The Importance of Dunnian Runoff in Atlantic Forest Remnants: An Integrated Analysis Between Machine Learning and Spectral Indices

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please see the attached file.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Moderate editing of English language required. The manuscript presents several words in Brazilian language! 

Author Response

Thank you for your valuable contributions. Almost all of your suggestions have been incorporated into the paper.

Please find attached the PDF file highlighting the changes that have been made.

Best regards,

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

A major revision is required to substantially enhance the quality of this manuscript, particularly the scientific soundness of the experimental design. My concerns were listed as follows, 

  In lines 47-52, please provide essential references to verify that few studies were conducted to investigate the importance of Dunnian runoff in forests.

  Figure 1 seems unclear. Please try to enhance the visual quality of this figure.

  In lines 122-125, why were two different satellite images dated on different dates used for this study? The Landsat-8 OLI/TIRES data has quite different band information from the PALSAR L-band image. How can the consistency of these two images be maintained? Why did the authors use them? 

  Figure 3 should show the terrain with DEM.

  Why use the Random Forest (RF) classifier rather than using the maximum likelihood algorithm or the other machine-learning-based classifiers? What are the advantages of this method?

  In section 2.2.2. Spectral Indices: NDVI and NDWI, it seems only one day's Landsat-8 data was used for this study. Why not use multiple data to perform the repeatable experiments? 

  In Table 1, "Randon" is misspelled.

  In section 3, only descriptive analysis was performed. What is the quantitative relationship between the predictor and response variables? 

  In Figure 12, the fitted regression models should present the p values or RMSE rather than the correlation coefficient.

  What are the innovative ideas behind this experimental design, the new findings, and the significant limitations? 

Author Response

Dear Reviewer,

I would like to thank you for your detailed and constructive comments. They were crucial in improving the manuscript, and most of your suggestions have been incorporated into the text.

Attached, you will find the revised version of the paper, along with a document containing detailed responses to your questions. Once again, thank you for your valuable contribution, and I kindly ask that you review the updated text after the corrections

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Comments - Questions (to be clarified in the text):

1. Abstract: Please, explain in parentheses what NDWI and NDVI mean!

2. General remark: Numbering of equations is needed.

3. In the equation for p_mk, what is l (y=K)?

4. Line 193: In which equation does Kappa index appear?

5. Section 4.1: (a) Please, explain the term "Hortonian flow"! (b) Which is the difference between Dunnian and Hortonian flow? (c) What does "Hortonian x Dunnian flow" mean?

6. Figure 10: Which index is on y-axis and which index is on x-axis?

7. Line 284: What does "the NDVI x NDWI values" mean?

8. Lines 383-384: Which is the difference between surface and Dunnian runoff?

9. References: Perhaps, the references in Portuguese should be translated into English (you can ask the editorial office).

10. See annotated manuscript! 

 

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Some "editorial" errors were annotated on the manuscript.

Author Response

Dear Reviewer,

Thank you very much for your valuable contributions, which were essential in improving this work. The observations and comments you provided have been incorporated into the text and can be found in the attached documents. Additionally, the updated version of the manuscript is now available for your review.

Best regards,

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Q1. Can the authors numerically label the findings in the abstract, e.g., (1), (2), etc.?

Q2. L30-71: Many points within these paragraphs need improvement, as follows:

- The paragraph introduces multiple topics (e.g., Atlantic Forest preservation, Dunnian runoff, satellite imagery) without clearly connecting them. The relationship between these topics needs to be clearer and more focused. Please revisit and improve.

- The transition from the discussion on the Atlantic Forest to Dunnian runoff is abrupt and lacks a clear connection. Please ensure that the significance of Dunnian runoff in the context of Atlantic Forest preservation is better explained.

- The term "Dunnian runoff" is introduced without sufficient background or explanation for readers unfamiliar with the concept. I recommend that the authors provide more context or a brief explanation earlier in the paragraph to improve clarity. Additionally, the paragraph mentions significant points, such as the reduction of the Atlantic Forest and the importance of Dunnian runoff, but lacks supporting data, references, or examples to substantiate these statements.

- I found that broad statements like "a critical gap in current environmental research" are made without strong evidence or references to justify them. I recommend reviewing at least two previous works and highlighting the current gap that appeared in these works.

- When highlighting the importance of Digital Terrain Model (DTM) and Digital Elevation Model (DEM) for identifying areas with greater slopes and flat areas in this work, please provide references from previous works (suggest 10.1016/j.ejrh.2022.101282 and 10.1007/978-981-19-1600-7_77).

- I found that the authors repeated the significance of vegetation and moisture correlation several times without adding new insights. Please revisit and carefully provide references for these statements.

- I believe that separating section 1.1. Background Dunnian Runoff in the Introduction section is not necessary. Please revisit and consider combining them.

- The literature review conducted is not sufficient and requires improvement. For example, the statement “They play a vital role in regulating water flow, filtering nutrients and sediments, and supporting aquatic biodiversity (L107-108)” discusses the role of surface runoff in the hydrological dynamics of watersheds, but it needs more references from previous works (suggest citing 10.1016/j.jenvman.2024.121375 and 10.1016/j.scitotenv.2024.174289). Additionally, the vital role in regulating water flow, filtering nutrients (10.1016/j.scitotenv.2024.175523) and sediments (10.1007/s00382-024-07319-7).

- Furthermore, please provide more references to strengthen the statement “(L112-113) This directly influences water availability, affecting both base flow and flood events in rivers” (suggest 10.3389/fenvs.2023.1304845 and 10.3850/978-90-833476-1-5_iahr40wc-p1339-cd).

