An Improved LandTrendr Algorithm for Forest Disturbance Detection Using Optimized Temporal Trajectories of the Spectrum: A Case Study in Yunnan Province, China
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors,
Your article aims to suggest an improved LandTrendr algorithm to detect different forest disturbances. The new algorithm improves the OA of the original LandTrendr significantly and for that reason, the research is especially valuable. However, besides the experiment itself, the MS has some flaws related to the text and figures that have to be addressed before acceptance.
General comments:
1. In the Introduction section, you should present the state-of-the-art methods for change detection! Referring primarily to articles older than 10 years is not preferable.
2. One of the main flaws of the Introduction is that you have generalized a lot without taking into account important specifics! You have to be careful what you write!
3. I suggest using complex disturbance instead of mutant disturbance! I have never met such a definition of disturbances before. It doesn't sound very clear.
4. I could not understand how you separated the different disturbance types. Please, explain it clearly in the Methods section!
5. The texts in the figures in general are too small and could not be read!
Minor comments:
Lines 35-37 - In general,…. - The role of forests in the global environment is diverse and crucial in many aspects. It doesn't sound good to generalize like that. Better replace in general with some more appropriate phrase!
Line 50 - can be - Repeated words!
Line 51 - Yes, this is a popular method for change detection and there are a lot of more recent publications to refer to. Please, cite some more recent research!
Line 78 - You cannot claim that a method introduced 14 years ago is new in such a dynamic sector as the satellite EO! Please, find a more appropriate phrase!
Lines 84 – 85 - You need to add some references to articles using this algorithm for such purposes on GEE!
Lines 86-89 - The above analysis does not reveal that! You did not compare individual algorithms but general approaches for change detection! In addition, you cannot claim that the algorithm is widely used with a single reference!
Lines 89-91 - References missing! Which are these many studies?
Lines 91-92 - Reference missing!
Please explain why the spectral trajectories of mutant forest disturbance events have not been characterized effectively!
Lines 98-99 - Reference missing!
Lines 113-114 - You have to explain that the precise mutant/complex forest disturbance using LandTrendr algorithm can be detected with the improvement you have suggested! There are a lot of methods, models, and algorithms with high effectiveness in disturbance detection. You have not evaluated and compared the effectiveness of all available algorithms!
Line 117 - The values of the vegetation index are smaller! However, it depends on the index used. Some indices show disturbances with increasing values. If you mean a specific index, you should explain this in the text!
Line 206 - This link is not active!
Lines 222 – 223 - Add a reference or if this statement is based on your observations, make that clear in the text!
Lines 502-503 - Again, you refer to one single paper and generalize!
Probably, the NBR was the optimal index amongst the other indices selected to be studied in this particular study. In other studies, the NBR is not so efficient for such purposes, compared to TCT-based indices for example.
Kind regards!
Author Response
We are truly grateful to the valuable comments and suggestions from the associate editor and the anonymous reviewers. Based on these comments and suggestions, we have made careful modifications on the manuscript. During the revision, we put the main efforts into improving the accuracy of the English language expression, the logic of language organization, and the lack of references in the abstract, introduction, and conclusion sections. Further, we answer the comments and suggestions of the reviewers point by point and item by item. For convenience, all changes made to the paper are marked out with red font.
Firstly, we have revised some descriptions of the introduction and methodology to make sure that the advantages of the proposed algorithm are clearly explained. Secondly, we increased the resolution and the size of the text of all the pictures in the article for better reading. Finally, we have revised some grammatical errors and inappropriate statements in the manuscript.
We hope that this revised version has properly addressed the issues of the reviewers and is more suitable for publication. Thank you very much.
Dear editors and reviewers,
Thank you very much for your letter about manuscript revision. We are very grateful for your comments and suggestions on the manuscript. According to the comments, many related sections have been improved. All the Commentss have been explained in detail as follows, the relevant contents have been amended in the revised manuscript, and the revised parts were marked out with red font.
The revised manuscript has been uploaded to the journal website for your approval in the following order, " forests-3157203-ori(0821)", "covering_letter", and " Figure".
Reviewer #1:
Thanks very much for the valuable comments and excellent suggestions, which are very helpful for us. According to each comment, the relevant content has been revised or explained as follows:
General comments:
Comments 1: In the Introduction section, you should present the state-of-the-art methods for change detection! Referring primarily to articles older than 10 years is not preferable.
Thank you for your scrupulous reminding. Sorry to ignore it. We have revised it in the introduction section. Please see the details of the introduction section in the revised manuscript (page1-3, lines 33-123).
