Use of Remotely Piloted Aircraft System Multispectral Data to Evaluate the Effects of Prescribed Burnings on Three Macrohabitats of Pantanal, Brazil

Round 1

Reviewer 1 Report

The authors analyzed the effects of prescribed burns at different periods of the season where this practice is performed in the tropical wetland biome of the Pantanal in Brazil, using images from a drone. They compare the times chosen for prescribed burn in 3 different macrohabitats present in the biome.

The study has some interest for the readers, but the quality of the presentation is insufficient and some fundamental steps of analysis have not been reported.

Major comments:

1. The English spelling requires corrections and the syntax and style should be thoroughly reviewed. Some sentences are redundant, others very long and hard to understand and several words are not adequate. Processing steps in the field of remote sensing are not adequately explained using standard language.

2. Fundamental steps of analysis seems to be missing from the manuscript. All results are based on models calculated from RPAS images, but the accuracy of such models (classification of landcover, NDVI, BAI) using ground-truth information should be performed. Without this accuracy assessment, the validity of all other results is compromised.

Specific comments:

Abstract:

- Specify the Pantanal biome and the location. For example: “ In 2020, the greatest drought in 60 years was recorded in the Pantanal, the World’s largest tropical wetland located mostly in Brazil, and almost one third of the biome was affected by fire”.

- line 28-30: can the information of this sentence be integrated in the previous one describing the three macrohabitats? As it is, it’s redundant.

Methodology:

- SVM classification: there is no mention on how the training samples were collected and how train/test/validation of the algorithm was performed. Sentences such as line 343-346 in Discussion are therefore not supported by the data and analyses of the study. The sentence that follows (line 346-349) mentions Kappa, but no such results is found in the manuscript.

- Fire severity: the description of remotely-sensed image manipulation and analysis should be described using a standard language (e.g. the various steps of georeferencing and co-registration of images, calculation of spectral indices and image differencing…). Moreover, no validation of the models produced was performed.

Study area:

- Figure 1: Red-colored plots are not clearly visible and some legend items are missing (M1,2,3). Please add them to legend or caption.

Discussion:

- line 351-356: This paragraph is not clear. What’s the connection of the change in the mentioned thermal bands with grasses being “conductors of flammability in tropical savannas”?

Other comments:

- There are Appendices in the Supplementary Information file. Appendices should be placed at the end of the main manuscript.

- Figure A2: it is not very clear what the map of the World is used for in the Flight Planning stage.

Author Response

Response to Reviewer 1  

Comments    

The authors thank the reviewer for these helpful comments and hope to have addressed them in the detailed comments below. 

Major comments: 

1. The English spelling requires corrections and the syntax and style should be thoroughly reviewed. Some sentences are redundant, others very long and hard to understand and several words are not adequate. Processing steps in the field of remote sensing are not adequately explained using standard language. 

The manuscript listed below has been translated/edited into proper English with respect to grammar, syntax, and semantics by highly qualified professional editors from Insight® publishing house. Certification of English evaluation is attached 

2. Fundamental steps of analysis seems to be missing from the manuscript. All results are based on models calculated from RPAS images, but the accuracy of such models (classification of landcover, NDVI, BAI) using ground-truth information should be performed. Without this accuracy assessment, the validity of all other results is compromised. 

 A quality analysis of the results obtained from the land cover and land use classifications for the different prescribed burning periods evaluated was performed, based on the accuracy estimated by statistical indicators, through a confusion matrix generating descriptive and analytical statistical indicators of overall accuracy and the Kappa index (analyses included in the methodology and results section). During the field tasks the three types of phytophysiognomies of each plot are verified (pre-fire condition), in addition to verifying the behavior of the fire passage over the vegetation (post-fire response) (the supplementary material gives the general description of the field tasks done). 

Specific comments: 

Abstract: 

- Specify the Pantanal biome and the location. For example: “ In 2020, the greatest drought in 60 years was recorded in the Pantanal, the World’s largest tropical wetland located mostly in Brazil, and almost one third of the biome was affected by fire”. 

