mean in Line 152? The introduction of watershed algorithm and the significance to do watershed segmentation?

2)        four parameters are calculated in section 2.3.2, the first three are easy to understand, what’s the relationships between the roundness and FSD, it is not reveled in the manuscript.

3)        Figure 6. (a) the CBIB_N region and (b) the CBIB_N region?

4)        as presented in section 3.2.2, 𝐿0 can be used to characterize the proportion of large floes. And a slight increase was observed in CBIB_S on 11 Sep., so is there any evidence to support the following statement?

5)        There are two R² is Table 3, and these two parameters are not discussed in the text.

6)        The time series is too short while analyze the relationships with temperature, it is better to extend it.

7)        the conclusion section is not well presented. It looks like a repetition of the results of analysis.

Author Response

Comments1: What does ‘oversemented regions’ mean in Line 152? The introduction of watershed algorithm and the significance to do watershed segmentation?

Response1:Thank you for your comment. In fact, 'oversegmented regions' is a spelling mistake, and the correct expression should be 'oversegmented regions'. Oversegmentation is a well known difficulty with watershed algorithm caused by noise and irrelevant contour elements in images[1]. 'Oversegmented regions' means that some of the complete floes in the image have been incorrectly segmented into many small pieces of ice, which can also be seen in attachment Figure A(a). We made the explanation in Line 158 and added Figure A(a) of the revised manuscript.

In attachment Figure A(b), there are still many adherent floes in the red boxes after dealing the images by thresholding and morphological methods. Therefore, a watershed segmentation algorithm is introduced in this paper to further segment these adherent floes. We explain the role of the watershed segmentation in Line 155 of the revised manuscript. The watershed segmentation method has been applied in many image segmentation studies[2–4].

Comments2:Four parameters are calculated in section 2.3.2, the first three are easy to understand, what’s the relationships between the roundness and FSD, it is not reveled in the manuscript.

Response2:The first three parameters area, perimeter and MCD are geometric parameters, while the fourth parameter roundness is a shape parameter. Although roundness is not used in the FSD study, roundness is an auxiliary explanation for the size of floes in this paper. When exploring the size of floes, a combination of geometric and shape parameters would be more comprehensive. The roundness of ice floes has been conducted by many FSD studies[5–7], which is a supplement to floe size. We give the above explanation in Line 241 of the revised manuscript.

Comments3:Figure 6. (a) the CBIB_N region and (b) the CBIB_N region?

Response3:Thank you for your comment.' (b) CBIB_N region' is a spelling mistake, and the correct expression is '(b) CBIB_S region'. We made a revision in Line 286 of the revised manuscript.

Comments4: As presented in section 3.2.2, L0 can be used to characterize the proportion of large floes. And a slight increase was observed in CBIB_S on 11 Sep., so is there any evidence to support the following statement?

Response4:On one hand, from June to September, the sea ice extent gradually deceases in the Arctic. Especially in the September, sea ice extent reaches the minimum value of the year[8]. As a result, the large floes in Figure B(you can see Figure B in attachment) cannot be formed by refreezing. On the other hand, There are studies that indicate the presence of ocean currents in the region[9,10]. Comparing the images of August 12th and September 11th, five large floes appeared suddenly, so we assumed that ocean currents brought floes to this area, and L0 increased as a result.We give the above explanation in Line 302 of the revised manuscript.

Comments5: There are two R² is Table 3, and these two parameters are not discussed in the text.

Response5:Thank you for your comment. The first R² is the fit evaluation metric for the upper-truncated power law function and the second R² is the fit evaluation metric for the Weibull function. To illustrate the fit of the two functions, we added the two R² to Table 3. R² are above 0.88, which is a better fit in total, except the upper-truncated power law function for the image of CBIB_N area on June 9th. There are few small floes on this day, which leading to irregularities in the decrease of the cumulative frequency curve. So the fitting effect is slightly worse.We added the results of the discussion on R² in Line 311 of the revised manuscript.

Comments6:The time series is too short while analyze the relationships with temperature, it is better to extend it.

Response6:We agree with you very much and a long time series will benefit the robustness of the relationships. Our original intention was also conducted such kind of investigations by using continuous images during a long time period. However, it was found to be very difficult to extend the time series because of limitations on the available high-resolution satellite imagery. For example, there is only one satellite image per month at the same location in the present MEDEA dataset. We have searched for all available satellite images of the Canadian Basin in the 2014 MEDEA dataset, the images used in this paper are already temporally complete and with the longest duration in the dataset. Additionally, although the time series is short, it still helps to explore the potential effect of temperature on FSD parameters especially in Arctic summer. And in several previous studies, satellite images from June to September have also been used to study the variations in FSD[7,11,12]. Finally, we agree that a longer time series will give better results than the present results. So we added a note in the last paragraph of the conclusion that this issue can be solved in the future with new datasets.

Comments7:the conclusion section is not well presented. It looks like a repetition of the results of analysis.

Response7:Thank you for your comment. The whole conclusion section has been completely revised accordingly to avoid repetitions of the results. Now it consists of three new paragraphs as below.

High-resolution satellite images from June to September 2014 obtained in two different study areas within the Canadian Basin of the Arctic Ocean were analysed, and the ice floe size parameters within the two areas were extracted via image processing techniques. The variations in the floe size distribution parameters were analysed. Finally, the effect of the mean monthly air temperature on the floe size parameters was explored, and the following conclusions were drawn:

Both the upper-truncated power law function and the Weibull function fit well with the cumulative frequency distribution of floe size. The fractal dimension D and the shape parameter r for the two regions increase and then decrease over time. Among which, D is often related to the environmental conditions and r represents the uniformity of ice floe distribution. A smaller value of D corresponds to a not so strong thermodynamic effect in June, while D is larger in August. The influence of air temperature on FSD parameters is explored by quantitative analysis, and both D and r displayed positive changes with air temperature. But the correlation between D and temperature is somewhat weaker than that for r. It is because that dynamic processes from wind and wave causing ice fracture are also possible factors influencing D.

The relationship between air temperature and FSD provides a potential parameterization of FSD for sea ice numerical modelings. Although, there are still some issues that need to be improved in future. High-resolution imagery of Arctic Sea ice at the same location during a long time period are still desired to obtain a robust conclusion on the present relationship. And method of deep learning can be also involved into the current image processing steps to reduce the possible errors induced by manual intervention. A deep insight into the environmental factors and physical processes affecting the variations in FSD should be also conducted if more concurrent measurements are available in future observations.

Author Response File: Author Response.docx