Comments 2: One of the main flaws of the Introduction is that you have generalized a lot without taking into account important specifics! You have to be careful what you write!
Response 2: Thank you for your scrupulous reminding. We have re-written and reorganized the introduction. Especially, compared with the original manuscript, the specifics of improved LandTrendr algorithm were described in detail. Please see the details of the introduction section in the revised manuscript (page1-3, lines 33-123).
In the revised introduction section, the central line is as follows:
- Significance and importance of forest disturbance.
- Introduction of main methods used in current studies of forest disturbance.
- The limitations of LandTrendr algorithm in complex forest disturbance mapping and the mechanism for solving these limitations are proposed in this paper.
- Research objectives.
Comments 3: I suggest using complex disturbance instead of mutant disturbance! I have never met such a definition of disturbances before. It doesn't sound very clear.
Response 3: Thank you for your good advice. We have replaced all ‘mutant disturbance’ with ‘complex forest disturbance types’ or ‘abrupt forest disturbance events’. Please see the details in the revised manuscript.
Comments 4: I could not understand how you separated the different disturbance types. Please, explain it clearly in the Methods section!
Response 4: Thank you for your scrupulous reminding. Sorry to ignore it. We have explained it in the method section. Please see the details in the revised manuscript (page 7, Lines 215-216).
Comments 5: The texts in the figures in general are too small and could not be read!
Response 5: Thank you for your scrupulous reminding. Sorry to ignore it. We have increased the resolution and the size of the text of all the pictures in the article. Please see the details in the revised manuscript.
Minor comments:
Comments 1: Lines 35-37 - In general,…. - The role of forests in the global environment is diverse and crucial in many aspects. It doesn't sound good to generalize like that. Better replace in general with some more appropriate phrase!
Response 1: Thank you for your excellent suggestion. We have removed it by checking the language of the context. Please see the details in the revised manuscript (page 1, Lines 37-38).
Comments 2: Line 50 - can be - Repeated words!
Response 2: Thank you for your scrupulous reminding. Sorry to ignore it. We have removed it. Please see the details in the revised manuscript (page 2, Line 52).
Comments 3: Line 51 - Yes, this is a popular method for change detection and there are a lot of more recent publications to refer to. Please, cite some more recent research!
Response 3: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text. Please see the details in the revised manuscript (page 2, Lines 54-56).
Comments 4: Line 78 - You cannot claim that a method introduced 14 years ago is new in such a dynamic sector as the satellite EO! Please, find a more appropriate phrase!
Response 4: Thank you for your scrupulous reminding. We have We have re-written and reorganized this section in the introduction. Please see the details in the revised manuscript (page 2, Lines 60-80).
Comments 5: Lines 84 – 85 - You need to add some references to articles using this algorithm for such purposes on GEE!
Response 5: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text. Please see the details in the revised manuscript (page 2, Lines 75-78).
Comments 6: Lines 86-89 - The above analysis does not reveal that! You did not compare individual algorithms but general approaches for change detection! In addition, you cannot claim that the algorithm is widely used with a single reference!
Response 6: Thank you for your scrupulous reminding. We have revised the language expression and added more references to support this view. Please see the details in the revised manuscript (page 2, Lines 75-79).
Comments 7: Lines 89-91 - References missing! Which are these many studies?
Response 7: Thank you for your scrupulous reminding. Sorry, this view is too absolute. It should be ‘there are still two limitations in LandTrendr that may hinder its applicability in forest disturbance mapping’ in here. It has been revised in the manuscript. Please see the details in the revised manuscript (page 2, Lines 79-80).
Comments 8: Lines 91-92 - Reference missing!
Please explain why the spectral trajectories of mutant forest disturbance events have not been characterized effectively!
Response 8: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text. Please see the details in the revised manuscript (page 2, Lines 84-85).
SG filter is a linear filter, it is difficult to capture the true change trends of abrupt signals with nonlinear. The disturbance signals from certain forest disturbance events, especially for forest disturbance with abrupt signals in the multi-temporal data, such as forest fire, deforestation, and urbanization, may be misjudged by the SG filter as data noise. These misjudges would introduce uncertainties in downstream complex forest disturbance detection.
Comments 9: Lines 98-99 - Reference missing!
Response 9: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text. Please see the details in the revised manuscript (page 2, Lines 96-97).