- line 28-30: can the information of this sentence be integrated in the previous one describing the three macrohabitats? As it is, it’s redundant. 

 The abstract has been modified to include the requested items: The study area located in the Pantanal biome. In 2020, the greatest drought in 60 years happened in the Pantanal. The fire affected almost one third of the biome. The objective of this study is to evaluate the effect of prescribed burnings made in 2021 on three macrohabitats (M1: natural grassland flooded with a proliferation of Combretum spp., M2: natural grassland of seasonal swamps, and M3: natural grassland flooded with a proliferation of Vochysia divergens) inside the SESC Pantanal Private Natural Heritage Reserve. Multispectral and thermal data analyses were conducted with Remotely Piloted Aircraft Systems in 1-ha plots in three periods of the dry season with early, mid, and late burning. The land use and land cover classification indicate that the predominant vegetation type in these areas is seasonally flooded grassland, except in zone three, which has a more diverse composition and structure, with the presence of arboreal individuals of V. divergem Pohl. 

 Methodology: 

- SVM classification: there is no mention on how the training samples were collected and how train/test/validation of the algorithm was performed. Sentences such as line 343-346 in Discussion are therefore not supported by the data and analyses of the study. The sentence that follows (line 346-349) mentions Kappa, but no such results is found in the manuscript. 

A change was made in the methodology section - SVM classification, including all training and validation steps. In addition to including the Overall accuracy and kappa results for each analysis plot (Results section 3.1). 

 - Fire severity: the description of remotely-sensed image manipulation and analysis should be described using a standard language (e.g. the various steps of georeferencing and co-registration of images, calculation of spectral indices and image differencing…). Moreover, no validation of the models produced was performed. 

A change was made in the methodology section - fire severity (section 2.5), improving the wording and including all processing steps and indicating that the verification of the plot areas was done after the burning during the PB fieldwork 

  Study area: 

- Figure 1: Red-colored plots are not clearly visible and some legend items are missing (M1,2,3). Please add them to legend or caption. 

Changes were made to figure 1, making the lines marking the different QP plots thicker and placing the legend over the three macrohabitats evaluated 

Discussion: 

- line 351-356: This paragraph is not clear. What’s the connection of the change in the mentioned thermal bands with grasses being “conductors of flammability in tropical savannas”? 

the change was made in the sentence: This spatial behavior reflected in grassland covers, which have the highest level of variation for post-fire events in PBs, thus  demonstrated that different forms of plant growth, such as grass, shrubs, and trees, presented varying levels of flammability, as well as revealing the dominant role that grasslands have as drivers of flammability in tropical savannas [80], by being the cover that is most affected by fire 

  Other comments: 

- There are Appendices in the Supplementary Information file. Appendices should be placed at the end of the main manuscript. 

We leave the appendix at the end of the manuscript. 

 - Figure A2: it is not very clear what the map of the World is used for in the Flight Planning stage. 

The world map is displayed in a general way as it corresponds to the interface of the flight planning application (Litchi for DJI Drones) 

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper evaluated the effect of prescribed burns(PB) in 3 macrohabitats through multispectral and thermal data analysis with Remotely Piloted Aircraft Systems(RPAS) in 1 ha plots, in three periods of the dry season with early, mid and late burning. The evaluation contains of classification of land cover by SVM algorithm, the thermal variation between pre-PBs and post-PBs, classification of fire severity through spectral indices. The research content of this paper is relatively comprehensive, and the method adopted is suitable and of practical significance. The evaluation of the effect of prescribed burns is meaningful for fire management.

There are some problems that deserve further consideration by the authors：

(1) In the Introduction section, the existing researches need to be carefully summarized and elaborated. The author introduced the research meaning, methods and purpose but lacked the summary of related researches.

(2) The box lines in 3 colors which marked different study areas are too thin and inconspicuous in Figure 1.

(3) In part of 2.4 Supervised classification, the introduction of classification algorithm is too concise and lacks technical details.

