Wildfire Susceptibility Mapping in Greece Using Ensemble Machine Learning
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsI've provided a more detailed breakdown of my thoughts but I have to say I don't think this work is ready, yet, for publication.
That stated aim - see line 126 - is the integration of the Ensemble Max and Ensemble Mean Metrics. These in turn related to many existing methods, but in the conclusion, the production of these methods is the leading thing.
The paper does not draw greatly on statistical methods and that's fine for the most part, but the primary message is pretty much a "place-holder" claim - here is some interesting work which needs to be taken further. As it is, it's interesting but not probative. It's not easy to evaluate just how good the improvements to existing methods really are. The message I get looking at the raster maps 11 to 15 is how similar they are. And simply noting that 83% (cf 70%) of ensemble Max models (cf ensemble Mean) were fires identified as being in areas of high or greater risks, may not mean that much unless there is some contrast between current and reported methods. If I've misssed this then my apologies.
There are sections with high numbers of typos and evidences of haste.
The fire risk problem cries out for good work, and this has all the seeds, but just needs some more tending.
Comments for author File: 
Comments.pdf
The actual use of English is fine. But as I said above, there is a lot of tidying up to do.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsDear editor,
The main goal of the manuscript is to use ensemble machine learning models to create 12 Wildfire Susceptibility Maps (WFSMs) for Greece, emphasising how they can be used as predictors. The handling of the manuscript in relation to the other sections has been nice and understandable. The necessary corrections and suggestions have been made to the manuscript.
The manuscript can be accepted after the corrections in the manuscript and below are made.
The article should be re-evaluated in terms of English and corrected by a native English speaker.
Fire data from the MODIS system needs to be tested and validated.
Stand structure is considered the most important factor affecting fire risk. The failure to consider this factor in the study is regarded as the biggest deficiency. Tree species, development stage and canopy cover are important stand parameters affecting fire risk.
Figure 4-9 combines and shows all figures in one figure on one page.
Write the formula using the formula function in Microsoft Office or etc., not a figure.
How did you classify the wildfire risk? Which method gives border values of each category
5 different fire risk values ​​should be given according to 5 different models.
Discuss the obtained results with the existing literature.
You should identify the shortcomings in this study and propose the next steps for improvement.
Comments for author File: Comments.pdf
The article should be re-evaluated in terms of English and corrected by a native English speaker.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
