Polar Cloud Detection of FengYun-3D Medium Resolution Spectral Imager II Imagery Based on the Radiative Transfer Model
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe paper entitled, ‘Polar Cloud Detection of FY-3D MERSI-II Imagery Based on the 6S Radiative Transfer Model’ present’s important work and results for an adequate field. After my thorough reading, I have not found any vague statements or blatant mistakes, which prompt me on a negative decision. I suggest the paper is good to publish in ‘MDPI Journals’. The results match the contemporary issues in a good way by addressing the lack of problems in previous report. Therefore, I am recommending it for publication, only after adding the below minor suggestions.
Title: too much abbreviations; simple it.
Page 1. Line 10. Replace ‘between cloudy and clear-sky pixels’ by ‘between the cloudy and clear-sky pixels’.
It is always awesome to arrange the keywords alphabetically for the effective readability of the work.
Line 19: replace ‘Varying degrees of cloud cover and shadows’ with ‘the variation of various degrees in cloud cover and shadows’
Line 32. Add ‘the’ before the line starts
Line 58: Zhu et al. proposed add reference no here as per MDPI format.
Line 63. What this mean (SVM) [? ]?
I do not like the introduction with few references. Add more brief history with adding past reference, then some novel work references also.
Line 90: what is the aim of this paper and what is new in this? What is the research gap? Write in a sentence or two
Figure 1: write lat and long on it
Write clear with all web links, when it is accessed. Some of them are not properly opening. Check all and update it
Figure 2: write lat and long on it
Figure 3: any unit of surface reflectance along y-axis?
Check the number of all equations carefully and re-check all equations
Figure6: write all details in the caption. What is ρ and ρ*? And what it showing
I do not like the discussion. Please put your results in perspective with other reports in the literature, explain significance of results and how they contribute to the overall state of knowledge or how they advance knowledge. What lacks in previous finding that are overcame in this paper? All these are discussions
Conclusions are also not as per scientific standard. Remove the unwanted discussion from conclusions and write concrete findings of this paper.
Check accuracy of referencing style and please discuss more articles among from international people literature.
Comments on the Quality of English LanguageThe English is ok. just need some minor english revisions
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe comments of “Polar Cloud Detection of FY-3D MERSI-II Imagery Based on the 6S Radiative Transfer Model” by Dong et al.
The cloud detection algorithm in satellite remote sensing is undoubtedly important. The authors create a monthly synthesized surface reflectance database by utilizing 8-day synthetic MODIS surface reflectance products (MOD09A1) and employing the 6S (the Second Simulation of the Satellite Signal in the Solar Spectrum) model for forward simulation. Through forward simulation of the correlation between apparent reflectance and surface reflectance across diverse conditions using the 6S radiative transfer model, the authors established a dynamic cloud detection model for the shortwave infrared channels. This method is valuable for satellite cloud detection algorithms, but there are still some issues in this article to clarify to improve the readability of the article.
1. In fact, in the last paragraph of the introduction, the key cloud detection technology problems to be solved in the paper are not clear enough. It is necessary to further clarify the method developed, to solve the scientific and technical problems of cloud detection, and to clarify who this method will attract attention.
2. What are the limitations of this approach? Where is the comparison with the results of the existing method cloud detection? What are the precautions in the North and South Pole application? Is it feasible to apply the cloud detection over the Tibetan Plateau in the third pole? That is, the problem of the adaptability of this method. Further clarity needs to be made.
3. The authors create a monthly synthesized surface reflectance database by utilizing 8-day synthetic MODIS surface reflectance products (MOD09A1) and employing the 6S (the Second Simulation of the Satellite Signal in the Solar Spectrum) model for forward simulation. So, what is the basis of the eight days of data? Why is it only 8 days of data? How to ensure its representativeness.
4. Section 3.2.2 can be further streamlined. Usually, if the same research or method is already available, just briefly elaborate and add references.
Comments on the Quality of English Language
1.It is recommended to polish the language expression of the correct article, and the current readability is weak(Such as: “Line 97-101.”)，I am not here to point out the specific problems line by line.