(4) In part of 3.1 Land use and land cover, the accuracy evaluation is neglected. The author just referenced other researcher's conclusion in part of Discussion that the SVM classification algorithm presents a high degree of accuracy. But systematic accuracy evaluation of the classification of land cover is essential.

(5) In part of 3.5, Table 5 is not intuitionistic enough due to the numerous data in the table, a figure which involves the percentage of different land cover in one severity level similar to Figure S11 can be added. It will be very helpful for discussing the association between fire severity and land cover.

(6) It is too general to explain the Figure 7 by 'Effect of burning on study macrohabitats'. Because the figure shows the burn area in 3 macrohabitats and the effect of burning is explained in above words. Besides, the author didn’t explain the curve changes in one macrohabitat, especially the curve of M2 that seemed different from the others.

Author Response

Response to Reviewer 2 

Comments   

The authors thank the reviewer for these helpful comments and hope to have addressed them in the detailed comments below. 

 The manuscript listed below has been translated/edited into proper English with respect to grammar, syntax, and semantics by highly qualified professional editors from Insight® publishing house. Certification of English evaluation. Please see the attachment.

. 

There are some problems that deserve further consideration by the authors： 

1. In the Introduction section, the existing researches need to be carefully summarized and elaborated. The author introduced the research meaning, methods and purpose but lacked the summary of related researches. 

Included research related to the article's theme, made changes in the introduction part of the manuscript 

1. The box lines in 3 colors which marked different study areas are too thin and inconspicuous in Figure 1. 

Changes were made to figure 1, making the lines marking the different PB plots thicker 

1. In part of 2.4 Supervised classification, the introduction of classification algorithm is too concise and lacks technical details. 

A change was made in the methodology section - SVM classification, including all training and validation steps. In addition to including the Overall accuracy and kappa results for each analysis plot (Results section 3.1) 

(4) In part of 3.1 Land use and land cover, the accuracy evaluation is neglected. The author just referenced other researcher's conclusion in part of Discussion that the SVM classification algorithm presents a high degree of accuracy. But systematic accuracy evaluation of the classification of land cover is essential. 

A quality analysis of the results obtained from the land cover and land use classifications for the different prescribed burning periods evaluated was performed, based on the accuracy estimated by statistical indicators, through a confusion matrix generating descriptive and analytical statistical indicators of overall accuracy and the Kappa index (analyses included in the methodology and results section). During the field tasks the three types of phytophysiognomies of each plot are verified (pre-fire condition), in addition to verifying the behavior of the fire passage over the vegetation (post-fire response) (the supplementary material gives the general description of the field tasks done). 

(5) In part of 3.5, Table 5 is not intuitionistic enough due to the numerous data in the table, a figure which involves the percentage of different land cover in one severity level similar to Figure S11 can be added. It will be very helpful for discussing the association between fire severity and land cover. 

Included the figure from the supplemental material section in the body of the manuscript, rewrote section 3.5, and changed the discussion to emphasize the association between fire severity and ground cover 

(6) It is too general to explain the Figure 7 by 'Effect of burning on study macrohabitats'. Because the figure shows the burn area in 3 macrohabitats and the effect of burning is explained in above words. Besides, the author didn’t explain the curve changes in one macrohabitat, especially the curve of M2 that seemed different from the others. 

The writing in the section explaining the effect of fire on macrohabitats has been redone 

Author Response File: Author Response.pdf

Reviewer 3 Report

Title: Remotely Piloted Aircraft Systems multispectral data to evaluate the effect of prescribed burns in three macrohabitats of Pantanal, Brazil

By: Harold E. Pineda Valles, Gustavo Manzon Nunes, Christian Niel Berlinck, Luiz Gustavo Gonçalves, and Gabriel and Henrique Pires de Mello Ribeiro

Review comments

The authors of this manuscripts utilized the Remotely Piloted Aircraft Systems to collect high resolution images both pre- and post-PB in three macrohabitats of Pantanal, Brazil, and to evaluate the effects of prescribed burns in terms of thermal status, vegetation cover, and burn severities. They demonstrated a practical role of UAV in monitoring fire effects with high spatial resolution. In general, the manuscripts were written not very well. Therefore, I suggest a major revision.

