Forest Fire Spread Monitoring and Vegetation Dynamics Detection Based on Multi-Source Remote Sensing Images
Round 1
Reviewer 1 Report
The article addresses a relevant and appropriate topic for the journal. However, I make some observations to improve the text.
1 - The introduction lacks a paragraph that describes a brief review of fire monitoring using multi-source remote sensing data, highlighting the difference between the present study and the others. There are different studies in this approach, for example:
Coops, N. C., Tompalski, P., Goodbody, T. R., Achim, A., & Mulverhill, C. (2022). Framework for near real-time forest inventory using multi source remote sensing data. Forestry: An International Journal of Forest Research. https://doi.org/10.1093/forestry/cpac015
Bolton, D. K., Coops, N. C., Hermosilla, T., Wulder, M. A., & White, J. C. (2017). Assessing variability in post‐fire forest structure along gradients of productivity in the Canadian boreal using multi‐source remote sensing. Journal of biogeography, 44(6), 1294-1305.
Kganyago, M., & Shikwambana, L. (2020). Assessment of the characteristics of recent major wildfires in the USA, Australia and Brazil in 2018–2019 using multi-source satellite products. Remote Sensing, 12(11), 1803.
2 - Bibliographic references are missing for the NBR (Key and Benson, 2005) and the dNBR (Brewer et al., 2005; Epting et al., 2005)
Key, C. H., & Benson, N. C. (2005). Landscape assessment: remote sensing of severity, the normalized burn ratio and ground measure of severity, the composite burn index. FIREMON: Fire effects monitoring and inventory system Ogden, Utah: USDA Forest Service, Rocky Mountain Res. Station.
Epting, J., Verbyla, D., & Sorbel, B. (2005). Evaluation of remotely sensed indices for assessing burn severity in interior Alaska using Landsat TM and ETM+. Remote Sensing of Environment, 96(3-4), 328-339.
Brewer, CK, W. J., & Redmond, R. L. (2005). Classifying and mapping wildfire severity: A comparison of methods. Photogrammetric Engineering & Remote Sensing, 71, 1311-1320.
3 – The topic “Variable Importance Analysis” should include bibliographic citations.
4 – What were the Random Forest Hyperparameters?
5 – What was done for Random Forest hyperparameter tuning?
6 – There are no bibliographic references for the NDVI (Rouse et al., 1973) and the VFC. What does the acronym VFC mean? Vegetation fractional coverage (Montandon & Small, 2008)?
Rouse Jr, J. W., Haas, R. H., Schell, J. A., & Deering, D. W. (1973). Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation (No. NASA-CR-132982).
Montandon, L. M., & Small, E. E. (2008). The impact of soil reflectance on the quantification of the green vegetation fraction from NDVI. Remote Sensing of Environment, 112(4), 1835-1845.
7 – The figures (1, 2, 3, 4, 5, and 6) have incomplete captions, which should be self-explanatory. It is important to describe the different frames that compose the figure (a, b, c, ...).
Minor corrections
Line 23 - “Data collected from the abovementioned were analyzed by a random forest algorithm, and the key factors, with their order of importance, affecting the spread of the two selected forest fires in Sichuan Province were identified.” I suggest “I suggest “The random forest algorithm analyzed the collected data and identified the main factors, with their order of importance, that affected the spread of the two selected forest fires in Sichuan Province”
Line 21-22 - “In this study, the meteorological, terrain, combustibles and human factors related to the fire were collected.” I suggest “This study collected the meteorological, terrain, combustibles, and human factors related to the fire.”
Line 26-27 - “The results of the study can provide effective information of the fires for Sichuan Province, and be a technical reference for fire spread monitoring, and analysis through remote sensing...” I suggest “The study's results can provide effective information on the fires in Sichuan Province and be a technical reference for fire spread monitoring and analysis through remote sensing…”
Line 53-55 “Forest fires threaten human security, wildlife habitat, regional economies, and global climate change [7-10], therefore, they have become severe natural disasters.” I suggest “Forest fires are severe natural disasters that threaten human security, wildlife habitat, regional economies, and global climate change [7-10].”
