the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
04 Jan 2021
04 Jan 2021
Substantial changes of gaseous pollutants and chemical compositions in fine particles in North China Plain during COVID-19 lockdown period: anthropogenic vs meteorological influences
- 1Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention, Department of Environmental Science & Engineering, Institute of Atmospheric Sciences, Fudan University, Shanghai, 200433, P.R. China
- 2Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science and Technology, Nanjing 210044, P.R. China
- 1Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention, Department of Environmental Science & Engineering, Institute of Atmospheric Sciences, Fudan University, Shanghai, 200433, P.R. China
- 2Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science and Technology, Nanjing 210044, P.R. China
Abstract. The rapid response to COVID-19 pandemic led to the unprecedented decreases of economic activities, thereby reducing the pollutant emissions. A random forest (RF) model was applied to determine the respective contributions of meteorology and anthropogenic emissions to the changes of air quality. The result suggested the strict lockdown measures significantly decreased primary components such as Cr (−201 %) and Fe (−154 %) in PM2.5, whereas the higher relative humidity (RH) and NH3 level, and the lower air temperature (T) enhanced the production of secondary aerosol including SO42− (47.2 %), NO3− (38.6 %), and NH4+ (22.7 %). Positive matrix factorization (PMF) result suggested that the contribution ratios of secondary formation (SF), industrial process (IP), biomass burning (BB), coal combustion (CC), and road dust (RD) changed from 35.2 %, 28.9 %, 19.4 %, 11.8 %, and 4.75 % before COVID-19 outbreak to 42.7 %, 20.5 %, 19.45 %, 9.80 %, and 7.56 %, respectively. The rapid increase of the contribution ratio derived from SF to PM2.5 implied the intermittent haze events during COVID-19 period were characterized with secondary aerosol pollution, which was mainly contributed by the unfavorable meteorological conditions and high NH3 level.
- Preprint
(1351 KB) -
Supplement
(1314 KB) - BibTeX
- EndNote
Rui Li et al.
Status: final response (author comments only)
-
CC1: 'Comment on acp-2020-1199', Markus Furger, 13 Jan 2021
I found the approach interesting and would like to see this article published. It shows a method to quantify the influence of meteorological parameters on pollutant time series and thus helps to better quantify the impact of COVID-19 lockdown measures on air pollution. There are, however, a few points that require better explanation or change before publication:
- A decrease of more than 100 % of a pollutant does not make sense. You cannot remove more than all the pollutant. The authors should reconsider which reference value they use to calculate the percent reduction. This relates to the abstract and section 3.2 (lines 193 ff), and probably also to the figures (see remark 3).
- The random forest approach (line 118) should be referenced with a citation to the literature. Why do you use this approach instead of other statistical methods?
- The observed or calculated decreases of pollutants in Fig. 2, 3 and 4 are not sufficiently explained. It is unclear what the trendlines (arrows) mean. The percent values do not explain the slopes of the curves, and it is unclear what the arrows should show. There is definitively more explanation needed to understand which concentrations are compared and used to calculate the reductions, either in the main text, in the figure captions, or in the supplement.
-
AC1: 'Reply on CC1', Hongbo Fu, 27 Mar 2021
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2020-1199/acp-2020-1199-AC1-supplement.pdf
-
RC1: 'Comment on acp-2020-1199', Anonymous Referee #1, 09 Feb 2021
Review of ACP-“Substantial changes of gaseous pollutants 1 and chemical compositions in fine particles in North China Plain during COVID-19 lockdown period: anthropogenic vs meteorological influences”
This paper analyzed changes of gaseous air pollutants as well as chemical compositions of PM2.5 based on the observational data in Tangshan in the North China Plain. A random forest model was applied to investigate the contributions of meteorology and anthropogenic emissions to the changes of air quality during COVID-19 period. PMF was further applied to determine changes of source contributions before and during the lockdown. The topic is interesting, however, there are several major concerns to be addressed before it can be considered for publication:
(1) the most important data used in this study is the observation data based on a supersite in Tangshan whereas the title is “North China Plain”. NCP covers large area with different topographical and meteorological conditions. Presenting the observation data at a single site is far enough to draw conclusion for a large and heterogeneous region. So the first limitation of this study whether this site is representative of the whole NCP region.
(2) Nine trace elements observed by Xact 625 are used in this study; however, Xact 625 can observe more than 20 species. Why did the authors only present 9 trace elements? What about the others?
(3) The authors used RF to quantify the influences from meteorology and emissions on air quality. Why do the authors use RF instead of other machine learning techniques? This should be clarified. In addition, how can the model results be evaluated, or in orther words, how do the authors demonstrate the robustness of their model results?
(4) When performing source apportionment with PMF, the major input data does not even include major chemical components like OC/EC. How will this influence the results?
Specific comments:
- Abstract should be rewritten; the major findings/answers to the question raised in the title should be present in the abstract. In addition, it is misleading to present results like Cr (-201%) and Fe (-154%). The decrease of more than 100 % of a pollutant is misleading.
- The NH3 measurement is made by GAC-IC, while Hg is observed with Xact 625. Quality assurance and quality control procedures and results should be well documented and provided at least as supporting information.
- In “Section 2.2 Deweathered model development”, the authors mainly described random forest; however, the deweathered technique with RF is not described. This should be clearly described in detail here. How are the meteorological conditions isolated by RF? How can the results be evaluated?
- Sulfate and NOx concentration decreased substantially after lockdown, while nitrate and ammonium increased. The authors explained that this might be due to the adverse meteorological conditions. This explanation is very weak. NO2 has been reduced by 62.8% while nitrate increased by 2.17%; is this attributable to the adverse meteorology? How did meteorological parameters change before and after the lockdown? Are there any changes in chemical reactions that are responsible?
- Some elements like Pb, Ca, Cr, Cu, Fe and Zn has been reduced by 7.44~91.5% while Hg, K, Ni increased by 20%, 0.08% and 1.17%, respectively. What’s the reason? The authors also mentioned that the slight increase of K might be linked with the unfavorable meteorological conditions. This explanation is still very weak and not convincible. The authors should give detailed data analysis regarding the changes of meteorological parameters to support their explanation.
- Line 194-196, the deweathered Hg concentration still kept stable increase by 18%, which is opposite to other trace elements. What’s the reason?
- Line 207, the deweathered Ca concentration decreased by more than 100%, it is hard to believe. Again, in Line 346, the Pb(-147%), Zn (-219%). This kind of description should be well clarified.
- Fig5 (C) some bars are not well shown, like Cr, Fe, Zn.
- Fig6-8 It does not make any sense to indicate DOY/Year is important or not in the prediction of gaseous pollutants using RF.
- There are many English grammar errors. The language should be polished thoroughly. For example, Line 130-131, “…were input into the model” there is grammar error in this sentence. Line 252, “both of ” should be “both”.
-
AC2: 'Reply on RC1', Hongbo Fu, 27 Mar 2021
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2020-1199/acp-2020-1199-AC2-supplement.pdf
-
RC2: 'Comment on acp-2020-1199', Anonymous Referee #2, 16 Feb 2021
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2020-1199/acp-2020-1199-RC2-supplement.pdf
-
AC3: 'Reply on RC2', Hongbo Fu, 27 Mar 2021
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2020-1199/acp-2020-1199-AC3-supplement.pdf
-
AC3: 'Reply on RC2', Hongbo Fu, 27 Mar 2021
Rui Li et al.
Rui Li et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 384 | 155 | 17 | 556 | 33 | 4 | 6 |
- HTML: 384
- PDF: 155
- XML: 17
- Total: 556
- Supplement: 33
- BibTeX: 4
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1