Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General comments:

Dear Authors,

I've read your article. It's important because of its focus on detecting and assessing forest cover in a selected area, which is particularly susceptible to human impact. Your study is based on a comparison of existing data on land cover and use, analyzing classes dedicated to forests. This study, conducted by collecting a database and developing transition matrices, is well done.

However, the article is rather chaotic and unevenly written in terms of textual description. I don't understand the division of the research into examining the area of one country and a series of selected, rectangular study plots designated in selected ecoregions. Perhaps you should either adopt countries or these ecoregions as divisions, or adopt an area of equal size within each ecoregion, representing its typical characteristics. For now, you should provide a justification for this choice in the text.

I have reservations about the presentation of the results and their illustrations—not the figures and tables themselves, but their arrangement in the text. Why, for example, are figures relating to Gabon placed in the text, while those relating to other research fields are in the appendix?

The text also requires improved formatting, as I have described below and in the comments to the text, particularly regarding the inconsistencies in the notation of numbers, percentages, abbreviations, etc.

• Study area:

In the study area description (page 3), five different study areas are used – the first is designated as "Gabon" (the entire country, its borders are also the range of study area no. 1), while the others are four different forest ecosystems (miombo, rain forest, montane forest, savanna). Why are they not all the same study areas, focusing on a specific vegetation formation? This is not explained. Furthermore, the entire Gabon is not covered by a single forest structure but lies within several ecoregions, as shown in Figure 1.

• Results:

There is no description in the materials and methods section of how the influence of homogenization and resampling on the results was assessed, but this is the first result described (3.1.).

Figure 3 (Page 8): What accounts for this difference in reclassification between the MODIS data (t1, p2) and other databases, particularly in the southern and western parts? Two areas with a very low share of forest area remain invisible.

The result figures (4, A6-A10) contain transition matrix – this term is not found in either the methodology or the results.

The results are presented sometimes in descending order (3.2.2.), sometimes in ascending order (3.1.2.) of the parameter, and sometimes without any ordering (3.1.1.; 3.1.5.).

The same type of result figures are sometime placed in the text (e.g. igure 5), sometimes in the Appendix (A1, A2, A3).

• Text formatting:

Double and multiple spaces left in the text.

Some messages or warnings left in French, perhaps copied?

In-text citations are placed in separate brackets where they follow one another – they should be listed in one parentheses, after a comma, or within a dash).

Some words and abbreviations are written differently in the article, e.g. LiDAR (LiDAR, lidar, Lidar, …).

Comments to the text:

1. Page 1, line 11: Terrestrial surfaces → terrestrial ecosystems.
2. Page 1, line 12: Tropical forests are threatened by human impact and climate changes everywhere, not especially in Africa.
3. Page 1, line 34: Double space between “[2].” and “However”.
4. Page 1, line 36: [3], [4], → [3, 4].
5. Page 2, line 45: Same as above.
6. Page 2, line 60: “Earth Observation” is too general term – Earth Observation methods of data acquisition?
7. Page 2, line 61-62: Multiple spaces and each reference to literature is in separate square brackets.
8. Page 2, line 78: Double space.
9. Page 2, line 84-85: You have written in other places “montane forests”, not “mountainous forests”.
10. Page 3, line 91: Error message in French, left in text sent to review.
11. Page 3, line 100: Forest…? Equatorial forest? Taiga forest?
12. Page 3, line 103: Typo: Congo Basin, not Congo Bassin.
13. Page 4, line 119: mountain forests → montane forests.
14. Page 5, line 170: 10 % → 10%.
15. Page 6, line 181: So, what is forest definition by FAO?
16. Page 7, line 204-205: This information should have been written much, much earlier in the text.
17. Page 8, line 234: The “window” should be included in the legend, not in the title of the figure 3.
18. Page 12, line 336-337: The definition should be in the materials and methods section, not in the discussion – I noted this in the comment on page 6.
19. Page 14, line 426: lidar → LiDAR.

Author Response

« I've read your article. It's important because of its focus on detecting and assessing forest cover in a selected area, which is particularly susceptible to human impact. Your study is based on a comparison of existing data on land cover and use, analyzing classes dedicated to forests. This study, conducted by collecting a database and developing transition matrices, is well done. »

Answer 1: We would like to express our gratitude for your insightful review of our paper.

