The Impact of the Beijing Winter Olympic Games on Air Quality in the Beijing–Tianjin–Hebei Region: A Quasi-Natural Experiment Study
Round 1
Reviewer 1 Report
The paper is written in good shape
, but I can see some drawbacks
here are some I will talk from a statistical point of view
1- Equation one is for the regression model, and it is a linear model
why do authors use a linear model? What if the relationship is nonlinear?
2- In Table 3. Baseline Regression Results (DID estimation). 367
what is the software used for these calculations?
I suggest sending the code for double check
3- Equation 3 contains some confusing variables. I suggest rechecking it
4- What is the need for the summation sign?
5- Figure 2. The results of the parallel trend test. (a) The parallel trend test for Ln (AQI); (b) Parallel trend 387 test for Ln (PM2.5).
Figure 2 needs some elaboration
6- Figure 3. The results of the placebo test. (A) Placebo test for Ln (AQI); (B) Placebo test for Ln (PM2.5).
figure 3 contains errors on the x-axis
The rest of the paper is very clear to me
I can accept the level of the language
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The study is scientifically sound, and the manuscript gives new information and worthy of publication. Besides, the procedures used to perform the work are quite appropriate. However, the criticism of this manuscript focuses on the design of the Introduction and Section 2. There is also unnecessary general information in the manuscript. That's why I recommend a major revision.
Major points:
INTRODUCTION
The statement of the problem should be better defined. For example, after stating why Beijing Winter Olympics is a threat in terms of air quality, the starting sentence in the Introduction should be used. There is not enough information about the negative effects of the winter Olympics on air quality.
It will be more effective if the information given in the 2.1 subsection under the title of literature review is included in the Introduction section. For this reason, it would be more appropriate to remove Section 2.1 and combine it with the Introduction part in terms of the design of the article.
Page 2: Lines 108-129 Paragraph does not make any contribution to the manuscript. It contains general information. So this should be removed.
Section 2.2: this gives a general information only. It has no contribution to the manuscript. This subsection should be removed.
Title of the 2.3 needs revision. I suggest “Air quality conditions in Beijing-Tianjin-Hebei Region”
It is more appropriate to include the first sentence of this section (lines:161-163) in the Introduction section.
Abstract:There is also this deficiency mentioned in the introduction section of the abstract section. This section should also be revised.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
While the impact analyses of Bejing Olympics on the air quality is not novel, the authors' application of a new statistical method of DID is unique for policy impact analyses on air quality. However, the key assumption by the authors that the economic indicators such as GDP affected the air quality is wrong. Such macroeconomic indicators can influence very long term air quality in a region, but not short-term as analyzed here. It is clearly evident that local traffic regulations decreased the mobile emissions that contributed to improvement in air quality.
The authors also found this out with no significant improvement in correlations by using the control variables in their DID model. Therefore, the reviewer suggest to rephrase some of the text to highlight the neutral impact of economic indicators.
Moreover, using traffic volume data along with engine type and fuel type data as control variables may be more reasonable. The reviewer suggest to analyse the impact of these variables with the PM2.5.
More specific comments below:
Line 77 - please explain briefly what is "quasi-natural experimental" in this context
Line 507 - Authors should explain how machine learning models are "advanced" and can improve precision. This seems to be a general statement without deep understanding of the difference between statistical models and the machine learning models.
Line 230 - this sentence is slightly wrong since other studies did successfully identified the impact. Probably authors would like to rephrase this sentence.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
Detailed comments
1. Literature Review
The general assumptions of DID models, their applications, evaluation and usefulness in air quality research are not described. There are also no relevant literature references.
Study of air quality conditions and prevention measures during the Beijing Winter Olympic Games, v. 165 - PM2.5 concentrations should be given in µg/m3.
The last paragraph of this section is imprecise. It is not known whether it refers to the past or the future. It is not specified what the authors consider to be “spatial spillover effects”.
2. Study Design
What does the contents of Chapter 3.1 have to do with Study design? It's unclear. In the chapter named “Study design”, the concept and research plan should be described, in accordance with the title.
3. 3.2. Research methods
The DID method (difference-in-differences models), its advantages and disadvantages as well as development in the researched area should be described in the Introduction chapter. There, the reader should be provided with information justifying the use of the method in the analysis.
The basic regression model (equation 1) has been described too casually. Please describe the implementation of this model using an example(s), selecting city(s) and year(s). Why was the Treati×Aftert variable introduced into the model? What is the function of this variable in the regression model?
