the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Dec 2023
Impact of Weather Patterns and Meteorological Factors on PM2.5 and O3 during the Covid-19 Lockdown in China
Abstract. The abnormal haze event in NCP (North China Plain) and the decline in ozone levels in SC (Southern China) from 21st January to 9th February 2020 have attracted public curiosity and scholarly attention during the COVID-19 lockdown. Most previous studies focused on the impact of atmospheric chemistry processes associated with anomalous weather elements in these cases, but fewer studies quantified the impact of various weather elements within the context of a specific weather pattern. To identify the weather patterns responsible for inducing this unexpected situation and to further quantify the importance of different meteorological factors during the haze event, two scenarios are employed. These scenarios compared observations to climatology averaged over the years 2015–2019 and the ‘Business As Usual’ (hereafter referred to as BAU) emission strength, using a novel structural SOM (Self-Organising Map) and ML (Machine Learning) models. The results reveal that the unexpected PM2.5 pollution and O3 decline from the climatology in NCP, North East China (NEC), and SC could be effectively explained by the presence of a double-centre high-pressure system. Moreover, the ML results provided a quantitative assessment of the importance of each meteorological factor in driving the predictions of PM2.5 and O3 under the specific weather system. These results indicate that temperature played the most crucial role in the haze event in NCP and NEC, as well as in the O3 decline in SC. This valuable information will ultimately contribute to our ability to predict air pollution under future emission scenarios and changing weather patterns that may be influenced by climate change.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(9288 KB)
-
Supplement
(11269 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(9288 KB) - Metadata XML
-
Supplement
(11269 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2425', Anonymous Referee #1, 15 Dec 2023
The authors adopt Self-Organizing Map (SOM) and Gradient Boosting Machine (GBM) to classify the weather patterns and identify key local meteorological factors contributing to the haze event characterized by the elevated PM2.5 levels in North China Plain and a slight decrease in ozone levels over Southern China (SC) from 21st January to 9th February 2020. The SOM classifies three types of synoptic-scale weather patterns during the first two months from 2015 to 2020 over China. Subsequently, the authors compare the changes in these three patterns in 2020 with the same period from 2015 to 2019, aiming to elucidate how synoptic patterns may facilitate the haze event in NCP and the decline in ozone levels over SC. On the other hand, the GBM reveals that temperature is the key local meteorological variable explaining the variations in PM2.5 over NCP and North East China (NEC). Surprisingly, it indicates that relative humidity (RH) plays a pivotal role facilitating ozone formation in SC, whereas temperature exerts a suppressing influence.
This work is well within the scope of ACP. The authors present sufficient results concerning the anomalies of PM2.5 and ozone during the study period. However, I’m skeptical about certain results of this study, particularly those derived from the GBM analyses. The acceptance of this manuscript is contingent upon the authors thoroughly validating the GBM results. In addition, several places in this manuscript require substantial improvement. I recommend the authors address the comments and concerns detailed below.
General comment:
After reading this manuscript, my initial impression is that the writing should be improved to reduce ambiguity. For instance, certain lines (e.g., Lines 50 to 52) could be revised to clearly state that the ozone decline is a regional phenomenon, rather than ambiguously suggesting it as a nationwide phenomenon during the study period. Furthermore, the decline is very small (see Fig. 2d), and this point should be clearly emphasized in the main text. Besides, the abbreviation “SC” needs to be consistent, as I’m not sure what SC specifically refers to in this study—Southern China (Line 9 in the abstract)? Southwestern China (Line 31)? Or Southern Coast (line 109)? The description of GBM in section 2.3 should include more details, as it is unclear how the GBM is implemented. For example, how is the cross-validation conducted? What are the hyperparameters?
In terms of the scientific aspect, I have two major concerns. First, the synoptic-scale weather patterns (SWPs) are classified and identified for the first two months of 2015 to 2020, whereas the haze event in NCP and ozone decline in SC occurred from 21st January to 9th February 2020. I’m afraid that the results of SWPs may be too broad to accurately interpret the study period (21st January to 9th February 2020). Second, the authors should validate the counterintuitive GBM results—RH facilitating and temperature suppressing ozone in SC. This does not align with the meteorological effects on ozone during summertime, as demonstrated by the studies of Li et al. (2019), Weng et al. (2022). I understand that there are differences in the study periods between this research and the other two; nonetheless, the fundamental mechanism by which meteorology influences ozone levels should be consistent. Arguably, the key meteorological factors may vary across seasons, but it is unlikely that temperature would suppress ozone formation. Therefore, I suggest that the authors undertake a more comprehensive discussion. For example, the authors could analyze the time series of temperature, RH together with ozone during their study period to investigate any unexpected correlations. It is also important to note that the impact of emission reductions may be equally important, particularly given the massive reductions during the study period.
