Geospatial Scenario Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Line 164: In Figure 1, the icon of the compass is too small. Please adjust the size of the compass icon.

 

Line 203: In Figure 2, you present the methodological framework for each stage of this study. However, for the "JOI" in the dynamic modeling section, there is no corresponding explanation throughout the text.

 

Line 242: In the final part "[43,46, 47, 48]", there is a missing space after "43," please modify this and check for similar errors.

 

Line 264: The introduction of the variables in formulas (1) and (2) is incomplete. No explanations are provided for terms such as "m", "i", and "j".

 

Line 346: In Formula (8), please provide corresponding explanations regarding the meanings of "j (xy)" and "j (V)".

 

Line 357: In Formula (10), no explanations are given for each symbol. Please provide corresponding explanations for them.

 

Line 379: In Table 1, "SSIL" appears. You did not explain it and used abbreviations such as "AGR" and "SIL" in the table, but used full names such as "AGRICULTURE" in the explanation. This is unreasonable and should be unified. The three-line table format in Table 1 is incorrect. Please make the necessary modifications.

 

Line 397: In Table 2, the three-line table format is incorrect, and the intervals between the contents in the table are inconsistent. Please make corresponding modifications and improvements.

 

Line 446: In Table 3, the three-line table format is incorrect. How is the annual conversion rate in the differences calculated?

 

Line 479: In Table 4, the three-line table format is incorrect. It is incorrect to use lowercase letters for "ff", "pas", etc. in the table but interpret them with "FF", "PAS", etc.

 

I was glad to read your manuscript:“Geospatial Scenarios Modeling with Cellular Automata: Land 2 Use and Cover Change in Southern Maranhão, Brazilian Savanna 3 (2020–2030)”. In the entire paper, it can be seen that you have done a lot of work throughout the process, but there are still some details that can be improved. I hope you can carefully review the details of the article and make timely adjustments to the inappropriate places. Finally, my opinion is a major repair.

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Author Response

Dear Reviewer 1,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (in red). Below, we present our point-by-point responses to the reviewers’ comments. (pdf file).

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

 

The manuscript has been reviewed, and my comments are as follows.

 

1 The keywords in the paper are too few. I suggest the author add some more.

 

2 How to validate the results of model simulation is crucial for ensuring the accuracy of future predictions.

 

3 Why choose this research area? I suggest the author add more explanations. Currently, many explanations remain at the technical level rather than scientific significance.

 

4 In this article, the time range used for modeling is too short. Why didn't the author choose data from a longer time range for modeling? This should improve the accuracy of the model.

 

5 In the introduction, some key references are still missing: Quantifying urban expansion from 1985 to 2018 in large cities worldwide, Stable classification with limited sample: Transferring a 30-m resolution sample set collected in 2015 to mapping 10-m resolution global land cover in 2017.

 

6 In the discussion section, I suggest the author add whether this research can be expanded to other areas in the future.

Author Response

Dear Reviewer 2,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (in red). Below, we present our point-by-point responses to the reviewers’ comments. (pdf file).

