|An et al. develop a high-resolution emissions inventory for I/SVOCs and perform chemical transport model simulations over the YRD region in China to study improvements in model performance and to understand source contributions to OA. The authors seemed to have responded suitably to the two sets of referee comments. I recommend publication of this manuscript to ACP. I have made some additional observations that the authors should consider before submitting a final version. |
1. Lines 165-193 and lines 246-262: I am disinclined to use the label ‘I/SVOC-P’ since IVOCs (assuming ‘I’ stands for IVOCs in the label) shouldn’t really be in the particle phase. The ratios the authors are using from Lu et a. (ACP, 2020) and likely the other sources are to determine SVOCs only. This issue was raised by referee #2
2. I would recommend the authors use a different name for the ‘IMPROVE’ simulation since it doesn’t convey how it is different from the ‘BASE’ simulation.
3. Section 3.2.2: This is an important section of the paper, but it seems like the authors overgeneralize the model-measurement comparison and give the model too much credit; I found this to be a common theme in the paper. For instance, while the model holds diurnal trends across seasons, it isn’t clear that the observations do the same, particularly in the winter months. It is easy to claim that the ‘model works’ but it comes at the cost of ignoring useful insight that the comparison sheds.
4. Too much precision (3/4 significant figures) in the numbers presented in the text?
5. Figure 8: What are the reasons for why the spatial distribution of SOA changes between seasons for the same source? I understand that this could be the case for biogenic VOCs but I am confused why one would expect this for mobile or industrial sources, emissions from which should be less sensitive to seasons.
6. While the emissions, chemistry, and SOA potential are much better understood for I/SVOC emissions from mobile sources, that isn’t necessarily the case for non-tailpipe sources of I/SVOCs (e.g., solvent use). This needs to be discussed in the final section of the paper.
7. For referee #2, comment #2, consider the work of Morino and coworkers: https://doi.org/10.1021/acs.est.8b01285 and https://doi.org/10.1016/j.atmosenv.2022.119319.
8. For referee #2, comment #13, consider adding details about the multigenerational oxidation scheme to the methods section.
1. Line 72: Given the hard ionization in the AMS, there are limits to how much source information can be extracted from AMS data.
2. Line 82: I don’t think that’s right. Material in volatility bins is moved as opposed to moving the bins themselves.
3. Line 83: It should be made clear that VBS is a framework to represent OA, i.e., the implementation of a VBS does not improve model performance. As the authors already know, it’s how the VBS is configured and parameterized is what makes it useful and powerful.
4. Line 110: ‘catching’?
5. Line 146: By definition, IVOCs are unlikely to partition to the particle phase at typical environmental conditions.
6. Line 162-163: What is an example of residential solvent use? What was the reason for not including biomass burning?
7. Line 226: missing ‘and’?
8. Line 311-313: Rephrase SO2 sentence.
9. Line 313: ‘In contrast’?
10. Section 3.1.1: What would be useful is for the authors to discuss why the IVOC-to-VOC ratios are different between sources.
11. Line 376-379: Lu et al. (ACP, 2018) claim that IVOCs from mobile sources are dominated by cyclic and branched alkanes so I take objection with this statement about aromatics being important (I do find it odd that Lu et al. (ACP, 2020) allot half of the IVOC range emissions to aromatics). More clarification is needed.
12. Table S1: Are the IVOCP4-6 the same as the IVOCP4-6-ALK in Lu et al. (ACP, 2020)?
13. Lines 448-455: The IVOC model comparison with observations is less than impressive. The model is higher in the summer (9 versus 6.8) and the model is lower in the winter (12.4 versus 18.2). Discussing if this is an emissions or chemistry (or both?) issue would be helpful.
14. Line 471: Effect of stronger photochemistry in the summer?
15. Figure S3: Why are there so few OC measurements to compare against? Or have these been averaged for the entire season? Is the model able to capture intra-season variability at each of the sites?
16. Figure 7: It needs to be made clear how the modeled species representing OA as well as the different AMS factors are aggregated to perform comparisons for ‘POA’ and ‘SOA’. My sense is that there are uncertainties in this aggregation and it is not self-evident that the AMS gets you clean POA and SOA data to compare model predictions against.