Landslide Susceptibility Assessment by Machine Learning and Frequency Ratio Methods Using XRAIN Radar-Acquired Rainfall Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript confronts the traditional LSM problem and employs several traditional ML models to compare with FR models, and although the methodology is not innovative enough, the overall structure is clear and the elaboration is meticulous. I recommend minor revision for publication.

1. Line 149 Figure 1. The upper-right map is incorrectly labeled in orange and needs to be redrawn.

2. Line 219 Instead of "Lithology", "geology" is used in other parts of the paper, and the expression of professional terms should be standardized.

3. Line 274 "[42]", not "[42)".

4. Line 362 How about no-landslide points, is the number of them same as landslide points?

5. Line 378 "distance to fault" in Figure 2(g), needs to be changed.

6. Line 725 Figure 19 is not "7 different LSM".

7. Line 753-755 The meaning of the preceding and following statements is opposite, need to improve English writing skills.

Author Response

The author response are as in the attached file. Thank you.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1.      The article doesn't conduct a correlation and importance analysis of all factors to ensure that they are all effective in landslide occurrence.

2.      The machine learning algorithms (random forests, logistic regression, artificial neural networks) and frequency ratios used in the article are already widely used in the field and lack a certain novelty.

3.      The title of the article indicates that the content of the article should not only compare the results between several machine learning methods, but also analyze the impact of the introduced XRAIN radar-acquired rainfall data, but it is not mentioned much in the result analysis.

Author Response

The author responses to the reviewer comments/suggestions are as attached.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

After reviewing the article Geosciences 3022634 I recommend its publication after minor revisions.  The contribution of the article is the application of high quality radar-based precipitation data and an automated parameter tuning methodology to ML approaches to generate Landslide risk maps.  The actual comparison of ML and the FR statistical approaches reflect the results found in earlier studies.  I have made several relatively minor suggestions in the attached file for the authors to consider to improve the manuscript.  I did find the manuscript a little long with some repetition of the methodology  in several places.  The different ML and FR approaches are now well established and their description could be shortened.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Relatively minor changes in word choice required.

Author Response

The author response to the reviewer comments/suggestions are as attached.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I am very happy to see that you have adopted some of my suggestions during the initial review, which has significantly improved the quality of the article. Based on your revision this time, I think the content of the article has been better optimized and improved. Therefore, I have proposed to the editorial department that your article will be accepted for publication.