 

« However, the article is rather chaotic and unevenly written in terms of textual description. I don't understand the division of the research into examining the area of one country and a series of selected, rectangular study plots designated in selected ecoregions. Perhaps you should either adopt countries or these ecoregions as divisions, or adopt an area of equal size within each ecoregion, representing its typical characteristics. For now, you should provide a justification for this choice in the text »

and

« In the study area description (page 3), five different study areas are used – the first is designated as "Gabon" (the entire country, its borders are also the range of study area no. 1), while the others are four different forest ecosystems (miombo, rain forest, montane forest, savanna). Why are they not all the same study areas, focusing on a specific vegetation formation? This is not explained. Furthermore, the entire Gabon is not covered by a single forest structure but lies within several ecoregions, as shown in Figure 1. »

And

« I have reservations about the presentation of the results and their illustrations—not the figures and tables themselves, but their arrangement in the text. Why, for example, are figures relating to Gabon placed in the text, while those relating to other research fields are in the appendix? »

And

« The same type of result figures are sometime placed in the text (e.g. igure 5), sometimes in the Appendix (A1, A2, A3). »

Answer 2: Thank you for your comments. To be clearer, we extended the analysis to cover the entire Congo Basin, and then described the results in terms of three ecoregions (see Table A 2):

1. i) Miombo: contain the TEOW regions with Miombo woodlands
2. ii) Savanna: contain the TEOW regions with forest savanna mosaic

iii) Rainforest: contain the TEOW regions with lowland, swamp or coastal forest

We have corrected the manuscript in accordance with your remarks. You can find the justification at lines 88-91: “The analysis was carried out over the Congo Basin and covered different ecoregions representative of the main tropical forest types in equatorial Africa, including: 1) moist forest, which is a region of central Africa highly affected by cloud [21]; 2) savanna forests and 3) miombo forests.”

According to these modifications, we have also changed the figures in the text to illustrate all the regions that were studied. (See Table A 2 and Figure 6).

 

« There is no description in the materials and methods section of how the influence of homogenization and resampling on the results was assessed, but this is the first result described (3.1.). »

Answer 3 : We agree with this comment and have added an explanation on line 204-205.

“We assessed the influence of homogenization and resampling by calculating the2mean area of forest and non-forest for each product at each resolution. “

 

« Figure 3 (Page 8) : What accounts for this difference in reclassification between the MODIS data (t1, p2) and other databases, particularly in the southern and western parts ? Two areas with a very low share of forest area remain invisible. »

Answer 5: If we have correctly understood, you mentioned the two bands of forest-savannah mosaic. The products differ in terms of classification in these regions. We list the classes observed in these regions below:

ESACCI: mainly shrubland

- GLCLU: Sparse vegetation

- HILDA+: pasture/rangeland

- TMF: Other land cover

- MODIS t1: Savannas, defined as tree cover of 10–30% (canopy > 2 m).

- MODIS p2: Open forests, defined as tree cover of 10–60% (canopy > 2 m).

The homogenisation required some compromises regarding the FAO definition of “forests”. For the MODIS products, we applied a tree cover of more than 10% as the definition of forest. Other products may consider these regions to have low vegetation based on height, while MODIS considers a canopy higher than 2 m to be a tree. This key aspect regarding the discrepancies in forest definition is discussed line 185-187.

“Information on tree cover and height was not present in all the datasets. We tried to match the FAO definition as closely as possible. However, some compromises had to be made (Table A 3-8)”

We also edited the text to emphasize this key issue in remote sensing products line 366-369 and 404-407

“Furthermore, the MODIS products include in their forest definition, tree higher than 2m. This leads to consider a larger proportion of the savanna as forest, given that the trees are shorter than those in the moist forest or miombo woodlands. “

“As shown by the MODIS products and the GLCLU over the savanna, the threshold for tree height also impacts the results (Table 2, Table A 4,5,7). By incorporating tree cover and height into the definition of future maps, users can select an appropriate threshold for their needs and the ecoregion of interest”

« The result figures (4, A6-A10) contain transition matrix – this term is not found in either the methodology or the results.»

Answer 6: We understand the confusion surrounding the matrix. These figures cannot be categorised as a transition matrix, as they do not represent the probability of transitioning from one state to another. To clarify this point we detailed the methodology by adding the term “confusion matrix” at line 208.

“Products were compared two by two for each common year (2005, 2010 and 2015) to produce a confusion matrix which is defined as follows: “NF/NF” was attributed if both products observed non-forest; “F/F” if both products observed forest and; “F/NF” if the “reference” product observed forest but the “compared” product did not; and NF/F, if the “compared” product observed forest but the “reference” product did not (Figure 2, 2b).”

 

« The results are presented sometimes in descending order (3.2.2.), sometimes in ascending order (3.1.2.) of the parameter, and sometimes without any ordering (3.1.1.; 3.1.5.). »

Answer 7: We rectified the results section by arranging all the presented parameters in ascending order.

 

« The text also requires improved formatting, as I have described below and in the comments to the text, particularly regarding the inconsistencies in the notation of numbers, percentages, abbreviations, etc. »

Answer 8: Thank you for your detailed feedback on the text, which we have taken into account.

