Optimization of Rocky Desertification Classification Model Based on Vegetation Type and Seasonal Characteristic

Round 1

Reviewer 1 Report

Dear Authors,

   Firstly, thank you very much for the manuscript submission to Journal of Remote Sensing. This research article present a solid foundation of their related methods and technical approach, the results are supportive. A few aspects needs further improvement, which I specified (but not limited to) as below:

  a) In Intro. Section, a short summary on your major contributions of work is missing; meanwhile, if no section on "Related Work", it is better to add a last paragraph on summary for the remainder of the other sections' work.

  b) In my point of view, the title and arrangement of Sections 2-3 need some adjustment: Section 2: "Related Work: Source Material / Theoretical Background"; Section 3: "Technical Approach"; meanwhile, shorten any of the unnecessary subsections on Section 3 (currently it covers almost 10 pages).

  c) Starting from Figure 3, the legends of all the subsequent figures may need uniform style of characters (i.e., Times New Roman, fontsize = 11/12); the resolution of some figures should be enhanced.

  d) I suggest using middle-alignment for the related formulas (not right-size aligned except for the numbers (X)).

  e) Regarding numerical results as tabulated in the Tables, please preserve all 3-valid decimals (i.e., 0.83 --> 0.830; similar with 0.79(0) and 0.82(0) in Table 2); besides, I think results on Tables 1-3 can be merged in one table. 

  f) Update Figure 9 with a title (adding specific explanations), and update the legends on Figure 10 by removing any possible occlusions.

  g) Conclusions: better to divide this section into 2-3 paragraphs including major summary of accomplished work, opening problems to be solved, and the proposed future study (or possible research orientations)

  h)  References: missing closely related publications in Years 2020, and it is preferable to add some more latest journal publications in lastest three Years; Double-check the citations on journals / conferences on qualification of publishing upon the agreed template formats.

  After the required edits, literal improvement and other related polishments, I suggest that this paper can be accepted after minor revision. Stay safe and best of luck to your future success!

With warm regards,

Yours faithfully,

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript describes the machine learning methods for optimization of rocky desertification classification model using data from Guizhou, China.

 

The manuscript needs significant improvement.

 

1. All acronyms in the titles must be replaced by entire words.

 

2. All figures with the maps must contain the geographical grid with the coordinates.

 

3. The description of all models must be shortened. These are already well-known models.

 

4. The authors must clearly explain the novelty of the work from a scientific point of view. Now it looks like a technical application of the models.

 

5. The authors need to justify the chosen time frame for the data series.

 

6. Since the data from China’s National Forest Continuous Inventory data (NFCI) of Guizhou is not available online the authors have to disclose the used data for example as an attachment to the paper. We have to follow the data transparency in science.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

see attached file

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

The article ‘Optimization of Rocky Desertification Classification Model Based on Vegetation Type and Seasonal Characteristic’ proposes the use of machine learning algorithm for rocky desertification models. Using the data obtained from the samples of 2010, and making use of three different models like Logistic Regression Model, Random Forest Model, and SVM models, they have tested the accuracies of classification. Combining factors like vegetation types and vegetation seasonal phases to their proposed approach, they have further improved the model accuracies.

The flow of article helps the readers understand the workflow, the dataset used for the research, formalization of different models used for the tests, as well as the results obtained. Figures are clear and have the required descriptions. The authors take into consideration several previous works related to rocky desertification classification and those focusing on the area of study.

However, I have a major remark regarding Figure 2. The authors talk about spatiotemporal analyses considering the different years. It is not completely clear to me how this step has been achieved, since the data came mainly from 2010 samples. The authors briefly talk about this multi-year analysis in the discussion section (section 5.4). Figure 2 does not clearly explain the arrows between the years. Are only file differences available? In that case, it is important to have the data from 2001, to obtain the data from 2005, and then 2010 etc. Spatiotemporal analyses are complex, especially when the focus is on understanding the changes. Are the authors focusing on changes using visualization techniques? Or are they using other calculations?

Minor errors

In abstract: After combined them, the -> . After combining them, the

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

While authors made significant progress in improving the manuscript I still can not recommend it for publication due to lack of data transperancy. 

The authors say:  "According to the regulation of the National Forestry and Grassland Administration of China, these survey data are confidential and can’t be disclosed considering safeness."

 If this is secret/commercial/military data you don't need to use it for publication in the open-access journal. 

There is no guarantee that the authors just compiled the results because nobody saw the real data. 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

I would like to thank the authors for taking into account my previous comments. They have added a small description in section 3.1 and updated the Figure 2 to clarify my remark.

Author Response

Thank you very much for your careful review and helpful comments, which indeed help us improve our manuscript accordingly.