Machine Learning in the Analysis of Carbon Dioxide Flow on a Site with Heterogeneous Vegetation

Round 1

Reviewer 1 Report

1) Using a list of grouped references is unhelpful for readers. It would be beneficial to provide a brief justification for each individual reference.

2) Check the quality of the figure. E.g. Fig. 4 doesn't seem very high quality to me.

3) References used in the text are focused only on the Asian part. It would be appropriate to work with results from other parts of the world as part of existing studies.

4) The novelty of your work in relation to existing articles requires a clearer elucidation. Offering more commentary on the strengths and limitations of prior publications is necessary. Moreover, a thorough assessment of the uniqueness and contribution of your article is essential.

5) It would be appropriate to provide readers with at least the approximate location of the monitored locality (see Fig. 1) within the country/continent.

6) The description of the methods used in section 2 is very vague.

7) Fig. 4b seems wrong. The content of the probability density (area under the curve) must be equal to 1, this does not seem to be met. Moreover, it is written that it is a distribution function, but that is not true, it is a probability density.

8) "It was found that 93% of the graph of the distribution of CO2 concentration during the growing season coincides with the normal distribution. During the absence of green vegetation, the distribution function of the experimental data is comparable to the Gaussian by 13%." I don't understand this claim, either the data comes from a normal probability distribution or it doesn't. Statistical tests can be used, e.g. the Lilliefors test.

9) The results from Fig. 5 to 7 must be supported by statistical tests.

10) Fig.8: Temperature and air humidity are probably related to each other, this can be seen by the fact that the measured area is not in the entire rectangle. Therefore, this choice of variables for the analysis is not very suitable.

11) There is a question as to whether the article is better suited to another magazine due to its focus - for example, Sustainability.

Author Response

Dear Reviewer,
Thank you for a very detailed description of the shortcomings of the article.

1) Using a list of grouped references is unhelpful for readers. It would be beneficial to provide a brief justification for each individual reference. - We have described the work of the research teams in detail, indicated in the links

2) Check the quality of the figure. E.g. Fig. 4 doesn't seem very high quality to me. - We changed the drawings

3) References used in the text are focused only on the Asian part. It would be appropriate to work with results from other parts of the world as part of existing studies. - We have added studies from other regions of the world

4) The novelty of your work in relation to existing articles requires a clearer elucidation. Offering more commentary on the strengths and limitations of prior publications is necessary. Moreover, a thorough assessment of the uniqueness and contribution of your article is essential. - We describe in more detail the novelty and significance of the work (see introduction and conclusion)

5) It would be appropriate to provide readers with at least the approximate location of the monitored locality (see Fig. 1) within the country/continent. - We have added information

6) The description of the methods used in section 2 is very vague. - We have added a description of materials and methods

7) Fig. 4b seems wrong. The content of the probability density (area under the curve) must be equal to 1, this does not seem to be met. Moreover, it is written that it is a distribution function, but that is not true, it is a probability density. - Figure 4b is correct according to the data. The problem lies in the range of values along the X-axis. We checked. We have corrected the rest of the comments in this paragraph.

8) "It was found that 93% of the graph of the distribution of CO2 concentration during the growing season coincides with the normal distribution. During the absence of green vegetation, the distribution function of the experimental data is comparable to the Gaussian by 13%." I don't understand this claim, either the data comes from a normal probability distribution or it doesn't. Statistical tests can be used, e.g. the Lilliefors test. - We made corrections to the text and used the Harke-Beer test.

9) The results from Fig. 5 to 7 must be supported by statistical tests. - We have made corrections

10) Fig.8: Temperature and air humidity are probably related to each other, this can be seen by the fact that the measured area is not in the entire rectangle. Therefore, this choice of variables for the analysis is not very suitable. - We have removed this analysis from the article.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors

I have gone through the manuscript and I find it suitable for publication. A finding of experimental study on CO2 concentration in regions of Russia are presented. The data is collected and machine learning methods are used to extrapolate the findings. Overall the findings are useful given the importance on reduction of green house gas emissions. 

There is room for improvement of the paper. The figures all of them are very poor quality and very small. They should be improved. Section 2 - materials and method is very small and not much details on the aspects of machine learning is provided in this section. The section needs to be revised. Section 3 of results is comprehensive but missing clear details on machine learning. 

Author need to discuss why machine learning is required for this task and what specific machine learning algorithms are used. Machine learning part needs to be clearly spelled out may be in a separate section.  

May be authors can also comment on the use of machine learning in other areas of science and technology and quote some reference for it. e.g.,

Pal, et.al -  Sensors. Basel : MDPI. ISSN 1424-8220. 2022, vol. 22, iss. 10, art. no. 3662, p. 1-13. DOI: 10.3390/s22103662.

and 

Pal, et.al - // Applied sciences. Basel : MDPI. ISSN 2076-3417. 2021, vol. 11, iss. 23, art. no. 11396, p. 1-13. DOI: 10.3390/app112311396.

Regards

 

Dear Authors

I have gone through the manuscript and I find it suitable for publication. A finding of experimental study on CO2 concentration in regions of Russia are presented. The data is collected and machine learning methods are used to extrapolate the findings. Overall the findings are useful given the importance on reduction of green house gas emissions. 

There is room for improvement of the paper. The figures all of them are very poor quality and very small. They should be improved. Section 2 - materials and method is very small and not much details on the aspects of machine learning is provided in this section. The section needs to be revised. Section 3 of results is comprehensive but missing clear details on machine learning. 

Author need to discuss why machine learning is required for this task and what specific machine learning algorithms are used. Machine learning part needs to be clearly spelled out may be in a separate section.  

May be authors can also comment on the use of machine learning in other areas of science and technology and quote some reference for it. e.g.,

Pal, et.al -  Sensors. Basel : MDPI. ISSN 1424-8220. 2022, vol. 22, iss. 10, art. no. 3662, p. 1-13. DOI: 10.3390/s22103662.

and 

Pal, et.al - // Applied sciences. Basel : MDPI. ISSN 2076-3417. 2021, vol. 11, iss. 23, art. no. 11396, p. 1-13. DOI: 10.3390/app112311396.

Regards

 

Author Response

Dear Reviewer,
Thank you for a very detailed description of the shortcomings of the article. We have made corrections in the article. They are highlighted in red.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors adequately addressed my comments. I recommend the paper for publication.

Reviewer 2 Report

Accept in present form. Changes have been made as recommended. I am satisfied with the changes. 

Thanks