Comments 10: Lines 113-114 - You have to explain that the precise mutant/complex forest disturbance using LandTrendr algorithm can be detected with the improvement you have suggested! There are a lot of methods, models, and algorithms with high effectiveness in disturbance detection. You have not evaluated and compared the effectiveness of all available algorithms!
Response 10: Thank you for your scrupulous reminding. Sorry, these sentences do not accurately express my true thoughts. We have rewritten and reorganized the reasons why the mapping accuracy of the LandTrendr algorithm for complex forest disturbance can be improved using the suggestion proposed by this paper. Please see the details in the revised manuscript (page 2, Lines 89-95).
The accuracy of forest disturbance mapping is limited by the quality of the temporal trajectory of the spectral index. Although the SG filter can effectively remove residual noise while maintaining the change trends of NBR time-series data. It is difficult to keep a good balance between data smoothing and preserving the changing trend when there are abrupt signals in the multi-temporal data. It can be seen from section 5.1 in the revised manuscript, that the spectral trajectory generated by the constrained SG filter algorithm well retains the changing trend of disturbance year, which was more consistent with the actual situation.
Comments 11: Line 117 - The values of the vegetation index are smaller! However, it depends on the index used. Some indices show disturbances with increasing values. If you mean a specific index, you should explain this in the text!
Response 11: Thank you for your scrupulous reminding. It should be “the spectral indices, like the normalized burn ratio (NBR) [47], which is high sensitivity to forest disturbance and recovery, were generally lower in the year of forest disturbance is lower than in the previous or following year”. Please see the details in page 3, Lines 108-111.
Comments 12: Line 206 - This link is not active!
Response 12: Thank you for your scrupulous reminding. We have deleted it in the revised manuscript. Please see the details in page 6, Line 195.
Comments 13: Lines 222 – 223 - Add a reference or if this statement is based on your observations, make that clear in the text!
Response 13: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text. Please see the details in the revised manuscript (page 7, Line 216).
Comments 14: Lines 502-503 - Again, you refer to one single paper and generalize!
Probably, the NBR was the optimal index amongst the other indices selected to be studied in this particular study. In other studies, the NBR is not so efficient for such purposes, compared to TCT-based indices for example.
Response 14: Thank you for your scrupulous reminding. Sorry to ignore it. It should be “NBR is one of the widely used spectral indices in forest disturbance mapping due to its high sensitivity to forest disturbance and recovery”. Meanwhile, we have added more references to support the idea of this paper. Please see the details in the revised manuscript (page 17, Line 497).
Thanks again for all of your good ideas and excellent suggestions, we appreciate it very much.
Thank you and best regards.
Yours sincerely,
Li He
E-mail: [email protected].
*Corresponding author:
Name: Liang Hong
E-mail: [email protected].
Reviewer 2 Report
Comments and Suggestions for AuthorsThe current manuscript provided an improved algorithm to detect forest disturbance. The writing is generally good, and the reviewer has a few comments:
1) Line 50: redundant “can be”
2) Line 55: a space is needed after [9]
3) Introduction: The two categories delineated in lines 48-85 are not entirely separate. Can the second category be considered a specialized application of the first one?
4) Lines 89-91: Please name a few studies that said this.
5) Line 148: Please watch for the significant figures. There are similar incidents throughout the manuscript.
6) Figure 1: Too small font. Please check other figures for similar problems.
7) Line 206: Not the right way to cite.
Comments on the Quality of English LanguageNo specific issues
Author Response
Thanks very much for so many valuable comments and excellent suggestions, they are very helpful to us. According to each comment, the relevant content has been revised or explained as follows:
General comments:
Comments 1: Line 50: redundant “can be”
Response 1: Thank you for your scrupulous reminding. Sorry to ignore it. We have removed it. Please see the details in the revised manuscript (page 2, Line 52).
Comments 2: Line 55: a space is needed after [9]
Response 2: Thank you for your scrupulous reminding. Sorry to ignore it. We have added a space on page 2 line 54.
Comments 3: Introduction: The two categories delineated in lines 48-85 are not entirely separate. Can the second category be considered a specialized application of the first one?
Response 3: Thank you for your scrupulous reminding. Sorry, we think these two categories are independent in terms of the number of remote sensing images and the detection methods. The change detection methods based on bi-temporal remote sensing images detected forest disturbance by comparing the spectral feature or change in land cover types. However, the temporal trajectory change detection from forest disturbance mapping using the changing trend in multi-temporal Landsat data. Please see the details in the revised manuscript (page 2, Lines 50-80).
Comments 4: Lines 89-91: Please name a few studies that said this.