The title of this manuscript should be revised. The phrase, “COVID-19 Lockdown” is too broad. This is because the authors focus on the period of haze event rather than the entire duration of COVID-19 Lockdown.
Specific comments:
Line 14: What are the two scenarios? Please clarify.
Line 16: I suggest specifying “Gradient Boosting Machine” instead of the generic term “ML (Machine learning)”.
Lines 16 to 17: This sentence is unclear. Why North East China (NEC) is abruptly introduced?
Lines 50 to 53: The later sentence, “In details, the substantial decrease…resulted in an increase in O3…” contradicts the “O3 downward trend” from the previous sentence.
Lines 70 to 71: This statement is too generic. Air quality is not necessarily improved under cyclonic weather systems. It is dependent on the types and positions of the cyclonic weather systems. For example, see Jiang et al. (2015), Wang et al. (2022).
Line 75: what does “100%” refer to here?
Line 155: Yes, predictions using conventional CTMs rely on accurate emission inventories. Besides, uncertainties can also be derived from the chemical mechanism (e.g., Knote et al., 2015; Weng et al., 2023).
Line 175: Please provide more detailed explanations regarding equation (6).
Lines 187 to 189: This sentence is ambiguous. If I understand the authors’ previous work correctly, the -0.9% decline of ozone is driven by the meteorological effect (Shen et al., 2022), not an observed -0.9% decline.
Line 200: Title of table 1 should be revised. I don’t think these are the mean averages.
Line 208: What does “climatology” refer to here?
Lines 244 to 245: Similar to lines 16 to 17, I find this sentence confusing, as I’m not sure which of NCP, NEC and SC has higher/lower PM2.5/ozone.
Line 247: what are “the days” here? The first two months? Same calendar days of the haze event? Similar issues can be found in the captions of Figure 5 to 10.
Line 292: “... is leading to a more moderate ozone increases ...” I suggest avoiding the term “leading to” here as it implies ozone increases are solely driven by meteorology. In addition, as mentioned in the general comment, I suggest giving detailed discussion regarding why negative temperature anomaly can contribute to ozone increases.
Line 321: It is unclear to me how this “business as usual” scenario is constructed. Please elaborate.
Line 341: Following my general comment above, could the higher wind speed be the predominant factor causing the decline in ozone in SC? If this is the case, why is this not depicted in Figure 11 (h)?
References:
Jiang, Y. C., Zhao, T. L., Liu, J., Xu, X. D., Tan, C. H., Cheng, X. H., Bi, X. Y., Gan, J. B., You, J. F., and Zhao, S. Z.: Why does surface ozone peak before a typhoon landing in southeast China?, Atmos. Chem. Phys., 15, 13331–13338, https://doi.org/10.5194/acp-15-13331-2015, 2015.
Knote, C., Tuccella, P., Curci, G., Emmons, L., Orlando, J. J., Madronich, S., Baró, R., Jiménez-Guerrero, P., Luecken, D., Hogrefe, C., Forkel, R., Werhahn, J., Hirtl, M., Pérez, J. L., San José, R., Giordano, L., Brunner, D., Yahya, K., and Zhang, Y.: Influence of the choice of gas-phase mechanism on predictions of key gaseous pollutants during the AQMEII phase-2 intercomparison, Atmos. Environ., 115, 553–568, https://doi.org/10.1016/j.atmosenv.2014.11.066, 2015.
Li, K., Jacob, D. J., Liao, H., Shen, L., Zhang, Q., and Bates, K. H.: Anthropogenic drivers of 2013–2017 trends in summer surface ozone in China, Proc. Natl. Acad. Sci. U. S. A., 116, 422–427, https://doi.org/10.1073/pnas.1812168116, 2019.
Shen, F., Hegglin, M. I., Luo, Y., Yuan, Y., Wang, B., Flemming, J., Wang, J., Zhang, Y., Chen, M., Yang, Q., and Ge, X.: Disentangling drivers of air pollutant and health risk changes during the COVID-19 lockdown in China, npj Clim. Atmos. Sci., 5, 54, https://doi.org/10.1038/s41612-022-00276-0, 2022.
Wang, N., Huang, X., Xu, J., Wang, T., Tan, Z. M., and Ding, A.: Typhoon-boosted biogenic emission aggravates cross-regional ozone pollution in China, Sci. Adv., 8, 1–9, https://doi.org/10.1126/sciadv.abl6166, 2022.
