From the journal Environmental Science: Atmospheres Peer review history

24-Feb-2023

Dear Dr Borlaza:

Manuscript ID: EA-ART-01-2023-000013
TITLE: Impact of COVID-19 lockdown on particulate matter oxidative potential at urban background versus traffic sites

Thank you for your submission to Environmental Science: Atmospheres, published by the Royal Society of Chemistry. I sent your manuscript to reviewers and I have now received their reports which are copied below.

I have carefully evaluated your manuscript and the reviewers’ reports, and the reports indicate that major revisions are necessary.

Please submit a revised manuscript which addresses all of the reviewers’ comments. Further peer review of your revised manuscript may be needed. When you submit your revised manuscript please include a point by point response to the reviewers’ comments and highlight the changes you have made. Full details of the files you need to submit are listed at the end of this email.

Please submit your revised manuscript as soon as possible using this link:

*** PLEASE NOTE: This is a two-step process. After clicking on the link, you will be directed to a webpage to confirm. ***

https://mc.manuscriptcentral.com/esatmos?link_removed

(This link goes straight to your account, without the need to log on to the system. For your account security you should not share this link with others.)

Alternatively, you can login to your account (https://mc.manuscriptcentral.com/esatmos) where you will need your case-sensitive USER ID and password.

You should submit your revised manuscript as soon as possible; please note you will receive a series of automatic reminders. If your revisions will take a significant length of time, please contact me. If I do not hear from you, I may withdraw your manuscript from consideration and you will have to resubmit. Any resubmission will receive a new submission date.

The Royal Society of Chemistry requires all submitting authors to provide their ORCID iD when they submit a revised manuscript. This is quick and easy to do as part of the revised manuscript submission process. We will publish this information with the article, and you may choose to have your ORCID record updated automatically with details of the publication.

Please also encourage your co-authors to sign up for their own ORCID account and associate it with their account on our manuscript submission system. For further information see: https://www.rsc.org/journals-books-databases/journal-authors-reviewers/processes-policies/#attribution-id

Environmental Science: Atmospheres strongly encourages authors of research articles to include an ‘Author contributions’ section in their manuscript, for publication in the final article. This should appear immediately above the ‘Conflict of interest’ and ‘Acknowledgement’ sections. I strongly recommend you use CRediT (the Contributor Roles Taxonomy, https://credit.niso.org/) for standardised contribution descriptions. All authors should have agreed to their individual contributions ahead of submission and these should accurately reflect contributions to the work. Please refer to our general author guidelines https://www.rsc.org/journals-books-databases/author-and-reviewer-hub/authors-information/responsibilities/ for more information.

I look forward to receiving your revised manuscript.

Yours sincerely,
Dr Nønne Prisle
Associate Editor, Environmental Sciences: Atmospheres

************

Reviewer 1

Comments
The manuscript titled ‘Impact of COVID-19 lockdown on particulate matter oxidative potential at urban background versus traffic sites’ is an interesting piece of work. The authors have presented the results of OP of PM assessed in 2 locations. ML technique was utilized to predict business as usual scenario for OP, PM and BC. I feel the authors have done a good work, however the language of the paper needs revision. I have raised several questions in the work. Please find specific comments below:
Specific comments:
• LINE NO: 110 - Any details of error percentage considered in OP quantification? Please mention
• LINE NO: 145 - With this limited set of the data points, how the model was run?
• LINE NO: 214 - For RF model, why 1-year lag of target variables was chosen, any specific reason?
• LINE NO: 215 – For ML, what parameters besides importance score were considered to determine the model performance
• LINE NO: 230 - Why the test R2 and RMSE did not show considerable improvements compared to training set in all the plots (S3, S17) is only the sample size an issue, or any other parameter influencing the results? Please explain
• Many places there is formatting error ‘Error ! Reference source not found’ please check

Reviewer 2

General comment
The paper reports an analysis of OP, measured with two acellular assays, in two sites to investigate potential effects of the COVID-19 lockdown on this parameter intended as a global indicator of potential health effects of atmospheric particles. The topic is interesting and relatively few papers have been published, up to now, on the effect of lockdown on OP values. It is suitable for the Journal and generally well written. However, there are a few aspects that are not completely clear and some details of the interpretation that should be revised (see my specific comments) before publication.