4. 3.3. Variable Settings
It is not explained why logarithm data?
5. 3.3.1. Explained Variable and Explanatory Variable
The description is incomprehensible. What is the explained variable, since the explanatory variables are AQI, PM2.5, NO2, and SO2?
6. 3.3.2. Control Variable
Some of the explanatory variables are called control variables. What is their uniqueness? Please describe.
7. 3.4. Data source
The data has been described very casually. Some of them have daily averaging, others yearly. How the data was implemented into the regression equation. Have they been converted? What has been done about the missing data problem? Why is the number of data for some variables 1296 and for others 126? There are no explanations.
8. 4. Empirical Results
4.1. Main results - Table 3. Baseline Regression Results (DID estimation).
What does Table 1 contain? The title is so general that it doesn't say anything. Regression coefficients? Why wasn't it written? For four different regression models? These models have not been described before, so I can only guess.
9. The following parts of the text are also unclear.
Major comments
In my opinion, the manuscript as presented cannot be published. The entire manuscript should be rewritten. The Authors should describe the methodology and the results and conclusions in such a way that they are understandable to a wide audience.
I encourage you to be concise in your text. All introductory information justifying the research, the choice of research goals and scope, as well as the choice of methods should be described in one, possibly concise, Introduction chapter. I don't want to impose the layout of the content on the Authors, but such approach is preferred in scientific articles. Therefore, I propose to remove repetitive and irrelevant information from the text and re-edit the Introductory chapter and delete chapter 2. Literature Review, introducing its arguments into the Introduction chapter.
The Methodology chapter should describe the researched area, raw data, preparation of data for analysis, modeling methodology - I suggest in this order.
The results should be presented in a clear and precise way.
Only when these conditions are met can the results, quality of discussions and conclusions be assessed.
In my opinion, the article requires editing by a person from outside the research team in order to give the text a form that is understandable to a wide audience. You should also work on the logical consistency of the text and avoiding repetitions and irrelevant information.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The paper is acceptable for me
Author Response
I am pleased that you find the revised manuscript acceptable for publication. Thank you again for reviewing the paper thoroughly and providing constructive feedback throughout the process. Your thoughtful comments have helped improve the work's quality, clarity, and accuracy.
I sincerely appreciate you sharing your expertise and guidance to strengthen the manuscript. Receiving your approval on the revisions is rewarding, as it indicates I have adequately addressed your previous concerns. I am grateful for your diligent review and feedback that have undoubtedly elevated the rigour of this research.
The paper will be significantly enhanced by the insights you have provided as a reviewer. I hope the published work effectively conveys my gratitude for your contributions to revising and strengthening the manuscript. Otherwise, thank you again for your time and helpful feedback throughout the review process. It is greatly appreciated.
Reviewer 2 Report
The authors made significant refinement to meet the requirements of a peer-reviewed paper. The MS can be accepted for publication in Sustainability.
Author Response
After addressing your thoughtful feedback, I am delighted to receive your positive feedback that the manuscript is now suitable for publication in Sustainability. It is gratifying to know my revisions have met the rigorous standards for publication in your esteemed journal.
I sincerely appreciate you recognizing the significant refinements made to improve the quality and rigour of the research. Your insightful critiques motivated me to enhance the manuscript substantially throughout the review process. This paper has benefited tremendously from your expert guidance.
Thank you again for investing your valuable time and energy to ensure this work is methodologically sound before publication. I am grateful for your diligence and high standards that have elevated the accuracy and value of my contribution. Otherwise, thank you again for your thorough, thoughtful, and constructive feedback. Learning from your expertise throughout the review process has been a privilege.
Reviewer 4 Report
I would like to thank the Authors for taking my comments into account in the preparation of the new version of the manuscript. He also suggests paying attention to two aspects, one terminological and the other practical, referring to the discussion of the results.
1. “The explanatory variables” or “the explained variables” (verse 242)? Please, look at “the explained variables” in Table 1 and be consistent.
2. It was decided to log the variables before creating the regression model. However, the practical significance of the results can be meaningfully assessed only after returning to the proper physical dimension. For example, the statement “the ln(AQI) decreased by about 0.287, and the ln (PM2.5) decreased by about 0.332” is incomprehensible because the given values refer to the logarithm values of AQI and PM2.5. - verses 361- 362 and others (v. 561-562). In order to communicate understandable conclusions, changes in the actual scale of AQI and PM2.5 values should be given.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
