Subtle Changes or Dramatic Perceptions of Air Pollution in Sydney during COVID-19
Round 1
Reviewer 1 Report
This paper evaluates some publicly available data of air quality in Sydney before and during the lockdown orders due to the coronavirus pandemic starting in March 2020. There is also some discussion about internet searching behavior. I found the paper disorganized and marginally compelling. Data are presented in confusing ways and analysis of the data presented is superficial. The overall intention of the study needs to be made more clear.
Some specific suggestions and questions
L 13 delete period after quality
L 31-32 I would delete the first sentence; 3 names are used for the same virus; choose one and be consistent throughout paper (maybe use this space to say that the virus is known by several names and indicate which will be used)
L 46 this is a concluding sentence given before the data are shown for Sydney and does not explain what part of the world is being described.
L 49 What does “in almost poetic terms” mean in this context? Is this a criticism?
L 55 “Air pollution studies during the pandemic are not always easy to evaluate, as these have often . . .” This sentence makes it seem like hundreds of papers have been written over a long time rather than acknowledging that the pandemic is a unique situation that is ongoing.
L 65-66 How long did the lockdown last after mid-March?
L 72 the phrase about journalism is odd and does not sufficiently prepare the reader for the very different analysis of the second half of the paper. It should be explained more here.
L 79 Why is data shown only through July? A justification related to the scope of the lockdown orders needs to be explicitly indicated here.
L 83-84 There is no context for why these sites were chosen, what differentiates them, how similar they are or how different.
L 92 delete ”is”
L 119 missing a verb
L 123 add “of” before NO2; does “Across Sydney” mean that all sites are averaged together? Is this really the best way to show the data?
L 127-128 what are the EXACT dates you are defining as “lockdown”? Actually, concentrations during the lockdown period look HIGHER! The figure is not the best way to compare and the basis of comparison to make any claim of differences needs to be more explicitly given. Several options are given by the authors but it does not appear that a consistent basis is used throughout the paper.
L 130-131 “frequent” “often” “frequently; too many times authors refer to trends in coverage without citing any specific references.
L 137-139 A side-by-side comparison in a table showing these data would be better than this text statement.
L 149 the meaning of “this” is unclear
L 155 many times the authors state a trend that is supposed to be apparent on the graphs in Figure 2 without any evidence and then dispute the trend with statistics. seems to me you can’t have it both ways. The timeseries in Fig 2 are fine for description, but any mention of trends has to be made with statistics and perhaps listed in a table. And the details of how the statistical comparison is made (averaging time period, etc.) has to be explicitly described.
L 168 The figures are For NO2 only, right? This needs to be stated.
L 170-174 I really don’t understand the point of these sentences. Are the authors attributing changes to lockdown? Are they disputing the “The Conversation” data? What about impact from weather, as noted earlier? The dates noted don’t match with any dates in Fig. 3
L 191-192 The inset to Figure 3g DOES NOT show anything clearly. The stats given are more informative. I think the inset is confusing.
L 200 I don’t understand the intent or importance of the statement “. . . although many accounts do not examine these reductions in detail.”. Delete.
L 208 typo. clarify meaning.
L 241 I don’t understand the x axis of Fig. 4, especially for (d) that seems to be reversed.
L 246 The traces in 4b and 4c would be easier to compare if they were put on the same graph
L251 subject-verb disagreement
L 254-258 I’m not sure what the point of this section is.
L 312 I really don’t know what to make of Fig. 5; How does the discussion of the second part of the paper tie in to the first part? Is there a general, cohesive message here? Am I missing it?
L 360-361 I fail to see the connection or significance of comparing search behavior with actual changes in air quality. I was left without a clear idea if the authors are claiming there WAS a noticeable change in air quality that can be attributed (at least statistically) with the lockdown and that the connection with public behavior (just google searches) can add any insight to air quality improvement. If the authors’ primary goal is to address the influence of policy making, as implied in the last sentence: “Successful regulation of air quality requires maintaining a delicate balance between our social perceptions and physical reality.” This theme needs to be more clearly stated throughout the paper and referred to at least in each part of the paper, but I am left with the question if such a question can be answered without the rigor of a social scientist on the paper to put these arguments in theoretical context or perhaps in a journal relating to policy making.
Author Response
Comments and Suggestions for Authors
This paper evaluates some publicly available data of air quality in Sydney before and during the lockdown orders due to the coronavirus pandemic starting in March 2020. There is also some discussion about internet searching behavior. I found the paper disorganized and marginally compelling. Data are presented in confusing ways and analysis of the data presented is superficial. The overall intention of the study needs to be made more clear.
