the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Nov 2023
Weakened aerosol-radiation interaction exacerbating ozone pollution in eastern China since China’s clean air actions
Hao Yang
Lei Chen
Hong Liao
Jia Zhu
Wenjie Wang
Abstract. Since China’s clean air action, PM2.5 air quality has been improved while ozone (O3) pollution has been becoming severe. Here we apply a coupled meteorology-chemistry model (WRF-Chem) to quantify the responses of aerosol-radiation interaction (ARI), including aerosol-photolysis interaction (API) and aerosol-radiation feedback (ARF), to anthropogenic emission reductions from 2013 to 2017, and their contributions to O3 increases over eastern China in summer and winter. Sensitivity experiments show that the decreased anthropogenic emissions play a more prominent role for the increased MDA8 O3 both in summer (+1.96 ppb vs. +0.07 ppb) and winter (+3.56 ppb vs. -1.08 ppb) than the impacts of changed meteorological conditions. The decreased PM2.5 caused by emission reduction can result in a weaker impact of ARI on O3 concentrations, which poses a superimposed effect on the worsened O3 air quality. The weakened ARI due to decreased anthropogenic emission aggravates the summer (winter) O3 pollution by +0.81 ppb (+0.63 ppb) averaged over eastern China, with weakened API and ARF contributing 55.6 % (61.9 %) and 44.4 % (38.1 %), respectively; this superimposed effect is more significant for urban areas during summer (+1.77 ppb). Process analysis indicates that the enhanced chemical production is the dominant process for the increased O3 concentrations caused by weakened ARI both in summer and winter. This study innovatively reveals the adverse effect of weakened aerosol-radiation interaction due to decreased anthropogenic emissions on O3 air quality; more stringent coordinated air pollution control strategies are needed for future air quality improvement.
- Preprint
(2571 KB) - Metadata XML
-
Supplement
(3905 KB) - BibTeX
- EndNote
Hao Yang et al.
Status: open (until 14 Dec 2023)
-
RC1: 'Comment on egusphere-2023-2393', Anonymous Referee #1, 20 Nov 2023
reply
General comments:
This paper mainly investigated the impacts of aerosol-photolysis interaction (API) and aerosol-radiation feedback (ARF) on the surface ozone concentrations under the background of China's clean air action (rapid anthropogenic emission reductions from 2013 to 2017).
The effects of API on ozone concentrations are not a new finding since I have found several previous studies already addressed it (Gao et al., 2022; Liu and Wang, 2020). However, I have not found any previous studies focused on the effects of ARF on ozone concentrations. Furthermore, the authors used the IPR methodology to investigate the contribution to O3 concentration variation from four processes (VMIX, CHEM, ADVH, ADVZ). In conclusion, I consider this paper valuable for publication, even if it has some limitations (as shown below).
- The absence of SOA formation and heterogeneous reactions in their simulations could be a limitation of this study; even the authors have sufficiently acknowledged this.
- Some parts/aspects are poorly elucidated, making it hard for me to understand.
A major revision is needed before it can be published in ACP.
Specific comments:
In my opinion, SOAs account for a substantial portion of total aerosols. Typically, in your research, the lack of consideration of SOA can truly affect the reliability of the results (the authors also mentioned that PM2.5 is underestimated in your model). I highly recommend the authors include SOA formation in their model.
Similarly, as the significant impacts of heterogeneous reactions on ozone concentrations mentioned by previous studies (Lou et al., 2014; Liu and Wang, 2020), I would expect the authors to include heterogeneous reactions in their models. If the authors have specific reasons for not including heterogeneous reactions in their models, those reasons need to be stated in the paper.
L160, you mentioned you fixed the meteorological field to the year 2013, can you explain how to achieve this? Can I understand that all *17M cases have exactly the same meteorological fields throughout 2017 simulation? However, I don't think all *17M cases should have the same meteorological fields, because you cannot investigate deltaO3_deltaARF_EMI if the meteorological fields are fixed in different cases. This needs to be explained more clearly in your paper.
L23-L25, you mentioned API and ARF. However, the API and ARF terminology is so abstract, making it hard for people to understand. It would help if you mentioned that API is related to the change in photolysis rates and ARF is related to the change of meteorological fields in your abstract.
L58, I think chemical species like CO and CH4 can also lead to the formation of O3.
L57-L62 The causal relationship between the following two sentences is not clear.
As a secondary air pollutant, troposphere O3 can be produced by nitrogen oxides (NOx = NO + NO2) and volatile organic compounds (VOCs) in the presence of solar radiation through photochemical reactions (Atkinson, 2000; Seinfeld and Pandis, 2006). - > Consequently, the concentration of O3 is closely related to changes in meteorological conditions and anthropogenic emissions (Wang et al., 2019; Liu and Wang, 2020a,b; Shu et al., 2020).
"solar radiation" is not directly related to "meteorological conditions", try to revise those sentences to make them more logical.
2.1 Model configuration: I recommend using a chart (like Table 1 in https://www.sciencedirect.com/science/article/pii/S1352231020307378 ) to summarize the model configuration.
