Impacts of urbanization on air quality and related health risks in a city with complex terrain

Zhan, Chenchao; Xie, Min; Lu, Hua; Liu, Bojun; Wu, Zheng; Wang, Tijian; Zhuang, Bingliang; Li, Mengmeng; Li, Shu

doi:https://doi.org/10.5194/egusphere-2022-486

Preprints

https://doi.org/10.5194/egusphere-2022-486

Preprints

04 Jul 2022

| 04 Jul 2022

Impacts of urbanization on air quality and related health risks in a city with complex terrain

Chenchao Zhan, Min Xie, Hua Lu, Bojun Liu, Zheng Wu, Tijian Wang, Bingliang Zhuang, Mengmeng Li, and Shu Li

Abstract. Urbanization affects air pollutants by urban expansion and emission growth, and thereby inevitably changes health risks of air pollutants. However, the health risks related to urbanization are rarely estimated, especially for cities with complex terrain. In this study, a highly urbanized city with severe air pollution and complex terrain (Chengdu) is selected to explore this issue. The effects of urban expansion are further compared with emission growth as air quality management is mainly to regulate emissions. Air pollution in Chengdu is mainly caused by PM_2.5 and O₃. PM_2.5 pollution tends to appear in cold months (November to February) due to the secondary circulation forced by complex terrain and the frequent temperature inversion, while O₃ pollution is likely to occur in warm months (April to August) because of high temperature and strong sunlight dominated by high-pressure systems. From 2015 to 2021, the annual total premature mortalities from all non-accidental causes (ANAC) attributed to PM_2.5 and O₃ exposure are 9386 and 7743, respectively. Based on the characteristics of PM_2.5 and O₃, six numerical experiments are conducted to investigate the impacts of urban expansion and emission growth on health risks. The results show that urban expansion causes an increase in air temperature and the boundary layer height, which is conducive to the diffusion of PM_2.5. Thus, surface PM_2.5 concentrations decrease by 11.7 μg m^-3 in January. However, the MDA8 O₃ concentrations increase by 10.6 μg m^-3 in July due to the stronger photochemical production and better vertical mixing during the daytime. Correspondingly, the total premature mortalities from ANAC attributed to PM_2.5 exposure decrease by 182 (6.9 %) in January, but those attributed to O₃ exposure increase by 203 (9.5 %) in July. As for the effects of emission growth, PM_2.5 and MDA8 O₃ concentrations can increase by 26.6 μg m^-3 and 4.8 μg m^-3 when anthropogenic emissions are taken into account. The total premature mortalities from ANAC attributed to PM_2.5 and O₃ exposure then increase by 424 (16.0 %) and 87 (4.1 %), respectively. The effect of urban expansion on health risks of PM_2.5 is about half that of anthropogenic emissions. Whereas the effect of urban expansion on health risks of O₃ can be 2.3 times that of anthropogenic emissions. This reminds us that the development of cities is also important for the urban air quality apart from the emissions reduction.

Received: 13 Jun 2022 – Discussion started: 04 Jul 2022

Download & links

Preprint (PDF, 6457 KB)

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 (6457 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

17 Jan 2023

Impacts of urbanization on air quality and the related health risks in a city with complex terrain

Chenchao Zhan, Min Xie, Hua Lu, Bojun Liu, Zheng Wu, Tijian Wang, Bingliang Zhuang, Mengmeng Li, and Shu Li

Atmos. Chem. Phys., 23, 771–788, https://doi.org/10.5194/acp-23-771-2023,https://doi.org/10.5194/acp-23-771-2023, 2023

Short summary

Chenchao Zhan et al.

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2022-486', Anonymous Referee #1, 20 Jul 2022
The paper presents WRF-Chem simulations to analyze changes in PM2.5 and O3 concentrations due to urbanization effects in the Chinese city of Chengdu. Two months are simulated, specifically January for PM2.5 and July for O3, and sensitivity experiments are performed for each of the two simulations to evaluate the impact of land-use changes associated with urban expansion and the impact of anthropogenic emissions. The authors also estimate the changes in premature mortality associated with the changes in air pollution concentrations. The study is well designed to differentiate between the meteorological changes associated with urbanization caused by land-use changes and the changes in emissions and is, in my opinion, a nice example of how a modeling study can be used to disentangle different factors. The paper is also overall very well written.

I have, however, some concerns related to the lack of information about the boundary-layer depth and vertical structure despite its supposedly large impact on the results (general comment 1) and the seemingly strong overestimation of PM2.5 in the model (general comment 2).