AGREED AND MADE MORE EMPHASIS IN THE INTRIDUCTION ON "It is additionally concerned with the relevance of social change and perceptions as portrayed on social media and within journalism "WE HAVE ALSO TRIED TO ADD EMPHASIS IN THE SECTIONS TO DRIVE HOME THE KEY ISSUES
Some specific suggestions and questions
L 13 delete period after quality
DONE
L 31-32 I would delete the first sentence; 3 names are used for the same virus; choose one and be consistent throughout paper (maybe use this space to say that the virus is known by several names and indicate which will be used)
YES BUT COVID-19 IS THE DISEASE AND coronavirus, SARS‑CoV‑2 ARE THE VIRUS, SO WE HAVE NOW MADE THIS CLEAR AND REMOVED THE VIRUS NAME AFTER ITS FIRST OCCRENCE
L 46 this is a concluding sentence given before the data are shown for Sydney and does not explain what part of the world is being described.
DONE
L 49 What does “in almost poetic terms” mean in this context? Is this a criticism?
YES CRITICAL SO WE HAVE ALTERED TO READ "welcomed with positive and sometimes poetic phrases "
L 55 “Air pollution studies during the pandemic are not always easy to evaluate, as these have often . . .” This sentence makes it seem like hundreds of papers have been written over a long time rather than acknowledging that the pandemic is a unique situation that is ongoing.
AGREE ALTHOUGH TO BE FAIR WE CHOSE THE WORD EARLY TO DESCRIBE THE RESEARCH. WE HAVE added "of an on-going crisis"
L 65-66 How long did the lockdown last after mid-March? GOOD POINT. IT IS DISCUSSED IN DETAIL LATER BUT WE HAVE ADDED "eased on 15 May"
L 72 the phrase about journalism is odd and does not sufficiently prepare the reader for the very different analysis of the second half of the paper. It should be explained more here. YES AGREED ALTERDED AS "It is additionally concerned with the relevance of social change and perceptions as portrayed on the in social media and within journalism and have become interested the enthusiasm with which air quality improvements have been greeted, even where changes have probably been small. "
L 79 Why is data shown only through July? A justification related to the scope of the lockdown orders needs to be explicitly indicated here.
A LITTLE MORE DETAIL ON LOCKDOWN ORDERS ARE GIVEN: "From 1 July, there was further easing, although it became rather more restrictive again on 17 July [3], because of an increase in cases, which meant for example, new rules for pubs being introduced." AND DETAIL IS AVAILABLE IN THE REFERENCE. WE HAVE ADDED AN EXPLANATION IN THE TEXT THAT THE FIRSTWAVE WAS OVER BY JULY SO IN THE RESULTS SECTION "March 2019 to July 2020, a point at which the first wave of infections was over,".
L 83-84 There is no context for why these sites were chosen, what differentiates them, how similar they are or how different.
THESE ARE MORE CENTALLY LOCATED IN SYDNEY AND NOW NOTED IN THE TEXT AS ""
L 92 delete ”is”
DONE
L 119 missing a verb
YES ADDED considered
L 123 add “of” before NO2; does “Across Sydney” mean that all sites are averaged together? Is this really the best way to show the data?
EMPHASISED THAT IT IS AVERAGED NOW. THERE ARE MANY WAYS TO DISPLAY THE DATA,. BUT GIVEN THAT WE LATER COMPARE THEN TO SATELLITE AVERAGESD IT SEEMED CONVENIENT.
L 127-128 what are the EXACT dates you are defining as “lockdown”? Actually, concentrations during the lockdown period look HIGHER! The figure is not the best way to compare and the basis of comparison to make any claim of differences needs to be more explicitly given. Several options are given by the authors but it does not appear that a consistent basis is used throughout the paper.
L 127-128 IS NOT MEANT TO DESCRIBE LOCKDOWN, BUT RATHER THE DATA IN THE FIGURE. OUR STATISTICAL COMPARISONS ARE FOR APRIL WHICH COVERS THE MOST STRINGENT PERIOD OF LOCKDOWN SO NOW ADD WORDS TO STRESS THAT ", during this period of stringent lockdown. "
L 130-131 “frequent” “often” “frequently; too many times authors refer to trends in coverage without citing any specific references.
AGREE TOO MANY! NOW READS "Air quality improvements under lockdown have often compared values with those of earlier years; frequently presented as satellite images [e.g. 6], which give a strong visual confirmation of change (block-of-four satellite images in Fig. 1)."
GOOD POINT ON THE NEED FOR A REFERENCE TO MEDIA. BUT IN THE END DECUIDED IT WAS BEST TO DELAY DISCUSSION OF JOURNALISM SO IT WAS MOVED LATER IN THE TEXT
L 137-139 A side-by-side comparison in a table showing these data would be better than this text statement.