Line 55 - “Reducing the occurrence of fires and reducing their damage” I suggest “Reducing fire occurrence and damage has”
Line 58-59 - “However, it is more difficult to measure the fires in remote areas and areas with steep terrain and inconvenient transportation” I suggest “However, transportation make measuring fires in remote areas and steep terrain more difficult.”
Line 67 - Put the meaning of the acronym NASA- National Aeronautics and Space Administration (NASA)
Line 69-70 - “limitations up to 16 days” I suggest “limitations of up to 16 days”
Line 71 - Put the meaning of the acronym MODIS - Moderate Resolution Imaging Spectroradiometer (MODIS)
Line 74 - “important role in acting as a buffer” I suggest “important role as a buffer”
Line 79-81 - “The Sentinel-2 satellite covers 13 spectral bands with ground resolutions of 10 m, 20 m, and 60 m. The Sentinel-2 is the only satellite in optical data that provides red-edge spectral band data [20].” However, the RapidEye Red Edge Band needs to be mentioned. Therefore, I suggest joining the two sentences not addressing the band's exclusivity “The Sentinel-2 satellite covers 13 spectral bands with ground resolutions of 10 m, 20 m, and 60 m and provides red-edge spectral band [20].”
Line 81-82 - “With emergence of” I suggest “With the emergence of”
Line 82- Put the meaning of the acronym GF - GaoFen (GF)
Line 88 - “satellite, the same area images” I suggest “satellite, same area images”
Line 89 - “Since GF-4 satellite” I suggest “Since the GF-4 satellite”
Line 97 – “the most used tools for fire occurrence and monitoring” I suggest “the most used fire occurrence and monitoring tools”
Line 106 - “of single” I suggest “of a single”
Line 107 - “increase observation frequency” I suggest “increase the observation frequency”
Line 110 - “of small number of images” I suggest “of a small number of images”
Line 122 - “There are few research” I suggest “Few research”
Line 135 - “the ability of using” I suggest “the ability to use”
Line 136 - “to collaborate monitor” I suggest “to collaborate in monitoring”
Line 143 - “The Muli County is located in the northwest of Liangshan. It is located between 100°03′ ~ 101°40′ E and 27°40′ ~ 29°10′ N.” I suggest “Muli County is located northwest of Liangshan between 100°03′ ~ 101°40′ E and 27°40′ ~ 29°10′ N.”
Line 147 - “meters with Pinus” I suggest “meters, with Pinus”
Line 149-150 - “Sichuan Province. It is located between 101°46′ ~ 102°25′ E and 27°32′ ~ 28°10′ N.” I suggest “Sichuan Province, between 101°46′ ~ 102°25′ E and 27°32′ ~ 28°10′ N.”
Line 161 - “satellite image data obtained by multiple” I suggest “satellite image data from multiple”
Line 251 -“the importance equation of feature X is” I suggest “the feature X importance equation is”
Line 253 - “Vegetation coverage, an important” I suggest “Vegetation coverage is an important”
Line 275 “The location and shape of the fire line was extracted” I suggest “The location and shape of the fire line were extracted”
Lines 293-294 - “sporadic fire spots was” I suggest “sporadic fire spots were”
Line 399 - “with the less than 10% humidity” I suggest “with less than 10% humidity”
Line 422 - “And due to artificial” I suggest “Moreover, due to artificial”
Line 473 - “information for the fires” I suggest “information about the fires”
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Your study highlights the critical need for application of multi-source datasets for effective fire monitoring. While this topic is of interest to the remote sensing community and would benefit fire managers, you paper fails in a number of areas:
1) Methods unclear: Your study is centered on the use of multi-source datasets. However, you failed to describe how each of the datasets were used. How was Sentinel/Planet/MODIS data used? How were the multiple scales of these data reconciled? In short, your study is not reproducible which is a vital aspect of scientific discourse.
2) No justification for fire severity classes. dNBR based fire severity was developed in the US, so might be translatable to Sichuan. You need to justify your choice here. Given that fire severity mapping was a large part of your study, the whole study falls through without such a vital justification.
3) General presentation of the paper is poor - You need to extensively revise your paper for readability and technical correctness.