General comments

The text refers repeatedly to the boundary layer height to explain, e.g., the temporal variations in PM2.5 (e.g., line 287, line 308), which makes of course perfect sense. These statements are, however, rather phrased as assumptions or hypotheses and the structure and development of the boundary layer is never discussed or even evaluated to make sure that the model actually captures this crucial factor correctly. In contrast to the authors’ statement (line 266), I would say that the model does not reproduce the diurnal cycle of PM2.5 well (Fig. 5). The correlation coefficient is not that high and the model also generally overestimates the observed values during most of the time. This could potentially be connected to a poor representation of the boundary layer.

I would recommend to start by comparing vertical profiles in the model with the radiosounding data to determine how well the model actually captures the vertical structure of the boundary layer. This is also commonly done before comparing the model output to surface observations to evaluate the larger scale first. This may even be helpful in better understanding the model performance for surface wind to see whether the large-scale flow is actually represented correctly.

Can it be shown that the temporal development of the boundary-layer depth and vertical structure can explain the modeled development in PM2.5 and O3 as discussed in the paper? Similarly, the text also hypothesizes about the impact of the urban heat island (line 339: “Urban land use can enhance surface heating leading to an increase in air temperature … resulting in and increase in the boundary layer height”). Can this actually be shown in the simulations?

Have the authors performed any sensitivity tests with different PBL parameterizations? They can have a large impact on the boundary-layer structure and depth, which could potentially have a large impact on the results.

As mentioned above, the model seems to overestimate PM2.5 relatively strongly, with a mean bias of 23.4 μg m^-3, which the authors consider to be small (line 264). This is almost the same magnitude as the monthly mean change related to anthropogenic emissions (26.6 μg m^-3), which the authors call significant (line 373), and twice the monthly mean change related to urban expansion. During individual days, the overestimation reaches even values of up to 200 μg m^-3(estimated from Fig. 5). I am somewhat concerned about the value of the results from the sensitivity tests considering the equally large model error and maybe that some processes affecting PM2.5 concentrations are not simulated correctly.

The paper reports very precise numbers for the changes caused by urbanization, e.g., the number of premature deaths resulting from PM2.5 (9386, 2609, 1321, and 1485 from ANAC, CVD, RD, and COPD) and the corresponding decreases of 424, 111, 55, and 56 without anthropogenic emissions. Considering that the equations given in Section 2.3 can likely provide only estimates and that I assume there must be large spatial variations in population density and air pollution throughout the city as well, which will affect the total premature deaths, I find the precision of these numbers somewhat misleading. Similarly, in the introduction a paper is cited by Liao et al. (2015) listing temperature and PM10 concentrations caused by urbanization with a resolution of 0.1ºC and 0.1 μg m^-3 while providing a range of more than 1ºC and more than 40 μg m^-3 and no information about the magnitude of the urban expansion.

I found it rather interesting that the model simulations for both PM2.5 and O3 show that very high concentrations can also be observed for large distances downstream of Chengdu. Assuming that there are also large populations in the vicinity of Chengdu, the overall impact would thus be even larger. I understand that an analysis of the effects around the city would be clearly beyond the scope of the paper, but I simply found this another interesting aspect.

Specific comments

Line 54: To which socioeconomic developments is the text referring?

Line 61: “the thermal contrast of the topography” is not entirely clear? Thermal circulations are caused by temperature gradients in the atmosphere.

Line 66: I would suggest to be more precise here and specify in which ways the diffusion conditions are more complicated.

Line 89: “Chengdu has the most complex terrain in the world”. This is a strong statement. How is that defined and determined?

Fig. 1: The figure shows SO2 emissions, but their role is never actually mentioned in the paper. There are also high-emission areas in the southeast corner of the figure. Are these other large cities?

Air quality and meteorological data: How many stations are there in Chengdu, i.e., the ones that are averaged for the analysis (line 129)? And how strongly do the locations differ?

Line 137: At which height is the lowest model level? Is the boundary layer also properly resolved during the night?

Line 175: The text says that β has units of (μg m^-3)^-1, which makes sense to have a non-dimensional exponent in eq. 5, but Table 3 mentions units of %.

Line 180 and Table 3: The references in the text and in the figure caption are not the same.

Lines 192 and 439: How are “pollution episodes” defined? Are these continuous periods, during which the air quality standard is exceeded, i.e., they can have different lengths? Does “pollution” always mean exceeding the standards (e.g., line 232) in this paper?

Line 213: What does “administrative division adjustment” mean?

Line 222: I am not sure how meaningful it is to calculate and discuss an increase or decrease between the first and last year if there is no real trend and the inter-annual variability is larger than the overall trend.

