Assessing the Added Value of Sentinel-1 PolSAR Data for Crop Classification

Round 1

Reviewer 1 Report

This study presented a comprehensive evaluation of adding value of Sentinel-1 PolSAR data for crop classification. In particular, this study had highlighted the significant contribution of Sentinel-1 PolSAR data in crop classification in areas with frequent cloud coverage and the effectiveness of the genetic algorithm in discovering the most informative features.

 

Overall, the topic is interesting that can be beneficial to the scientific community. However, this manuscript could benefit from significant editing before publication.

 

Specific comments:

1. Line 75: The reviewer suggests adding a table, which can clearly show the research trend of previous studies.

2. Line 187: The reviewer proposes providing detailed research objectives at the end of the introduction section.

3. Line 260: Could you add more information about H/A/α polarimetric decomposition technique?

4. Line 266: The reviewer is confused about equation 3. Could you provide more detailed content?

5. Line 495: In the result section, the reviewer got lost in searching the main result.

6. Line 621: What is the meaning of 23 H/A/α?

7. Line 713 (Figure 10-11): Please explain why these variables can be selected as the important variables. Please provide the detailed contributions of the important (or optimal) variables in this study. Please provide images with finer spatial resolution. Why did Feb obtain the substantially lower frequency in the feature importance assessment?

8. Line 734: Please add the limitations in this study.

 

9. Line 745: For improving the readability, the reviewer suggests splitting the conclusion content into three or four points (use (1)…, (2)…, etc.) related to the main result.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have carried out a complex study, with numerous data and well-chosen and implemented calculation methods. The chosen algorithms are explained in detail, justifications are provided for each implemented step. The present paper refers to numerous publications that have appeared recently, proving the extensive documentation conducted by the authors in carrying out this study. The results are presented and discussed accordingly.

Author Response

We would like to the reviewer for the positive feedback and the kind words.

Reviewer 3 Report

Main contribution of this study should be addressed in conclusion. It is good to distinguish between what will be written in the introduction and conclusion. Likewise, it is not desirable to move on to a discussion without a result after method.  There should be a part corresponding to the result in the content by arranging other sections.

In Figure 1, it is not known which figure is the higher scale figure; since many people read manuscripts online, it is necessary to improve the quality of all figures.

Comparing classification results according to the presence or absence of cloud is a very interesting. However, although results are the classification based on satellite images such as sentinel, there is no maps; so it is difficult to confirm whether the classification is actually processed well enough. At least, the map of the case before and after applying the polsar data with the highest accuracy should be added.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have replied my questions and revised this manuscript, so I suggest that it can be accepted.

Reviewer 3 Report

All of comments were well reflected. I suggest this manuscript to be published after the final grammar check and text editing.