An Alternative Method for the Generation of Consistent Mapping to Monitoring Land Cover Change: A Case Study of Guerrero State in Mexico

Round 1

Reviewer 1 Report

The paper explicates a just reasonable approach to land cover change mapping based on the pre-identified zones of change.
The authors very correctly listed the shortcomings of the existing technique for land cover supervised classification. However, their proposed approach is not free of the same spatial/thematic classification errors. Moreover, the zones of stable land cover, which are excluded from consideration by the authors, can also be identified improperly.
The assignment of additional derived images is not justified. For example, why are the principal components was selected but the independent components was not? I recommend to describe in detail the selection of additional data layers.
The author's chosen "debugging" term is improper. In remote sensing instead of it the  "post-classification" or "clumping" terms are generally used. The engaged principle of reclassification based on  mapping unit minimum area is rather primitive. In my opinion, reclassification should be carried out taking into account the ground classes' topology, for example, as it is done in [Piestova et al., 2021, http://ceur-ws.org/Vol-2864/paper14.pdf].
Within the method's description the authors mentioned the generating a segmentation layer of the input image, but did not explain at what stage of the analysis it is used and why.
The weak point of the results' validation is the same ground-truth data applying to classify an image of another time (-5 years time-lag) and another sensor (Landsat 5 TM vs Landsat 8 OLI). Therefore, the estimates obtained are likely to have a systematic error. In future work, the authors are encouraged to apply an end-to-end confidence assessment for classes and changes in all processing stages. One possible way to perform such confidence assessment is the probability propagation technique, as for instance in [Hoberg & Rottensteiner, 2010, https://www.isprs.org/proceedings/XXXVIII/part7/a/pdf/53_XXXVIII-part7A.pdf].
There are some formal inaccuracies in the paper:
It seems the authors' affiliations 1 and 3, as well as 4 and 5 are the same;
line 39: "earth’s" should be replaced with "Earth’s";
line 124: "Aster" should be replaced with "ASTER" as an abbreviation;
line 125: "spectral ranges" should be replaced with "spectral bands" as usual in remote sensing;
reference 28: "The most detailed portrait of Earth" should be changed "GLOBCOVER - The most detailed portrait of Earth".

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

This study presents solid and interesting results. Only a couple of minor revisions needed before publication.

1. Figure 2, please change yellow polygons color. Put a, b, and c’s locations in the satellite image.
2. Please change fonts and size of the diagraph ---- it is hard to read in the current format.
3. Line 279: Please give more details of how you gathered ground-truth data.
4. Figure 7. Please make legend larger.
5. Line 617. I would like to see a short paragraph to discuss what caused the change in land cover type.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

Thank you for reviewing opportunity. The manuscript titled “An alternative method for the generation of consistent mapping to monitoring land cover change” fits the aims and scopes of Land. Proposing a method for land cover mapping is highly helpful for land use/cover change analysis and assessing its effects on biodiversity and ecosystem services. On the whole, the structure is clear (Figure 3) and the work is interesting. Some suggestions are as follows.

1. Title. This study proposed a method. So it’s a method study, not a traditional empirical research. In general, the authors firstly propose a method, and then they select a typical area to conduct an empirical research using the proposed method. So, the title “An alternative method for…..——a case study of Guerrero State in Mexico” maybe better.
2. Structure. Similar suggestions to Title. In my opinion, the first half is methodology and the second half is case study. The case study is indispensable, which is used to prove the feasibility of the method.
3. Figures and Tables. Total 13 figures and 8 tables, a little too many. In other words, these figures and tables maybe a big challenge to readers. If possible, try to delete some of them or move some of them to supplementary materials. e.g., is figure 2 necessary? Only keep the most important figures/tables.
4. Captions. For example, Figure 1. Caption “Study area”, which is not informative, can be better. Maybe use the name of the study area? Please check this issue for whole manuscript, particularly for figure/table captions. Figure 3. General flowchart of the proposed methodology for what?
5. Resolution. Figure 3 is very important, but the resolution is poor and font size is small. Figure 4 is good. Please revise figure 3 as clear as figure 4.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

This is a nice paper with some flaws that should be removed or contained in a revised version. My impression is that authors are providing a good technical overview, but the problem in this paper is that I would see some more revisions and additions concerning technical details for non-technical readers that are not able to manage remote sensing practices. For instance, I would see a specific explaination of Tasseled Cap methodology and similar alternatives and why you have selected this method. Additionally, you have used spatial decomposition in Principal Components which is OK with TC, however, more traditional analyses concerning Linear Fisher Discriminant Analysis Functions can be used to pattern recognition in remote sensing. I would see a broader discussion of technical issues in this paper. I believe readers need more in this perspective. Thank you for your efforts.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 5 Report

The state of land cover/land use is essential to scientists and land managers. This paper developed an alternative method for the generation of consistent mapping to monitoring land cover change. The results show that the proposed method is viable and fully automatable. The generated LULC map is accurate and robust, and the boundary consistency was visually superior to the conventional classified map. However, there are still some details that need to be further modified and supplemented by the authors.

 

1.  

Line 293: The right parenthesis may be missed.

Line 502: The bold italic word “map Vs.” would be spelling mistakes. Please check it.

 

2. Most results are presented in the form of table such as Table 7, using figures to show them is a better choice.

 

3. There are many error references in the content, such as Line 88 and 90, Line 422-434……Please recheck the manuscript.

 

4. The base date is 21st January 2016 and the test date is 24th February 2011, when it is winter in Northern Hemisphere. Please discuss and explain the influence of season appropriately.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 4 Report

Thank you for your revision. I am rather happy about the corrections but I don't feel completely OK with the article. Ithink it is too descriptive and poorly connected with global literature.

Please expand literature review.

Please clarify originality and novelty of the approach

Please give a conclusion in terms of true contribution to science

Please go beyond the latent description of phenomena, give the importance of your results for international readers

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 5 Report

I am satisfied with the response. The paper can be accepted.

Author Response

Thank you very much. We appreciate you again for your valuable comments.