Table 4: Are these numbers calculated for individual days and then summed over the whole year? Section 2.3 does not say for which time period the given equations are applicable. It is mentioned that equation 5 is applied to MDA8 O3, but it does not say either over which time period the PM2.5 values have to be averaged.

Line 235: “cold westerly winds from the north” seems to be a contradiction. Are the winds from the west or from the north?

Line 239: Why do the humidity differences cause the inversion? Actually, it seems that the temperature decreases with height, i.e., there is no inversion, but only a stable layer.

Fig. 3: The 850-hPa level should be lower than the Tibetan Plateau. So what do the plotted fields over the Tibetan Plateau actually show? The areas above 850 hPa should best be masked in the plots. Also, are these monthly averages?

Fig. 4: (I) What do the black dots to the right of the subfigures (where one usually finds the wind arrows) mean? (ii) Are the profiles from the radiosoundings or also from the NCEP analysis as Fig. 3? Are they monthly means? (iii) Why are both 00 and 12 UTC shown if the differences are never discussed?

Line 269: A comparison of RH may not actually say much about the model performance in simulating humidity correctly because it combines the effects of humidity and temperature. It might be better to use, e.g., mixing ratio or specific humidity if the observations are available to convert the units.

Fig. 5

Are the measurements averages over all stations in the city as suggested in the previous section? If so is the model output equally averaged over all urban grid cells or over all grid cells closest to the stations?

The figure and figure caption do not say which of the shown datasets (black dots or colored lines) is the model and which the observations. Based on a statement in the text that the model underestimates RH, I assume that the observations are the dots.

Wind speed, RH, and temperature seem to have a resolution of only 1 m s^-1, 5 or 10%, and 1ºC, respectively, which makes it really difficult to compare the diurnal cycles because, e.g., wind speed remains basically constant for hours. This will also affect the quantitative comparison (correlation coefficient). Is this the actual resolution of the data?

Why is wind direction not shown?

Line 292: “mountain-valley breezes” – Where can they be seen in the figure? Chengdu seems to be located in a northeasterly flow directed towards the mountains all day and night. How does that agree with diurnal circulations?

Line 294: It is very difficult to see in the horizontal cross sections whether there are actually converging flows, because the figures show the flow at the first model level (I assume so because the figure caption does not say), which means very large elevation differences. The shown wind field is about 2-3 km higher in the leftmost part of the figure than in the right part. Based on the figures, one could also argue that the high concentrations are simply located at the foot of the mountains, i.e., they are trapped by the terrain.

Line 298: Is this really a secondary circulation that is forced by the terrain? Could it not be the upper-level westerly flow that is shown in Fig. 3? This would also agree with the absence of this “secondary circulation” in July, when a weak-gradient situation prevails at upper levels.

Technical comments, typos, etc.

Several of the figure axis labels are very small and thus almost impossible to read, e.g., axes labels in Fig. 2 (in particular the days of the week, but also the months), Fig. 4., Figs. 6-11.

For past events, the past tense should be used: e.g., “increased by 24.2%” (line 214). Similarly, lines 217, 232 and 438. There may be others.

Lines 25 and 30: It would be good to add that the mentioned changes in PM2.5 concentrations are based on monthly means.

Line 26: MDA8 is not defined

Line 55: “increases in anthropogenic emissions”

Line 60: The reference list contains only Qian et al. (2022).

Line 67: “notable pollution episodes”

Line 67: Chow et al. (2013) is not included in the reference list.

Line 71: I don’t understand the meaning of the sentence “Although the principles …”

Line 91: It would be really helpful to label these terrain features in Fig. 1a?

Fig. 1: It would be helpful to indicate the location of Chengdu in subfigure (b). I assume it corresponds to the red location in the center with high emission values?

Line 115: “Data and methods” might be a more common section title.

Line 200: “The high temperature and strong sunlight contribute to the elevated …”

Line 189: “two crucial air pollutants that account for air pollution”. This is a somewhat awkward sentence. “that account for air pollution” can easily be removed.

Line 237: “to the west of Chengdu”

Line 240: “cold air”

Line 314: “and mixing” (remove “well”)

Line 360: “can increase” suggests that this is a maximum value. Or is the text also referring to a monthly average as for PM2.5? It would actually help to always add “monthly mean” when talking about monthly mean values (e.g., line 404) to avoid misunderstandings.

Fig. 10: Is there a reason why the color scale is not adjusted to show the whole range? Based on the color bar it looks like the highest values are around 10 μg m^-3, whereas the text actually mentions a monthly average of 26.6 μg m^-3.

