A Noise De-Correlation Based Sun Glint Correction Method and Its Effect on Shallow Bathymetry Inversion
Round 1
Reviewer 1 Report
Despite good writing and presentation, several critical data/discussions are missing from the article :
* the method seems to work well for the elimination of whitecaps but looking at figure 3H it does not correct from sun glint. In this field, the reference of [Kay, 2009 https://www.mdpi.com/2072-4292/1/4/697/htm] can help to understand if the method is clearly suitable for the elimination of sun glint or whitecaps.
* the adaptation of hyper-spectral images to multi-spectral images can have an impact on the parameters mu and eta respectively equal to 2 and 0.015 (l.180). No discussion on this subject even if the adaptation from the hyper-spectral image to the multi-spectral image can be considered as an interesting scientific contribution.
*Field bathymetry is always difficult to obtain, however Sentinel-2 data is free and easily downloadable, which can significantly increase the amount of data and produce usable results. With 3 processed images, the number of points can be quite large. However, Sun glint correction must be robust to the acquisition configuration and to the atmospheric conditions, which implies processing a larger number of images. For this type of study, we can concentrate on a single geographical site but with a higher number of images.
* The above study involves taking into account the tidal information related to each site.
* The uncertainties of the ground truth data must be clearly presented in order to understand the importance of improving the estimated bathymetry.
Author Response
Dear Reviewer,
Thank you very much for pointing out the problems with manuscript " A Noise De-Correlation Based Sun Glint Correction Method and Its Effect on Shallow Bathymetry Inversion". Based on your valuable comments, we have carefully revised the manuscript and answered the questions in detail.We wrote the response in the attachment.
Sincerely,
Aijun Cui.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper proposes a sun glint correction method based on noise de-correlation (ND-SGC) which is not affected by the essential characteristics of the water body itself and does not require any auxiliary data. ND-SGC method retains intrinsic spectral information useful for bathymetric inversion of remote sensing images while removing sun glint. The results are important for the shallow bathymetry remote sensing. However, there are some minor issues to be improved.
Minor comments.
1.Points in different color in Figure 7(e) and (f) are not very clear. What's the threshold between the sun glint pixels and the sun glint-free pixels? In some cases, they are really close in space. How to separate them?
2. The message shown in Figure 11 is not clear. It's better to remove Figure 11.
Author Response
Dear Reviewer,
Thank you very much for pointing out the problems with manuscript " A Noise De-Correlation Based Sun Glint Correction Method and Its Effect on Shallow Bathymetry Inversion". Based on your valuable comments, we have carefully revised the manuscript and answered the questions in detail. We wrote the response in the attachment.
Sincerely,
Aijun Cui.
Author Response File: Author Response.docx
Reviewer 3 Report
This paper proposes an ND-SGC method to remove the effect of sun glint and compares it with two traditional methods; at the same time, the influence on bathymetric inversion based on remote sensing is analyzed. The research is interesting and the manuscript can be published after some minor revision.
1. In formula 11, what is the meaning of T, transpose of matrix?
2. The characteristics should be the same; for example, LW and RW have the same meaning as the remote sensing reflectance.
3. In the spectral fidelity analysis, three indexes are used to assess the spectral fidelity, namely, the CC and the larger CC value of the spectral correlation coefficient indicate better spectral fidelity; the error and the greater the error value indicate less distortion of the image spectrum after correction; and the spectral angle and the closer the value is to 0, the smaller the spectral angle is and the smaller the spectral distortion is. However, in Figure 4, the ND-SGC method has less error than the other two methods.
4. In line of 354, it is confuse to directly say that the ND-SGC method has improved the accuracy of bathymetric inversion in the extremely shallow water region to some extent, since only the MRE of the Log-Linear and Stumpf models before sun glint correction are listed. However, in Figure 6(A), the result of the original image is similar to that of the ND-SGC method, indicating that the results of the original images are also acceptable. However, the Hedley and Goodman methods reduce the accuracy of bathymetric inversion. Is the original image a radiometric and atmospheric corrected image?
5. In line 376, ‘this may be due to the influence of the substrate, where the precondition of nonzero reflectance in the NIR band does not hold leading to overcorrection in the two conventional models’, consider the short wave infrared instead of replace the NIR? The Worldview-2 and Sentinel-2 images have shortwave infrared bands.
6. In Figure 8, the error of the ND-SGC method is larger than that of the other two methods, which is the advantage of the ND-SGC method in Dongdao. We checked the sentence in line 455.
7. In line of 430-431, it say that ‘After the sun glint correction, the difference between the accuracy of bathymetric inversion of the sun glint pixels and sun glint-free pixels is not significant, indicating that sun glint can affect the bathymetric inversion’, this conclusion seems irrelevant to the argument, to reach this conclusion, a difference comparison should be made between the original image and the image after the sun glint correction.
8. In Figure 7, subplots a and d represent the same location, Dongdao Island, but the results are not consistent.
9. In figure 6, (A) the water depth of Taiping Island ranges from 0 ~ 2 m; (B) Tiexian Jiao ranges from 4 ~ 11 m; (C) Dongdao Island ranges from 2 ~ 20 m. the water depth only >11 m should be considered separately, may be have interesting result?
10. I figure 9, the number of the validate data of different methods is different?
Author Response
Dear Reviewer,
Thank you very much for pointing out the problems with manuscript " A Noise De-Correlation Based Sun Glint Correction Method and Its Effect on Shallow Bathymetry Inversion". Based on your valuable comments, we have carefully revised the manuscript and answered the questions in detail. We wrote the response in the attachment.
Sincerely,
Aijun Cui.
Author Response File: Author Response.docx
