Delineating Polynya Area Using Active and Passive Microwave Sensors for the Western Ross Sea Sector of Antarctica

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Can be accepted in present form.

Author Response

Thanks for the feedback and we are happy you liked it in the current form.

Reviewer 2 Report (Previous Reviewer 2)

I have read the revised version and the cover letter proposed by authors and the revised version is much improved. I also appreciated author's response and addressed all the comments. Related with GLCM parameters computations, the revised version is attached with appendix and the appendix title should be corrected as 'GLCM parameters computations used in the paper'.  

Author Response

We thank you for the positive comments on the revised paper. As suggested, the title of the “Appendix” is changed to “GLCM parameters computations used in the paper”, can be identified in page 22 of 25 (first line) of the ‘track changes’ document. 

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

This study derived Ross Sea polynya areas from Sentinel-1 C-band images by three methods (manually delineated, rule-based and texture-based classification) and compared with coarser resolution data (NSIDC SIC & AMSR2 SIC). The paper gave some commonsense results and lack of innovation. For the SAR image processing methods, the conclusion “the rule-based approach was more accurate than the texture-based approach” which can be drawn without reading the mns. Also the comparison methods are quite qualitative and no much information delivered to readers. 

Reviewer 2 Report

The paper is related with polynya area detection using synthetic aperture radar and comparison with the result retrieved by passive microwave radiometer data (PMR) ,which is crucial to understanding ocean-atmoshphere interaction and sea ice production in polar area. The paper proposed two algorithms (rule-based supervised and texture based clustering) to classify polynya area from sea ice using Sentinel-1 C-band SAR data and gives some interesting results. The paper is well organized and the proposed methods will be interested by scientific community. I suggested the paper could be accepted after major modifications. Some suggestions are summarized as follows,

Major modifications:

1, The paper proposed two methods to detect polynya area from SAR data, one the rule-based algorithm and the other is texture-based algorithm.

The rule-based algorithm is just threshold (-11 -4dB) method, I think the threshold criteria is not full discussed and especially how to obtain this threshold range. Futhermore, the threshold should be associated with radar frequency, incidence angle and etc. It is better to give more information related with this semi-supervised technique. 

The other algorithm associates with texture-based clustering technique. The paper did give some explanations about this method based on GLCM, but I could not find the necessary parameter computation method (about 6 parameters) and also GLCM will need window size, orientation direction (which will depend on the geometric correction also). Furthermore, the method also give 5 clusters (line 256) and I hope to know how to determine the polynya finally becasue all unsupervised method has similar problem.

Some minor modifications:

1,  Line 180, 'a threshold40'  here 40 should be reference number? or threshold value?  Similarily, line 182 , 'studies3' the number 3 has similar problem.

2, Figure caption problem. 

   Figure 2 all background image should be SAR data. Please add necessary information in the figure caption. Figure 2-8 has similar problems. 

   All other informations have been noticed in the attached file for reference.

3, Some reference information is not completed as shown in the attached file.

Comments for author File: Comments.pdf

Reviewer 3 Report

These authors assessed the performances of polynya area determination by PMW SIC threshold and SAR-based algorithm, compared to manually delineated polynya areas in many cases (polynya/non-polynya, TNBP/MCM/RSP). This manuscript is generally well-written and logical. This manuscript can be accepted for this journal with a few questions. i) It is questionable whether the manually delineated polynya can be accurate enough to be the reference. ii) Polynya generally includes areas of frazil ice and open water. However, this study did not consider the frazil iced area at all. An explanation of this is required. iii)  Although TNBP and MCM do not differ significantly in terms of scale, there is a difference in the performance of the S1 algorithm. What is the reason? Could it be because of a manual delineation problem? iv) There are many errors in spelling and notation in this manuscript. Therefore, a closer look is recommended.