General comments

Zheng and Xie present a method for assessing the ozone forming potential of a series of volatile organic compounds measured in the Sichuan Basin, China.  The method is based on observation of ambient concentrations of VOCs, followed by estimates of what is then referred to as “initial” VOCs, intended to represent the emitted amounts of these VOCs or the amount produced by secondary production in the atmosphere.  Ozone formation potentials are then calculated using literature maximum incremental reactivities based on either observed or corrected VOCs.

The manuscript is clearly written and the figures are of good quality for presentation.  It will be of interest to the readership of ACP.

The major comment is that the classification of VOCs by chemical functional group rather than sources may lead to some errors.  This is especially true for alkenes, but may also pertain to the oxygenates. Anthropogenic alkenes should be treated differently from biogenic alkenes (mainly isoprene), since these VOCs have very different sources.  Division into these categories would make the analysis methods also self-consistent, although it may require some change in methodology for isoprene itself.

English language comment: The paper uses past tense frequently where present tense make more sense for English writing.  If possible, suggest proofreading by a native English speaker.  As one example, the first line of the manuscript (line 24) is better write as “Volatile organic compounds are key species in ozone formation”.

Specific comments

Line 28: The terminology of “initial VOCs” to indicate VOC emissions is somewhat confusing.  Would suggest a different label for these compounds to make this clear, as in “emitted VOCs”.

Line 81: replace “moving” with “removing” and “particulate matters” with “particulate matter”

Line 97-98: The “NO₃ or O₃ exposure method” is not defined.  How is this used to estimate emissions ? 

Line 117 and equation 1: Benzene and isoprene are not co-emitted species, so no relationship or emission ratio would be expected between these species.  Furthermore, these species have very different diel emissions profiles, with isoprene not emitted at night but strongly peaked during daytime.  The indicated dependence of the benzene to isoprene ratio is therefore not likely to be applicable to this pair.  A different approach would be required to relate the concentrations of these species that lack a common emission source.  Other pairs, such as benzene – styrene, or benzene – butadiene that have common anthropogenic sources should be amenable to this method.

Line 271 and figure S10: It does not necessarily make sense to calculate an OFP during nighttime hours.

Line 309: Isoprene rather than Ioprene

Having hourly concentrations of 99 VOCs and OVOCs at three different field sites in the Sichuan Basin, China, Zheng and Xie calculate the ozone formation potential (OFP) for these species using the maximum incremental reactivity (MIR) method. However, unlike previous studies that have calculated OFPs using the observed/measured concentration of the VOC or OVOC, the authors correct these concentrations to estimate what the concentrations of these species would have been before being consumed or produced in the atmosphere (something they refer to as "initial" concentrations). The authors show that using the "initial" concentration of a species can impact the relative importance of that species in forming ozone.

The paper would be appropriate for inclusion in ACP after the following comments are addressed:

Specific Comments

- For Eqs. 1, 3, and 5, there should be a brief explanation for each on why benzene and isoprene, benzene and cis-2-butene, and ethylbenzene and m,p-xylenes are paired to calculate the NO₃, O₃, and OH exposures, respectively. For example, in the NO₃ exposure, benzene and isoprene are paired, but they have completely different emission sources and do not seem well correlated in Figure S4, so explaining why they are paired would be beneficial to explain for the reader.

- Fig S4: The initial emission ratios between benzene and isoprene were obtained from Figure S4 for use in Eq. 1. However, the authors should explain why the fit slopes and their associated uncertainties do not appear to accurately represent the underlying data shown. For example, I would have expected a higher slope at Meishan than what is presented. What is the R-squared for each of those fits? How much does the magnitude of the slope ultimately affect the calculation of the NO₃ exposure?

- Fig S5: Similar comment as above for Figure S4 .

- Line 298: I agree that there appears to be a diurnal variation of OFP at Chengdu in Figure S10, but there does not appear to be one in Deyang (particularly in Panel (a) of Figure S10) even though the text says there is one. Sunlight intensity is mentioned as a possible reason, but total OFP at night is similar to that of the day in Figure S10a. Could the authors explain what is happening here?

- Lines 312-314: Comparison is only made to the Song et al. (2018) paper, but Lines 39-41 suggest multiple studies reporting OFP for Chengdu and Deyang. I would like the authors to compare their results with those other cited studies as well in addition to the Song et al. paper.

- Comment on Abstract and Conclusion: The text for the abstract and conclusion are almost identical. I would suggest that the authors consider using these sections to reiterate how their modified method for calculating OFP is novel compared to past approaches. Is there a result that particularly stands out that the prior approach of simply using the observed/measured concentrations to calculate OFP missed?

Technical Corrections

- In Section 2.2, refer to the upcoming sections (2.2.1, 2.2.2, etc.) as they are described in the text.

- Line 137: [m-alkenes_j] is mentioned in the text, but Eq. 2 uses [a-alkene_j]. Change one to be consistent with the other.

- Line 149: Same comment as Line 137. Use either a or m to be consistent.

- Figure 2 caption: Change wording of caption to denote that ambient VOC concentrations are colored black and initial VOC calculations are colored red. The (a) and (b) notation does not make sense.

- Line 262: Say OVOCs instead of oxygenated VOCs to be consistent.

- Line 290: Mention that this is for Meishan.

- Line 291: Should it say Figure S9ef?

- Line 298: A space is needed between 403.80 and μg m^-3

- Line 309: Ioprene should be isoprene.

- Lines 315-318: consider moving this paragraph to the conclusion.

- Check that appropriate verb tense is used throughout the manuscript. Numerous sentences were written in past tense when the present tense would be more appropriate.