Line 407: “assess” instead of “access”

Line 545: This reference is not cited in the manuscript.
Citation: https://doi.org/10.5194/egusphere-2022-486-RC1
- AC1: 'Reply on RC1', Min Xie, 19 Oct 2022
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-486/egusphere-2022-486-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2022-486-AC1
RC2:
'Comment on egusphere-2022-486', Anonymous Referee #2, 05 Sep 2022

General comments:

Accelerated urbanization across the globe has resulted in considerable changes in land surface parameters and subsequently affect meteorological conditions and air pollutant levels. In this study, Zhan et al. reveal the changes in criteria air pollutants between 2015 and 2021 and probe the environmental consequence of rapid urbanization over a typical megacity situated in southwestern China, Chengdu. They also quantify the premature mortalities attributed to exposure to ozone and PM2.5. This paper is well-written and presents results that would be interesting to the air quality modeling community from a practical perspective. I have several concerns that the authors should consider when revising the manuscript, as listed below. I recommend publication after the following comments are adequately addressed.

Specific comments:

1) The literature review could be better. The authors only provide an example of urbanization impacts focused on the YRD by Liao et al. (2015), while extensive studies have been focusing on identifying the effects of urbanization on the regional meteorological phenomena and air quality (including the Beijing-Tianjin-Hebei area and the Pearl River Delta). Furthermore, recent studies have widely acknowledged the critical role of urbanization in altering air quality in Chengdu [Wang et al., 2021, 2022a]. Thus, these studies should be discussed and properly cited.

2) Line 34-36: This sentence is a bit vague and I genuinely don’t understand this sentence - please rephrase.

3) Figure 1: It seems that the shapefile used by the authors (NCL default shapefile) is wrong. Please check. Also, please clarify the source of SO2 emission for subplot (b). Moreover, it would be valuable to provide the boundary of Chengdu city in Figure 1 for making it clear.

4) Line 190: The criterion for PM2.5 and MDA8 ozone is a bit vague. I believe that it is annual PM2.5 concentrations less than 75μg/m3 and MDA8 ozone less than 160μg/m3. Please clarify the time period for these metrics.

5) Line 193-196: The authors apply the annual mean MDA8 ozone concentrations for illustrating the variations of ozone across Chengdu over time. However, annual mean MDA8 ozone is not a meaningful metric as wintertime low MDA8 ozone would pull low ozone levels. In general, it is recommended for using the warm season (April-September) MDA8 ozone average (see Wang et al., (2022b)) or the 90th percentile of MDA8 ozone (which is based on Chinese NAAQS GB3095-2012).

6) It would be better to use “heat maps of (a) daily average PM2.5 and (b) MDA8 O3 concentrations” rather than “distribution” for the caption of Figure 2.

7) Line 219-220: “the premature mortalities due to O3 fluctuate.” This is an incomplete sentence. Please check.

8) Line 221: “annual average” might be “7-year average”. Please check.

9) Line 275-278: Is the WRF-Chem model performance comparable with prior studies over Chengdu (or Sichuan Basin)? It would be valuable to briefly compare the model performance with previous studies (Yang et al., 2021; Wu et al., 2022) for demonstrating the robustness of model results.

10) Line 369-370: This sentence is a bit vague and I genuinely don’t understand this sentence - please rephrase.

11) Line 385-395: The authors attribute the ozone changes in Chengdu to the Ozone-NOx-VOCs regime but do not provide any details about the formation regime. A comprehensive discussion on the underlying mechanism of the VOCs-limited ozone regime in urban Chengdu is needed (Wang et al., 2022a).

References

[1] Wang, H., et al. (2021). Impact of different urban canopy models on air quality simulation in Chengdu, southwestern China. Atmospheric Environment, 267, 118775. https://doi.org/10.1016/j.atmosenv.2021.118775

[2] Wang, H., et al. (2022a). Impact of Urbanization on Meteorology and Air Quality in Chengdu, a Basin City of Southwestern China. Frontiers in Ecology and Evolution, 10, 845801. https://doi.org/10.3389/fevo.2022.845801

[3] Wang, Y., et al. (2022b). Long-term trends of ozone and precursors from 2013 to 2020 in a megacity (Chengdu), China: Evidence of changing emissions and chemistry. Atmospheric Research, 106309. https://doi.org/10.1016/j.atmosres.2022.106309

[4] Wu, K., et al. (2022). Drivers of 2013–2020 ozone trends in the Sichuan Basin, China: Impacts of meteorology and precursor emission changes. Environmental Pollution, 300, 118914. https://doi.org/10.1016/j.envpol.2022.118914

