Progress and Limitations in the Satellite-Based Estimate of Burnt Areas

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I carefully read the manuscript. I found the research interesting, but I have some concerns that I would like to ask the authors to clarify. Please, find the comments attached. Best regards.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

There are some grammatical mistakes and repetitions throughout the whole manuscript. I would suggest the authors keep a coherent vocabulary throughout the text. For example, the authors refer to the first analysis conducted using "long timeseries", "long-time" or "long-term". The same abbreviations are introduced multiple times, often remaining unused.

Author Response

See attached file

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Overall, the research builds interesting analyses, but not enough on the aspect of testing with several indicators that have been mentioned, especially on the comparison of the use of remote sensing databases. Because research needs to provide detailed implications to bring out the novelty of research findings. It is necessary to build a more explicit framework of study results and comparative findings. This includes strengthening the virtues of using remote sensing in the detection of burnt areas, especially applied on pre- and post-event images.

 

Thus, there are several things that need to be done and revised,

 

Whether there is a significant difference in comparing the values of specific spectral indices such as NBR, BAI, and MIRBI, including the use of databases (EFFIS, ESACCI, Copernicus, FIRMS, etc.) this needs to be confirmed in the analysis and discussion. This can also be briefly mentioned in the abstract and conclusion as continuity in this paper.

 

It needs to be confirmed that why Sentinel-2 is considered effective in detecting burnt land, is there no obstacle in cloud cover (0% of cloud cover?), what about Sentinel-3, or a combination of both? This has not been mentioned in this paper, of course, this is important regarding the novelty of this research.

 

If Sentinel-2 has enough resolution to detect burnt land, then does it have enough resolution to detect burned land based on the database used?

 

How is the value generated in each database in detecting burnt land? And how does each database compare in cloud cover constraints? Is it good enough and precise?

 

Technically, how is this data analysed? Does it combine all databases to produce accurate burnt land detection, or is it limited to the comparison of each database?

For example, in the method section: on p. 5 line 169 it is mentioned that this paper covers the years 2019, 2020, and 2021, is the data taken based on the trend of the month of burnt occurrence? And does the database have a limited range of data available? For example, the last 5 years or the last 10 years?

Including table 2 page 5 and line 167, notes need to be given on the table.

 

On page 5 lines 195 to 197, it states that "in the 'local' analysis we left out the COPER (Copernicus Land Service burnt areas) dataset since it seems significantly less accurate than others (see following results)" here it has not been mentioned in detail about the inaccuracy of the data, in what context?

 

In the discussion section, the table is very necessary to give important notes on the main points of the table mentioned.

 

On page 8 line 243, is there any particular factor that makes the CGLS dataset the most relevant compared to the other three databases? And has it been tested in different areas and years? including providing important notes on figures 3 and 4.

 

It needs to be explained that, the advantage of comparative database analysis in detecting the accuracy and size of burnt land is more effective and efficient if multiple data sources are used? Or how to validate the resulting dataset?

 

Figure 13 visualised the distribution of the burned areas according to the three datasets, if referring to the figure the value results are significantly different, it is necessary to explain the data comparison, or make important notes on the figure.

Including Figure 14, does the dataset come from 2018, considering that the data limit mentioned in this research is 2019, 2020, and 2021?

 

Finally, I would like to see more on how this article can compare the availability and accuracy of the databases, especially comparing the values of specific spectral indices of each database, given that remote sensing is one of the most efficient and effective tools in the field of land fires, including the detection of land before and after events. Does the use of Sentinel-2 with the dataset used in this study provide accurate resolution and pixels in the detection of burnt land? This needs to be answered in the conclusion section.

Comments on the Quality of English Language

Minor editing of the English language required

Author Response

See attached file.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The authors present a study that aims to compare various burned area products in Italy, with a focus on the detection and extent of burned areas. While the topic is interesting, the manuscript's overall clarity is lacking, making it challenging to follow. 

Several issues need addressing. The Materials and Methods section is not clear, and the Results and Discussion section brings part of the information that should be in the methodology. There is a lack of a clear statement regarding the study area since the information provided is indirect (lines 153). Additionally, the authors solely focused on determining whether a burned area was detected or not, without considering the cases of non-burned areas being incorrectly classified as burned areas.

Additionally, the manuscript lacks a clear demonstration of novelty and a compelling justification for publication in the remote sensing journal. Therefore, I do not recommend this manuscript for publication in its current form, and I suggest a thorough revision of the entire manuscript.

 

Here are specific comments for improvement:

 

Line 22-23: Where is the study area?

Abstract: The abstract should include some quantitative information to give readers an overview of the results.

Lines 104-109: The objectives of the study need to be more clearly defined.

Materials and Methods: this section should read more like a scientific paper and less like a report. Avoid using excessive hyphens and consider using a flowchart to better illustrate the procedural steps. The manuscript lacks a clear statement about the study area. Please provide a concise description of the study area.

Lines 136-151: Summarize the information in these lines, and consider whether Table 3 can sufficiently explain the datasets.

Table 2: If the table does not provide important information, consider omitting it or incorporating the crucial information into the text.

Lines 173-179: Summarize the information in a more concise manner.

Lines 186-187: Clarify whether Sentinel-2 data were also resampled to 250m.

Lines 195-197: Ensure that any mention of the COPER product is cited before it is introduced.

Lines 203-208: Information presented here should be cited in the methodology section.

Lines 308-309: Provide quantitative data to support the presented results.

Figures: Enhance the quality of the figures and include coordinates and other essential information for clarity.

Line 562: Avoid using speculative language like "probably" in the conclusion without proper evidence.

 

This comprehensive revision should address these concerns and improve the manuscript's overall quality and clarity.

Author Response

See attached file.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Overall the paper attempts to address an important issue of error in burned area datasets. The story gets a little muddled by including long-term, single year (and land cover map differences between source). This is at least partially due to the long list of acronyms, some of which are not used consistently (e.g. CGLSBA = CGLS/COPER = CGLS = COPER).

 

There are two studies here, each of which may require its own communication. The comparative analysis between burned area sources is reasonably informative, though I'd suggest a few  more map examples. Comparison against ground data seems the most informative, however it's not clear that the mismatch between ground data and satellite products is really meaningful depending on the scale and purpose of the ground data collection. Does CUFA include all fire starts? Even if the start doesn't develop into a fire of a certain size? In which case, one would not expect that it be detectable from overhead imagery. Provide a better description of CUFA - what is collected 'on the ground'? What is the minimum mapping unit?

 

I believe these issues need to be resolved before this manuscript is suitable for consideration.

Other comments:

• Use acronyms consistently (even in map legends)
• Use consistent style in the maps
• Line 36 "The long term maps of burned areas allow to assess trends in 36 the fire regimes" and 38 "allows to estimate"
• Line 190 "…tools in Matlab" - which tools
• Line 344 "land cover map" specify don't speculate

Comments on the Quality of English Language

The English is OK, though "burnt" v.s "burned" seems a bit awkward.

Author Response

See attached file.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The paper has been corrected and responded to according to my previous comments, therefore, in my opinion, it can be published in the current form

Author Response

Thank you.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript is improved, but still too long in my opinion. I appreciate that English is not the authors native language. The manuscript still needs work for completeness (table 2 is not complete, etc.)

Comments on the Quality of English Language

I respect that the authors are not native English speakers, but the manuscript has extensive need for improvement it English style, grammar, and spelling. I'd suggest an English editing service.

Author Response

The English of the paper has been further reviewed.