General comments

1.     In the introduction part, the previous works on the effects of prescribed burning are not well illustrated. More descriptions are needed.

2.      In section 3.3, taking the climatological variables in the prescribed burn moment as one of the results is quite strange. I suggest using these data to explain your results in the discussion part.

Specific comments

1.     Line 3: change “effect” to “effects” because prescribed burning could lead to multiple effects on ecosystems.

2.     Line 41: what does “pine lineages” mean?

3.     Line 141: update “due to” by “and”, add a comma after plain, and change the second “the” to “The”.

4.     Line 166-169: when did the pre-PB flights carry out?

5.     In Figure 2, the legend of the left chart should be M1,M2,M3, not B1, B2, B3.

6.     Line 203, does the field cover mean grass cover?

7.     In figure 3, I suggest change the legend to tree, grass, shrub, or to forest land, grassland, shrubland to keep consistency.

8.     In figure 5, there lacks caption to the Y-axis, and what’s the unit of value? Also, I suggest moving the early, mid, and late to below of the X-axis.

9.     Line 265, add a comma after fire.

10.   In figure 7, there is no caption for Y-axis.

11.  Line 322-323, I guess you meant area burned not burn severity.

12.  Table 5, update 1,2,3 by M1, M2, M3 respectively.

Author Response

Response to Reviewer 3 

Comments   

The authors thank the reviewer for these helpful comments and hope to have addressed them in the detailed comments below. 

 The manuscript listed below has been translated/edited into proper English with respect to grammar, syntax, and semantics by highly qualified professional editors from Insight® publishing house. Certification of English evaluation, Please see the attachmen. 

  General comments  

1. In the introduction part, the previous works on the effects of prescribed burning are not well illustrated. More descriptions are needed. 

Research related to the theme of the article has been included, a change is made in the introduction part of the manuscript 

 In section 3.3, taking the climatological variables in the prescribed burn moment as one of the results is quite strange. I suggest using these data to explain your results in the discussion part. 

emphasis was made on the description in the discussion part of the climatic variables, but they were mentioned in the results section when the variables were taken at the moment of making the prescribed choices 

Specific comments 

1.     Line 3: change “effect” to “effects” because prescribed burning could lead to multiple effects on ecosystems. 

The comment is accepted, and the respective modification is made:  Remotely Piloted Aircraft Systems multispectral data to evaluate the effects of prescribed burnings on three macrohabitats of Pantanal, Brazil 

2.     Line 41: what does “pine lineages” mean? 

"Pine lineages" refers to the different branches or genetic lineages within the pine tree family (Pinaceae). Pines are a clear example of species shaped by fire. Pines show the current oldest evidence of fire as an evolutionary pressure. Pausas, J.G. Evolutionary Fire Ecology: Lessons Learned from Pines. Trends Plant Sci 2015, 20, 318–324, 464 doi:10.1016/J.TPLANTS.2015.03.001.  

1. Linha 141: atualize “devido a” por “e”, adicione uma vírgula após simples e troque o segundo “o” por “O” 

Change was made in the sentence. 

1. Line 166-169: when did the pre-PB flights carry out? 

Prior to the operation of the PBs, an aerial photogrammetric survey was carried out using a RPAS with a multispectral sensor attached. The early and mid-burns occurred on July 14, 15, and 16 and September 15, 16, and 17, respectively, while the late burns occurred on October 26 and 27, 2021 between 09:00 and 15:00. The post-fire flights took place one or two days after the PBs. The field work for the PBs is described in the supplementary material (Appendix A). 