Here are my per line comments:
Line 23: Consider replacing "by" with "with" or "using"
Line 34: "...effects were different in the fire areas depending on fire severity." It might be useful here to clearly state how vegetation recovery varies with fire severity e.g. "high vegetation recovery was associated with low ..."
Line 50: Doubt how useful this statement is. Given our long history with fire, its debatable if fire occurrences are incredible. Please consider revising - choose better descriptors.
Line 61: "a waste of resources". Consider replacing this with "costly"
Line 67: Landsat data are not considered a high-resolution imagery. Revise accordingly or justify.
Line 70-71: Confusing statement, please revise. What is the binary star system?
Line 73: Replace "conduct" with "enable" or "support".
Line 82: GF? Please insert full name before abbreviating.
Line 86 - 87: This statement needs justification or a qualification! Is G better than QuickBird or WordView or Planet?
Line 100 - 101: It is possible to map burned areas using a single image. So, please justify this statement. Must be revised too.
Line 114: "fire areas"? You mean burned areas or fire scars? Please use correct terminology. Please check this issue across your paper.
Figure 1: Provide more descriptive caption that highlights the three parts in the figure.
Line 177: Change "radiation" to "radiometric".
Line 199: "combustible"? Meant "fuel"?
Line 211: "the geospatial data cloud". Capitalize where necessary, as in Table 2.
Table 3: How are these thresholds justified for Sichuan?
Line 237 - 238: Not clear what the dependent variable is here. Please clarify.
Line 258 - 260: Insert citation for VFC.
Section 3.1.1. Why is this section under results? It is not a result of any of your analyses, right? Looks more suited for the study area section.
Figure 3: Either get rid of the background map or the images. Both convey the same message. Change "fire" in caption to "fire scar" or "burned area".
Figure 10: This figure could be combined with Figure 6 to emphasize relationship between fire severity and VFC.
Line 358: Change "significant" to "crucial" or "vital".
Line 380: dNBR is not a technology, please revise for technical correctness.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
Reviewer’s Report on the manuscript entitled:
Forest fire spread monitoring and vegetation dynamics detection based on multi-source remote sensing images
The authors investigated the mountain fires occurred in Muli County, on March 28, 2020, and in Jingjiu Township on March 30, 2020, in Sichuan Province, using multi-source satellite remote sensing image data images. I found the manuscript well-written and the results interesting and useful. I have some comments for further improvement.
1) In the method section, please add a flowchart of your monitoring process.
2) Line 77. Some MODIS bands have spatial resolution of 250m. Please clarify this.
3) Other machine learning techniques for forest fire monitoring can be mentioned in the Introduction as well, such as Long Short Term Memory (LSTM)
https://doi.org/10.3390/fire5010013
and Fuzzy login models:
https://doi.org/10.3390/f12081005
4) At the end of the Introduction and by referring to the Section numbers, please describe how the rest of the manuscript is organized.
5) Figure 1. Please enlarge the numbers of the scale bar and latitudes and longitudes. Please ensure that the font size is the same for all the texts and numbers.
6) Line 180. Was there any cloud/shadow in the images? If so, which masking algorithm you used. Please describe here briefly.
7) Lines 198-205 and Line 220. Some more recent references can be added here. For example, the authors of the following paper used statistical inference to forecast potential forest fire risk by considering the factors, such as altitude, slope, distance from roads and rivers, land use/land cover, rainfall, and temperature:
https://doi.org/10.3390/su14073881
8) In the captions of Figures 3 and 5, please say the date of the fire (day, month, year). Also, it would be nice to enlarge the font size a bit (legend, on scale bar, etc.).
9) Figure 6. Please increase the font size.
10) Line 472. Please mention the limitations of this study.
Thank you for your contribution
Regards,
Regards,
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
I would like to congratulate the authors who made all the suggested corrections, improving the understanding of the text.
Reviewer 2 Report
Your latest manuscript is a significant improvement from the initial version. You have also largely addressed my comments on the previous version. Therefore, I would like to recommend tour manuscript for publication.
Reviewer 3 Report
I would like to thank the authors for addressing my comments. The manuscript is improved and I recommend acceptance. Please carefully proofread the manuscript. I found some grammar issues:
Line 78. Grammar issue. Please replace "due to" with "since"
Line 189. "Flowchart" is one word.