Specific comments
There are several instances in my pdf with “Error! Reference not found” that should be corrected.

Page 3. Better to say only “…both combustion sources on the OP levels in ambient air.” Like in the abstract.

Lines 25-32. I would also mention that some studies focalised on ultrafine particles and on size distributions mentioning the recent work of Conte et al. (Atmospheric Environment 295 (2023) 119559). This work, also done in an urban background site, showed a similar effect of biomass burning increase that limited to reduction of pollution levels due to the lockdown.

Line 36. Better even in background sites.

It should be better to clarify if the GRE site is urban or urban background because there are contrasting things written in different parts of the paper. According to the discussion it seems to be not very influenced by traffic so likely it is an urban background.

Lines 44-53. I would suggest to mention the recent work of Guascito et al. (Journal of Hazardous Materials 448 (2023) 130872) that provides a quite extensive overview of comparison of acellular OP (with DTT assay) and in vitro toxicological effects.

Line 110 better was rather than is.

Section 2.2 is quite limited in the information regarding methodological approaches and it would be better to improve it. For example, it should be stated if the OP analysis at the two sites were done following the same protocols in terms of extraction from filters, chemical reactions, concentrations of reagents and so on. Tris-HCl was used as in the protocol of Cho et al (2005)? Another aspect that is not clear to me is why absorbance during DTT assay was measured every 10 minutes (i.e. 4 points in 30 minutes) and those in AA assay every 4 minutes (likely with a total time of 32 minutes)? Is there a reason why not using the same number of points? Eventually this information could be provided in the supplementary.

Looking at the summarised results in Table 1. I would suggest to use only one threshold for p (i.e. 0.05). In addition, it seems that both OPDTT in Bern has a different behaviour looking at PM10 and PM2.5. An increase in PM2.5 and a decrease in PM10. This does not happen for OPAA. This aspect should be interpreted. In addition, most of the differences observed for OP are not statistically significant, especially in comparison with predicted values. I would suggest to clearly state in the conclusions that, actually, the results for OP are variable for the two sites and it does not seems that the effect of lockdown is clearly discernible. The sentences on lines 307-311; Lines 329-332; the first and second points of the summary (section 4) should reflect this not clear results for OP and that the mentioned changes are seen only in one size fraction.

Line 350. The increase of biomass burning does not seem to happen at both sites.

The third point of the summary section is likely an over-interpretation of results. I do not see that there is a clear conclusion, from the data presented, that prediction with RF models produce more realistic results. This point should be made mainly as a possibility rather than a certainty.

Impact of COVID-19 lockdown on particulate matter oxidative potential at
urban background versus traffic sites

Authors’ response

We would like to thank the referees for their time to evaluate our manuscript and for their positive and constructive feedbacks, which helped us improve the quality of the paper. Our point-by-point responses to the comments are presented below (in blue) and changes have been made accordingly in the revised version of the manuscript (provided along with the present response document).

Reviewer #1:
The manuscript titled ‘Impact of COVID-19 lockdown on particulate matter oxidative potential at urban background versus traffic sites’ is an interesting piece of work. The authors have presented the results of OP of PM assessed in 2 locations. ML technique was utilized to predict business as usual scenario for OP, PM and BC. I feel the authors have done a good work, however the language of the paper needs revision. I have raised several questions in the work. Please find specific comments below:

Line 110: Any details of error percentage considered in OP quantification? Please
Mention
Response: Thank you for this comment. In Section 2.2, it was mentioned that a positive control test is performed in every experiment using a 1,4-naphthoquinone (1,4-NQ) solution for both the DTT and the AA assays. All positive control tests yielded a <3% coefficient of variation (%) for both assays. To look into the uncertainty levels, all samples were also analysed in triplicate with a coefficient of variation ranging between 0 and 10% for each assay.

Line 145: With this limited set of the data points, how the model was run?
Response: Thank you for this clarification. Table S1 summarizes the number of data points available for each air quality parameter. In fact, there is a good number of data points in each air quality parameter (even up to 6885 for PM2.5 in the BERN site). The lowest number of data points can be found in OP in the PM10 (n=273) and PM2.5 (n=177) fraction in the BERN site. Generally, more data points can further increase the accuracy of a model. For this study, a hyper-parameter tuning step was used to make sure the best model performance was reached with the available dataset. We have found that good model performances were reached based on the model performance metrics reporting the root mean square error (RMSE) and R-squared value (r2) in the training and test sets, as presented in Figures S4 to S17.