Response 4: Thank you for your excellent suggestion. Sorry, this view is too absolute. It should be ‘there are still two limitations in LandTrendr that may hinder its applicability in forest disturbance mapping’ in here. It has been revised in the manuscript. Please see the details in the revised manuscript (page 2, Lines 79-80).
Comments 5: Line 148: Please watch for the significant figures. There are similar incidents throughout the manuscript.
Response 5: Thank you for your scrupulous reminding. Sorry to ignore it. We have increased the resolution and the size of the text of all the pictures in the article. Please see the details in the revised manuscript.
Comments 6: Figure 1: Too small font. Please check other figures for similar problems.
Response 6: Thank you for your scrupulous reminding. Sorry to ignore it. We have increased the resolution and the size of the text of all the pictures in the article. Please see the details in the revised manuscript.
Comments 7: Line 206: Not the right way to cite.
Response 7: Thank you for your scrupulous reminding. Sorry to ignore it. We have modified the cited format on page 6 Line 195.
Thanks again for all of your good ideas and excellent suggestions, we appreciate it very much.
Thank you and best regards.
Yours sincerely,
Li He
E-mail: [email protected].
*Corresponding author:
Name: Liang Hong
E-mail: [email protected].
Reviewer 3 Report
Comments and Suggestions for AuthorsComments
Abstract: Overall, the abstract needs to be restructured. The following points need to be made clear: the objective of the paper, the methodology, and the main results. These elements should be well described in the abstract. Additionally, I suggest removing the last sentence of the abstract, as the authors did not assess these points in the paper. How can it be concluded that, based on remote sensing, diseases were the main disturbance factors?
In addition to restructuring the abstract, I also suggest addressing the following aspects:
- Clarity and Cohesion: The abstract should be clear and cohesive. Some sentences seem confusing or poorly structured. For instance, the sentence “...ignores the quality of the time trajectory of the spectrum and its changing trend...” could be rephrased for greater clarity.
- Excessive Technical Details: The abstract includes technical details that may not be necessary, such as the specific description of the "constrained Savitzky-Golay (SG) filter" algorithm. In the abstract, it is preferable to focus on the impact and relevance of the methodology, leaving technical details for the body of the paper.
- Study Motivation: While the text mentions the importance of detecting forest disturbances, it would be helpful if the abstract more clearly highlighted the motivation for improving LandTrendr. Why is it important to address the mentioned limitations?
- Relevance and Contribution: The abstract should also clearly state the original contribution of the study. Why is the proposed improvement to the algorithm significant? How does it advance the state of the art?
- Conclusion: The abstract lacks a conclusion that synthesizes the study’s findings and broader implications. After presenting the results, a final sentence that emphasizes the importance of the work and possible future applications would be beneficial.
1. Introduction
- In lines 33 and 34, it is important that the authors cite studies for each mentioned natural factor.
- In line 34, the authors only mention deforestation and urbanization. I suggest including other human-caused problems, such as pollution and climate change. I recommend reading the following articles:
- Dos Santos, A. M., et al. (2021). Deforestation drivers in the Brazilian Amazon: assessing new spatial predictors. Journal of Environmental Management, 294, 113020.
- da Silva, C. F. A., et al. (2023). Spatial modelling of deforestation-related factors in the Brazilian semi-arid biome. International Journal of Environmental Studies, 80(4), 1021-1040.
- Trigueiro, W. R., et al. (2020). Uncovering the spatial variability of recent deforestation drivers in the Brazilian Cerrado. Journal of Environmental Management, 275, 111243.
- In line 41, when discussing the regional scale, it would be interesting to also address local and global scales.
- The sentence in line 52 requires a citation.
- In line 61, when mentioning forest fires and deforestation, the authors should include citations.
- In line 63, the sentence requires a reference.
- In line 99, the authors state that forest disturbance types in large areas are complex and diverse. Please elaborate on this point.
- The paragraph between lines 123 and 135 is confusing, and the study’s objective is unclear. By presenting a general objective and then listing four specific objectives, the text creates a sense of repetition and lack of cohesion. I suggest restructuring the paragraph, clearly and directly highlighting the main objective of the study. The methodological steps mentioned should be removed, as they do not make sense as objectives.
2. Study Area and Datasets
2.1 Study Area
- In line 143, a citation is needed.
- In the study area description, I suggest that the authors present the existing phytophysiognomy type and cite the source.
- It is important to provide the population of the study area and cite the corresponding source.
