the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Peculiar COVID-19 effects in the Greater Tokyo Area revealed by spatiotemporal variabilities of tropospheric gases and light-absorbing aerosols
- 1Center for Environmental Remote Sensing (CEReS), Chiba University, Chiba, 2638522, Japan
- 2Graduate School of Environmental Studies, Nagoya University, Nagoya, 4640064, Japan
- 3Department of Physics, Universidad de Santiago de Chile, Santiago, 3363, Chile
- 1Center for Environmental Remote Sensing (CEReS), Chiba University, Chiba, 2638522, Japan
- 2Graduate School of Environmental Studies, Nagoya University, Nagoya, 4640064, Japan
- 3Department of Physics, Universidad de Santiago de Chile, Santiago, 3363, Chile
Abstract. This study investigated the spatiotemporal variabilities in nitrogen dioxide (NO2), formaldehyde (HCHO), ozone (O3), and light-absorbing aerosols within the Greater Tokyo Area, Japan, the most populous metropolitan area in the world. The analysis was based on total column, partial column, and in situ observations retrieved from multiple platforms and additional information obtained from reanalysis and box model simulations. This study mainly covers the 2013–2020 period, focusing on 2020, when air quality was influenced by the coronavirus disease 2019 (COVID-19) pandemic. In 2020 overall, NO2 concentrations were reduced by about 10 % annually, with reductions exceeding 40 % in some areas during the pandemic state of emergency. Light-absorbing aerosol levels were also reduced for most of 2020, while smaller fluctuations in HCHO and O3 were observed. Moreover, the degree of weekly cycling of NO2, HCHO, and light-absorbing aerosol levels was significantly enhanced in urban areas during 2020. The latter changes were unprecedented in recent years and potentially related to coincident reduced mobility in Japan, which, in contrast to other countries, was anomalously low on weekends in 2020. This finding suggests that, despite the lack of strict legal restrictions in Japan, widespread adherence to recommendations designed to limit the spread of the pandemic caused modification of common habits, resulting in unique air quality changes.
- Preprint
(5224 KB) -
Supplement
(18748 KB) - BibTeX
- EndNote
Alessandro Damiani et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-110', Anonymous Referee #2, 04 May 2022
The paper by Damiani et al. is well structured and well written, with English of high quality. The paper has high-quality and informative figures. Combining different type of measurements for multiple species with model outputs and weather information provides a very complete record of changes in composition during lock-down, weekends and end-of-year holidays. I am in favour of publishing this paper after my major and minor comments have been addressed by the authors.
Major comment:
In general I am of the opinion that the list of references does not well reflect the detailed studies conducted to document the COVID-19 lockdown impact on air pollution levels in the past two years. The authors could add reviews on this topic, like Gkatzelis et al., https://doi.org/10.1525/elementa.2021.00176 and add some extra citations about the interaction between ozone, NOx and aerosol during the lockdowns. The authors remark that "many studies" on the relation COVID-19 and air quality have been conducted in the past two years, including results for the country of Japan. The authors should cite more extensively papers discussing the East-Asia region to provide the reader with a good overview on what is already published on COVID-19. Starting from this the authors should subsequently indicate what is new in the present work, and how this complements the earlier studies.
Minor remarks:
Abstract :
l15: “NO2 concentrations". It would be good to mention if this refers to surface, lower troposphere, column or all. Same for aerosol.
l18: Maybe better remove "in recent years", or do the authors mean that this happens both in 2021 and 2020?