Line 214: For RF model, why 1-year lag of target variables was chosen, any specific
reason?
Response: Generally, the features used in random forest models are known explanatory variables. The use of a 1-year lag of the target variable attempts to represent levels of the target variable in the past when no lockdown restrictions were in place. Figures S1 to S3 shows the importance of each feature in the RF model. Each tree of the random forest calculates the importance of a feature according to its ability to increase the accuracy of the model prediction.

Line 215: For ML, what parameters besides importance score were considered to
determine the model performance
Response: In section 2.3, it was mentioned that the model performance was evaluated by calculating the root mean square error (RMSE) and R-squared value (r2). The optimal models were chosen based on their RMSE (as low as possible) and r2 (as high as possible) on the testing set.

Line 230: Why the test R2 and RMSE did not show considerable improvements compared
to training set in all the plots (S3, S17) is only the sample size an issue, or any other parameter
influencing the results? Please explain
Response: The dataset was split into training and testing sets by randomly assigning 70% of the input data as training set and the rest of the 30% as testing. As the RF model is trained on the training set, it is expected that the model performance is higher in the training set. The number of data points could be the main issue in OP for the BERN site. However, we cannot rule out any other factor possibly influencing the target variable that has not been included in the list of features used in the model and could well lead to better results.

Many places there is formatting error ‘Error ! Reference source not found’ please check
Response: Thank you for this comment. These errors were corrected in the revised manuscript.

Reviewer #2:
The paper reports an analysis of OP, measured with two acellular assays, in two sites to investigate potential effects of the COVID-19 lockdown on this parameter intended as a global indicator of potential health effects of atmospheric particles. The topic is interesting and relatively few papers have been published, up to now, on the effect of lockdown on OP values. It is suitable for the Journal and generally well written. However, there are a few aspects that are not completely clear and some details of the interpretation that should be revised (see my specific comments) before publication.

There are several instances in my pdf with “Error! Reference not found” that should be corrected.
Response: Thank you for this comment. These were errors were corrected in the revised manuscript.

Page 3. Better to say only “…both combustion sources on the OP levels in ambient air.” Like in the abstract.
Response: We appreciate this feedback. However, we would like to clarify that the combustion sources discussed here were estimated from black carbon measurements.

Lines 25-32. I would also mention that some studies focalised on ultrafine particles and on size distributions mentioning the recent work of Conte et al. (Atmospheric Environment 295 (2023) 119559). This work, also done in an urban background site, showed a similar effect of biomass burning increase that limited to reduction of pollution levels due to the lockdown.
Response: Thank you for this suggestion. It has been included in the revised manuscript.

Line 36. Better even in background sites.
Response: Thank you for this suggestion. It has been updated in the revised manuscript.

It should be better to clarify if the GRE site is urban or urban background because there are contrasting things written in different parts of the paper. According to the discussion it seems to be not very influenced by traffic so likely it is an urban background.
Response: Line 15 mentions that the GRE site is indeed an urban background site. This is also specifically mentioned in the title of the article.

Lines 44-53. I would suggest to mention the recent work of Guascito et al. (Journal of Hazardous Materials 448 (2023) 130872) that provides a quite extensive overview of comparison of acellular OP (with DTT assay) and in vitro toxicological effects.
Response: Thank you for this suggestion. This reference has been added in the list of work related to OP as an emerging health-based metric of PM exposure.

Line 110 better was rather than is.
Response: Thank you for this suggestion. It has been updated in the revised manuscript.

Section 2.2 is quite limited in the information regarding methodological approaches and it would be better to improve it. For example, it should be stated if the OP analysis at the two sites were done following the same protocols in terms of extraction from filters, chemical reactions, concentrations of reagents and so on. Tris-HCl was used as in the protocol of Cho et al (2005)? Another aspect that is not clear to me is why absorbance during DTT assay was measured every 10 minutes (i.e. 4 points in 30 minutes) and those in AA assay every 4 minutes (likely with a total time of 32 minutes)? Is there a reason why not using the same number of points? Eventually this information could be provided in the supplementary.
Response: Thank you for this suggestion. We have now mentioned that the protocol performed was applied on samples collected on both sites.
Apologies for the confusion. We wanted to cite Cho et al. (2005) on the interaction representing the consumption of DTT as the capacity of PM to generate ROS. Our protocol is based on Calas et al. (2018). This has been corrected in the revised manuscript.
For AA assay, no titration is needed, making it possible to measure absorbance on a smaller time resolution (every minute is also possible). For DTT assay, we do need to titrate, and a 10-minute time step is necessary. We have added some specifics of the protocol for clarity, as suggested by the reviewer.