2. Line 63 [?] why.
3. Line 82, Which article is this article?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsIt is difficult to distinguish clouds, snow, and ice in remote sensing images, especially for polar regions. This paper presents a dynamic threshold cloud detection algorithm that relies on simulation and statistical techniques. This method effectively enhances the accuracy of cloud detection in polar regions and improves the discrimination between ice/snow and clouds. But there are some issues in the text.
1. The Tabel.2 in Line 167 should be Table.3
2. When abbreviations first appear, they should indicate their full name. such as GLCM，FOVs，NECP，AERONET，AOD in Line 63,179,186,188,198.
3. Citation display error. such as Line 63,371.
4. 1.64m in Line 251 should be 1.64μm.
5. The same symbol expresses different meanings. such as the ρ in Line 345 and Line 347 .
6. The N real − cloud in Line 402 and N real − clear in Line 403 did not appear in the equations.
7. In the display section of the result map, there are many colors displayed in the pseudo color image, which cannot clearly represent clouds, snow, and other ground objects. It is hoped that the author can annotate the ground objects represented by each color.
8. According to the figure 8\9\10, the distinction between clouds and snow/ice is mainly based on the differences in 1.64um of spectral band. Can a judgment threshold be given for use?
9. Although the experimental results have high accuracy, how can we determine the true values of cloud in the image? How to solve mixed pixels in medium resolution?
10. In the experiment, only comparison with MOD35 was conducted, and the comparative experiments were not sufficient. Can there be more comparative experiments with other methods (such as some machine learning methods).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsI recommend the authors to add more references, presenting similar works.
After checking the English language and editing the article could be accepted for publication.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for AuthorsThis study proposes a new cloud detection algorithm based on the 6S radiative transfer model, applied to 1000m resolution polar satellite imagery, which shows improved thin cloud detection over traditional thresholding methods and MODIS products. Although further validation is needed with larger sample sizes and ground reference data to assess performance on multi-layer and mixed-phase clouds, and consider the impacts of seasons and surface types, the paper demonstrates the potential of physics-based approaches for polar cloud detection. The efficient computation also makes operational implementation feasible. Future work will evaluate the fusion of multi-source data and 3D models to further improve accuracy. Overall, this is a valuable first exploration that provides cloud analytics to support polar climate research and points the way toward algorithmic improvements. The overall structure and logic of this paper are reasonable and clear. However, there are still some deficiencies or questions that need to be addressed, as follows:
1. There is a lack of sufficient cloud Lidar truth data to validate the accuracy of the algorithm's cloud detection. The validation sample size is small, which may introduce biases in the results.
2. This study only compared with MODIS products, and did not compare with other satellite sensor products. The impacts of different seasonal conditions and surface types on cloud detection were also not considered. It is recommended that validation results under different conditions (e.g. four seasons) be added.
Specific comment:
1. The introductory part of the Introduction contains repetitive information. It is suggested to simplify it. For example, the first paragraph about the ISCCP project can be removed and start with the importance of cloud detection directly.
2. The sentence "regardless of the type of cloud detection method" in Line 42 is repetitive in meaning and can be deleted.
3. The two sentences in Lines 43-45 have similar grammar structure and can be combined into one sentence.
4. The causal relationship "given the scarcity of ground observation stations in polar regions" in Line 50 can be emphasized.
5. The reference number [?] in Line 63 needs to be completed.
6. The transition phrase "Accordingly this paper" in Line 81 needs to be smoother. "enables the establishment of" in Line 86 can be simplified to "establishes". "improves the discrimination between" in Line 89 can be changed to "distinguishes". The logical relationships between sentences can be made more explicit to enhance paragraph flow. Reference formats should be unified.
8.Figure 3 depicts the spectral reflectance curves of cloud and snow, but the source of the plotted data is not cited. Please add a reference for the origin of the cloud and snow spectral data shown in Figure 3.
9.The captions of Figures 8-10 do not include the observation time and latitude/longitude range. "RGB: 167" is not rigorous and it is suggested to modify it to "False color (band1, 6, 7)". Legends distinguishing clear sky and cloud should be added under MOD35 and Result to differentiate them.
Comments on the Quality of English LanguageMinor editing of the English language is still needed.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 5 Report
Comments and Suggestions for AuthorsThe authors have addressed all of my concerns.