TABLE ADDED ****
L 149 the meaning of “this” is unclear AGREE CHANGED TO "The CO data is assembled from the 11 monitoring sites is rather patchy, but "
L 155 many times the authors state a trend that is supposed to be apparent on the graphs in Figure 2 without any evidence and then dispute the trend with statistics. seems to me you can’t have it both ways. The timeseries in Fig 2 are fine for description, but any mention of trends has to be made with statistics and perhaps listed in a table. And the details of how the statistical comparison is made (averaging time period, etc.) has to be explicitly described.
WE DON'T REALLY MEAN TO SAY THAT TRENDS ARE APPARANT ON THE GRAPHS. WE WROTE :"Even with NO2, the most sensitive marker of change, the declines seen in ground level
measurements under lockdown may be subtle" HOWEVER, IT IS TRUE THAT THESE MAY BE DETECTED STATISTICALLY, HOWEVER WE ARGUES THIS ARE NOT ESPECIALLY CONVINCING.
THE STATISTICAL ANALYSIS IS FOR DAILY AVERAGES FOR APRIL SO WE HAVE TRIED TO ADD EMPHISIS TO THIS IN THE MODIFIED TEXT.. ADDED THE SITE AVERAGE IN THE METHOD IN MORE DETAIL TO THE TEXT WE HAVE ADDED AN TABLE AS AN APPENDIX TO CLARIFY THE STATISTCS
L 168 The figures are For NO2 only, right? This needs to be stated.
INDEED NO2 ONLY THIS IS NOW STRESSED ALONG FOR THE REASONS WHY WE USED ONLY NO2.
L 170-174 I really don’t understand the point of these sentences. Are the authors attributing changes to lockdown? Are they disputing the “The Conversation” data? What about impact from weather, as noted earlier? The dates noted don’t match with any dates in Fig. 3
SATELLITE DATA IS LIMITED IN ITS SPAN SO WE HAD NO CHOICE OVER THE PERIODS OF DATA AVAILABLEITY, BUT NOTE THAT IN THE MODIFIED TEXT ****.
WE HAVE SIMPLIFIED THE POINT MAYE BY THE CONVERSTAION AS: ". These were well-reported in the media e.g. The Conversation [39]. However these changes were not universal, and by contrast The Conversation noted that NO2 increased by 20% for Newcastle, with the country’s largest concentration of coal-burning heavy industry, by 40% for Melbourne, a sprawling city strongly dependent on cars, whereas Perth showed no real change. "
L 191-192 The inset to Figure 3g DOES NOT show anything clearly. The stats given are more informative. I think the inset is confusing.
DELETED
L 200 I don’t understand the intent or importance of the statement “. . . although many accounts do not examine these reductions in detail.”. Delete.
CHANGED TO "many accounts do not present detailed values for the reductions. "
L 208 typo. clarify meaning.
CHANGED TO "The tabulation found in Kumar et al [46] suggest that papers treating on emission reduction may take measurements of concentration or column density reduction as the indicator, rather than determine emissions independently. Many perhaps seeing reductions under lockdown as inevitable. "
L 241 I don’t understand the x axis of Fig. 4, especially for (d) that seems to be reversed.
BELIEVE THESE CORRECT BECAUSE ONLY THE FIRST HALF OF 2020 WAS AVAILABLE WHEN THE PAPER WAS WRITTEN. I HAVE ADDED A NOTE TO THE CAPTION
L 246 The traces in 4b and 4c would be easier to compare if they were put on the same graph
DONE
L251 subject-verb disagreement
YES GOOD POINT CHANGED
L 254-258 I’m not sure what the point of this section is.
UNDERSTAND SO HAVE JOINED THE TRANSPORT AND ENERGY PARAGRAPHS ADDED A SUMMARY SENTENCE "Overall the traffic patterns and their emissions changed in Sydney during lockdown, but energy generation shows low emissions though these are at great distance. As a public heath response, "
L 312 I really don’t know what to make of Fig. 5; How does the discussion of the second part of the paper tie in to the first part? Is there a general, cohesive message here? Am I missing it?
GOOD POINT WE FAILED TO INTRODUCE SECTION 3.5 , so have started the the introductory paragraph differently.