- Regarding the maps, the study area map should be more prominently displayed. Currently, the map has poor resolution, and the scale does not allow for clear visualization of the area. I suggest separating and presenting a single map of the study area, including China and neighboring countries, as well as highlighting the study area with a title, legend, scale, orientation, and marginal information. Then, the other maps can be presented. For the rainfall map (C), I suggest using a blue color palette, as the current palette is not aesthetically pleasing.
Topic 2.2:
- I suggest that the authors create a table with details of each scene, year by year. Additionally, it would be helpful to create another table with sensor details, including resolution and source.
Topic 2.3:
- When mentioning an applied method, it is important to explain it in the methods section.
Topic 2.4:
- I suggest introducing a table.
- A DOI should not be cited in the text (line 206). The authors should cite the author, not a link.
3. Methods
- Overall, I suggest that the authors include more citations, as there is a noticeable lack of literary support.
4. Results
- In line 383, the authors mention "other factors." What are these factors?
- In Figure 8, what do a1, a2, a3, and a4 represent? The meaning of each should be clarified.
- In Figure 9, a legend explaining what each color represents should be added.
5. Discussion
- Overall, the discussion section does not correspond to a true discussion but rather continues the description of the results. I suggest that the authors move this section to the results and create a discussion section where they compare their main results with international literature, discussing the implications of their findings.
6. Conclusion
- The conclusion section needs to be rewritten. Although the authors state that the study improved the LandTrendr algorithm, this claim needs to be revised.
- Another point is that the authors claim that natural factors, such as climate change, natural disasters, and soil erosion, were the main drivers of forest disturbances in this region. Did the authors verify this through the study?
- The conclusion should be based on the results and, most importantly, should address the research objective proposed at the beginning. Thus, the authors should rewrite the conclusion, clearly presenting the main results and their implications. Additionally, I suggest that the authors include recommendations for future work.
Comments on the Quality of English Language
The text needs English corrections.
Author Response
Thanks very much for so many valuable comments and excellent suggestion, they are very helpful for us. According to each comment, the relevant content has been revised or explained as follows:
Abstract: Overall, the abstract needs to be restructured. The following points need to be made clear: the objective of the paper, the methodology, and the main results. These elements should be well described in the abstract. Additionally, I suggest removing the last sentence of the abstract, as the authors did not assess these points in the paper. How can it be concluded that, based on remote sensing, diseases were the main disturbance factors?
In addition to restructuring the abstract, I also suggest addressing the following aspects:
- Clarity and Cohesion: The abstract should be clear and cohesive. Some sentences seem confusing or poorly structured. For instance, the sentence “...ignores the quality of the time trajectory of the spectrum and its changing trend...” could be rephrased for greater clarity.
- Excessive Technical Details: The abstract includes technical details that may not be necessary, such as the specific description of the "constrained Savitzky-Golay (SG) filter" algorithm. In the abstract, it is preferable to focus on the impact and relevance of the methodology, leaving technical details for the body of the paper.
- Study Motivation: While the text mentions the importance of detecting forest disturbances, it would be helpful if the abstract more clearly highlighted the motivation for improving LandTrendr. Why is it important to address the mentioned limitations?
- Relevance and Contribution: The abstract should also clearly state the original contribution of the study. Why is the proposed improvement to the algorithm significant? How does it advance the state of the art?
- Conclusion: The abstract lacks a conclusion that synthesizes the study’s findings and broader implications. After presenting the results, a final sentence that emphasizes the importance of the work and possible future applications would be beneficial.
Response 1: Thank you for your excellent suggestion. We have rewritten and reorganized the abstract. The overall structure has been adjusted. We have introduced the basic idea of improving LandTrendr for forest disturbance mapping on page 1 lines 13-19, rather than the detail of the method. Such revisions highlight the original contribution of this paper. Please see the details in the revised manuscript.
- Introduction
Comments 1: In lines 33 and 34, it is important that the authors cite studies for each mentioned natural factor.
在第33行和第34行,重要的是作者引用了每个提到的自然因素的研究。
Thank you for your scrupulous reminding. Sorry to ignore it. We have added references to the text for each mentioned natural factor. Please see the details in the revised manuscript (page 1, Lines 35-36).
Comments 2: In line 34, the authors only mention deforestation and urbanization. I suggest including other human-caused problems, such as pollution and climate change. I recommend reading the following articles:
- Dos Santos, A. M., et al. (2021). Deforestation drivers in the Brazilian Amazon: assessing new spatial predictors. Journal of Environmental Management, 294, 113020.