Figure 1: The time axis (x-axis labels) in panel (a) is difficult to interpret: 2020.4 seems to coincide with the end of May. Would be useful to have 12 major ticks with months "Jan", "Feb" etc. For panel (b) could you please indicate that the period 7 April - 25 May was used. How is the 0% level determined?
l90: "In this study, we apply an integrated approach ..". See my general comment: why is this study unique, and what new result(s) are obtained?
l105: "FWHM = 0.4 nm at 357 and 476 nm". Why mention these two wavelengths instead of saying something like ""FWHM = 0.4 nm for this wavelength range". Is there a large change in FWHM as a function of wavelength?
l106: "wavelength calibration was performed daily to account for .. signal degradation" ? Do you mean "radiometric calibration" ?
l113: "relative humidity over water ". Why "over water"?
l115: "This procedure is expected to better account". Can this be tested, e.g. by comparing the four measurements?
ll119: "but sampled at higher accuracy". Please explain.
l127: "we used the NO2 and HCHO datasets". Please provide the processor versions of both datasets.
l128: "interpolated over a regular grid of 0.1 × 0.1°". Why was this done? One extra interpolation step will potentially degrade the comparison, adding extra representativity uncertainty.
l132: "Screening of TROPOMI NO2 data involved retaining data with a quality flag (QF) value higher than 0.5 and a cloud fraction (CF) lower than 0.2." The README file of TROPOMI suggests the removal of data with a quality value < 0.75. Why did the authors use a different filtering? Why is the cloud fraction limit different from OMI?
l187: I assume that wind, PBL height and temperature are also available in the CAMS reanalysis data record? Why do the authors use also MERRA? Does this have advantages over CAMS?
l196: Could you please explain what "transit stations" means. Is this bus and train only? Road traffic would be more relevant for emissions I guess. Does the transit station class scale well with the number of cars and trucks?
l206: "were estimated to be described" Please reformulate.
l215: I found this part very difficult to understand. At which altitude was the wind speed sampled? Is it the 10m wind, or PBL averaged wind, or something else? What is high, and what is low wind? "we computed NO2 as the difference between the composite values of days with high and low wind speed." Please explain the logic behind this. What does this difference represent? Is the difference plotted in Fig. 3, or the TROPOMI NO2 value itself?
l221. What are "communication routes"? Do you mean "transportation routes"?
l240: "does not align the urbanized region" -> does not align with the urbanized region
l242: "application of cloud screening ". The filtering of the data follows the TROPOMI readme file. Does this remark mean that an additional or reduced cloud screening was applied on top of the standard filtering? Please explain what was done. "somewhat different": what is the reference here?
l246: "sensitivity" please rephrase or explain.
l258: "recovered" A strange word for PBL ozone. "increased" would be better.
l270: "assimilate satellite observations of tropospheric NO2" CAMS is adjusting concentrations, which implies that the impact of the assimilation is expected to be relatively short, and a short range (12h or 1 day) forecast is expected to differ only slightly from a run without NO2 satellite data assimilation. What kind of CAMS product was used? Is it the analysis or the short range forecast? (may be good to mention this in 2.1.6)
l302: Figure 5c is a bit unclear. What are the steps between the red and black contours? Is it OMI (suggested by the caption) or TROPOMI (suggested by the text) based? It may be useful to introduce a separate figure for the 5-c panel.l411-419: The absence of an ozone weekend effect is indeed somewhat surprising. I was wondering if a more clear signal is found when only winter or summer months are selected? One expects more titration in winter, and more formation in summer.
Section 4 discussion: This section lists the main conclusions, but could be extended by listing shortcomings and with suggestions for future improvements and outlook on new datasets to be explored in the future (e.g. new satellite missions).
-
RC2: 'Comment on acp-2022-110', Anonymous Referee #1, 26 May 2022
The authors focus on the spatiotemporal variability of gases and light-absorbing aerosols in the Greater Tokyo Area during the COVID19 lockdown and the resulting changes in mobility. In general, I find this manuscript to be of interest for publication and appropriate for Atmospheric Chemistry and Physics. The manuscript is well written and would benefit from some additional details in the methods and discussion sections. Consequently, I can only recommend this paper for publication after major revisions.
Major comments
The content of the manuscript needs to be backed by more references. I have mentioned some instances under minor comments.
The authors also need to provide an overview of other studies which have investigated the impact of COVID19 lockdown. The authors should also compare their findings with existing literature (Cooper et al., 2022, Miyazaki et al., 2021).