5.     In Figure 2, the legend of the left chart should be M1,M2,M3, not B1, B2, B3. 

Changes were made in figure 2 

6.     Line 203, does the field cover mean grass cover? 

the sentence was changed and left in the text pattern (Grass cover) 

7.     In figure 3, I suggest change the legend to tree, grass, shrub, or to forest land, grassland, shrubland to keep consistency. 

Changes were made in figure 3 

8.     In figure 5, there lacks caption to the Y-axis, and what’s the unit of value? Also, I suggest moving the early, mid, and late to below of the X-axis. 

Changes were made in figure 5, and include the caption for Y-axis, and we make the adjustment in X-axis. 

9.     Line 265, add a comma after fire. 

Change was made in the sentence  

10.   In figure 7, there is no caption for Y-axis. 

Changes were made in figure 7, and include the caption for Y-axis. 

11.  Line 322-323, I guess you meant area burned not burn severity. 

Change was made in the sentence: The early plot of M2 recorded the highest area burned 

12.  Table 5, update 1,2,3 by M1, M2, M3 respectively. 

The respective update is performed in table 5.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors provided a revised version of their manuscript. The quality of the science and methods have improved.

Despite a professional editing was performed, a big problem still persists with the quality of the presentation, in particular the structure of the sentences and organization of topics in the different paragraphs. It is highly suggested to have a scientist familiar with the methods and the project, with good English writing capabilities to fix the manuscript to reach publication standards.

I present here a few suggestions to get started; however, much more is needed.

Line 22-23: Connect the three sentences. Ex. “The study area is located in the Pantanal biome, which in 2020 experienced the greatest drought in the past 60 years, affecting almost one third of the biome.”

Line 31: connect the sentences. Ex. “...with percentages above 73%, except for zone three, which […] and structure, characterized by the presence of…”.

Also, the term “zone three” is not adequate in the abstract (the reader has no idea what and where this is). Maybe use “one assessed area” or other.

Line 32: add these words “the behavior of the thermal range showed differentiation between pre-...”

Line 43: change to “...acts as an evolutionary pressure on several living species, among which pine lineages.”

Line 44: change to “Ignitions can have natural origins [4] or can be caused by anthropogenic actions…

Line 61-63: rewrite as follows: “The Brazilian Pantanal is a seasonally flooded sedimentary basin surrounded by plateaus, mountains, and high hills, covering an area of 150,355 km² [18]. Located at the center of the South American continent, it is considered the largest seasonal wetland in the world [19, 20].”

Line 67: change to “In 2020, the worst drought of the past 60 years was recorded in the region, with an...”

Line 74: remove “4”

Line 78: Change “scenario” with “outcome”

Line 103: first appearance of the acronym “RPAS”. Add the full words.

Line 104-109: The same sentence is repeated and two different references are indicated. The third sentence (line 107) and the fourth (line 108) need to be connected with the previous.

Line 117-118. This sentence is not ideal here. Explanation of spectral analysis is explained later in line 130-132.

Line 139: place the reference at the end of the sentence.

Line 139: Change to “This study aims at evaluating…”

Line 144: change to“...by the means of three analysis plots set up within each macrohabitat, corresponding to early, mid, and late PBs, respectively.”

Line 146: change “an” to “a”.

Line 152: change to “It is located…”

Line 157: remove the “,”

Line 186: Here you mention an Appendix A to be found in the supplementary material. As I mentioned in my previous review, appendices are additional materials added at the end of the manuscript (it is clearly indicated in MDPI article templates). Instead, supplementary materials are presented in a separate file.

Also, I don’t see the Appendix at the end of the manuscript but it is in the Supplementary file. Moreover, Appendix A includes supplementary figures. Therefore these materials are mixed up while they should be clearly distinct and in different files.

Table 1: Correct “μ” to “μm” in the caption legend

Line 212: change to “presents”

Line 215: change “data classification” with “landcover classification”

Line 217-219: This sentence is badly formulated and I think it is also not necessary. It could be deleted. As a consequence, the next time ROI is mentioned, it should be indicated with its full name.

Line 244: the end of the sentence need to be fixed.