 

 

Text formatting:

« Double and multiple spaces left in the text. » : The issue has now been rectified.

« Some messages or warnings left in French, perhaps copied? » :

An error occurred when the MDPI template was applied. It referred to Figure 1. This has now been rectified.

 

« In-text citations are placed in separate brackets where they follow one another – they should be listed in one parentheses, after a comma, or within a dash). » :

We changed the Zotero citation type in order to address this issue.

 

Some words and abbreviations are written differently in the article, e.g. LiDAR (LiDAR, lidar, Lidar, …). : The issue has now been rectified.

 

Comments to the text:

1. Page 1, line 11: Terrestrial surfaces → terrestrial ecosystems.

Replaced in the text

2. Page 1, line 12: Tropical forests are threatened by human impact and climate changes everywhere, not especially in Africa .

The sentence « especially Africa » was deleted

3. Page 1, line 34: Double space between “[2].” and “However”.

The issue has now been rectified.

4. Page 1, line 36: [3], [4], → [3, 4].

The issue has now been rectified with Zotero

 

5. Page 2, line 45: Same as above.

The issue has now been rectified with Zotero

 

6. Page 2, line 60: “Earth Observation” is too general term – Earth Observation methods of data acquisition?

“Earth Observation (EO)” was replaced by “Earth Observation (EO) data”

 

7. Page 2, line 61-62: Multiple spaces and each reference to literature is in separate square brackets.

The issue has now been rectified with Zotero

 

8. Page 2, line 78: Double space.

The issue has now been rectified.

 

9. Page 2, line 84-85: You have written in other places “montane forests”, not “mountainous forests”.

The issue has now been rectified.

 

10. Page 3, line 91: Error message in French, left in text sent to review.

The error occurred due to the application of the MDPI template. It referred to Figure 1, the issue has now been rectified.

 

11. Page 3, line 100: Forest…? Equatorial forest? Taiga forest?

The term « forest » was replaced by « rainforest »

 

12. Page 3, line 103: Typo: Congo Basin, not Congo Bassin.

The issue has now been rectified.

 

13. Page 4, line 119: mountain forests → montane forests.

The issue has now been rectified.

 

14. Page 5, line 170: 10 % → 10%.

The issue has now been rectified.

 

15. Page 6, line 181: So, what is forest definition by FAO?

The definition is already provided at lines 172-173. However, we have added the justification 'tree cover >10%' to line 184 for clarity.

 

16. Page 7, line 204-205: This information should have been written much, much earlier in the text.

The treatment over the years was unclear. We added additional information at line X.

 

17. Page 8, line 234: The “window” should be included in the legend, not in the title of the figure

We have amended the legend of the figure for clarity.

 

18. Page 12, line 336-337: The definition should be in the materials and methods section, not in the discussion – I noted this in the comment on page 6.

The definition is already provided at line 170-171 of the material and method section

 

19. Page 14, line 426: lidar → LiDAR.

The issue has now been rectified.

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript compares seven LULC datasets for forest cover assessment in equatorial Africa. While the research tries to address a relevant topic for conservation planning, I believe several concerns limit its contribution:

Technical limitations--The method applied relies primarily on basic resampling and zonal statistics operations. Although sophisticated techniques aren’t always necessary, the current approach feels insufficient for a research paper’s scope and impact.

Method issue--The data processing workflow lacks precision. For example, for each dataset, did you sum all the images from 2001 to 2019 and then take an average? For GLCLU, which has one image every five years, did you just add the four images together and average them? The single comparison table requires better justification in the methods section.

Professional presentation--Multiple formatting errors reduce readability, including non-English (French) text (line 91), inconsistent spacing (lines 34, 57, 61-62, 78), lack of superscript formatting (line 96). Also, In Equations 1 and 2, the values should be multiplied by 100%; otherwise your percentage calculation doesn’t make sense. That’s a very basic math concept.

Research significance--Regarding the significance of this research, the authors’ explanation is very general. How exactly will future LULC product be combined with SAR imagery and LiDAR data in the future to increase forest mapping accuracy? Aside from a bit of help for cartography, what about the broader environmental or biological significance? There’s very little on this.

Overall, I think this paper has quite a few problems that need to be addressed.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

No more comments. The authors spent a lot time and efforts to conduct this research, though the implication of the study is quite limited.

Author Response

Thank you very much for giving us the opportunity to revise, for a second time our manuscript. Compared to the previous version, an important effort was made on making clearer that a temporal analysis was performed. This led to the inclusion of two new figures in the main text and four new figures in the supplementary which helped us strengthening the discussion and results sections.