Looking at the summarised results in Table 1. I would suggest to use only one threshold for p (i.e. 0.05). In addition, it seems that both OPDTT in Bern has a different behaviour looking at PM10 and PM2.5. An increase in PM2.5 and a decrease in PM10. This does not happen for OPAA. This aspect should be interpreted. In addition, most of the differences observed for OP are not statistically significant, especially in comparison with predicted values. I would suggest to clearly state in the conclusions that, actually, the results for OP are variable for the two sites and it does not seems that the effect of lockdown is clearly discernible. The sentences on lines 307-311; Lines 329-332; the first and second points of the summary (section 4) should reflect this not clear results for OP and that the mentioned changes are seen only in one size fraction.
Response: Thank you for the feedback. The threshold for p-values have been set at p≤0.05 (*) and p≤0.01 (**). This has been clarified in the table caption. We deem useful to indicate both significance levels. In fact, the differences in median percentage change during the lockdown period are discussed in detail in section 3.3. Furthermore, Figure 4 and Figure 5 elaborates on the distribution of the data to dig deeper on the actual differences across the datasets (historical, observed, and predicted). It was also mentioned that in an urban background site, changes in the PM mass concentration levels and a decrease in OP (assessed by AA assay) were found to be less pronounced. This decrease was not found in the OPDTT likely due to the sustained contributions from wood burning sources in the urban background site during the lockdown period. The changes are not only seen in one size fraction as presented in Table 1 and Figure 4.

Line 350. The increase of biomass burning does not seem to happen at both sites.
Response: We agree that biomass burning contributions (estimated by BCwb) did not increase in both sites. This take-away was specifically mentioned to be associated with “urban background site during the lockdown period”.

The third point of the summary section is likely an over-interpretation of results. I do not see that there is a clear conclusion, from the data presented, that prediction with RF models produce more realistic results. This point should be made mainly as a possibility rather than a certainty.
Response: We appreciate this comment. We have updated this point that now reads as:

“The RF modelling technique provided a good estimate of a BAU scenario, allowing for a more representative assessment of the changes of pollution levels during the lockdown period.”

27-Mar-2023

Dear Dr Borlaza:

Manuscript ID: EA-ART-01-2023-000013.R1
TITLE: Impact of COVID-19 lockdown on particulate matter oxidative potential at urban background versus traffic sites

Thank you for submitting your revised manuscript to Environmental Science: Atmospheres. I am pleased to accept your manuscript for publication in its current form. I have copied any final comments from the reviewer(s) below.

You will shortly receive a separate email from us requesting you to submit a licence to publish for your article, so that we can proceed with the preparation and publication of your manuscript.

You can highlight your article and the work of your group on the back cover of Environmental Science: Atmospheres. If you are interested in this opportunity please contact the editorial office for more information.

Promote your research, accelerate its impact – find out more about our article promotion services here: https://rsc.li/promoteyourresearch.

We will publicise your paper on our Twitter account @EnvSciRSC – to aid our publicity of your work please fill out this form: https://form.jotform.com/211263048265047

How was your experience with us? Let us know your feedback by completing our short 5 minute survey: https://www.smartsurvey.co.uk/s/RSC-author-satisfaction-energyenvironment/

By publishing your article in Environmental Science: Atmospheres, you are supporting the Royal Society of Chemistry to help the chemical science community make the world a better place.

With best wishes,

Prof. Nønne Prisle
Associate Editor, Environmental Sciences: Atmospheres

Reviewer 1

Though the authors have tried to substantiate the difference in R2 and RMSE between test and training set of data. I feel the difference is so large in many cases and cant be acceptable.

Reviewer 2

Authosr revised the paper and improved it sufficiently to be accepted for publication.