Line 128 There may have been a version change in TROPOMI products in this time period. If yes, the authors should briefly mention this and discuss how did they go about it. Also, what do the authors mean by ‘interpolation’ here? Do they mean ‘oversampling’? How have the uncertainties been considered? Are the TROPOMI columns shown in Figure 3 and referred to in line 235 error-weighted averages?
Line 240-245 I would recommend the authors to also perform a sensitivity check if they considered median and error-weighted mean (if not already) of the TROPOMI HCHO columns and see if the interpretation of the results changes.
Can the authors use existing literature to comment on the relative contribution of biogenic and anthropogenic sources to HCHO and interpret the current findings of no apparent changes in TROPOMI HCHO?
Figure 5 It is difficult to read and interpret sub-figure (a) because of the image resolution. Also, the figure caption needs to describe that the NO2 mean is shown in grey. Can the authors consider an alternative way to present the data in this figure? For example, have the data as a table (supplementary material) and only show the top 10 or 20 cities in a figure (main text). The contours in sub-figure (c) are also complex to interpret as the color legend is only for population density. Please describe in the caption how the contours should be interpreted.
Minor comments
Lines 24-34 The introductory paragraph lacks references. More references for health effects of NO2 such as Achakulwisut et al. (2019), and for NO2 trends using satellite data such as Vohra et al. (2022).
Lines 35-40 Have the emissions of ozone precursors significantly decreased worldwide? I do not suspect the same in Asia and Africa. The authors should mention whether this refers to any city or a larger region and if both VOCs and NOx emissions are decreasing? It would be a good idea for authors to add details of the studies cited.
Line 91 Reference for HCHO as a proxy for VOCs.
Line 112 Which year are the MAX-DOAS measurements from? Refer to line 185.
Line 113 Any reference to support this?
Line 133 Any reference to support this? Why is the cloud fraction criteria different from OMI? Is it possible to assess the impact on the results if this cloud fraction threshold were to be changed to 0.3?
Line 177 OMI overpass time earlier is 13:40 LT and here it is 1:30pm. The authors should use consistent values and formats.
Line 212-216 Confusing. Please rephrase.
Line 240 Is there an increasing trend in CH4 which could be playing a role here?
Line 249-250 There needs to be some discussion around what the value of this ratio is. What is the transition regime you have been considering given this varies with both space and time (Duncan et al., 2010)? Add references too.
Line 275 Were the CAMS measurements for 2020 not available at the time of analysis? If they are available now, they should be included in the study.
Line 289 Earlier in line 145, the OMI data record is mentioned as 2005-2019.
Line 290-291 How many cities have been removed because of Friday being a rest day? The authors should list them either in the main text or in the supplementary for completeness. Also, if a lot of cities have been removed, the authors can consider the weekdays to be from Monday to Thursday and compare with Sunday for the weekend effect.
Line 305 What are the “two contour lines”?
Line 412 How does this look compared to findings from Miyazaki et al. (2021)?
Figure 3 The color scale for the last column should be reversed. Warm colors should indicate positive values and cool colors negative.
Figure 4 Caption text “Results are shown as percentage changes with respect to the 2013–2019 average (left and central panels) and 2019 (right panel).” is confusing. Please rephrase.
Figure 6 OMI total columns are referred to in the figure but tropospheric columns in the caption. Please correct as needed.
For data products (OMI/TROPOMI/MAX-DOAS, etc), please include URL stating when they were last accessed or point to references if data not publicly available, so that potential users can use these.
References
Achakulwisut et al., doi: 10.1016/S2542-5196(19)30046-4, 2019.
Cooper et al., doi:10.1038/s41586-021-04229-0, 2022.
Duncan et al., doi: 10.1016/j.atmosenv.2010.03.010, 2010.
Miyazaki et al., doi: 10.1126/sciadv.abf7460, 2021.
Vohra et al., doi:10.1126/sciadv.abm4435, 2022.
Alessandro Damiani et al.
Alessandro Damiani et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 262 | 45 | 8 | 315 | 22 | 3 | 3 |
- HTML: 262
- PDF: 45
- XML: 8
- Total: 315
- Supplement: 22
- BibTeX: 3
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1