L 360-361 I fail to see the connection or significance of comparing search behavior with actual changes in air quality. I was left without a clear idea if the authors are claiming there WAS a noticeable change in air quality that can be attributed (at least statistically) with the lockdown and that the connection with public behavior (just google searches) can add any insight to air quality improvement. If the authors’ primary goal is to address the influence of policy making, as implied in the last sentence: “Successful regulation of air quality requires maintaining a delicate balance between our social perceptions and physical reality.” This theme needs to be more clearly stated throughout the paper and referred to at least in each part of the paper, but I am left with the question if such a question can be answered without the rigor of a social scientist on the paper to put these arguments in theoretical context or perhaps in a journal relating to policy making.
WE HAVE TRIED TO STRESS THIS MORE CAREFULLY AND WHILE WE HAVE SOCIAL SCIENCE COLLEAGUES THEY WERE BUSY WITH THEIR OWN PROJECTS.
TRUE WE EXAMINED ONLY GOOGLE TRENDS AND THE Barbieri DATA, BUT WE HAVE TRIED TO MAKE OUR ASSESSMENT CLEARER IN THE TEXT.
THE LAST SENTENCE IS MERE SPECULATION, BUT PERHAPS AS THE LAST SENTENCE OF A PAPER SPECULATION ABOUT THE POLICY IMPLIVATIONS OF OUR WORK CAN BE ALLOWED, ALTHOUG WEAKENED TO: "Successful regulation of air quality likely requires maintaining a delicate balance between our social perceptions and physical reality."
Reviewer 2 Report
This article is of scientific merit in the current climate and is well written. Overall, I do not think any significant changes should be made except for a rereading and focus on the English as there are some errors and awkward phrasing (examples provided below).
- line 13 - full stop after air quality should be changed to a comma
- line 23 - the smoke from bushfires (in) late 2019
- line 142 - as some 90% of the data is available... awkward phrasing
- line 227-229 - the discussion of activity levels is awkward
Open Review
(x) I would not like to sign my review report
( ) I would like to sign my review report
English language and style
( ) Extensive editing of English language and style required
( ) Moderate English changes required
(x) English language and style are fine/minor spell check required
( ) I don't feel qualified to judge about the English language and style
Yes Can be improved Must be improved Not applicable
Does the introduction provide sufficient background and include all relevant references?
(x) ( ) ( ) ( )
Is the research design appropriate?
(x) ( ) ( ) ( )
Are the methods adequately described?
(x) ( ) ( ) ( )
Are the results clearly presented?
(x) ( ) ( ) ( )
Are the conclusions supported by the results?
(x) ( ) ( ) ( )
Comments and Suggestions for Authors
This article is of scientific merit in the current climate and is well written. Overall, I do not think any significant changes should be made except for a rereading and focus on the English as there are some errors and awkward phrasing (examples provided below).
Author Response
line 13 - full stop after air quality should be changed to a comma DONE
line 23 - the smoke from bushfires (in) late 2019 DONE
line 142 - as some 90% of the data is available... awkward phrasing GOOD POINT CHANGED TO "As more of the data is available (~90%) the comparison is more reliable than for NO2."
line 227-229 - the discussion of activity levels is awkward AGREED CHANGED TO: "Australians spend: 25% of the time working and commuting, ~10% on recreational activities, while 15% at home (social media, TV etc), 30% sleeping, with domestic duties and recreation taking 20% [51]. "
Reviewer 3 Report
Dear Editor, Dear Authors,
the manuscript by Brimblecombe and Lai discusses and compares the scientific evidence and the public perception of the supposed reductions of air pollution in Australia during the COVID-related lockdown and more extended period of restrictions. This is a very interesting case study because of the record-breaking wildfires that ravaged Australia just before the COVID crisis, so producing a "high background" to which compare the supposed pollutant reductions due to COVID-related restricted mobility, social life and industrial activities.
I have truly enjoyed reading this manuscript; the Authors undoubtedly know how to interest the reader. This is not a neglectable point because writing an interesting paper is truly an art.
Nevertheless, the manuscript is not perfect, in my opinion and a few points have to be clarified before the manuscript is acceptable for publication in Environments. As I'm more expert of Atmospheric Sciences than Psychometry/Sociology, the core of my comments addresses Sections 3.1-3.3. One more substantial comment is that I would have really liked a slightly stronger connection of the discussions of Sections 3.1-3.3 and 3.4-3.5. The impact of the previous wildfires is discussed in Sections 3.4-3.5 (in terms of "perception" of AQ improvements) but probably not sufficiently in Sections 3.1-3.3 (in terms of "scientific evidence" of AQ improvements). Overall, I think my comments can be probably defined "a large group of Minor Comments", but I finally decided to give a Major Revision recommendation, so to give the Authors a bit more time to address these many small issues. Please find my specific comments in the following.
My best regards,
Specific comments:
Title: The Title is probably too "geographically" generic, condidering that the manuscript addresses one specific region of interest/case study (Australia). Please consider the possibility to mention "Australia" in the title.