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Using remote sensing data, Random Forest classification, and a cellular-automata model, this paper simulates land-use and land-cover change trends in the southern Maranhão portion of the Brazilian Cerrado biome through 2030, revealing the marked impact of agricultural expansion on natural vegetation. The study integrates multi-source geospatial data with machine-learning techniques, delivering high-accuracy land-use classification and plausible future-scenario simulations that demonstrate strong methodological rigor and regional applicability. The manuscript is well written, but only minor revisions are needed to present the findings more clearly to readers.
Line 24: The phrase “achieving an accuracy exceeding 45%” is vague: it does not specify which metric is reported (e.g., Kappa, overall accuracy), and a value as low as 45% is unlikely to meet common mapping standards, yet the word “exceeding” may mislead readers into regarding it as satisfactory.
Lines 123-124: This is a critical point, yet the authors do not elaborate on how the existence of these "legally available for conversion" lands affects the reliability of the simulation results or their policy implications.
Line 130: The stated research objectives are overly generic; they fail to clarify the study’s originality in methodology, regional focus, or policy implications.
Lines 227-228: These two parameters are the most critical in multi-resolution segmentation, yet the authors offer no justification for choosing these particular values. Were they selected subjectively, or do they have support from the literature? This directly affects the fundamental quality of the image objects.
Line 232: Although 77 features are mentioned, the exact band/index combinations used to derive them are not disclosed, so the feature space cannot be reconstructed. Generating a large feature set easily invites the “curse of dimensionality” and overfitting. The manuscript does not indicate whether any feature-selection step (e.g., importance ranking) was applied to remove redundant or irrelevant variables; relying solely on StandardScaler normalization is insufficient.
Line 404: This is a very strong prior assumption, yet it lacks adequate justification. Why 90 % rather than 80 % or 95 %? This parameter decisively shapes the spatial pattern of the simulation; a sensitivity analysis should be conducted to test its impact.
Line 410: Likewise, no rationale is given for selecting 1.5. This parameter governs the shape of newly formed patches (1 = most compact, 2 = most linear), and its value appears to have been set subjectively.
Line 432: This is one of the most critical flaws in the paper. A fuzzy similarity of 0.45 is extremely low, indicating large discrepancies between the simulated and actual maps. Labeling this value as “acceptable” is unsubstantiated and falls well below the common benchmarks reported in comparable studies, directly undermining the credibility of the entire simulation.
Validation relies solely on fuzzy similarity; other widely accepted metrics such as the Figure of Merit (FoM) or Kappa simulation are not employed, preventing a comprehensive assessment of model performance.
Line 458: This is an extremely anomalous and counter-intuitive finding in a Cerrado agricultural frontier. The authors dismiss it with a single phrase—“gained area primarily from savanna formations”—and offer neither in-depth explanation nor validation. Could it be a classification error (e.g., regrowth or specific cropland mis-labelled as forest)? Or an artefact of image date or phenological differences? By glossing over this critical issue they seriously undermine the credibility of all subsequent analyses.
Line 467: The long paragraph (“These figures indicate that...”) merely repeats the numbers already shown in Table 4 without distilling any deeper insight and is therefore redundant.
Line 620: This is an important finding, but it is not discussed in depth. The effectiveness of these protected areas and the potential pressures they face should be further explored rather than simply stated as a fact.
The discussion remains largely descriptive (“what happened”) and lacks a deeper exploration of the “why.” For example, why is proximity to existing cleared plots the dominant driver, whereas roads and protected areas show little effect? What underlying land-speculation behaviors, infrastructure-development patterns, or governance failures does this reflect?
Lines 735-737: The conclusion should not reiterate methodological minutiae (e.g., “high degree of automation”); instead, it must summarize what deeper insights or superior outcomes the method delivered. This sentence feels superficial and distracts from the core task of a conclusion.

Author Response

Dear Reviewer 3,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (in red). Below, we present our point-by-point responses to the reviewers’ comments. (pdf file).

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript can be of a practical value. It is better to connect it more to the field significance of the result and display samples of images.

Pages 3 and 4 - objectives can be made more visible than are. It is better to provide a special paragraph.

There is an essential aspect to focus on geomatic approaches and after on the regional problem and solutions. 

Some samples of details from satellite imagery can help in understanding the regional context and the land cover features.

Please check all geographical names to occur on the maps.

Methodology

Is it enough data on a two images to  predictict land use changes in study area?

What about the calibration of the images? 

Special features encountered on OBIA approach can be explained by providing some detail samples. It is interesting and useful to present the improvements and the adaptations of the methods and of algorithms. Segmentation is a part of the methodology needing more attention...

Field validation need more description. Some images can help.

Results and Discussion

These sections can be separated. First is focused on a review of results and second on the interpretation of results in order to extract the most significant aspects.

There is a big volume of statistical data which is good but the spatial expression is necessary.

Figure 3 maps are rather general. It is better to focus on some case studies with complementary features. Figure 4 showed two areas but these can be mapped in detail.

There are a lot of geographical names. These can be checked on the maps as well.

Figure 5 maps are rather general and needs some details and focused explanations/interpretations on the predicted LULC class changes.

Some more calibration wth similar approaches from literature can be helpful.

More detailed mapping and terrain connections are needed.

Conclusion

Extract more visible the practical significances of the approach, after identifying the pluses and the minuses of the employed methodology.

Author Response

Dear Reviewer 4,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (in red). Below, we present our point-by-point responses to the reviewers’ comments. (pdf file).

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Line 23: The results presented in the text are ambiguous, lacking specific accuracy values or ranges. Please clearly state the overall accuracy or Kappa value.

Line 232: Figure 2 is visually clear, but the flowchart section features small text, unclear step descriptions, and cluttered connection lines. Adjusting font size or splitting into separate figures is recommended.

Line 242: Redundant dataset description; recommend deletion.

Line 277: Abbreviation not explained. The text uses “ACP” but the more common abbreviation “PCA” (Principal Component Analysis) should be used, with the full name provided at first mention. Additionally, this paragraph is overly brief. No explanation is provided regarding how many principal components were retained after PCA, nor the cumulative variance explained by these components. This prevents readers from assessing whether the dimensionality reduction process was reasonable and whether sufficient information was preserved for subsequent classification.

