Effectiveness of Green Infrastructure Location Based on a Social Well-Being Index
Round 1
Reviewer 1 Report
The title and the content of the paper are not so coherent: The focus is not the urban scale, rather the regional one.
It is an interesting approach, but it is not clear why some parameters were chosen instead of others.
The word ecology, ecological, etc. shall be changed because ecosystems are not properly considered or at least not clearly and superficially: GI buffer, GI scenic, GI connection seams arbitrary parameters if there is not a proper description or explanation of what these kinds of GI are and how were calculated/generated.
Scenario Analysis and Discussion should be definitely expanded because otherwise the practical aspect of the work is missing, giving too much space to theory: how these indexes should help policy and people liveability?
Please find detailed comments in the .pdf attached.
Comments for author File: 
Comments.pdf
Author Response
Thank you for giving us the opportunity to respond to the review comments on the manuscript “Investigate of Effective Green Infrastructure Location for Sustainable Urban Space.”
For the response, please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
Review Report:
- What are the drivers of model prediction and city planning with respect to green infrastructure? Mention about the drivers in the result section.
- The authors mentioned about 3 research questions but in the result and conclusion section, answers to only two questions are clear. For example, the result section does not make it clear about the residents, liveability. Add a paragraph discussing the effect on residents' liveability
- Line 186-188: what does the conventional time zone of the entire data mean? What does it signify for the selection of the selected time period?
- The paper can benefit from the inclusion of univariate analysis of the variables for Seoul and Gyeonggi-do.
- The motivation for selecting a random forest regressor to build an ML model is not strongly convincing, as there are methods in ML for data pre-processing to handle outliers, imbalances, etc. Have the authors considered adapting to methods like SVM, or Neural Network, and comparing the performance measures to demonstrate a better model selection scenario?
- Line 321 [REF] - reference missing.
- Inline to FAIR guidelines (https://www.nature.com/articles/sdata201618), it will be beneficial to include the experiment setup details - like, the programming language used, ML implementation toolkit (e.g. - scikit learn in python), and parameters/arguments defined for the ML algorithm.
- Accuracy alone is not a good measure to evaluate the performance of an ML model, as often a higher accuracy can be because of overfitting the model. Inclusion of parameters like MCC, Balanced Accuracy, Kappa value, Sensitivity, Specificity, ROC curve, etc. can give a clearer picture.
Author Response
Thank you for giving us the opportunity to respond to the review comments on the manuscript “Investigate of Effective Green Infrastructure Location for Sustainable Urban Space.”
For the response, please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
Investigation of Effective Green Infrastructure Location for Sustainable Urban Space
General comments
This paper aims to investigate the effectiveness of Green Infraestructure in urban areas that can enhance social welfare benefits for urban sustainability.
The topic is of interest to readers. However, in my opinion, the manuscript has to be thoroughly reviewed and some deficiencies corrected in order to be published.
The paper at times becomes long and tedious, boring the reader. I invite the authors to try to give the manuscript a dynamism that allows it to hold the reader's attention. It is necessary to generally improve the graphics and the definition of formulas and add a discussion section where the results of this investigation are compared with other investigations carried out (even if they are on a smaller scale), also drawing conclusions from this analysis . There are a lot of research published about this topic in MDPI journals.
I now attach a list of more specific recommendations
Specific comments
Abstract should be 200 words maximum. Consider reducing it, since it is too long.
Line 52. Can you update reference 3?
Line 122. ML (machine learning) should be defined the first time is cited.
Line 137. Random Forest regressor is less susceptible to scales than others. Add reference to prove it.
Figure 2. Add a scale bar to both figures. Located small graph is mute. Add names to it.
Table 1 should be placed after it is cited (line 188).
Equation 2. “(Σ?…” , The parenthesis seems to be not necessary.
Line 237. OECD (Organisation for Economic Co-operation and Development) should be defined the first time is cited.
Line 271. “Equation (5) represents the” …. Urban Ecological Index? It seems something is missing.
