Wildfire Prediction Model Based on Spatial and Temporal Characteristics: A Case Study of a Wildfire in Portugal’s Montesinho Natural Park

Round 1

Reviewer 1 Report

This study applied several classic machine learning models to predict wildfire risk. This study is okay as a student practice work. In other words, this manuscript does not provide much useful information. Some comments are as below:

1. The wildfire data is imbalanced. How did this study deal with it?

2. The section structure is messed up. Section 3 is not results, which should be in Section 2. Section 4 is not a discussion, part of that should be in Section 2, and the others are results.

3. In figure 1, the color is better changed to represent altitude.

4. figure 2 does not reveal any useful information, which just shows a standard data processing flow. It is better to remove it.

5. Figure 3-6 needs a lot of help. The current figures do not emphasize what you want.

6. Why does KNN can be a baseline in this study? What is the ground truth data for the model validation?

7. The second line in Table 1 should be 1<= Y <= 9.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper is well written and explains the study well. That said, the claimed 'discovery' of a spatio temporal pattern in wildfires is facile at best.

See attached.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This article deals with wildfire prediction with a Spatio-temporal approach. Although the title of the article is attractive, the authors have not performed well in presenting the methods and results of the article. Therefore, I am giving the authors a second chance to be considered for the next round of reviewing if the following corrections are made. My comments are as follows:

Abstract:

1. The full name of the models (XGBoost) should be stated. It has not been well talked about the materials and time of the fires.

introduction:

2. In the introduction, research innovation is not well expressed.

3. What is the reason for using these algorithms? These algorithms have been used in previous forest fire research.

4. The role of k means algorithm is not well defined.

Materials and Methods:

5. The quality of the figures should be improved.

6. Explain the research methodology.

7. State the date of the fire.

8. You mentioned a large number of prediction models along with evaluation criteria in the abstract, and you should explain these methods in this section.

Result:

9. The input and output of machine learning modeling data are not well explained.

10. The quality of all figures should be improved.

11. What percentage of data is used for training and what percentage is used for testing?

12. What is the setting of hyperparameters in these algorithms?

13. The results of Figures 5 and 6 are given on the map of the study area.

14. Prepare the final prediction map using these algorithms.

15. Where are the results of the K-means algorithm stated?

16. Validation of result of K-means algorithm?

17. Figure 7 is related to the results of which algorithm?

18. The discussion section should be stated separately from the results section.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

This manuscript uses a famous open-source database that is wide-used for machine learning practices. The current introduction needs more help to explain why this work benefits the academic community.

Author Response

Thank you for your work and your valuable comments. Our team has carefully considered your valuable comments and has made changes in the latest manuscript.

Point 1: This manuscript uses a famous open-source database that is wide-used for machine learning practices. The current introduction needs more help to explain why this work benefits the academic community.

Response 1: Thank you for your valuable comments. We made minor changes to the "Abstract", "Introduction", and "Conclusions" to highlight the problem we are solving and the innovation of our work.

Reviewer 3 Report

The authors have improved the article, so my decision is to accept the article.

Author Response

Thank you very much for your time and effort in reviewing the paper, and we wish you a pleasant life and successful career!

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.