L13 : ''…quality, especially…''
L32-33 : ''The coronavirus…shortly after'' : This calls for a reference, even if "only" from newspapers (but there should be scientific papers with precise information on the evolution of medical knowledge of coronavirus)
L33-34: This sentence also needs a literature reference from medical journals.
L55-56: The expression "often reflected" also calls for references. Is it something observed before?
L55-62: In all this section, a discussion of concurrent primary versus secondary pollutants is probably lacking (see Shy and Brasseur, 2020)
L67-67: ''…because of bushfires late in 2019 [38]'': There are now scientific papers available for Australian wildfires 2019-20 that can be cited ([38] is a newspaper article), e.g.: https://www.nature.com/articles/s43247-020-00022-5
Section 2 and more in general throughout the whole text: Please define all chemical formulae (e.g.: "...sulphur dioxide (SO2)...") at first occurrence and then use the formulae only. Same for acronyms, e.g. ''PM''
L85-87: Please state explicitly your criteria for using/not using a station.
Figure 1: OMI observations inset is definitely very small. Please rearrange Figure 1 to have slightly larger OMI maps. Also please put some reference values and units on the color bar.
L89: “…shows the…”
L90: “Aura/OMI”
L119-120: “The results presented…”, I don’t understand the sentence
L120-121: “Please define "social factors"”
Section 3.1: As it is well shown in Fig. 2, it would be interesting to briefly discuss also the impact of Australian Fires 2019-20 on these time series (increased values and variabilities of pollutants concentrations and isolated peaks, see other comments). This would allow a link to Sect. 3.4-5
Section 3.1: A recent important paper is lacking here: https://www.sciencedirect.com/science/article/pii/S0048969720339486. I suggest to cite and use it in the discussion of this section.
L127-128: ''…and with a smaller variance…”: Please consider that large variabilities of pollutants are usually observed at proximal locations downwind "natural emitting plumes" (volcanoes, fires), as described here: https://www.nature.com/articles/s41598-020-71635-1. To my eyes, the timeseries of Fig. 2 for October/November 2019-February/March 2020 are very clearly associated to exposure to the peak emissions from bushfires 2019-20, due to large values of pollutants and the mentioned variability.
L137-139: Maybe more details might be given for this (a figure in the Supplementary Material?)
L156-157: “Ozone, expected to increase…”: It actually depends on the chemical regime and then, also, e.g. VOCs. Please smooth and/or complete this statement.
Figure 2: I would slightly modify this figure:
1) please extend the "Lockdown" indication to all panels (e.g., it is hard to compare what shown in panel e with indication in panel a only);
2) would you have some metrics to indicate the period impacted by fires? (e.g., there is a clear peak on CO for the period of activation of the fire-induced pyroCb described in Khaykin et al., 2020)
L164: “Tropomi”: Please note this is (partly) an acronym, so should be spelled "TropOMI"
L166: please note that TropOMI spectral range allows the observations of a number of other species, so please change wording to account for this ("...of, e.g., O3, NO2, ...")
L166-167: “at a maximum 3.5 km x 5.5 km horizontal resolution”
L167: “…although typically…”: This is not typical. There are many TropOMI-based works where individual observations are used. Please correct.
L175: For consistency, if "Aura OMI" is used, "Sentinel 5P TropOMI" should be used
L177: “…mean of NO2 across…”: Please clarify that this is "surface concentration"
L178-180: “The inset…April 2020”: Please mention what's satellite and what's surface in situ. It might look redundant but, in my opinion, not (re)mentioning it may lead to misunderstandings.
L181-182: “…they are also integrated…large areas…”: This is not actually true: one can take individual satellite overpasses and match with ground-based observations with some ad-hoc matching criteria. What's true, instead, is that in situ and satellite observations have different spatio-temporal resolution (including vertical and horizontal spatial resolution). Please correct.
L181: “air column”: I suggest "atmospheric column"
L182-183: “Thus, they fail…”: this is due to spatial dilution and poor temporal sampling of satellite observations. Please clarify.
L190: “…a distinct seasonality…”: Which is NO2 distinct seasonality?
L196-197: “…but the way…impression of change.”: This conclusive sentence is very important but not very clear to me. Please clarify.
Section 3.3: A word missing in the title?
L208: “this so this”: Is there an error here?
Table 1: Having percent emissions directly in the table, for every type of source and every pollutant, would be a great add.
L274-275: This is probably the precise place where it can be discussed how the vertical and horizontal smoothing can give false perceptions
L275: “…further integrating over many days”: Again, this is not always the case.
