the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Feb 2024
Biomass burning sources control ambient particulate matter but traffic and industrial sources control VOCs and secondary pollutant formation during extreme pollution events in Delhi
Abstract. Volatile organic compounds (VOCs) and particulate matter (PM) are major constituents of smog. Delhi experiences severe smog during post-monsoon season, but a quantitative understanding of VOCs and PM sources is still lacking. Here, we source-apportioned VOCs and PM, using a high-quality recent (2022) dataset of 111 VOCs, PM2.5, and PM10 using positive matrix factorization. Contrasts between clean-monsoon and polluted-post-monsoon air, VOC source fingerprints, molecular-tracers, enabled differentiating paddy-residue burning from other biomass-burning sources, which has hitherto been impossible. Fresh paddy-residue burning and residential heating & waste-burning contributed the highest to observed PM10 (25 % & 23 %), PM2.5 (23 % & 24 %), followed by heavy-duty CNG-vehicles 15 % PM10 and 11 % PM2.5. For ambient VOCs, ozone, and SOA formation potentials, top sources were petrol-4-wheelers (20 %, 25 %, 30 %), petrol-2-wheelers (14 %, 12 %, 20 %), mixed-industrial emissions (12 %, 14 %, 15 %), solid fuel-based cooking (10 %, 10 %, 8 %) and road construction (8 %, 6 %, 9 %). Emission inventories tended to overestimate residential-biofuel emission (>2) relative to the PMF output. The major source of PM pollution was regional biomass burning, whereas traffic and industries governed VOC and secondary pollutant formation. Our novel source-apportionment method quantitatively resolved even similar biomass and fossil-fuel sources, offering insights into both VOC and PM sources affecting extreme-pollution events. It represents a notable advancement over current source apportionment approaches, and would be of great relevance for future studies in other polluted cities/regions of the world with complex source mixtures.
- Preprint
(2695 KB) - Metadata XML
-
Supplement
(2424 KB) - BibTeX
- EndNote
Status: open (extended)
-
RC1: 'Comment on egusphere-2024-501', Anonymous Referee #1, 21 Mar 2024
reply
General comments
The manuscript presents a positive matrix factorization analysis of a consequent PTR-ToF-MS dataset, to which PM2.5 and PM10 data were added. The general outline, scope and main conclusions are very clear. The results are interesting, and each of the 11 obtained factors is thoroughly described, backed up with external data and source profiles, and well explained.
However, I do feel that the methodology is not described enough. There should be more details on how the uncertainties were calculated, what were the uncertainties for each compound and their range. Also, more information is needed on the PMF approach of adding PM2.5 and PM10 data to the VOC dataset, what were the steps leading to the solution (how many runs, how was the base case chosen if these runs gave different solutions, …), were there any challenges with this approach? In the abstract you mention “our novel source apportionment method”, but it is not very clear in the paper how novel or different it is.
Also, you mention that the factors are stable in the bootstrap repetitions; however, the uncertainties of the model in Figure 3 seem quite important. Also, the contribution of factors (i.e. paddy, residential) for PM2.5 and PM10 changes a lot when the number of factors varies, suggesting they may not be very stable. Do you have other information to back up the factors’ stability (i.e., low timeseries correlations between the factors)? How do the scaled residuals change when increasing the number of factors or between different runs?The comparison of the PMF output with emission inventories results needs more justification. If I understand correctly, PMF results are concentrations and seem to be directly compared to emissions, which are different quantities and on different scales. The conclusions drawn here seem too strong (i.e. lines 536-539). Also, please justify why the PMF results are more correct than model outputs? (i.e. when you state that sources are under-/over-estimated in the models)
Be more concise when you present the description of the factors, the fact that all the values and VOC m/z are written in the main text makes it tedious to read. Use only VOC names (or formula if unclear what the compound is, but the m/z are already all listed in Table S1). Also, delete all the concentration and % values in the main text if they are already on the figures, except if it is useful to emphasize the point (example in line 632: “a considerable portion of the PM10 (18%) and PM2.5 (28%)”). Same for log10C0, find a clearer way to present them. Another option would be to put the extensive description of factors in SI and a summary and interpretation in the main text.
Specific comments/questions
Line 87-88: I would suggest adding a map of the receptor site with the surroundings (i.e. roads, industries, agriculture…), and referencing it when needed.
Line 108: I think it would be worth summarizing the main differences between the 3 wind sectors (in terms of typology, specificity, and later on results).
Section 3.1 & Figure 3: I would suggest putting Figure 3 in supplementary and replacing it with only this study’s factors profiles in concentration (instead of normalized). In text 3.1, I would add the R correlation (of profile and/or diurnal cycle) of this study’s factors with the mentioned reference factors to justify the factors’ interpretation.
Sections 3.1 & 3.2: Since you have a dedicated subsection for the comparison of the sources with references, you don’t have to repeat them when describing each factor.
Figure 3: How were the displayed compounds chosen for this graph? And please use the compounds’ names so that it is clearer.
I would suggest adding Figure S3 in the main text as it is referenced a lot, and that way you don’t need to put the % in the main text.
Line 224: “The source identity of the PMF factors was confirmed by matching the normalized PMF factor profiles with normalized source fingerprints”. Could you add more detail about this, did you check the R correlations? Or was it just by visually comparing them?
Line 235-236: Did you measure the Munirka furniture market and Dhobighat at Akshar Dham samples? If not, could you add their reference?