Line 260: Also this sentence need rephrasing

Line 284: change “presents” with “obtained” and delete “different”

Line 290-291: connect the sentences: “...(Figure 4), by comparing the pre-fire and post-fire images.”

Line 292: change to: “The three land use and land cover classes typical of the Pantanal landscape can be clearly distinguished, as well as the natural formations of …”

Line 299: change “shoes” with “shows”

Line 300: Because it’s prior to disturbance, “temperature for the PBs” doesn’t sound appropriate. I guess you mean the temperature of the plots.

Line 303: remove “degree”

Line 308: remove “of the sensor”

Line 310: change ”,” with “.”

Line 313-316: this sentence should be rephrased. What about: “The areas covered by tree individuals show little variation in temperature after the prescribed burnings because tree crowns are basically not affected by the fire, while areas covered with grasses show the highest temperature variation because of the bare soil highly exposed to solar radiation.

Line 321-322: the second part of the sentence is out of place. Rephrase or make it a standalone sentence.

Line 337: change “effect” with “from”

Line 340: the phrase “less than one day had no rain” is very unclear

Line 340-343: Separate in two sentences.. Replace “,” with “.” at line 341. Second sentence: “Where tree cover was higher, the effect of fire was lower.

Line 345-347: Rephrase the sentences. “Figure 7 shows the total burned area for each vegetation type during the three PB periods in the three studied macrohabitats.”

Line 352-356: This sentence is too long. Maybe you can separate the last part, from the end of line 354 after the comma. “In this case, the main fire management objective of reducing fuel load of grasses while conserving the arboreal cover was successfully achieved.”

Alternatively, you can delete this sentence because it fits better in the Discussion section than in the Results section.

Line 370: Rephrase to “...the highest area burned, leaving only 0.55% of its surface unburned.”

Line 371-372: “...highest burn severity, covering 20.23% of its total area (Figure 9).”

Line 388-389: confusing sentence that needs rephrasing. You are repeating the spatial resolution advantage.

Line 395-396: This sentence is redundant. You can complete the previous sentence by adding “, as confirmed by the Kappa index.”

Line 398: change “classed” with “considered”

Line 399: This is not clear, you mean visual identification or coupled with classification algorithms?

Line 405: This is unclear. The thermal variation of grassland is between pre- and post-fire or grasslands of different plots have different thermal values after the fire? The latter is the meaning from the way you write, but I guess it’s not what you want to say...

Line 405: change “demonstrated” to “demonstrating”

Line 404 -408: I still don’t understand the link between thermal difference post-fire (check my remark before) and the flammability of grasses. What makes sense to me is the analysis of water-stress before the fire. We can see that grass cover is “warm” suggesting higher water stress, therefore probably drier and more prone to burn. The sentence should be rephrased to make it clear.

Line 419: move the “,” after “here”

Line 426: Start the sentence with “Based on historical measurements from two meteorological stations […], located close to the study area, a correlation can be found with the catastrophic events of the year 2020, where irregular...”

Line 431: move the references after “fire spread”

Line 434: references should be merged

Line 448: change “They” with “These invasive species”

Line 474-476: the two sentences can be merged and rephrased. “The use of RPAS provides information with high detail, and spectral analysis allows for an accurate evaluation of the effects of fire, including prescribed fire for management purposes, on different vegetation types and structures.”

Line 478: Change to “Prescribed burns performed in the early stage of the dry season are the best option…”

Line 280: “What do you mean by “inputs”?

Line 488- 490: This last sentences comes out all of a sudden. If you want to discuss this statement, it should be done in other parts of the manuscript, such as the Discussion or the Introduction. It is not a conclusion derived from your study.

Reviewer 2 Report

The revision of the introduction section is incomplete and the issue of inadequate literature review persists. The revision of the remaining sections is adequately comprehensive.

Reviewer 3 Report

I could not find the revisions on the revised manuscripts as there has no indications, especially the two general comments.  I have no time to compare the original one with the revised one. hence, I have no choice for its acceptance or not.