Citation: https://doi.org/10.5194/egusphere-2022-486-RC2
- AC2: 'Reply on RC2', Min Xie, 19 Oct 2022
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-486/egusphere-2022-486-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2022-486-AC2

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2022-486', Anonymous Referee #1, 20 Jul 2022
The paper presents WRF-Chem simulations to analyze changes in PM2.5 and O3 concentrations due to urbanization effects in the Chinese city of Chengdu. Two months are simulated, specifically January for PM2.5 and July for O3, and sensitivity experiments are performed for each of the two simulations to evaluate the impact of land-use changes associated with urban expansion and the impact of anthropogenic emissions. The authors also estimate the changes in premature mortality associated with the changes in air pollution concentrations. The study is well designed to differentiate between the meteorological changes associated with urbanization caused by land-use changes and the changes in emissions and is, in my opinion, a nice example of how a modeling study can be used to disentangle different factors. The paper is also overall very well written.

I have, however, some concerns related to the lack of information about the boundary-layer depth and vertical structure despite its supposedly large impact on the results (general comment 1) and the seemingly strong overestimation of PM2.5 in the model (general comment 2).

General comments

The text refers repeatedly to the boundary layer height to explain, e.g., the temporal variations in PM2.5 (e.g., line 287, line 308), which makes of course perfect sense. These statements are, however, rather phrased as assumptions or hypotheses and the structure and development of the boundary layer is never discussed or even evaluated to make sure that the model actually captures this crucial factor correctly. In contrast to the authors’ statement (line 266), I would say that the model does not reproduce the diurnal cycle of PM2.5 well (Fig. 5). The correlation coefficient is not that high and the model also generally overestimates the observed values during most of the time. This could potentially be connected to a poor representation of the boundary layer.

I would recommend to start by comparing vertical profiles in the model with the radiosounding data to determine how well the model actually captures the vertical structure of the boundary layer. This is also commonly done before comparing the model output to surface observations to evaluate the larger scale first. This may even be helpful in better understanding the model performance for surface wind to see whether the large-scale flow is actually represented correctly.

Can it be shown that the temporal development of the boundary-layer depth and vertical structure can explain the modeled development in PM2.5 and O3 as discussed in the paper? Similarly, the text also hypothesizes about the impact of the urban heat island (line 339: “Urban land use can enhance surface heating leading to an increase in air temperature … resulting in and increase in the boundary layer height”). Can this actually be shown in the simulations?

Have the authors performed any sensitivity tests with different PBL parameterizations? They can have a large impact on the boundary-layer structure and depth, which could potentially have a large impact on the results.

As mentioned above, the model seems to overestimate PM2.5 relatively strongly, with a mean bias of 23.4 μg m^-3, which the authors consider to be small (line 264). This is almost the same magnitude as the monthly mean change related to anthropogenic emissions (26.6 μg m^-3), which the authors call significant (line 373), and twice the monthly mean change related to urban expansion. During individual days, the overestimation reaches even values of up to 200 μg m^-3(estimated from Fig. 5). I am somewhat concerned about the value of the results from the sensitivity tests considering the equally large model error and maybe that some processes affecting PM2.5 concentrations are not simulated correctly.

The paper reports very precise numbers for the changes caused by urbanization, e.g., the number of premature deaths resulting from PM2.5 (9386, 2609, 1321, and 1485 from ANAC, CVD, RD, and COPD) and the corresponding decreases of 424, 111, 55, and 56 without anthropogenic emissions. Considering that the equations given in Section 2.3 can likely provide only estimates and that I assume there must be large spatial variations in population density and air pollution throughout the city as well, which will affect the total premature deaths, I find the precision of these numbers somewhat misleading. Similarly, in the introduction a paper is cited by Liao et al. (2015) listing temperature and PM10 concentrations caused by urbanization with a resolution of 0.1ºC and 0.1 μg m^-3 while providing a range of more than 1ºC and more than 40 μg m^-3 and no information about the magnitude of the urban expansion.

I found it rather interesting that the model simulations for both PM2.5 and O3 show that very high concentrations can also be observed for large distances downstream of Chengdu. Assuming that there are also large populations in the vicinity of Chengdu, the overall impact would thus be even larger. I understand that an analysis of the effects around the city would be clearly beyond the scope of the paper, but I simply found this another interesting aspect.

Specific comments

Line 54: To which socioeconomic developments is the text referring?

Line 61: “the thermal contrast of the topography” is not entirely clear? Thermal circulations are caused by temperature gradients in the atmosphere.

