Volatile Organic Compound Composition of Urban Air in Nairobi, Kenya and Lagos, Nigeria

Round 1

Reviewer 1 Report

I recommend this paper for publication. Despite the small number of samples collected and limitations regarding temporal and spatial coverage, given the sparcity of measurements taken in this part of the world, this work is timely and relevant.

There are however some important changes that should be addressed in order for it to be made suitable for publication, as well as several details that must be included:

The abstract does not convey the main findings or conclusions of the report (there is no mention of OFP, the exposure calculations or that sources were investigated based on VOC correlation). The introduction is too narrow in scope, focussing almost entirely on BTEX. As a result the achievements that arise from this work are somewhat diluted.  (I do not think it is difficult to convince any environmental scientist that BTEX measurements are important!) But a paragraph or two on the significance of other VOC measurements (their toxicity, their atmospheric reactivity, their utility, etc.) would be most welcome  - as well as the benefits of GCxGC over other analytical methods that are more commonly deployed for studies such as these.

Some important details are missing from the sampling and analytical methods alongside other minor corrections – please see my comments below.

Abstract:

“Not only are several of the VOCs toxic, they will also have indirect effects on air quality 27 by increasing the incidence and severity of photochemical episodes, thus causing further detrimental health effects.”

Consider removing or rephrasing as at present it's not clear whether these areas are in a NOx or VOC limited regime.  Please see my final comment.

Include the number of VOCs quanitified.

LINE 82: repetitive, please review.

LINE 143: please clarify, were the samples collected twice a day over the two week period? Include the total number of samples for clarity / consistency.

Line 147: I am not sure 'commerical environment' will be well understand and needs defining. Consider rephrasing throughout. 

LINE 162: Did the authors run any tests to check the stability of the tubes in transit. If so, any details of the results would be most welcome.  Were the tubes chilled in transit? Also, were the tubes stored before analysis and so everything was analysed in one batch or were multiple batches analysed.  Can the authors provide insight into how the storage and shipping may affect stability on the tubes and the presence of interferents? Perhaps from the field blanks?

LINE 165: Please include the solvent.

LINE 172: Briefly describe the standard (e.g., number of compounds, including alkanes, aldhydes etc. In 95:5 water: methanol - Sigma Aldrich are not necessarily going to sell this product forever).  Also, quote the ISO to which this CRM adheres. Please provide details of how the concentrations were calculated? Using the FID?

LINE 182: this is a little confusing.  Were the two reference samples the alkane and the aromatic calibration standards? If so, this should be re-phrased as it is implied that a single reference solution was made by combining the two commercial standards in methanol.  If not, are you referring to the reference solution and the calibration mix?

If it's the former, when was the calibration solution used (every 10 samples or at the start or end of each sequence)?

LINE 240: repetitive, please review.

LINE 270: The compound classes you label here are not the same as the groupings later on. Consider rephrasing.  Most of them are single compounds.  Are the aromatics within each box all structural isomers? (the term 'aromatics' is a little vague). 'Aromatic hydrocarbons' is pretty unambiguous in this case.

Do you know what the broad peak is at around 57 min (/1.5 min) and the tailing peak at 60 min (2.7 min)?

LINE 275: 75 compounds is a large number.  Which is great...but I guess more than were included in the calibration mix.  'Definitively identified' is a strong statement....how was this done (on RT or by match hit)? Please explain how the extra VOCs were quantified (e.g., by closest eluting proxy or by response factor)?

I count 74 compounds listed in table 2.

LINE 288: please clarify, on a single day or on different days?

LINE 448 (and paragraph starting line 371): It is difficult to assess the influence of secondary pollutant formation on the (averaged) proportion of compounds analysed, or the correlation between compounds, because the sampling timeframe remains unclear.  It is therefore difficult to compare the two cities without as it is not known whether the sampling regime was well-balanced between them.  Were samples taken in both the morning and evening in Nairobi? What about in Lagos?  (Or were they taken throughout the day in the latter)?