L125-L127, have you applied meteorological nudging? See above, I am not sure how you fix the meteorological fields to 2013 or 2017 when running the model.
L151, you mentioned the biogenic emissions are calculated online by MEGAN. Have you coupled the MEGAN model with WRF-Chem dynamically? Please ascertain whether the biogenic emissions are calculated online or offline by MEGAN.
L166, can you explain which aerosol optical properties are turned to zero?
L200-202, you mentioned "To avoid potential deviations caused by long-term model integration, each simulation is re-initialized every eight days". I was confused about why re-initialize the simulation every eight days can avoid potential deviations. What do you mean "potential deviations"? Can you explain this more?
L214-217 I feel confused about how many sites are operated by China National Environmental Monitoring Center (CNEMC)? You mentioned "1296 sites", does this number refer to the number of total sites of CNEMC or the number of sites chosen in your research? Moreover, are there really 1296 points (sites) on Figs. 2a and 2c?
Figure 2 shows the simulated results of which case? (BASE_17E17M?) You need to specify this point in L251 and Fig. 2.
Why there are less points on Figs. 3a and 3d than Fig. 2? Please explain.
L221, if possible, I recommend explaining more about IPR in your paper.
Table 2, how many sites are used for Table 2 (1296 sites?)?
L284-285, you mentioned NOx-limited and VOCs-limited regions, I recommend that you could add a figure (based on your simulation results) like Fig. 5 in https://www.sciencedirect.com/science/article/pii/S1352231013000514 to your supplement, to show different O3-sensitive regions on the map.
L290-292, the meteorological effects are comparable or larger or smaller than emissions effects? This should be mentioned.
L429-431, you mentioned "multi-pollutants coordinated emissions control strategies", can you specify this and give more details? Liu and Wang, 2020 suggested that "to reduce O3 levels in major urban and industrial areas, VOC emission controls should be added to the current NOx-SO2-PM policy". Does your research have similar insights, or can you make other recommendations that could help policymakers?
Technical corrections
L211, "353 stations" - > "353 meteorological stations"
Figure S5, "from 2013to" - > "from 2013 to"
Figure 6, "on the right side of each panel" - > "on the upper right side of each panel"
Data and code availability should be added.
Reference:
Gao, J., Li, Y., Xie, Z., Hu, B., Wang, L., Bao, F., and Fan, S.: The impact of the aerosol reduction on the worsening ozone pollution over the Beijing-Tianjin-Hebei region via influencing photolysis rates, Sci. Total Environ., 821, 153197, https://doi.org/10.1016/j.scitotenv.2022.153197, 2022.
Liu, Y. and Wang, T.: Worsening urban ozone pollution in China from 2013 to 2017 – Part 2: The effects of emission changes and implications for multi-pollutant control, Atmospheric Chem. Phys., 20, 6323–6337, https://doi.org/10.5194/acp-20-6323-2020, 2020.
Lou, S., Liao, H., and Zhu, B.: Impacts of aerosols on surface-layer ozone concentrations in China through heterogeneous reactions and changes in photolysis rates, Atmos. Environ., 85, 123–138, https://doi.org/10.1016/j.atmosenv.2013.12.004, 2014.
Citation: https://doi.org/10.5194/egusphere-2023-2393-RC1 -
RC2: 'Comment on egusphere-2023-2393', Anonymous Referee #2, 21 Nov 2023
reply
This study examines the role of aerosol-radiation interaction (ARI), decomposed into aerosol-photolysis interaction (API) and aerosol-radiation feedback (ARF) on surface ozone concentration in China. Surface ozone increased remarkable in eastern China, contrasting the dramatic decline of PM2.5 concentrations. It is therefore necessary to investigate the reasons for the ozone increase. The study found that reduced ARI due to decreased PM concentrations contributes to ozone production, with API playing a more important role than ARF. The regional differences are also briefly discussed. I think this is a nice study that is helpful in understanding the recent ozone increase in China. I only have a few minor comments.
- A previous study seemed to indicate that chemical processes associated with PM2.5 reduction, i.e., reduced removing rate of hydroperoxy radicals, is the main reason for the ozone increase in eastern China (Li et al., 2019, PNAS). I wonder how this effect compare to the ARI discussed in this study?
- In the WRF-Chem experiments, the authors zeroed off aerosol optical properties to exclude ARF. I wonder if aerosol microphysical properties are still included? This may affect cloud properties and still impact the radiation budget.
- Section 3.2, model evaluation: why not also evaluate VOCs, which is also an important precursor for ozone?
- Line 87 and associated discussions: Does ARI always suppress O3 formation? Could the change the meteorological variables through ARF increase O3 concentration, say by reducing RH or increasing regional transport?
- I suggest the authors discuss more about the summer-winter differences. Wintertime has much less radiation and lower temperature, so ARI is in general much lower. In summer, meteorology seems to make large contributions than emission changes (Figure 4, left column), what might be the reason?
- Figure 4: model seems to significantly underestimate the ozone change in BTH for summer (Figure 4a2). This area experienced the most ozone increases in the past decade. So it is important for the model to correctly represent ozone trend in this region. What might be the reason for this significant bias?
