Estimation of Broadleaf Tree Canopy Height of Wolong Nature Reserve Based on InSAR and Machine Learning Methods

Round 1

Reviewer 1 Report

Dear authors,

In general, I think the paper has interesting results and could be published. However, the quality of English needs to be improved in some parts (mainly abstract, introduction, and methods).

General comments:

- Clearly add the research gap, research question, and objective(s) at the end of the introduction.

- Please also briefly describe the limitation of the study, computational complexity, and future direction in the last paragraph of the discussion.

- Conclusion: Conclude research constraints and guidelines for future research.

 

Specific comments:

- L 1: The title should be informative. Where is the study area? Natural forest or plantation?  Broad-leaved or coniferous?

- L 26: Add more words to the keywords section.

- L 250: It would be better to mention latitude, first, then longitude.

- L 255: Change "3544 m" to "3544 m a.s.l."

- L 296: Same as the previous comment.

- L 492-496: Add some related new references, for example:

** https://doi.org/10.3390/rs13214282

** https://doi.org/10.3390/rs14061453

 

Author Response

Dear Reviewers and Editors，

Thank you for your kind comments concerning our manuscript.

We have revised the manuscript according to the reviewer's comment. For responses to specific comments, please see the attachment.

We really appreciate the reviewers for their valuable comments and suggestions, those comments are valuable for revising and improving our paper with important guiding significance.

Thank you again for your time!

Best regards,

Sincerely,

Xinyi Liu

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors

     The new version of your article: "Estimation of tree canopy height based on InSAR and machine learning methods" presented several changes compared to the second version. The new organization of results and discussions favors a better understanding of the steps implemented to carry out the work. It is noticed that an arduous work of review and inclusion of new elements to the article make it suitable for publication.

    The new arrangement of the figures in the results and the division into subchapters of the discussions make it easier for the reader to understand the analyzes carried out.

    I would just like to make one more suggestion: based on the results achieved, what recommendations for future studies would you add to the conclusions?

      I conclude my review and congratulate you for all the work done and for the new version presented for the article.

 

Respectfully,

Author Response

Dear Reviewers and Editors，

Thank you for your kind comments concerning our manuscript.

We have revised the manuscript according to the reviewer's comment.Enclosed below are the revised points including point-by-point response to the comments made by the reviewers.

Point 1: based on the results achieved, what recommendations for future studies would you add to the conclusions?

Response 1: Thank you for your valuable suggetion. We added these in the conclusion section.

This study used sentinel-1 repeat-pass InSAR and two machine learning methods to estimated broadleaf forest canopy height. In future studies, we suggest to further explore the applicability of single-pass InSAR in tree height estimation and extended this approach to coniferous forest and plantations, and developed machine learning models with higher inversion accuracy.

We really appreciate the reviewers for their valuable comments and suggestions, those comments are valuable for revising and improving our paper with important guiding significance.

Thank you again for your time!

Best regards,

Sincerely,

Xinyi Liu

Round 2

Reviewer 1 Report

The work has improved considerably. I think it can be accepted in its current state.

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

Dear Authors:

