Estimating the Impact of COVID-19 on the PM2.5 Levels in China with a Satellite-Driven Machine Learning Model
Round 1
Reviewer 1 Report
The manuscript describes China's atmospheric pollution during the COVID pandemic. The article is very clear about the methodology applied and the description of the results. I believe it can be published with minimal corrections.
I suggest to the authors to quote in the captions of the figures the software used to describe the spatial distribution.
In the conclusion section, the authors could expand what future research they could generate with the article.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This is the third time I am looking at this paper.
The contents and presentation are now fine, and can be published as it is.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The impact of COVID-19 lockdown on regional air quality is studied on the basis of satellite observations because the ground measurements of PM2.5 concentrations do not offer comprehensive spatial coverage especially in suburban and rural regions. A machine learning method with satellite aerosol remote sensing data, meteorological fields and land use parameters as major predictor variables is developed to estimate spatiotemporally resolved daily PM2.5 concentrations in China. The study period consists of a reference year (November 1, 2018 – April 30, 2019) and a pandemic year (November 1, 2019 – April 30, 2020), with six modeling months in each year. Each period was then divided into period 1 (November and December), period 2 (January and February) and period 3 (March and April). The reference year model obtained a 10-fold cross-validated R2 (RMSE) of 0.79 (17.55 ug/m3) and the pandemic year model obtained a 10-fold cross validated R2 (RMSE) of 0.83 (13.48 ug/m3) for daily PM2.5 predictions. The prediction results show high PM2.5 concentrations in the North China Plain, Yangtze River Delta, Sichuan Basin and Xinjiang Autonomous Region during the reference year. PM2.5 levels were lowered by 4.8 ug/m3 during the pandemic year comparing to the reference year and PM2.5 levels during period 2 decreased by 18%. The Southeast region was affected most by the COVID-19 outbreak with PM2.5 levels during period 2 decreased by 31%, following by the Northern Yangtze River Delta (29%) and Pearl River Delta (24%).
General comments
The objective of this study is that spatiotemporally resolved daily PM2.5 estimates over 5 km × 5 km grid cells across China allow epidemiologists to further quantify the health benefits of reduced air pollution with higher accuracy. Unfortunately, there is no information if satellite observations are possible during all overpasses i.e., the availability of satellite observations is not presented. That means the data basis cannot be evaluated. But this is necessary to review if the study objective is met. It also not presented how the different data basis of satellite observations and ground-based observations for daily averages is operated. These weaknesses reduce the information content of the study results as mentioned in the chapter Discussion: “One limitation of this study is the incomplete spatial coverage due to cloud and snow cover, especially in northeastern China, which may introduce region-specific sampling biases when estimating mean PM2.5 levels in each period.”
The final sentence in the chapter Conclusion is really a conclusion. It would be helpful to explain it in more detail.
The paper addresses relevant scientific questions. The paper is presenting novel concepts, ideas and tools.
The scientific methods and assumptions are valid and clearly outlined so that substantial conclusions are reached.
The description of experiments and calculations are sufficiently complete and precise to allow their reproduction by fellow scientists.
The quality, information of the figures and tables is good.
The related work is completely cited.
Title and abstract reflect the whole content of the paper.
The overall presentation is well structured and clear. The language is fine.
The abbreviations and units are generally correctly defined and used.
Specific Comments
The capture information of figures and tables must be improved to understand the figures and tables without reading the text.
Technical corrections
Line 167: pattern instead of pastern
A lot of doi are missing in the references.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