Line 66: I would suggest to be more precise here and specify in which ways the diffusion conditions are more complicated.

Line 89: “Chengdu has the most complex terrain in the world”. This is a strong statement. How is that defined and determined?

Fig. 1: The figure shows SO2 emissions, but their role is never actually mentioned in the paper. There are also high-emission areas in the southeast corner of the figure. Are these other large cities?

Air quality and meteorological data: How many stations are there in Chengdu, i.e., the ones that are averaged for the analysis (line 129)? And how strongly do the locations differ?

Line 137: At which height is the lowest model level? Is the boundary layer also properly resolved during the night?

Line 175: The text says that β has units of (μg m^-3)^-1, which makes sense to have a non-dimensional exponent in eq. 5, but Table 3 mentions units of %.

Line 180 and Table 3: The references in the text and in the figure caption are not the same.

Lines 192 and 439: How are “pollution episodes” defined? Are these continuous periods, during which the air quality standard is exceeded, i.e., they can have different lengths? Does “pollution” always mean exceeding the standards (e.g., line 232) in this paper?

Line 213: What does “administrative division adjustment” mean?

Line 222: I am not sure how meaningful it is to calculate and discuss an increase or decrease between the first and last year if there is no real trend and the inter-annual variability is larger than the overall trend.

Table 4: Are these numbers calculated for individual days and then summed over the whole year? Section 2.3 does not say for which time period the given equations are applicable. It is mentioned that equation 5 is applied to MDA8 O3, but it does not say either over which time period the PM2.5 values have to be averaged.

Line 235: “cold westerly winds from the north” seems to be a contradiction. Are the winds from the west or from the north?

Line 239: Why do the humidity differences cause the inversion? Actually, it seems that the temperature decreases with height, i.e., there is no inversion, but only a stable layer.

Fig. 3: The 850-hPa level should be lower than the Tibetan Plateau. So what do the plotted fields over the Tibetan Plateau actually show? The areas above 850 hPa should best be masked in the plots. Also, are these monthly averages?

Fig. 4: (I) What do the black dots to the right of the subfigures (where one usually finds the wind arrows) mean? (ii) Are the profiles from the radiosoundings or also from the NCEP analysis as Fig. 3? Are they monthly means? (iii) Why are both 00 and 12 UTC shown if the differences are never discussed?

Line 269: A comparison of RH may not actually say much about the model performance in simulating humidity correctly because it combines the effects of humidity and temperature. It might be better to use, e.g., mixing ratio or specific humidity if the observations are available to convert the units.

Fig. 5

Are the measurements averages over all stations in the city as suggested in the previous section? If so is the model output equally averaged over all urban grid cells or over all grid cells closest to the stations?

The figure and figure caption do not say which of the shown datasets (black dots or colored lines) is the model and which the observations. Based on a statement in the text that the model underestimates RH, I assume that the observations are the dots.

Wind speed, RH, and temperature seem to have a resolution of only 1 m s^-1, 5 or 10%, and 1ºC, respectively, which makes it really difficult to compare the diurnal cycles because, e.g., wind speed remains basically constant for hours. This will also affect the quantitative comparison (correlation coefficient). Is this the actual resolution of the data?

Why is wind direction not shown?

Line 292: “mountain-valley breezes” – Where can they be seen in the figure? Chengdu seems to be located in a northeasterly flow directed towards the mountains all day and night. How does that agree with diurnal circulations?

Line 294: It is very difficult to see in the horizontal cross sections whether there are actually converging flows, because the figures show the flow at the first model level (I assume so because the figure caption does not say), which means very large elevation differences. The shown wind field is about 2-3 km higher in the leftmost part of the figure than in the right part. Based on the figures, one could also argue that the high concentrations are simply located at the foot of the mountains, i.e., they are trapped by the terrain.

Line 298: Is this really a secondary circulation that is forced by the terrain? Could it not be the upper-level westerly flow that is shown in Fig. 3? This would also agree with the absence of this “secondary circulation” in July, when a weak-gradient situation prevails at upper levels.

Technical comments, typos, etc.

Several of the figure axis labels are very small and thus almost impossible to read, e.g., axes labels in Fig. 2 (in particular the days of the week, but also the months), Fig. 4., Figs. 6-11.

For past events, the past tense should be used: e.g., “increased by 24.2%” (line 214). Similarly, lines 217, 232 and 438. There may be others.

Lines 25 and 30: It would be good to add that the mentioned changes in PM2.5 concentrations are based on monthly means.