L288-289: Isn't this a good place to make a cross-reference to Figure 2?
L293-295: “The tragedy…effects[61]”: Not clear to me. Please clarify.
L315 and 332: Which "extinction coefficient"? "Aerosol extinction coefficient"? With which instrument?
Conclusions: Also in the Conclusions, please use "NO2" or "nitrogen dioxide" consistently in the text. Same for other chemical species and acronyms.
L355: Again, how vertical columns and spatial smoothing change perception?
Author Response
Dear Editor, dear Authors,
the manuscript by Brimblecombe and Lai discusses and compares the scientific evidence and the public perception of the supposed reductions of air pollution in Australia during the COVID-related lockdown and more extended period of restrictions. This is a very interesting case study because of the record-breaking wildfires that ravaged Australia just before the COVID crisis, so producing a "high background" to which compare the supposed pollutant reductions due to COVID-related restricted mobility, social life and industrial activities.
I have truly enjoyed reading this manuscript; the Authors undoubtedly know how to interest the reader. This is not a neglectable point because writing an interesting paper is truly an art.
Nevertheless, the manuscript is not perfect, in my opinion and a few points have to be clarified before the manuscript is acceptable for publication in Environments. As I'm more expert of Atmospheric Sciences than Psychometry/Sociology, the core of my comments addresses Sections 3.1-3.3. One more substantial comment is that I would have really liked a slightly stronger connection of the discussions of Sections 3.1-3.3 and 3.4-3.5. The impact of the previous wildfires is discussed in Sections 3.4-3.5 (in terms of "perception" of AQ improvements) but probably not sufficiently in Sections 3.1-3.3 (in terms of "scientific evidence" of AQ improvements). Overall, I think my comments can be probably defined "a large group of Minor Comments", but I finally decided to give a Major Revision recommendation, so to give the Authors a bit more time to address these many small issues. Please find my specific comments in the following.
My best regards,
Specific comments:
Title: The Title is probably too "geographically" generic, condidering that the manuscript addresses one specific region of interest/case study (Australia). Please consider the possibility to mention "Australia" in the title.
YES... WE HAVE TRANSFERRED SYDNEY FROM THE KEYWORDS TO THE TITLE AND PUT AUSTRALIA IN KEYWORS
L13 : ''…quality, especially…''
DONE
L32-33 : ''The coronavirus…shortly after'' : This calls for a reference, even if "only" from newspapers (but there should be scientific papers with precise information on the evolution of medical knowledge of coronavirus)
GOOD POINT SO REFERENCE "Leung, K., Wu, J. T., Liu, D., & Leung, G. M. (2020). First-wave COVID-19 transmissibility and severity in China outside Hubei after control measures, and second-wave scenario planning: a modelling impact assessment. The Lancet."
L33-34: This sentence also needs a literature reference from medical journals.
AGAIN CITED THE EXCELLENT REFERERENCE FROM THE LANCET BY Leung, et al. (2020).
L55-56: The expression "often reflected" also calls for references. Is it something observed before?
YES SO REFERENCED THE CRITIQUE: Heederik, D. J.; Smit, L. A.; Vermeulen, R. C. Go slow to go fast: a plea for sustained scientific rigour in air pollution research during the COVID-19 pandemic. Eur Respir J. 2020 ,56: 2001361.
L55-62: In all this section, a discussion of concurrent primary versus secondary pollutants is probably lacking (see Shy and Brasseur, 2020)
AGRREE AND KNOW THE REFERENCE WELL, SO HAVE EXTENDED THAT AND ADDED SOME FURTHER TEXT HERE TO ELABORATE THE INTEREST IN SECONDARY POLLUTANTS AND THE MORE RECENT REFERENCE OF FAN et al "Peak air pollution concentrations might remain [32] and increased ozone levels that have accompanied the declines in nitrogen oxides (NOx) have been widely discussed [33-35]. Enhanced ozone has also offered the possibility that secondary aerosol concentrations would additionally increase, reminding us that emission and pollution is often linked by a complicated chemistry [5,35]. Some additionally suggest that emissions might increase as public transport is restricted [36]. "
L67-67: ''…because of bushfires late in 2019 [38]'': There are now scientific papers available for Australian wildfires 2019-20 that can be cited ([38] is a newspaper article), e.g.: https://www.nature.com/articles/s43247-020-00022-5
EXCELLENT SUGGESTION
Section 2 and more in general throughout the whole text: Please define all chemical formulae (e.g.: "...sulphur dioxide (SO2)...") at first occurrence and then use the formulae only. Same for acronyms, e.g. ''PM''
DONE
L85-87: Please state explicitly your criteria for using/not using a station.
