Review Reports

Round 1

Reviewer 1 Report

This review is focused on the article "Forecasting Amazon rain-forest deforestation using a hybrid machine learning model" by David Dominguez et al. The article describes the analysis of historical and future deforestation of the Amazon forest.

In general, I found the approach of the article very important, However, I consider the structure of the paper need to be logically arranged and need to pay clear attention to represent the results. 

Please find below major and minor suggestions to address by the Authors before publication.

Major suggestions

The manuscript is not carefully following the Journal guidelines and I highly recommended to the Authors carefully follow the guidelines and rearrange the paper for proper format of Sustainability format.

Methodology - This part is well written but I believe that it is providing more theoretical information and it needs present in a shorter format.

It is important to provide a description of input data separately by changing Chapter 2 to “materials and method” as the journal also follows the same procedure. 

When the paper is structured I feel a lack of results in the manuscript. A scientific paper should present enough results than the theoretical approaches.

Recommended to rearrange the figures and tables in with a logical flow and to improve the readability and clarity.

It is very important to provide the level of prediction accuracy of the deforestation model through the ML. 

It is good that authors have used recent citations but still, I feel a lack of citation in the paper.

 

Minor suggestions

It is recommended to move Figure 1 into the section 2.1

The figures can be improved to help readers by adding labels to the panels, such as (a), (b), etc instead of “Left and right panel”,  the can then be referred to in the main text easily (Figure 6 (a)). This is standard practice, and it will be easier for readers to see the labels than the names of places in the panels.  

Table 3. What is the unit shown in the Table? Authors must provide information very carefully.

 

Author Response

Please see the attachment, pages 1 to 2. Thank you for your thoughtful revision.

Author Response File: Author Response.pdf

Reviewer 2 Report

The present study aims to carry out an analysis of the Amazon rain-forest deforestation, which can be analyzed from actual data and predicted by means of artificial intelligence algorithms. The subject addressed is interesting and within the scope of Sustainability.  Suggested revisions:

Abstract

• Please add quantitive results of annual deforestation area increasing in the abstract.

Introduction

• Please add a workflow graph after the last paragraph of the introduction.

Methodology

• How do you discern the fake and real data in the augmentation section?
• The results of augmentation should be added to the results section.

Results

• Please add a legend in Figure 10.
• I think that scatter plots of predicted and observed data together R squared values should be added to the results section to show the ability of the presented method in modeling deforestation, and then results of prediction are presented.
• The results section is too short.
• I suggested that the results of standelon model of LSTM for modeling of deforestation be added to results section. It is not nessessery to considered static data in standalone LSTM. This proves that your hybrid method was better than the standelon methods.

Another revisions:

• Please add graphical abstract to the manuscript.
• Please add a highlight to the manuscript.

Author Response

Please see the attachment pages 2 to 4. Thank you for your thoughtful revision.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The improvements I suggested have been made.