     His article: "Research on tree canopy height inversion based on InSAR and UAV - taking spruce and larch as examples" presents a relevant theme for forest inventory activities. However, the presented text needs a deep revision because in several parts the text is confusing. Review punctuation, agreement, and understanding of the ideas presented. It would be important to clearly present the objectives of the work in chapter 1. They are presented in a less objective way.
     I present 42 less relevant comments pointed out directly in the digital file. I request your special attention on the following points:
    1) Abstract: review the writing because it does not allow an understanding of the study developed, methods applied, and results achieved. Some methods and results are shown but the text is confusing.
    2) Line 31: “general optical remote sensing techniques such as multispectral, hyperspectral, multi-angle, synthetic aperture radar (SAR), and laser detection and ranging (LiDAR) techniques” the text is confusing and suggests that SAR and Lidar are optical techniques. Review the writing.
   3) Line 49: “and can interact with the interior of vegetation canopy”. This characteristic depends on the wavelength used by the sensor. To review.
   4) Line 134: Display the DEM information used.
   5) Line 148: “Field measurements included: coordinate values, elevation, slope, aspect... LAI”. Do you measure in-field elevation, slope, and aspect? If yes, detail this measurement and technique employed. In the case of LAI, did you use the Laicor instrument?
   6) Line 155: Please detail the characteristics of the aerial cover: longitudinal and lateral cover, flight height, flight direction, GSD. Did you collect GNSS observations/ground control points for accurate processing?
   7) Line 173: Please detail the algorithm used to filter the point cloud by classifying it into points on and above the terrain.
   8) Line 174: What is the criterion used to define the spatial resolution of the DSM/DTM?
   9) Improve wording from line 184 to line 200. The text is confusing in some parts.
   10) Please detail the pre-processing steps applied to Sentinel images 1. In this case, was the Snap application used?
   11) Explain the terms presented in equations 1, 2, 3, and 4. Please cite the bibliographic references.
   12) Line 234: Density (unit/area)? 0.7 to 1 tree? Is it possible to show the UAV sample image in Figure 2?
   13) Line 244: “The CHM values ​​generated from the UAV images were used as the true tree heights of the sample sites”. Please explore the related issue that in this case the data was generated from the passive sensors and this has some implications in forest areas.
   14) Line 253: Please calculate the correlation between observations to reinforce the analysis produced.
   15) Detail the “sync function” presented in the work.
   16) Figure 10: use letters in the Figure and in the legend to detail the figure.
   17) Line 290: Explain the criteria used to choose the random forest and multilayer perception, regression models.
   18)Figures 13, 15, 18, and 20: use letters in these figures to facilitate the discussion presented after the figures.
   19) Conclusions: Review and adapt the conclusions to the study objectives and results achieved.

    I conclude by congratulating them for their work and hope that my observations will be useful in improving the presented article.

Respectfully,

Comments for author File: Comments.pdf

Author Response

Dear Editor and reviewer:

Thank you for your letter and for the reviewers’ comments concerning our manuscript entitled” Research on tree canopy height inversion based on InSAR and UAV - taking spruce and larch as examples”(ID:forests-1642959). Those comments are all valuable and very helpful for our paper,as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval.

We appologized for the confusing English writing .This manuscript was edited for proper English language, grammar, punctuation, spelling, and overall style by one or more of the highly qualified native English speaking editors at NativeEE. The NativeEE Company specializes in editing and proofreading scientific manuscripts for submission to peer-reviewed journals.

This manuscript has been revised through review mode. The main corrections in the paper and the responds to the reviewer’s comments is in the attachment.

Thank you and best regards

Your sincerely

First author

Name: Xinyi Liu

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript tried to improve the tree canopy height inversion efficiency by integrating SAR, UAV and machine learning technology. The data processing method is fine, but the English writing and organization of the paper are far away from publishment standard. Therefore, I suggest the author reorganize their paper and submit again.

 

1. L19, 24, 25 et al. CO2 Format error.
2. L17-L28. Dual carbon target is important, but it is not very relevant to your research. You need to keep focus on the implement of tree height in “indispensable for ecological process and forest carbon cycle models”. And point out the necessary of tree height inversion.
3. L30-35. Lack of references.
4. L46-48. Inconsistent views. UAV has higher spatial resolution than SAR.
5. L36-L58. Please keep focus on TREE HEIGHT. UAV + SAR is not the innovation of your research.

6.L64. What are X/C/L/P-bands?

7.L33, L40 why two different abbreviations?

8.L46 SAR. Full name

9.L59-L71. Here, you should clear the methods in canopy height observation, and what is/are the remain questions. And better merged to paragraph one.

10. L70 InSAR. Full name

11.L73 Why suddenly start InSAR introduction? Lack of logic.