Weng, X., Forster, G. L., and Nowack, P.: A machine learning approach to quantify meteorological drivers of ozone pollution in China from 2015 to 2019, Atmos. Chem. Phys., 22, 8385–8402, https://doi.org/10.5194/acp-22-8385-2022, 2022.
Weng, X., Li, J., Forster, G. L., and Nowack, P.: Large Modeling Uncertainty in Projecting Decadal Surface Ozone Changes Over City Clusters of China, Geophys. Res. Lett., 50, e2023GL103241, https://doi.org/https://doi.org/10.1029/2023GL103241, 2023.
Citation: https://doi.org/10.5194/egusphere-2023-2425-RC1 -
AC2: 'Reply on RC1', Fuzhen Shen, 15 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2425/egusphere-2023-2425-AC2-supplement.pdf
-
AC2: 'Reply on RC1', Fuzhen Shen, 15 Apr 2024
-
RC2: 'Comment on egusphere-2023-2425', Anonymous Referee #2, 24 Jan 2024
General comment:
This manuscript classified different types of synoptic-scale weather patterns during the first two months from 2015 to 2020 over China. Based on the ML (Machine Learning) models, the authors provided a quantitative assessment of meteorological factors in driving the predictions of PM2.5 and O3 under the specific weather system. The authors provided useful information about the anomalies of PM2.5 and O3 during the study period. This study is well within the scope of ACP. However, the discussions of relative results from the ML analyses were not well be demonstrated. While there is a need for minor revision, particularly in the discussion sections. I suggest that this paper could be published in the journal of ACP in case of the comments is addressed by the authors.
- Highlight the new findings of this study. The authors should demonstrate the creative results, especially to differentiate those in previous studies. I think, studies on meteorological effects on driving the predictions of PM5 and O3 have been widely obtained. The authors should introduce more studies about them, and their comparisons with each other should be summarized and discussed in the “Discussion” part.
- Analytical method appeared adequate; however some key procedural and QA/QC details are missing. Please provide more detailsin the manuscript, including the time resolutions of field and reanalysis data, and the uncertainties of ML analysis.
- Gradient Boosting Machine (GBM) was selected for a quantitative assessment of meteorological factors in driving the predictions of PM5 and O3. Does the authors try to compare it with other models, such as random forest, etc.
- Since the authors focus on the COVID-19 lockdown period, and how about the influence of emission reductions on the anomalies of PM5 and O3?
Overall, this paper was well organized, but I still find some explanations for lack of evidence. Please try to improve it.
Specific comments:
Lines 16 to 17: Please clarify the sentence.
Line 75: What does the “100%” mean?
Lines 187 to 189: Please clarify the sentence.
Lines 244 to 245: This sentence is unclear.
Citation: https://doi.org/10.5194/egusphere-2023-2425-RC2 -
AC1: 'Reply on RC2', Fuzhen Shen, 15 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2425/egusphere-2023-2425-AC1-supplement.pdf
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2425', Anonymous Referee #1, 15 Dec 2023
The authors adopt Self-Organizing Map (SOM) and Gradient Boosting Machine (GBM) to classify the weather patterns and identify key local meteorological factors contributing to the haze event characterized by the elevated PM2.5 levels in North China Plain and a slight decrease in ozone levels over Southern China (SC) from 21st January to 9th February 2020. The SOM classifies three types of synoptic-scale weather patterns during the first two months from 2015 to 2020 over China. Subsequently, the authors compare the changes in these three patterns in 2020 with the same period from 2015 to 2019, aiming to elucidate how synoptic patterns may facilitate the haze event in NCP and the decline in ozone levels over SC. On the other hand, the GBM reveals that temperature is the key local meteorological variable explaining the variations in PM2.5 over NCP and North East China (NEC). Surprisingly, it indicates that relative humidity (RH) plays a pivotal role facilitating ozone formation in SC, whereas temperature exerts a suppressing influence.
This work is well within the scope of ACP. The authors present sufficient results concerning the anomalies of PM2.5 and ozone during the study period. However, I’m skeptical about certain results of this study, particularly those derived from the GBM analyses. The acceptance of this manuscript is contingent upon the authors thoroughly validating the GBM results. In addition, several places in this manuscript require substantial improvement. I recommend the authors address the comments and concerns detailed below.