Figure 5:
I would suggest enlarging (by the x axis) the timeseries plot, to make them easier to read. You should keep the same order of the factors as in description (& throughout the paper). What do the lines/shaded areas for the diurnal cycles represent (mean, median…)?3.2.2. There is a mention that this factor may not be always fresh, which I found interesting, you could add a few words at the end of the paragraph about the fresh/aged nature of the factors based on all the information.
3.2.3. Some of these compounds (i.e. aromatics) can also be associated with cooking activities (e.g. Crippa et al (2013), doi.org/10.5194/acp-13-8411-2013).
3.2.4. You could add one sentence about the interpretation of VOCs (i.e. methanol and ethanol) for this factor.
3.2.5. & 3.2.6. Add a sentence (or change existing text) to highlight the differences between 2-wheeler & 4-wheeler factors.
Line 422: Interesting! Could this last sentence mean that part of PM2.5 for this factor would be SOA?
Line 432-433: Do you have references for this last statement?
3.2.9. Interesting, the last sentence suggests a possible link of the OVOCs with SOA?
3.3. It’s a little tedious to read with all the emission values, please select when it is truly important to have them.
Lines 500-505: A map could be useful here as well.
Line 536-539: “our PMF results indicate that the actual emissions are slightly smaller than those” “our PMF estimates fall in between those of the EDGARv6.1 inventory and the REASv3.2.1 inventory” I don’t understand how you come to these conclusions, did you calculate emissions out of the PMF concentrations? If yes, please state. If not, I don’t think you can directly compare PMF results and emissions, only in terms of contributions to the total “measured” compounds for each method.
Line 551: “The EDGARv6.1 inventory significantly underestimates PM2.5 & PM10 from agricultural activities” Please backup this statement with a map for example to justify that agricultural emissions should be high.
Line 554-556: There were any more results available from FINNv2.5? “between 15th and August and 26th November 2021 alone” please clarify, was it 15/08-26/11? Then it’s the same length as the current dataset…
Table S1: You could add calculated uncertainties and detection limits here. Also, if the “Sr. No” numbers are not used, you can delete them from the table. Are the “Mean” and range values here the detection limits or the averaged concentrations throughout the campaign?
Table S2 & S3: Same comment about the “Sr. No”.
Figure S1: Are these figures referenced in the paper?
Technical corrections
Throughout the paper, add · in units (ex µg·m-3)
Title: There shouldn’t be an abbreviation in the title, please use volatile organic compounds instead of VOC.
Line 16: There is a repetition of the word “using”, please change.
Line 23: Replace “(<2)” by “at least by a factor of 2”.
Line 36: Please reformulate “continues to add”.
Line 70: Delete the first “source” in “quantify the source contribution of the different sources”.
Line 80: Delete “:” in the title and check all the titles.
Line 113: “in blue” aren’t there other colours used on the graph too?
Line 116: Correct to “solar radiation as photosynthetically active radiation (PAR)”.
Line 119: Please add the dates of monsoon and post-monsoon seasons.
Line 151-152: The structure of the sentence seems wrong, please correct.
Line 180-181: There is a repetition of the word “model”, please change.
Line 190: “T” to delete at the beginning of the paragraph.
Line 180-181: There is a repetition of the word “using”, please change.
Line 194: Change to “The secondary organic aerosol production (SOAP)” in small case.
Line 196 & 197: Replace NOx with NOX and check this throughout the paper.
Line 197-199: This sentence is a bit unclear.
Line 219: Replace “while” by starting a new sentence with “In addition,”.
Line 220: Replace “are” with “were” and check that it is the right tense throughout the paper.
Line 247: Delete “,” in “(Fig. 4 a & d) were petrol”.
Figure 4: “Photo”, “P2W” & “P4W” could be written in full name.
Line 252: Delete “,” between “both” & “paddy”.
Line 293: Put “-3” in superscript.
Line 286 & l288: Delete “,” in “A recent study in Punjab indicated that” and “increased by 0.027 and 0.047 µg·m-3 respectively”.
Line 357: There is a repetition of the word “identified”, please change.
Line 62: I would suggest deleting the sentence “this is consistent with our results”, as “confirms” in line 358 already suggests this.
Line 368-369: Keep “µg·m-3)” on the same line.
Line 383: Delete the first “source” in “The source fingerprint of this source”.
Line 397: Correct the start of the sentence to “This factor contributes on average more than 30 µg·m-3”
Line 397-398: The second part of the sentence, “due to…”, to reformulate and you could reference the added map of surroundings.
Line 399: Add space in “NO (R=0.7)” and correct “CH4”.
Line 402 & 404: Once you have written full MTBE and MT, abbreviation is fine. For monoterpenes, you can also write only full name.
Line 403: There are 2 “,” after “acetaldehyde (1.2 µg·m-3)”.
Line 415-418: This part is a little difficult to read, cf general comment about writing all the values.
Line 438: Use “acetone + propanal” as before.
Line 452-460: This part is quite difficult to read and understand, cf general comment about writing all the values.
Line 531-532: Keep “y-1” on the same line.
Line 558: Delete “to” in “Our PMF results reveal that to agricultural”.
Line 608: “two criteria air pollutants” do you mean “critical”?
Line 622: What is EDGARv6.1 better than in this sentence?
Line 635: Add “in Delhi”: “Despite including the most comprehensive set of organic species in Delhi to date”
Line 644: Add “,” after “that”
Line 651: Replace “till date” by “to date”
Citation: https://doi.org/10.5194/egusphere-2024-501-RC1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 301 | 54 | 14 | 369 | 26 | 8 | 8 |
- HTML: 301
- PDF: 54
- XML: 14
- Total: 369
- Supplement: 26
- BibTeX: 8
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1