WE USED ONLY THE STATIONS IN THE CENTRAL PARTS OF SYDNEY. THIS IS NOW MENTIONED IN THE TEXT
Figure 1: OMI observations inset is definitely very small. Please rearrange Figure 1 to have slightly larger OMI maps. Also please put some reference values and units on the color bar.
DONE
L89: “…shows the…”
DONE
L90: “Aura/OMI”
DONE
L119-120: “The results presented…”, I don’t understand the sentence
CHANGED TO "The results presented here consider concentration changes over from March 2019 to July 2020..."
L120-121: “Please define "social factors"”
AGREE - CHANGED TO "with perceptions and mobility changes "
Section 3.1: As it is well shown in Fig. 2, it would be interesting to briefly discuss also the impact of Australian Fires 2019-20 on these time series (increased values and variabilities of pollutants concentrations and isolated peaks, see other comments). This would allow a link to Sect. 3.4-5
YES GOOD POINT ADDED AS "The bushfires from November 2019 had a detrimental effect on air quality [43], with a set of increasingly higher and more frequent peaks in PM2.5 over Sydney through November, becoming almost continuous in December. Figure 2c suggests a small respite towards the end of December, but further increases over the last few days of the year, which are also seen in the record from TropOMI described by Khaykin et al [38]. They observed a peak in the aerosol absorbing index off the East Australian Coast (150°–155 E, 20°–40° S) on 22 December with a gap and then a further peak on 31 December, declining through the first week of January. The ground level measurements of Ulpiani et al [43] detail the changes in Sydney across the period 20 December 2019 until 13 January 2020, and show slight decreases in PM10 over the Christmas period, but higher concentrations into the New Year. While the spatial extent and averaging times between the ground level and satellite measurements are different, they show a satisfying agreement. " AND LINKED LATER AS "The dramatic changes must have made it easy to see lockdown as a time of cleaner air, a case well supported by satellite imagery and Fig. 2b, where we can see rapid changes between bad days and days that were less affected."
Section 3.1: A recent important paper is lacking here: https://www.sciencedirect.com/science/article/pii/S0048969720339486. I suggest to cite and use it in the discussion of this section.
THIS HAS NOW BEEN ADDED AS REFERENCE 13 AND REFERRED TO IN SECTION 3.1.
L127-128: ''…and with a smaller variance…”: Please consider that large variabilities of pollutants are usually observed at proximal locations downwind "natural emitting plumes" (volcanoes, fires), as described here: https://www.nature.com/articles/s41598-020-71635-1. To my eyes, the timeseries of Fig. 2 for October/November 2019-February/March 2020 are very clearly associated to exposure to the peak emissions from bushfires 2019-20, due to large values of pollutants and the mentioned variability.
THE VARIANCE ISSUE HAS TROUBLED US FOR SOME TIME AND WAS A PROBLEM IN AN EARLIER PAPER ON WUHAN. IT IS PROBABLY BEST DROPPED AT THIS POINT, SO WE DELETED THE NOTE.
L137-139: Maybe more details might be given for this (a figure in the Supplementary Material?)
WE HAVE REWRITTEN THIS A LITTLE AS THE LACK OF RELIABILITY HAS BEEN OVERSTATED AND OF COURSE WE CHECKED IN VARIOUS WAYS AS NOTED. "Comparisons between the two years may be unreliable because some 20% of the data are missing. However, we were reassured that an additional station-by-station analysis examining only those days where >15 hours of data is available, showed that the median daily values were lower in 2020, for only 9 of the 15 sites."
L156-157: “Ozone, expected to increase…”: It actually depends on the chemical regime and then, also, e.g. VOCs. Please smooth and/or complete this statement.
INDEED SO RECAST THE SENTENCE AS "Ozone, shows a slight, though barely significant (p1~.09) increase from 14.2±3.3 to15.4±3.5 ppb (Fig. 2e) as often observed because of decreasing NOx emissions under lockdown."
Figure 2: I would slightly modify this figure:
1) please extend the "Lockdown" indication to all panels (e.g., it is hard to compare what shown in panel e with indication in panel a only);
GOOD POINT NOW DONE
2) would you have some metrics to indicate the period impacted by fires? (e.g., there is a clear peak on CO for the period of activation of the fire-induced pyroCb described in Khaykin et al., 2020)
GOOD POINT NOW DONE
L164: “Tropomi”: Please note this is (partly) an acronym, so should be spelled "TropOMI"
GOOD POINT NOW DONE
L166: please note that TropOMI spectral range allows the observations of a number of other species, so please change wording to account for this ("...of, e.g., O3, NO2, ...")