General comment:
After reading this manuscript, my initial impression is that the writing should be improved to reduce ambiguity. For instance, certain lines (e.g., Lines 50 to 52) could be revised to clearly state that the ozone decline is a regional phenomenon, rather than ambiguously suggesting it as a nationwide phenomenon during the study period. Furthermore, the decline is very small (see Fig. 2d), and this point should be clearly emphasized in the main text. Besides, the abbreviation “SC” needs to be consistent, as I’m not sure what SC specifically refers to in this study—Southern China (Line 9 in the abstract)? Southwestern China (Line 31)? Or Southern Coast (line 109)? The description of GBM in section 2.3 should include more details, as it is unclear how the GBM is implemented. For example, how is the cross-validation conducted? What are the hyperparameters?
In terms of the scientific aspect, I have two major concerns. First, the synoptic-scale weather patterns (SWPs) are classified and identified for the first two months of 2015 to 2020, whereas the haze event in NCP and ozone decline in SC occurred from 21st January to 9th February 2020. I’m afraid that the results of SWPs may be too broad to accurately interpret the study period (21st January to 9th February 2020). Second, the authors should validate the counterintuitive GBM results—RH facilitating and temperature suppressing ozone in SC. This does not align with the meteorological effects on ozone during summertime, as demonstrated by the studies of Li et al. (2019), Weng et al. (2022). I understand that there are differences in the study periods between this research and the other two; nonetheless, the fundamental mechanism by which meteorology influences ozone levels should be consistent. Arguably, the key meteorological factors may vary across seasons, but it is unlikely that temperature would suppress ozone formation. Therefore, I suggest that the authors undertake a more comprehensive discussion. For example, the authors could analyze the time series of temperature, RH together with ozone during their study period to investigate any unexpected correlations. It is also important to note that the impact of emission reductions may be equally important, particularly given the massive reductions during the study period.
The title of this manuscript should be revised. The phrase, “COVID-19 Lockdown” is too broad. This is because the authors focus on the period of haze event rather than the entire duration of COVID-19 Lockdown.
Specific comments:
Line 14: What are the two scenarios? Please clarify.
Line 16: I suggest specifying “Gradient Boosting Machine” instead of the generic term “ML (Machine learning)”.
Lines 16 to 17: This sentence is unclear. Why North East China (NEC) is abruptly introduced?
Lines 50 to 53: The later sentence, “In details, the substantial decrease…resulted in an increase in O3…” contradicts the “O3 downward trend” from the previous sentence.
Lines 70 to 71: This statement is too generic. Air quality is not necessarily improved under cyclonic weather systems. It is dependent on the types and positions of the cyclonic weather systems. For example, see Jiang et al. (2015), Wang et al. (2022).
Line 75: what does “100%” refer to here?
Line 155: Yes, predictions using conventional CTMs rely on accurate emission inventories. Besides, uncertainties can also be derived from the chemical mechanism (e.g., Knote et al., 2015; Weng et al., 2023).
Line 175: Please provide more detailed explanations regarding equation (6).
Lines 187 to 189: This sentence is ambiguous. If I understand the authors’ previous work correctly, the -0.9% decline of ozone is driven by the meteorological effect (Shen et al., 2022), not an observed -0.9% decline.
Line 200: Title of table 1 should be revised. I don’t think these are the mean averages.
Line 208: What does “climatology” refer to here?
Lines 244 to 245: Similar to lines 16 to 17, I find this sentence confusing, as I’m not sure which of NCP, NEC and SC has higher/lower PM2.5/ozone.
Line 247: what are “the days” here? The first two months? Same calendar days of the haze event? Similar issues can be found in the captions of Figure 5 to 10.
Line 292: “... is leading to a more moderate ozone increases ...” I suggest avoiding the term “leading to” here as it implies ozone increases are solely driven by meteorology. In addition, as mentioned in the general comment, I suggest giving detailed discussion regarding why negative temperature anomaly can contribute to ozone increases.
Line 321: It is unclear to me how this “business as usual” scenario is constructed. Please elaborate.
Line 341: Following my general comment above, could the higher wind speed be the predominant factor causing the decline in ozone in SC? If this is the case, why is this not depicted in Figure 11 (h)?
References:
Jiang, Y. C., Zhao, T. L., Liu, J., Xu, X. D., Tan, C. H., Cheng, X. H., Bi, X. Y., Gan, J. B., You, J. F., and Zhao, S. Z.: Why does surface ozone peak before a typhoon landing in southeast China?, Atmos. Chem. Phys., 15, 13331–13338, https://doi.org/10.5194/acp-15-13331-2015, 2015.
Knote, C., Tuccella, P., Curci, G., Emmons, L., Orlando, J. J., Madronich, S., Baró, R., Jiménez-Guerrero, P., Luecken, D., Hogrefe, C., Forkel, R., Werhahn, J., Hirtl, M., Pérez, J. L., San José, R., Giordano, L., Brunner, D., Yahya, K., and Zhang, Y.: Influence of the choice of gas-phase mechanism on predictions of key gaseous pollutants during the AQMEII phase-2 intercomparison, Atmos. Environ., 115, 553–568, https://doi.org/10.1016/j.atmosenv.2014.11.066, 2015.