12.L105. “At present…”  What about the missing technology development of 18 years from 2005?

13. Lack of introduction of machine learning technology used in your study, and why you choose ML
14. L108-125. Introduction is not abstract, please do not confuse about them. Omit the result in these sentences, and shorten them.
15. Figure1. Source of DEM?
16. L142. What are the meaning of description the plot based on community species composition and nature of trees?
17. L149. What is the basis for dividing by 1.5m. Too redundancy doing the description of field measurements.
18. Figure2. It can be seen from the figure that the samples are on both sides of the road, and the distribution is not uniform (there is only one bamboo forest plot). How do you deal with the problem of inconsistent canopy characteristics?
19. Section 2.2 sample sites data processing. I cannot see any contents of “DATA PROCESSING”

20.Section 2.3. Lack of datetime of UAV image acquire.

21. L170 Why need are with high forest depression?
22. L170-L174. Rephrase these sentences. What are DSM and DTM?
23. Figure 3. What is the meaning of the path without images? Do not just show your original data.
24. L176. How do you get and process Sentinel-1A data.
25. L179. “According to the principle of InSAR inversion of forest structure parameters” Missing references.
26. L180. “ignoring…” Why can you ignoring these information?

27.L184-L192. Do not describe these common information in method, it belong to introduction.

28.Figure 5. This figure is completely un-acceptable.

29. Lack of Machine leaning method, and implementation
30. Section 2.5. Why do you describe sampling plot again? How do you get these parameters, link forest densities, tree crown width, etc… Why take 20m as sampling spacing? The most important thing this Why do you show the tree height result in this section, rather than in Result section?
31. L249,253. “The following figure…” “From the fugure…” Which figure(s)?
32. Figure 8 and Figure 9 describe the same meaning, you need merge them. Missing of name of x-axis in Figure9.
33. All figures in the manuscript are basically either unclear or small front size. The authors should improve them.
34. I stop here, because the confusing writing of M&M section, and I believe there are massive problems in the following parts. The authors should carefully reorganize the manuscript.

Author Response

Dear Editor and reviewer:

Thank you for your letter and for the reviewers’ comments concerning our manuscript entitled” Research on tree canopy height inversion based on InSAR and UAV - taking spruce and larch as examples”(ID:forests-1642959). Those comments are all valuable and very helpful for our paper,as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval.

We appologized for the confusing English writing .This manuscript was edited for proper English language, grammar, punctuation, spelling, and overall style by one or more of the highly qualified native English speaking editors at NativeEE. The NativeEE Company specializes in editing and proofreading scientific manuscripts for submission to peer-reviewed journals.

This manuscript has been revised through review mode. The main corrections in the paper and the responds to the reviewer’s comments is in the attachment.

Thank you and best regards

Your sincerely

First author

Name: Xinyi Liu

Author Response File: Author Response.pdf

Reviewer 3 Report

Regarding the MS ID: forests-1642959 Entitled: Research on tree canopy height inversion based on InSAR and UAV - taking spruce and larch as examples

the paper seems very interesting and totally matches the scope of the Journal and I would like to accept it after some minor revisions: 1- low quality of Fig. 3 2- show the significant differences in Fig. 9 3- please deduce the number of the figures. 

Author Response

Dear Editor and reviewer:

Thank you for your letter and for the reviewers’ comments concerning our manuscript entitled” Research on tree canopy height inversion based on InSAR and UAV - taking spruce and larch as examples”(ID:forests-1642959). Those comments are all valuable and very helpful for our paper,as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval.

We appologized for the confusing English writing .This manuscript was edited for proper English language, grammar, punctuation, spelling, and overall style by one or more of the highly qualified native English speaking editors at NativeEE. The NativeEE Company specializes in editing and proofreading scientific manuscripts for submission to peer-reviewed journals.

This manuscript has been revised through review mode. The main corrections in the paper and the responds to the reviewer’s comments is in the attachment.