- da Silva, C. F. A., et al. (2023). Spatial modelling of deforestation-related factors in the Brazilian semi-arid biome. International Journal of Environmental Studies, 80(4), 1021-1040.
- Trigueiro, W. R., et al. (2020). Uncovering the spatial variability of recent deforestation drivers in the Brazilian Cerrado. Journal of Environmental Management, 275, 111243.
Response 2: Thank you for your good advice. Sorry to ignore it. Yes, human-caused problems, such as pollution and climate change, also affect the occurrence of forest disturbance. We have added other human-caused problems as the divers of forest disturbance. Please see the details in the revised manuscript (page 1, Line 37).
Comments 3: In line 41, when discussing the regional scale, it would be interesting to also address local and global scales
Response 3: Thank you for your excellent suggestion. We have modified it to “local and global scales” on Page 1, Line 42.
Comments 4: The sentence in line 52 requires a citation.
Response 4: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references on Page 2, Line 51.
Comments 5. In line 61, when mentioning forest fires and deforestation, the authors should include citations.
Response 5: Thank you for your scrupulous reminding. Sorry to ignore it. We have reorganized this section of the current methodology review. Please see the details in the revised manuscript (page 1, Lines 50-58).
Comments 6: In line 63, the sentence requires a reference.
Response 6: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references on Page 2, Line 60.
Comments 7: In line 99, the authors state that forest disturbance types in large areas are complex and diverse. Please elaborate on this point.
Response 7: Thank you for your scrupulous reminding. Sorry, these sentences do not accurately express my true thoughts. It should be “The characteristics of the temporal trajectory under the different forest disturbance types (e.g., abrupt, gradual, and multiple) may be quite different. The temporal trajectory of abrupt forest disturbance, like wildfire, changes rapidly, while forest degradation changes slowly.” Please see the details in the revised manuscript (pages 2-3, Lines 98-101).
Comments 8: The paragraph between lines 123 and 135 is confusing, and the study’s objective is unclear. By presenting a general objective and then listing four specific objectives, the text creates a sense of repetition and lack of cohesion. I suggest restructuring the paragraph, clearly and directly highlighting the main objective of the study. The methodological steps mentioned should be removed, as they do not make sense as objectives.
Response 8: Thank you for your excellent suggestion. We have rewritten and reorganized the research objectives. Please see the details in the revised manuscript (page 3, Lines 117-123).
- Study Area and Datasets
2.1 Study Area
Comments 1: In line 143, a citation is needed.
Response 1: Thank you for your scrupulous reminding. Sorry to ignore it. We have added references on Page 3, Line 131.
Comments 2: In the study area description, I suggest that the authors present the existing phytophysiognomy type and cite the source.
Response 2: Thank you for your excellent suggestion. We have added some description of phytogeographic divergence on page 3, Lines 134-138.
Comments 3: It is important to provide the population of the study area and cite the corresponding source.
Response 3: Thank you for your excellent suggestion. We have added some description of the population of the study area on page 3, Line 139.
Comments 4: Regarding the maps, the study area map should be more prominently displayed. Currently, the map has poor resolution, and the scale does not allow for clear visualization of the area. I suggest separating and presenting a single map of the study area, including China and neighboring countries, as well as highlighting the study area with a title, legend, scale, orientation, and marginal information. Then, the other maps can be presented. For the rainfall map (C), I suggest using a blue color palette, as the current palette is not aesthetically pleasing.
Response 4: Thank you for your excellent suggestion. We have readjusted the resolution and color palette of Figure 1.
Topic 2.2:
Comments 1: I suggest that the authors create a table with details of each scene, year by year. Additionally, it would be helpful to create another table with sensor details, including resolution and source.
Response 1: Thank you for your good advice. We have added a table to introduce the detailed information of the Landsat sensor used in this paper. For more detailed information about this Comments, please see Table 1.
Topic 2.3:
Comments 1: When mentioning an applied method, it is important to explain it in the methods section.
Response 1: Thank you for your excellent suggestion. We have explained in detail the method used and the reasons for using it on page 5, Lines 168-172.
Topic 2.4:
Comments 1: I suggest introducing a table.
Response 1: Thank you for your good advice. We have added a table to introduce detailed information on the data source of validated samples used in this paper. For more detailed information about this Comments, please see Table 2.
Comments 2: A DOI should not be cited in the text (line 206). The authors should cite the author, not a link.
Response 2: Thank you for your scrupulous reminding. Sorry to ignore it. We have modified it on Page 6, Line 196.