Li, K., Jacob, D. J., Liao, H., Shen, L., Zhang, Q., and Bates, K. H.: Anthropogenic drivers of 2013–2017 trends in summer surface ozone in China, Proc. Natl. Acad. Sci. U. S. A., 116, 422–427, https://doi.org/10.1073/pnas.1812168116, 2019.
Shen, F., Hegglin, M. I., Luo, Y., Yuan, Y., Wang, B., Flemming, J., Wang, J., Zhang, Y., Chen, M., Yang, Q., and Ge, X.: Disentangling drivers of air pollutant and health risk changes during the COVID-19 lockdown in China, npj Clim. Atmos. Sci., 5, 54, https://doi.org/10.1038/s41612-022-00276-0, 2022.
Wang, N., Huang, X., Xu, J., Wang, T., Tan, Z. M., and Ding, A.: Typhoon-boosted biogenic emission aggravates cross-regional ozone pollution in China, Sci. Adv., 8, 1–9, https://doi.org/10.1126/sciadv.abl6166, 2022.
Weng, X., Forster, G. L., and Nowack, P.: A machine learning approach to quantify meteorological drivers of ozone pollution in China from 2015 to 2019, Atmos. Chem. Phys., 22, 8385–8402, https://doi.org/10.5194/acp-22-8385-2022, 2022.
Weng, X., Li, J., Forster, G. L., and Nowack, P.: Large Modeling Uncertainty in Projecting Decadal Surface Ozone Changes Over City Clusters of China, Geophys. Res. Lett., 50, e2023GL103241, https://doi.org/https://doi.org/10.1029/2023GL103241, 2023.
Citation: https://doi.org/10.5194/egusphere-2023-2425-RC1 -
AC2: 'Reply on RC1', Fuzhen Shen, 15 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2425/egusphere-2023-2425-AC2-supplement.pdf
-
AC2: 'Reply on RC1', Fuzhen Shen, 15 Apr 2024
-
RC2: 'Comment on egusphere-2023-2425', Anonymous Referee #2, 24 Jan 2024
General comment:
This manuscript classified different types of synoptic-scale weather patterns during the first two months from 2015 to 2020 over China. Based on the ML (Machine Learning) models, the authors provided a quantitative assessment of meteorological factors in driving the predictions of PM2.5 and O3 under the specific weather system. The authors provided useful information about the anomalies of PM2.5 and O3 during the study period. This study is well within the scope of ACP. However, the discussions of relative results from the ML analyses were not well be demonstrated. While there is a need for minor revision, particularly in the discussion sections. I suggest that this paper could be published in the journal of ACP in case of the comments is addressed by the authors.
- Highlight the new findings of this study. The authors should demonstrate the creative results, especially to differentiate those in previous studies. I think, studies on meteorological effects on driving the predictions of PM5 and O3 have been widely obtained. The authors should introduce more studies about them, and their comparisons with each other should be summarized and discussed in the “Discussion” part.
- Analytical method appeared adequate; however some key procedural and QA/QC details are missing. Please provide more detailsin the manuscript, including the time resolutions of field and reanalysis data, and the uncertainties of ML analysis.
- Gradient Boosting Machine (GBM) was selected for a quantitative assessment of meteorological factors in driving the predictions of PM5 and O3. Does the authors try to compare it with other models, such as random forest, etc.
- Since the authors focus on the COVID-19 lockdown period, and how about the influence of emission reductions on the anomalies of PM5 and O3?
Overall, this paper was well organized, but I still find some explanations for lack of evidence. Please try to improve it.
Specific comments:
Lines 16 to 17: Please clarify the sentence.
Line 75: What does the “100%” mean?
Lines 187 to 189: Please clarify the sentence.
Lines 244 to 245: This sentence is unclear.
Citation: https://doi.org/10.5194/egusphere-2023-2425-RC2 -
AC1: 'Reply on RC2', Fuzhen Shen, 15 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2425/egusphere-2023-2425-AC1-supplement.pdf
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 323 | 121 | 23 | 467 | 42 | 12 | 21 |
- HTML: 323
- PDF: 121
- XML: 23
- Total: 467
- Supplement: 42
- BibTeX: 12
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Michaela I. Hegglin
Yue Yuan
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(9288 KB) - Metadata XML
-
Supplement
(11269 KB) - BibTeX
- EndNote
- Final revised paper