Deforestation as the Prominent Driver of the Intensifying Wildfire in Cambodia, Revealed through Geospatial Analysis
Round 1
Reviewer 1 Report
The study has analyzed time-series MODIS data for assessing wildfires in Cambodia. The article has a relatively clear context, but there is some missing info in the manuscript in various sections like product names and resolutions. The introduction needs better literature. Validation of results is missing. The current form of the manuscript needs major improvement. The specific comments are given below.
Major Comments:
1) Abstract: L18: Clearly state which years “over the last two years”.
2) Introduction: It is not properly contextualized on different approaches of wildfire detection using satellite data (satellite indices to ML algorithms) and set-up drivers that cause wildfires. A few suggested articles can help in this regard.
10.1016/j.apgeog.2022.102867
DOI: 10.1038/s41598-022-23697-6
3) Fig 1B. mention clearly what are NFS and FS
4) In L104: Which MODIS product is used? Write the name of the product.
5) Table 1: As temporal resolutions are the same, they can be mentioned once in the text. But add spatial resolution in place of temporal.
6) Forest loss (or deforestation) variable: how it is differentiated between loss due to fire and loss due to other sources? Clarity is missing on the use of this variable. It seems that authors assumed all losses are due to deforestation
7) R2 by superscript R2 throughout the ms
8) Explain Fig 3 along with the negative and positive residuals
9) Fig 4 is missing as after fig 3, I can see fig 5
10) How results based on GWR are validated?
11) What if MODIS data is substituted by “Fire CCI” is provided by Climate change initiative (CCI) project? How this is going to change Figure 5
12) Write the modern statistics such as Machine Learning algorithms for fires studies under discussion
13) Add limitations of the study under discussion.
minor editing required
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors want to investigate the driver of fires in Cambodia across time and space. But the authors fails to bring relevant candidate drivers to the investigation. The main problem comes in the title were deforestation is announced as my main driver for fire, but then the variable that describe it is forest loss, for admission of the authors, include also forest loss by fire. This logical circularity cannot be resolved with a citation, but authors need to make an effort to separate deforestation from forest lost by fire. Otherwise the article do not make any contribution to the causal link between deforestation and fire. If the problem are farmer that want to enlarge rice paddies ( as said by the authors) a driver could be the presence of rice paddies in the 20km frame
Minor:
1)How many fires compared to total were excluded by excluding incomplete macrogrid?
2)This would allow to see how much is loss. How many fires could be caused by problem across the border?
3)Gives a references to all sources of information including Terra-Climate
4)for climate variable why use average across year, climate variable impact at the time of the fire. MODIS gives the exact date for each fire and so exact date climate variable should be used
5)in fig1C mark what is the extent of fire season, for example by color coding the months
6)line 294“ that this increase may be attributed to recent, abnormally heavy rainfall and
drought events that have rendered fires more sensitive to precipitation patterns”
Low rain it is easy to connected to high fire but the contrary not. And indeed standard coefficient is negative in table 2. So why to discuss about abnormally high rainfall? If the GWR has a change in sign for precipitation it should be duly noted and even better mapped. It is an interesting pattern to be investigated
My main perplexity is the use of the word "grid" to refer to single cell within the grid. I had to reread some sentences because not immediately clear.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors have improved the overall quality of the manuscript. However, I expected there should be some discussion on Point 11 (Fire CCI) and Point 12 (Machine Learning algorithms for fires studies). Currently, they are missing from the article.
moderate edit required
Author Response
We appreciate the reviewer's feedback.
We added the discussion of Fire CCI and machine learning algorithms in the discussion section.
Author Response File: 
Author Response.docx