Q3. Please improve section 2.1. Materials as it is poorly written in its current form. I recommend that the authors provide the datasets used in a table with references, spatial and temporal resolutions, etc. Additionally, why use Landsat-08 instead of Sentinel-1 or -2? For example, Sentinel-1 has a higher temporal resolution with a revisit time of 6 days at the equator, making it more suitable for applications that require frequent observations. Also, due to its lower spatial resolution, Landsat (30 m) may not be ideal for detecting small-scale land cover changes, fine-grained urban studies, or detailed agricultural monitoring, which the authors intend to reveal in this work.

Q4. Why Random Forest (RF) and not another method? The authors do not provide a thorough discussion of the limitations of the RF model, such as its sensitivity to the quality of input data and potential overfitting. I recommend a more critical evaluation of the model’s performance and its implications for the results. Additionally, RF is heavily dependent on the quality and representativeness of the training data. If the training dataset is biased or not comprehensive, the model may misclassify land use types, leading to inaccurate classification. In this case, the authors used Landsat imagery, which I believe should be strengthened to make the work stand out from many others.

Q5. Please be aware that the format of Table 1 in the current manuscript might not be accurate. Carefully check this point. In addition, please enlarge Figures 2 to 7 and 9.

Q6. While the conclusion suggests broad implications for watershed management and environmental conservation, the findings are based on a single case study. The authors do not adequately justify how these conclusions can be generalized to other regions or forest types. Please revisit carefully and ensure the conclusions are supported by the current findings.

Q7. In the discussion section, there is a lack of discussion regarding the uncertainties in the data and analysis, particularly regarding the accuracy of the satellite-derived indices and the classification results. Please provide this information.

Q8. Please include a section on future work.

Q9. In conclusion, please highlight the main findings with a brief description (suggest highlighting qualitative results), but keep them short, direct, and concise. The current form is lengthy and difficult to follow.

Q10. Please numerically label the main findings, e.g., (1), (2), etc. 

Author Response

Dear Reviewer,

Thank you very much for your comments, which were instrumental in improving the quality of the article. I have carefully reviewed and addressed each of your suggestions, incorporating the necessary changes to enhance the manuscript's clarity and overall contribution. Please find my responses to your remarks, as well as the updated version of the article, which I hope is now more in line with your valuable feedback.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The Authors addressed my concerns comprehensively and carefully. I have appreciated the new additions: in the Introduction, on the description of the area under study, and the clarification about the limitations of this study. For these reasons my recommendation is "Accept in present form". However, the manuscript still needs several refinements in terms of style and English language. For instance, Figures 1 and 2 are not yet perfect. I hope these refinements will be made at the proofreading stage; but I'm confident, knowing well the professionalism of the Editorial team.

PS I would suggest that Authors avoid (in the future) the use of tracking changes in pdf documents in their responses to Reviewers. The Authors should comment separately each Reviewer's concern and highlight the changes they have made in the paper.

 

Comments on the Quality of English Language

The manuscript needs a thorough review in terms of style and English language.

Author Response

Dear Reviewer,

We sincerely appreciate your comments, which have been instrumental in improving the quality of our manuscript. We would like to address the following points:

1. Figures: We acknowledge the current limitations of the figures, mainly due to page size constraints. However, we are committed to enhancing the quality and clarity of the images while adhering to the journal's editorial guidelines. We will work on optimizing the resolution and resizing them appropriately to ensure the data presented is more legible.

2. Tracking System: We apologize for the confusion regarding the submission of responses. This was our first experience using the MDPI system as authors, and we encountered some initial challenges with the platform. We appreciate your understanding and will be more attentive in future steps to avoid similar errors.

We remain available for any further suggestions and hope the revisions meet your expectations.

Reviewer 2 Report

Comments and Suggestions for Authors

I am pleased to read this revised manuscript. Compared with last version, this version is substantially enhanced and meets the criteria for publication. Congratulations to the authors.

Author Response

Dear Reviewer,

We greatly appreciate your valuable comments and suggestions, which have been essential in improving the quality of our manuscript. Your observations allowed us to reflect on important aspects of the work and make adjustments that strengthened both the clarity and relevance of the presented results.

We remain available for any further clarifications or adjustments that may be required.

Reviewer 3 Report

Comments and Suggestions for Authors

Comments - Questions (to be clarified in the text):

1. Equation (5): What does I(y=k) mean?

2. See annotated manuscript!

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

Thank you very much for your insightful comments, which have enhanced the quality of our manuscript.

Regarding Equation 02, we consulted the source (https://scikit-learn.org/stable/modules/tree.html#mathematical-formulation). The backslash operator () represents the set difference operation, i.e., Qm∖Qmleft(θ)Q_m \setminus Q_m^{left}(\theta)Qm​∖Qmleft​(θ). We have made the necessary adjustments to clarify this notation.

As for Figures 4, 5, 6, and 7, we have resized them, but due to the journal's layout constraints, they may still not be fully legible. To address this, we will send you the original-sized figures as an attachment for better visualization.

We also corrected the alignment of the equation numbers, adjusted the references that were in uppercase, and revised line 325. Additionally, in reference 31, we clarified that INMET refers to the National Institute of Meteorology (Instituto Nacional de Meteorologia).

Moreover, we adjusted the size of Figure 13 for better formatting.

Lastly, regarding Equation 5, we would like to clarify that I(y=k) is an indicator function that takes the value 1 if the sample y belongs to class k and 0 otherwise. In other words, this function helps count how many samples in dataset Q_m belong to class k

Once again, we appreciate the opportunity to revise and improve the manuscript. We remain available for any further clarifications.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for the revision.

Author Response

Dear Reviewer,

We greatly appreciate your valuable comments and suggestions, which have been essential in improving the quality of our manuscript. Your observations allowed us to reflect on important aspects of the work and make adjustments that strengthened both the clarity and relevance of the presented results.

We remain available for any further clarifications or adjustments that may be required.