GOOD POINT NOW DONE
L166-167: “at a maximum 3.5 km x 5.5 km horizontal resolution”
GOOD POINT NOW DONE
L167: “…although typically…”: This is not typical. There are many TropOMI-based works where individual observations are used. Please correct.
YES POINT OUT THAT THIS CAN BE USEFUL
L175: For consistency, if "Aura OMI" is used, "Sentinel 5P TropOMI" should be used
YES DONE
L177: “…mean of NO2 across…”: Please clarify that this is "surface concentration"
YES IMPORTANT CLARIFICATION
L178-180: “The inset…April 2020”: Please mention what's satellite and what's surface in situ. It might look redundant but, in my opinion, not (re)mentioning it may lead to misunderstandings. THE INSET WAS DELETED AT THE REQUEST OF ANOTHER REFEREE
L181-182: “…they are also integrated…large areas…”: This is not actually true: one can take individual satellite overpasses and match with ground-based observations with some ad-hoc matching criteria. What's true, instead, is that in situ and satellite observations have different spatio-temporal resolution (including vertical and horizontal spatial resolution). Please correct.
YES INDEED AND MADE THE POINT ABOUT THE RESOLUTION ISSUE.
L181: “air column”: I suggest "atmospheric column"
DONE
L182-183: “Thus, they fail…”: this is due to spatial dilution and poor temporal sampling of satellite observations. Please clarify.
CHANGED TO "Thus, they may not capture just how variable ground-level concentrations may be (Fig. 3f). "
L190: “…a distinct seasonality…”: Which is NO2 distinct seasonality?
ADDED AS: "show a distinct seasonality (i.e. higher in winter than summer). "
L196-197: “…but the way…impression of change.”: This conclusive sentence is very important but not very clear to me. Please clarify.
REWRITTEN "This may have arising from the saturated colours associated with published images, and perhaps also through the smooth integration of NO2 across time and space. These might enhance an impression of change."
Section 3.3: A word missing in the title?
YED COMMA "Emissions, human mobility and electricity production"
L208: “this so this”: Is there an error here?
INDEED SHOULD READ "years to generate, it represents "
Table 1: Having percent emissions directly in the table, for every type of source and every pollutant, would be a great add.
DONE
L274-275: This is probably the precise place where it can be discussed how the vertical and horizontal smoothing can give false perceptions
YES AND DONE AS "However, satellite images represent column densities, and in the case of media presentations during COVID-19 were smoothed over large areas across many days [e.g. 39]. This means although these gave a good sense of longer-term change in emissions, the published satellite imagery might not represent day-to-day changes in perceived visibility, or even less the hour-long ground level exposure to NO2 relevant to health concerns. Nevertheless, the vivid nature of the imagery often looks so convincing it may inhibit extended reflection on the subtly of the changes and the high degree of variation."
L275: “…further integrating over many days”: Again, this is not always the case.
GOOD POINT, SO WE HAVE MADE THIS SPECIFIC TO THE MEDIA PRESENTATIONS OF THE COVID-19 DATA
L288-289: Isn't this a good place to make a cross-reference to Figure 2?
ABSOLUTELY AND DONE!!
L293-295: “The tragedy…effects[61]”: Not clear to me. Please clarify.
QUITE RIGHT IT WAS VERY CONFUSED, SO CHANGED AS: "Although the impact of COVID-19 can be enhanced by pollution it has also been true, as often reported in the media, it has led to some improvements in the natural world, with more wild-flowers and animals, so there may be both positive and negative effects [61]. "
L315 and 332: Which "extinction coefficient"? "Aerosol extinction coefficient"? With which instrument?
IT IS DERIVED FROM VISUAL RANGE, SO NOW MENTIONED AT THIS POINT IN THE CAPTION. IT WOULD BE A BIT ODD TO GHIVE THE INSTRUMEMNTS AS THIS IS FROM THE MONITORING NETWORK, SO THEN WE WOULD HAVE TO DESCRIBE ALL THE INTRUMENTS USED BY THE EPA.
Conclusions: Also in the Conclusions, please use "NO2" or "nitrogen dioxide" consistently in the text. Same for other chemical species and acronyms.
DONE
L355: Again, how vertical columns and spatial smoothing change perception?
CHANGED AS "The extent of perceived improvement in air quality may have been emphasised by the widely distributed satellite imagery from around the globe. False-colours may have added drama to the change, especially given that satellite column densities have a spatio-temporal resolution that differs from ground level. "
Round 2
Reviewer 3 Report
Dear Authors,
I've really enjoyed reading your manuscript. Thank you for taking into account my comments. In my opinion, the manuscript is now ready for publication.
My best regards and happy 2021,