Figure 4. It is not clear when correlation between two variables is red or blue and why. Does it have to do anything about Urban Ecological Index?
Line 307. Equation (5) should be (6). Line 343 equation (6) should be (7) and so on.
Line 321. [REF]. Include the right number of this reference. Check the rest of the references in case there were any more errors
Line 362. Variables in the equation should be defined.
Figure 6. Explain what SHAP values are, e.g,. add a reference.
Figure 7. Add a color scale value so that you can interpret the results. Distance scale bar is also required to interpret sizes and distances on maps.
Line 428. Discussion section is very poor. “Authors should discuss the results and how they can be interpreted in perspective of previous studies and of the working hypotheses” Other research in this section has not been compared and it is mandatory . Table 2 seems to be more results than a discussion table. May be this part could fit better in result section.
Line 472. Table 2 title should be on next page (16).
This section should answer the 3 questions you ask in the introduction section (lines 119-122). Future research is also lacking.
Comments for author File: Comments.pdf
Author Response
Thank you for giving us the opportunity to respond to the review comments on the manuscript “Investigate of Effective Green Infrastructure Location for Sustainable Urban Space.”
For the response, please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Novelty. The paper is well in line with the emerging use of open data in urban and territorial planning, highlighting the potentiality of their use for a proper implementation of Green Infrastructure at the urban scale. Nevertheless, the green part of the infrastructure is considered a mere number and green qualitative assessment is neglected.
Style. English is not always clear and understandable. Also, there are parts which are very well written (the methodologies for instance) and part which are not (the introduction for instance). Please try to unify the stile of writing through the text. I strongly recommend a final English revision to improve the text readability.
Structure. I believe that the section 2 and 3 shall be merged into a kind of material and methods section, or the subchapter 3 have another title: methodologies of what? Also, scenario analysis is not introduced in the methods, and it is presented after the results and together with the discussions. This is very confusing.
Comments for author File: Comments.pdf
Author Response
We appreciate your thorough considerations and edits for our research. Based on the comments that you provide us in the manuscript (with notes), we accepted all the suggested changes and revised/updated the script based on your comments.
Please see the attachment for the detailed responses.
Thank you.
Best regards,
Authors
Author Response File: 
Author Response.docx
Reviewer 3 Report
Investigation of Effective Green Infrastructure Location for Sustainable Urban Space
The manuscript has solved some of the problems of the previous version. However, in my opinion, the discussion section has hardly been modified and it is one of the poorest sections of the manuscript since it does not compare results with other similar research (even if they are on a smaller scale). There are many papers from the MDPI editorial that can be revised to enhance this discussion. Like I said in previous review, discussion section is very poor. “Authors should discuss the results and how they can be interpreted in perspective of previous studies and of the working hypotheses” Other research in this section has not been compared and it is mandatory. Table 2 seems to be more results than a discussion table. This part could fit better in result section.
On the other hand, there are still some typographical errors, for example in the new line 430 there is a line break that must be eliminated and Equation 2. “(Σ?…” , The opening parenthesis seems to be not necessary.
Comments for author File: Comments.pdf
Author Response
We appreciate your thorough considerations of our research. Please see the attachment for the revised manuscript with track changes and responses.
Author Response File: Author Response.pdf
Round 3
Reviewer 3 Report
Authors have followed the recommendations proposed in the previous review, improving the quality of the manuscript. In my opinion it would be ready to be published
Minor comments. Number conclusión sección
Author Response
We appreciate your thorough considerations and edits for our research. Based on the comments that you provide us, we accepted all the suggested changes and revised/updated the script based on your comments.
Below we have responded to some of your open comments and suggestions in turn. We have placed your original comments and notes in the manuscript in italics, followed by our response in red color.
The authors have followed the recommendations proposed in the previous review, improving the quality of the manuscript. In my opinion, it would be ready to be published
Minor comments. Number conclusión sección
Response:
Based on your comments, the numbering of the conclusion section is updated with the appropriate style.
Thank you very much for your reviews.
Best regards,
Authors
