An Air Quality Modeling and Disability-Adjusted Life Years (DALY) Risk Assessment Case Study: Comparing Statistical and Machine Learning Approaches for PM2.5 Forecasting
, Ferhat Karaca 1,3, Aisulu Torezhan 2 and Egemen Avcu 4,5
Round 1
Reviewer 1 Report
The authors have made a concise overview of the topic and a brief reference to existing literature. They have indicated the main task of the paper among its motivation. Finally, they have pointed out the key message and the potential benefits of their work. In general, the text is well structured and has clearly defined topics. The abstract is a good guide for what follows. More or less all fundamental theory details that are needed are discussed and concluding remarks are sufficient.
Some comments for improvement:
1. While the objectives of the research work are clearly stated in the last paragraph of the 1st section, its novelty and impact are not described. The authors should clarify the novelty of the work among its added value.
2. As a general drawback, I could say that there is limited reference to similar approaches (e.g. [1],[2]). For example in [1, the authors describe the technical details and limitations of a real-time data collection system. Besides in [2] authors tried to provide some answers to two distinct questions: (1) whether these technologies are fit for the various purposes envisaged and (2) how far these technologies and their applications have progressed to provide answers and solutions.
[1] Spandonidis, C., Tsantilas, S., Giannopoulos, F., Giordamlis, C., Zyrichidou, I., and Syropoulou, P., 2020. Design and Development of a New Cost-Effective Internet of Things Sensor Platform for Air Quality Measurements. Journal of Engineering Science & Technology Review, 13(6).
[2] L. Morawska et al., “Applications of low-cost sensing technologies for air quality monitoring and exposure assessment: How far have they gone?,” Environ. Int., vol. 116, no. April, pp. 286–299, 2018.
3. Did authors perform any kind of correction (e.g. using ambient temperature and/or relative humidity) to their reporting values for PM?
4. Why did the authors perform the statistical analysis? Do all the findings are already known in the literature? What outcome is novel?
5. What is the purpose of Figure 3? Is it needed?
6. Do the authors have a description of the metrics referred to in table 5?
7. Did the authors perform any kind of data pre-processing (outlier removal, smoothening, etc)? If yes what approach did they follow?
8. Authors could provide in a more illustrative way the outcome of the comparisons.
9. Authors could enhance the conclusions section such that it better describes both the outcome of the research and the added value it provides. Restrictions and limitations could also be added.
10. In general, authors could refine the figures and tables following the journal template.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
See attached file
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors have presented a study on air quality on the effect of PM2.5 on population health in Astana, Kazakhstan based on 2 monitoring stations in this city. The measurements include PM2.5, PM10, SO2, CO, NO2, and NO as well well as some meteorological parameters.
The AIRQ+ tool was used to access the health impact in term of DALY (Disability-adjusted Life years) lost. The authors also conducts a study on models for forecasting PM2.5 based on multiple linear regression (MLR) and Random-Forest (RF) which is one of the machine learning methods. I am not clear what the connection between conducting a predicting PM2.5 model and the health impact study when the PM2.5 measurements at the two stations are already obtained. The authors need to clarify the link between the 2 separately approaches in this manuscript, in particular what is the aim of the study. If there are 2 objectives than it needs to make it in the abstract and introduction regarding the aim of the study.
The results are expected and informative regarding the study of air quality and health impact in Kazakhstan, a region that is of interests to researchers on air quality. In addition to the location of the 2 monitoring stations shown in the map, for people who is not familiar to the area, it would be helpful to show the population distribution in this city where most people are living. There are many graphs and tables in the manuscript. Some of these can be put in the Appendix or supplementary materials.
Overall, I recommend the manuscript be accepted for publication with minor revision.
Some specific comments
(1) Line 207: "in S5 but not in S6". Use "at" rather than "in"
(2) Line 250: "Finally, T and WD had a moderate but negative correlation with PM2.5". How 's about WSP. The correlation with WSP is also negative as shown in Table 3
(3) Line 259: "Four predictive models for PM2.5 concentrations have been established for S5 and S6". Why these 4 models ?. On what basis they were chosen. Need to clarify.
(4) Line 305: "These findings are in parallel to recent studies..". It is clearer to change this to "These findings are in agreement with those of recent studies..."
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors managed to cover previous comments