A typical photochemical episode would lead to elevated oVOCs by mid-afternoon.  Averaging two samples in this case may not make sense. A quick comparison of the concentrations obtained from the two sampling times could be made to see if a significant difference exists between them (relative to e.g., alkanes).

LINE 460: Low correlation coefficients may reflect values below the LOQ. (e.g., pyrene – can you please comment on this?  Otherwise, this would be an interesting result).

Figure 5: There are 62 compounds included in the heatmap, how were the compounds selected? And please describe the statistical method.

LINE 602. Please take care when comparing OFPs from studies which measure different compounds – highlight these differences in as much detail as you can (see comment below) if you intend to draw conclusions from comparative OFPs. Alternatively, could you re-calculate the total OFP based on the same compounds (if you are able to determine this from Van Langenhove et al., 2013).

LINE 615.  You mention formaldehyde etc., but it reads a little odd to do this and to not also point out that very few unsaturated aliphatic compounds were measured (including compounds which have some of the highest IRs, such as the butenes and pentenes).   

LINE 649.  No need to discuss POCPs if you have already introduced OFPs.

LINE 748: Which commonly measured pollutants are you referring to (BTEX are commonly measured and are of course VOCs)? If you mean NOx, Ozone, SO₂ and PM, name them explicitly.

LINE 759: Your point here is essentially that VOCs can be very useful in determining sources of pollution (I agree with that, as can the chemical composition of PM) and with this you should also mention their role in determining the photo-chemical age of air masses.

But in terms of air quality policy, what I think is missing in this report is mention of how more VOC measurements are needed, alongside measurements of radicals, in order to determine which species limit ozone production and therefore where future mitigation strategies must be focussed.

Author Response

Thank you for your detailed and constructive comments on our manuscript, please attached our response to each of the points you raised.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments to the manuscript atmosphere-1387816: Volatile organic compound composition of urban air in Nairobi, Kenya and Lagos, Nigeria

The present manuscript deals with VOC emissions in Nairobi and Lagos at different sites.  The analytical data underline the statements given in the text. The content fits to the Journal. English is fine. Hence, the manuscript should be published after taking the following aspects into considerations and adding references where necessary.

• Line 79-82: It might be helpful to compare the emissions of vehicles (perhaps differentiated between diesel and gasoline) with other typical emission sources like solid waste combustion, biogas combustion…. Maybe this comparison can be added as supplementary materials or concise as additional subchapter.
• Line 84: Is there a difference in the VOC spectrum of diesel engines and gasoline engines?
• In Figure 1 with exception of acetone and benzene only lipophilic compounds are present. Are there no moderate lipophilic compounds like cyclohexane, hexane, heptane, MTBE or similar? If it is so, please explain their disappearance. In line 337-338 the presence of oxygenates is mentioned. In Line 355-357 the answer is given – quite late. Hence, a note should be given directly to the Figure to explain the spectrum as it is.
• Line 356-357: It is mentioned that highly volatile low-chain compounds are also present in the air as previously described as typical emissions of engines. It should be mentioned that highly volatile VOCs are also formed during incineration / electrification of biogas. A very helpful literature reference which should be added is:
  • DOI:1016/j.jclepro.2019.04.258

Suggestion: “…lighter than pentane, so their detection was not possible in this study by the used analytics, but their presence in substantial quantities in roadside air as well as combustive emissions of industrial sites has been shown previously (Huang, Ling et al. 2015; Dobslaw et al. 2019).

• Figure 5b is incomplete due to page change.
• Line 664-674: It should be pointed out whether benzene is originally present in the gasoline and set free by incomplete ignition or formed by chemical reactions during incomplete ignition. In first case, what is the range of concentrations in the gasoline. In second case, incomplete ignition is the main source and might also explain differences during European studies and the present one. In that way average age of cars used in these cities should be estimated and compared to European values.

Author Response

Thank you for your constructive comments on our manuscript, please find attached our response to each of the points you raised.

Author Response File: Author Response.docx