Thank you and best regards

Your sincerely

First author

Name: Xinyi Liu

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Authors

The new version of your article has included all my suggestions. The reassessment task was facilitated by consulting the cover letter carefully prepared by you. However, I only ask for two more adjustments:

1)What is the criterion used to define the spatial resolution of the DSM/DTM? Usually, information such as point density, and precision, among others is used to calculate the size of the GSD. Was this calculation performed in your case?

2) In the conclusions: "While the tree height was estimated by the coherent magnitude method, it was seldom affected by the topographic factor". This topographical factor could be removed from the interferogram using an unwrapping technique. Could you provide a recommendation for this procedure to be applied?

I end my review by congratulating you on the new version of the article.

Respectfully.

Author Response

Dear Reviewers and Editors， 

Thank you for your kind comments concerning our manuscript.

We have revised the manuscript according to the reviewer's comment.Enclosed below are the revised points including point-by-point response to the comments made by the reviewers.

Response 1:

Thank you for your valuable suggestions.

We define the ground sampling distance based on the vertical image GSD calculation formula. The calculation is done using the sensor single image size, flight altitude, and camera focal length in UAV aerial photography.

The spatial resolution of DSM and DTM also follows the standard of ground sampling distance, and uses GSD as the image resolution of raster data.

Response2：

Thank you for your valuable advice.

It is feasible to use topographic radiation correction or unwrapping technique to remove topographic influence, but the scattering process of forests in complex forest areas is very complicated and the complete scattering process mechanism cannot be controlled at present, so it is not possible to remove the topographic influence completely and effectively.

The advantage of using the coherent amplitude method is that body scattering dominates when short-wavelength radar waves (e.g., c-band, x-band) interact with the forest, and the surface scattering contribution can usually be ignored, thus avoiding the complexity caused by topographic factors.

We really appreciate the reviewers for their valuable comments and suggestions, those comments are valuable for revising and improving our paper with important guiding significance.

Thank you again for your time!

Best regards,

Sincerely,

Xinyi Liu

Reviewer 2 Report

I can see the efforts the author have done to improve their manuscript. However, the manuscript still seems like chaos. Maybe the authors try to solve too much questions in one paper, because although this version supplemented many information of methods, this part still lack of focus. The scientific manuscript better to follow pathway of question-method-result-discussion, that is when you state ONE question in Introduction, you need solve it with ONE section in M&M, then show the readers with ONE section in Results, and finally, discuss uncertainties or do comparison in ONE section in Discussion. The work did by authors should be accept, but the writing of manuscript must be re-organized.

Title. The title is too big for your study, as there are no contents about the inversion results, such as spatiotemporal distribution of canopy height. Maybe title related to method comparison is more suitable for this study.

L152-157. Bring in some references for supporting your point of view.

L292. What is HH polarization?

L296 repeated period

L313-322 Point out what kind of ML method you choose? And why?

L329 repeated period

L323-L325 I believe the problems stated in the two sentences are the critical scientific questions you try to solve in this study. And I can see the amounts supplement of introduction to several methods. However, the Introduction section still lack of focus. I suggest the authors should re-organize the logic of this part and figure out what should remain and what should omit.

L326 Too suddenly bring out the research area.

L327-L335 Rephrase these sentences, make your methods more logic

L453. Reference?

Figure 1 “(a) (b)” Better inside the subfigure.

Results. Split the description of your result and discussion.

Discussion. Lack of focus. Lack of uncertainty analysis.

Author Response

Dear Reviewers and Editors，

Thank you for your kind comments concerning our manuscript.

We have revised the manuscript according to the reviewer's comment. For responses to specific comments, please see the attachment.

We really appreciate the reviewers for their valuable comments and suggestions, those comments are valuable for revising and improving our paper with important guiding significance.

Thank you again for your time!

Best regards,

Sincerely,

Xinyi Liu

Author Response File: Author Response.docx