Line 26: MDA8 is not defined

Line 55: “increases in anthropogenic emissions”

Line 60: The reference list contains only Qian et al. (2022).

Line 67: “notable pollution episodes”

Line 67: Chow et al. (2013) is not included in the reference list.

Line 71: I don’t understand the meaning of the sentence “Although the principles …”

Line 91: It would be really helpful to label these terrain features in Fig. 1a?

Fig. 1: It would be helpful to indicate the location of Chengdu in subfigure (b). I assume it corresponds to the red location in the center with high emission values?

Line 115: “Data and methods” might be a more common section title.

Line 200: “The high temperature and strong sunlight contribute to the elevated …”

Line 189: “two crucial air pollutants that account for air pollution”. This is a somewhat awkward sentence. “that account for air pollution” can easily be removed.

Line 237: “to the west of Chengdu”

Line 240: “cold air”

Line 314: “and mixing” (remove “well”)

Line 360: “can increase” suggests that this is a maximum value. Or is the text also referring to a monthly average as for PM2.5? It would actually help to always add “monthly mean” when talking about monthly mean values (e.g., line 404) to avoid misunderstandings.

Fig. 10: Is there a reason why the color scale is not adjusted to show the whole range? Based on the color bar it looks like the highest values are around 10 μg m^-3, whereas the text actually mentions a monthly average of 26.6 μg m^-3.

Line 407: “assess” instead of “access”

Line 545: This reference is not cited in the manuscript.
Citation: https://doi.org/10.5194/egusphere-2022-486-RC1
- AC1: 'Reply on RC1', Min Xie, 19 Oct 2022
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-486/egusphere-2022-486-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2022-486-AC1
RC2:
'Comment on egusphere-2022-486', Anonymous Referee #2, 05 Sep 2022

General comments:

Accelerated urbanization across the globe has resulted in considerable changes in land surface parameters and subsequently affect meteorological conditions and air pollutant levels. In this study, Zhan et al. reveal the changes in criteria air pollutants between 2015 and 2021 and probe the environmental consequence of rapid urbanization over a typical megacity situated in southwestern China, Chengdu. They also quantify the premature mortalities attributed to exposure to ozone and PM2.5. This paper is well-written and presents results that would be interesting to the air quality modeling community from a practical perspective. I have several concerns that the authors should consider when revising the manuscript, as listed below. I recommend publication after the following comments are adequately addressed.

Specific comments:

1) The literature review could be better. The authors only provide an example of urbanization impacts focused on the YRD by Liao et al. (2015), while extensive studies have been focusing on identifying the effects of urbanization on the regional meteorological phenomena and air quality (including the Beijing-Tianjin-Hebei area and the Pearl River Delta). Furthermore, recent studies have widely acknowledged the critical role of urbanization in altering air quality in Chengdu [Wang et al., 2021, 2022a]. Thus, these studies should be discussed and properly cited.

2) Line 34-36: This sentence is a bit vague and I genuinely don’t understand this sentence - please rephrase.

3) Figure 1: It seems that the shapefile used by the authors (NCL default shapefile) is wrong. Please check. Also, please clarify the source of SO2 emission for subplot (b). Moreover, it would be valuable to provide the boundary of Chengdu city in Figure 1 for making it clear.

4) Line 190: The criterion for PM2.5 and MDA8 ozone is a bit vague. I believe that it is annual PM2.5 concentrations less than 75μg/m3 and MDA8 ozone less than 160μg/m3. Please clarify the time period for these metrics.

5) Line 193-196: The authors apply the annual mean MDA8 ozone concentrations for illustrating the variations of ozone across Chengdu over time. However, annual mean MDA8 ozone is not a meaningful metric as wintertime low MDA8 ozone would pull low ozone levels. In general, it is recommended for using the warm season (April-September) MDA8 ozone average (see Wang et al., (2022b)) or the 90th percentile of MDA8 ozone (which is based on Chinese NAAQS GB3095-2012).

6) It would be better to use “heat maps of (a) daily average PM2.5 and (b) MDA8 O3 concentrations” rather than “distribution” for the caption of Figure 2.

7) Line 219-220: “the premature mortalities due to O3 fluctuate.” This is an incomplete sentence. Please check.

8) Line 221: “annual average” might be “7-year average”. Please check.

9) Line 275-278: Is the WRF-Chem model performance comparable with prior studies over Chengdu (or Sichuan Basin)? It would be valuable to briefly compare the model performance with previous studies (Yang et al., 2021; Wu et al., 2022) for demonstrating the robustness of model results.