Line 314: Variable definitions in Equations 1 and 2 are unclear. Variables xi+ and x+i are defined incorrectly. The Kappa formula is expressed non-standardly; the form in the text is its expanded version. It is recommended to first present the standard form before substituting values.

Line 401: Equation 4 is identical to Equation 5, constituting redundancy.

Line 401: Equations 6 and 7 contain errors and should be rewritten.

Line 413: Variable descriptions in the equation are inconsistent, with one variable having multiple interpretations. Please standardize.

Line 430: Equation 9 has formatting errors; please correct.

Line 573: The original text repeats the statement “It is worth noting that the classification process was refined, generally eliminating confusion between classes.” Delete the redundant sentence and retain only one instance.

Line 596: In Table 4, while conversion rates between categories are presented, the discussion remains fragmented. A comprehensive analysis is lacking to identify which native vegetation type (FS, FC, FF) faces the greatest conversion pressure and the primary conversion to which human land use (PAS, AGR).

Line 659: In Figure 4, the legend “High/Low Probability” lacks specific numerical ranges. Add probability ranges or grading explanations.

Line 742: Table 6 uses “*” symbols without explanation in the legend and lacks a three-line table format. Clarify in the table notes that “*” denotes significance.

Lines 887: The policy recommendations section is weak, merely mentioning support for sustainable resource management and planning in general terms. More targeted and actionable policy suggestions should be proposed based on the specific findings of this study.

The paper lacks sufficient comparison with studies from other tropical savannas or agricultural expansion frontiers globally. We recommend incorporating cross-regional comparisons in the discussion to explore whether the identified drivers and change patterns are universal, thereby enhancing the study's global relevance.

Redundant expressions such as “It is worth noting that...” appear in multiple sections. We suggest replacing these with more concise alternatives like “Notably,” or stating the facts directly.

I was pleased to review your manuscript:“Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)”. The overall structure and content are sound. It is evident that you have invested significant effort throughout the process, though some details warrant refinement. I encourage you to meticulously review the article's specifics and promptly adjust any inappropriate sections. Ultimately, my recommendation is for substantial revision.

Author Response

Dear Reviewer 1,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (in red). Below, we present our point-by-point responses to the reviewers’ comments.

 Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

The revised manuscript can be accepted.

Best wishes,

Author Response

Dear Reviewer 2,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics. The feedback received was extremely valuable for improving the quality of our work.

Authors

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Line 280: The text explicitly mentions “seven statistical indices,” but actually lists eight, creating a numerical discrepancy. We recommend verifying and aligning the “stated quantity” with the “actual number of items listed” to ensure consistency.Line 409: The logical symbol for “not” is inconsistent. The logical symbol for “not” in Equations 4, 5, and 10 is inconsistent. We recommend standardizing it to a single logical symbol for “not” to ensure uniformity of symbolic notation throughout the equations.
Line 438: Incomplete sentence semantics. The sentence ends with “S (C)” without subsequent components, resulting in semantic discontinuity and incomplete expression.
Line 441: Inconsistent symbols between text and formulae. The text uses “W+” and “W-”, while the corresponding formulae use “W⁺” and “W⁻”. Format inconsistency is noted; standardize both symbol formats.
Line 471: Table formatting errors. Multiple tables throughout the text exhibit formatting irregularities. Recommend adjusting all tables according to the document's unified formatting standards—including borders, alignment, row/column dimensions—to ensure visual consistency.
Line 619: Figure 4 labeling and symbol error
The compass in Figure 4 lacks the “N (North)” directional label. Additionally, expressions like “0,75” appear with commas, suggesting potential decimal point errors. Recommendation: Add the “N” direction label and review for corrections.
Line 665: Inconsistent Capitalization in Category Names
The first letter of “pasture” is lowercase, while other category names in the text use uppercase initial letters, creating formatting inconsistency. Recommendation: Standardize capitalization to match other category names.
Line 702: Abbreviation list lacks introductory text. The list below the table begins abruptly without any introductory statement, violating formatting standards. Add an introductory phrase to enhance document consistency.
Line 1210: Reference formatting and full-text review. The reference format at this location does not comply with standards. Adjust this reference according to specified guidelines and conduct a comprehensive review of all references throughout the document to ensure uniform formatting.

Author Response

Dear Reviewer 1,

We would like to thank you for the comments and suggestions provided on our manuscript entitled "Geospatial Scenarios Modeling with Cellular Automata: Land Use and Cover Change in Southern Maranhão, Brazilian Savanna (2020–2030)", submitted to Geomatics.

We appreciate the reviewers’ continued efforts and the opportunity to further improve our manuscript during this third round of revision. The manuscript has been carefully revised, and the suggestions were incorporated whenever possible into the updated version (highlighted in red). Below, we present our point-by-point responses to comments.

Authors

Author Response File: Author Response.pdf