- Methods
Comments 1: Overall, I suggest that the authors include more citations, as there is a noticeable lack of literary support.
Response 1: Thank you for your excellent suggestion. We reorganized the language expression of the method section and adopted more literature to support the basic idea proposed in this paper. Please see the details in the revised manuscript (pages 7-8, Lines 214-237).
- Results
Comments 1: In line 383, the authors mention "other factors." What are these factors?
Response 1: Thank you for your scrupulous reminding. Sorry to confused it. It should be “the year of forest disturbance shows a significant spatial heterogeneity in Yunnan province” on page 12, Line 374.
Comments 2: In Figure 8, what do a1, a2, a3, and a4 represent? The meaning of each should be clarified.
Response 2: Thank you for your excellent suggestion. We have clarified listed the meaning of different pronouns on Page 13, Line 385.
Comments 3: In Figure 9, a legend explaining what each color represents should be added.
Response 3: Thank you for your good advice. We have added a legend to explaining in Figure 9.
- Discussion
Comments 1: Overall, the discussion section does not correspond to a true discussion but rather continues the description of the results. I suggest that the authors move this section to the results and create a discussion section where they compare their main results with international literature, discussing the implications of their findings.
Response 1: Thank you for your excellent suggestion. Sorry, we didn’t introduce the meaning clearly. In this section, we want to discuss the effect of the quality of temporal trajectory and the constraints adopted in the SG filter on forest disturbance mapping in the improved LandTrendr algorithm proposed in this paper. Please see the details in the discussion section of revised manuscript.
- Conclusion
Comments 1: The conclusion section needs to be rewritten. Although the authors state that the study improved the LandTrendr algorithm, this claim needs to be revised.
Another point is that the authors claim that natural factors, such as climate change, natural disasters, and soil erosion, were the main drivers of forest disturbances in this region. Did the authors verify this through the study?
The conclusion should be based on the results and, most importantly, should address the research objective proposed at the beginning. Thus, the authors should rewrite the conclusion, clearly presenting the main results and their implications. Additionally, I suggest that the authors include recommendations for future work.
Response 1: Thank you for your excellent suggestion. We have rewritten and reorganized the conclusion section according to your advice mentation above. Please see the details in the revised manuscript (pages 17-18, Lines 508-519).
Thanks again for all your good ideas and suggestions, we appreciate it very much.
Some revised parts not mentioned in the text were also marked out with red font.
Thanks again for all of your good ideas and excellent suggestions, we appreciate it very much.
Thank you and best regards.
Yours sincerely,
Li He
E-mail: [email protected].
*Corresponding author:
Name: Liang Hong
E-mail: [email protected].
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsDear, Authors!
Thank you for taking into account all my remarks and recommendations! The article has been greatly improved. Good work!
Kind regards!
Author Response
Comments 1: Dear, Authors!
Thank you for taking into account all my remarks and recommendations! The article has been greatly improved. Good work!
Kind regards!
Response 1: Thank you very much for your recognition, we strive to do better in our future work.
Thank you and best regards.
Yours sincerely,
Li He
E-mail: [email protected].
*Corresponding author:
Name: Liang Hong
E-mail: [email protected].
Reviewer 3 Report
Comments and Suggestions for AuthorsGeneral comments
Overall, the authors have done an excellent job. However, further improvements are needed before the article can be published. Below are my recommendations:
Abstract: The authors did not clearly present the objective and methodology. I suggest that the authors rewrite parts of the abstract. Example: The aim of the study is to carry out, verify...
In addition, the authors should make the methodology clear in the abstract.
Introduction: As requested in the previous review, the authors should deepen the discussion on the regional and global scale.
2.1. Study Area: The following sentence is unclear: "According to a 2021 government work report, the population and vegetation coverage of Yunnan Province are approximately 47.2 million [52] and 65.04%." The vegetation covers 65.04% of what? I recommend rewriting for clarity.
Figure 1A: The map still lacks good cartographic quality. I suggest the authors remove the toponymic information of neighboring countries and keep only China, as the scale does not allow for good visualization. Additionally, the authors should provide information on the Datum, the coordinate system used, and the data sources for the map.
Discussion: The authors insist on maintaining the discussion as it is. Still, this section should involve an analysis of the results compared to other studies, not just a description of the results. If the section is titled "Discussion," then there should be a discussion based on different studies. Currently, the section titled "Discussion" is more of a description of the results. I think the authors should revise this section to include comparisons with other studies.