- Finally, the effects of API and ARF may not be independent, i.e., there may be nonlinear interaction between the two effects. This should be noted and discussed.
Citation: https://doi.org/10.5194/egusphere-2023-2393-RC2 -
RC3: 'Comment on egusphere-2023-2393', Anonymous Referee #3, 24 Nov 2023
reply
The manuscript focuses on the aerosol-radiation interaction (ARI), discussing how this process has changed in the context of the abrupt aerosol decrease in East China during 2013-2017, and evaluates its contribution to the recent ozone increase in China. ARI is divided into aerosol-photolysis interaction (API) and aerosol-radiation feedback (ARF), with the WRF-Chem model used to quantify these impacts. The authors have found non-negligible ozone increase resulting from the aerosol decrease through the API and ARF processes, which has implications for the synergistic control of aerosol and ozone. This is an interesting topic and I believe it can make a novel contribution to the community. However, several important aspects need to be addressed before it can be published in ACP.
General comments:
- The study focuses on aerosol-radiation interaction (ARI), which is split into two parts: the direct aerosol impact on radiation through scattering and absorbing (API) and the subsequent feedback on meteorology (ARF), with both influencing ozone concentrations. However, the Introduction Section could do a better job at breaking down these concepts. A detailed explanation of the distinctions between API and ARF would aid comprehension. Also, elucidating the specific ARF-related meteorological variables and their influences on ozone concentrations would be beneficial. Regarding the cited papers, such as Hong et al. (2020) and Zhu et al. (2021), the authors may consider including additional information about which ARF-related meteorological factors have been identified as important in affecting ozone concentrations.
- In Section 3.2, could the authors talk more about how well the model is doing in reproducing the observed decrease in PM5 levels from 2013-2017. This analysis is crucial for assessing whether the model’s effectively capturing the weakening of ARI.
- Section 4 needs to be better organized for clarity. I’ve outlined some areas for consideration:
- The titles suggest Section 4.1 should focus on ΔO3_MET and ΔO3_EMI, while 4.2 should be devoted to ΔO3_ΔARI_EMI. However, there is content overlap since 4.1 also examines ΔO3_ΔARI_EMI, which obscures the distinctions between the two subsections.
- Section 4.1 discusses ΔO3_MET, ΔO3_EMI, and ΔO3_ΔARI_EMI at sparse polluted grids (so-called urban areas) while 4.2 talks about ΔO3_ΔARI_EMI in term of regional averages. It is unclear why the discussion about ΔO3_MET and ΔO3_EMI focuses only on urban polluted regions. Also, the rationale for addressing urban ΔO3_ΔARI_EMI prior to regional averages is not evident, particularly when urban results mirror the regional ones, though more pronounced. I recommend relocating the OBS-SIM ozone change comparison from Section 4.1 to Section 3.2 (to combine it with PM5 change evaluation) and discussing regional ΔO3_ΔARI_EMI before the urban analysis.
- Section 4.3 and Figure 7 are quite similar to Section 4.2 and Figure 5. Please consider merging Sections 4.2 and 4.3.
- Could the authors explain why ΔO3_ΔARI_EMI displays a much steeper spatial gradient in summer compared to winter (Fig. 5), whereas the PM5 change suggest the opposite pattern (Fig. S8)? How does meteorology contribute to this discrepancy? Moreover, why does summertime ΔO3_ΔARI_EMI exhibit both positive (e.g., NCP) and negative (e.g., Shandong province) values, even though the PM2.5 decreases universally?
- From my understanding, the reduced impact of ARI on ozone is a component of the anthropogenic impact on ozone, since the reduction in ARI results from changes in anthropogenic emissions. However, the phrasing in Lines 396-398 and abstract (specifically the use of “superimposed”) suggest that ΔO3_ΔARI_EMI is and additional, separate effect rather than being nested within the broader anthropogenic impact on ozone. Please clarify.
- In the Abstract, needs to explicitly clarify that the numbers presented are derived from different analysis. Lines 28-29 are for sparse polluted grids, while Lines 33-35 are for regional averages. Otherwise, readers may erroneously interpret the ratio between the numbers in Lines 33-35 and Lines 28-29 as the contribution of ARI to the total anthropogenic impacts.
Specific comments:
- Line 61, natural emissions are also an important precursor source. Please clarify.
- Section 3.2, it should be “Fig. 2” instead of “Figs. 2”. Similar typos are found in other places, e.g., Line 290, 302, 348. Please check.
- Line 293, delete “will”.
- Lines 310-312 and figure 4, please clarify in the figure caption that ARI_EMI can be obtained by summing the bars of API_EMI and ARF_EMI.
- Lines 353-354 and figure 5, the numbers mentioned in the text are inconsistent with those presented in the figure. Please correct.
- Figure 6, the first x-axis label should be “ARI” instead of “ALL”.
Citation: https://doi.org/10.5194/egusphere-2023-2393-RC3
Hao Yang et al.
Hao Yang et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 176 | 51 | 13 | 240 | 15 | 5 | 5 |
- HTML: 176
- PDF: 51
- XML: 13
- Total: 240
- Supplement: 15
- BibTeX: 5
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1