10) Line 369-370: This sentence is a bit vague and I genuinely don’t understand this sentence - please rephrase.

11) Line 385-395: The authors attribute the ozone changes in Chengdu to the Ozone-NOx-VOCs regime but do not provide any details about the formation regime. A comprehensive discussion on the underlying mechanism of the VOCs-limited ozone regime in urban Chengdu is needed (Wang et al., 2022a).

References

[1] Wang, H., et al. (2021). Impact of different urban canopy models on air quality simulation in Chengdu, southwestern China. Atmospheric Environment, 267, 118775. https://doi.org/10.1016/j.atmosenv.2021.118775

[2] Wang, H., et al. (2022a). Impact of Urbanization on Meteorology and Air Quality in Chengdu, a Basin City of Southwestern China. Frontiers in Ecology and Evolution, 10, 845801. https://doi.org/10.3389/fevo.2022.845801

[3] Wang, Y., et al. (2022b). Long-term trends of ozone and precursors from 2013 to 2020 in a megacity (Chengdu), China: Evidence of changing emissions and chemistry. Atmospheric Research, 106309. https://doi.org/10.1016/j.atmosres.2022.106309

[4] Wu, K., et al. (2022). Drivers of 2013–2020 ozone trends in the Sichuan Basin, China: Impacts of meteorology and precursor emission changes. Environmental Pollution, 300, 118914. https://doi.org/10.1016/j.envpol.2022.118914

Citation: https://doi.org/10.5194/egusphere-2022-486-RC2
- AC2: 'Reply on RC2', Min Xie, 19 Oct 2022
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-486/egusphere-2022-486-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2022-486-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Min Xie on behalf of the Authors (19 Oct 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (27 Oct 2022) by Thomas Karl

RR by Anonymous Referee #1 (08 Nov 2022)

ED: Publish subject to minor revisions (review by editor) (16 Nov 2022) by Thomas Karl

AR by Min Xie on behalf of the Authors (22 Nov 2022) Author's response Author's tracked changes Manuscript

ED: Publish as is (30 Nov 2022) by Thomas Karl

AR by Min Xie on behalf of the Authors (07 Dec 2022) Manuscript

Journal article(s) based on this preprint

17 Jan 2023

Impacts of urbanization on air quality and the related health risks in a city with complex terrain

Chenchao Zhan, Min Xie, Hua Lu, Bojun Liu, Zheng Wu, Tijian Wang, Bingliang Zhuang, Mengmeng Li, and Shu Li

Atmos. Chem. Phys., 23, 771–788, https://doi.org/10.5194/acp-23-771-2023,https://doi.org/10.5194/acp-23-771-2023, 2023

Short summary

Chenchao Zhan et al.

Viewed

Total article views: 535 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
404	118	13	535	2	3

HTML: 404
PDF: 118
XML: 13
Total: 535
BibTeX: 2
EndNote: 3

Views and downloads (calculated since 04 Jul 2022)

Month	HTML	PDF	XML	Total
Jul 2022	135	39	6	180
Aug 2022	51	6	0	57
Sep 2022	29	10	3	42
Oct 2022	58	20	4	82
Nov 2022	61	13	0	74
Dec 2022	46	24	0	70
Jan 2023	24	6	0	30
Feb 2023	0
Mar 2023	0
Apr 2023	0
May 2023	0
Jun 2023	0
Jul 2023	0
Aug 2023	0
Sep 2023	0
Oct 2023	0

Cumulative views and downloads (calculated since 04 Jul 2022)

Month	HTML	PDF	XML	Total
Jul 2022	135	39	6	180
Aug 2022	51	6	0	57
Sep 2022	29	10	3	42
Oct 2022	58	20	4	82
Nov 2022	61	13	0	74
Dec 2022	46	24	0	70
Jan 2023	24	6	0	30
Feb 2023	0
Mar 2023	0
Apr 2023	0
May 2023	0
Jun 2023	0
Jul 2023	0
Aug 2023	0
Sep 2023	0
Oct 2023	0

Viewed (geographical distribution)

Total article views: 460 (including HTML, PDF, and XML) Thereof 460 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 30 Oct 2023

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (6457 KB)
Metadata XML

Short summary

With the development in urbanization, urban land use and anthropogenic emissions increase, which deteriorates urban air quality and then health risks of air pollutants. In this study, we systematically evaluate the impacts of urbanization on air quality and the corresponding health risks in a highly urbanized city with severe air pollution and complex terrain. This work fills a gap in the health risks caused by urbanization and can provide valuable insight for air pollution strategies.


Total:	0
HTML:	0
PDF:	0
XML:	0