Figures 10, 11, and 12: These figures represent results and should not be included in the discussion section, as this confuses the reader. Results should be presented in the results section, while the discussion should be reserved for interpreting those results.
Conclusion: As suggested in the first round of reviews, the authors should include a paragraph with recommendations for future work.
Comments on the Quality of English LanguageThe authors need to do a minor revision of the English.
Author Response
Dear editors and reviewers:
Thank you very much for your letter about manuscript revision. We are very grateful for your comments and suggestions on the manuscript. According to the comments, we have made careful modifications on the manuscript many related sections have been improved. During the revision, we put the main efforts into improving the accuracy of the English language expression and the logic of language organization. Further, we answer the comments and suggestions of the reviewers point by point and item by item. For convenience, the relevant contents have been amended in the revised manuscript, and the revised parts were marked out with red font.
Firstly, we have revised some descriptions of the Abstract to make sure that the advantages of the proposed algorithm are clearly explained. Secondly, we re-drawn the spatial disturbance map of the study area in Figure 1. Finally, we have revised some grammatical errors and inappropriate statements in the manuscript.
We hope that this revised version has properly addressed the issues of the reviewers and is more suitable for publication. Thank you very much.
Comments 1: Overall, the authors have done an excellent job. However, further improvements are needed before the article can be published. Below are my recommendations:
Abstract: The authors did not clearly present the objective and methodology. I suggest that the authors rewrite parts of the abstract. Example: The aim of the study is to carry out, verify...
In addition, the authors should make the methodology clear in the abstract.
Response 1: Thank you for your excellent suggestion. We have re-written the objective and methodology in the abstract section. Please see the details in the revised manuscript (Page 1, Lines12-21).
Comments 2: Introduction: As requested in the previous review, the authors should deepen the discussion on the regional and global scale.
Response 2: Thank you for your scrupulous reminding. Sorry to neglect it. We have re-originated the research background and the research objectives of introduction. Please see the details in the revised manuscript (Pages 1-3, Lines35-123)
Comments 3: 2.1. Study Area: The following sentence is unclear: "According to a 2021 government work report, the population and vegetation coverage of Yunnan Province are approximately 47.2 million [52] and 65.04%." The vegetation covers 65.04% of what? I recommend rewriting for clarity.
Response 3: Thank you for your scrupulous reminding. Sorry to confused it. We have modified it to “According to a 2021 government work report, the population of Yunnan Province is approximately 47.2 million, and the forest area accounts for about 65.04% of the total area.” Please see the details in the revised manuscript (Page 3, Lines136-137)
Comments 4: Figure 1A: The map still lacks good cartographic quality. I suggest the authors remove the toponymic information of neighboring countries and keep only China, as the scale does not allow for good visualization. Additionally, the authors should provide information on the Datum, the coordinate system used, and the data sources for the map.
Response 4: Thank you for your good advice. We have removed the toponymic information of neighboring countries and keep only China in Figure 1a. Meanwhile, we have added the reference of data source for these maps. Please see the details in the revised manuscript (Page 4, Lines144-148)
Comments 5: Discussion: The authors insist on maintaining the discussion as it is. Still, this section should involve an analysis of the results compared to other studies, not just a description of the results. If the section is titled "Discussion," then there should be a discussion based on different studies. Currently, the section titled "Discussion" is more of a description of the results. I think the authors should revise this section to include comparisons with other studies.
Figures 10, 11, and 12: These figures represent results and should not be included in the discussion section, as this confuses the reader. Results should be presented in the results section, while the discussion should be reserved for interpreting those results.
Response 5: Thank you for your scrupulous reminding. We have re-written and re-organized the section results and discussion. Specifically, Figures 10, 11, and 12 have been moved to the Result section. Also, we have discussed the mapping accuracy difference between the forest disturbance mapping generated by iLandTrendr and the existing Global Forest gain and loss product in the discussion section. Please see the details in the revised manuscript (Page 17, Lines 506-511)
Comments 6: Conclusion: As suggested in the first round of reviews, the authors should include a paragraph with recommendations for future work.
Response 6: Thank you for your excellent suggestion. We have added some recommendations for future work in section conclusion. Please see the details in the revised manuscript (Page 18, Lines 541-544)
Thanks again for all of your good ideas and excellent suggestions, we appreciate it very much.
Thank you and best regards.
Thank you and best regards.
Yours sincerely,
Li He
E-mail: [email protected].
*Corresponding author:
Name: Liang Hong
E-mail: [email protected].
E-mail: [email protected].