General Comments:

Interesting work with clear and significant conclusions, but requires more detail and explanation in some areas. For example, more context and clarity are required in the introduction and methods sections. and further experimental details to aid reproducibility. Some edits required to make the language used more fluent and precise.

Specific Comments:

Introduction:

Line 45 – “No one instrument can capture all VOCs out there and even when they can be measured there is information missing on identification and properties (Yuan et al., 2017;Wang et al., 2014)”. This sentence needs expanding into a paragraph explaining what these species, instruments, and measurement methods are. Which VOC species can’t currently be measured/haven’t been measured and using which techniques?

Line 47 – “By now, emission inventories of VOCs used in air quality models only include the VOC species that can be measured”. This sentence needs some clarity. Would suggest changing `by now` to `currently` and giving examples of which emissions inventories and VOC species are being discussed.

Line 70 – “The inclusion of the missing VOCR can help to improve the model performance in simulating photochemistry processes”. Clarify which model.

Line 90 – “Given that the missing VOCR accounts for a large part of total VOCR”. This should probably read `could potentially account for`, or something similar, as this has not been determined yet.

Method:

More details required on the experimental procedure. For example, how and where the instruments were deployed. Were continuous measurements taken from 26^th Sept to 30^th Oct? Were the GC, PTR, and custom-built instrument run simultaneously?

PTR - Was this run in selected ion monitoring mode? If so, which reagent and product ions were selected? (Would suggest including a table of these in the SI.) Other missing details include drift tube pressure, temperature, and voltage, etc.

Include details of GC–MS/FID parameters.

Line 119 – Can the correction curve be supplied in the SI?

Line 134 - `A total of 56 NMHCs species were measured`, provide information in SI.

Line 142 - `A total of 31 VOCs were calibrated using either gas or liquid standards`, provide information in SI.

Line 143 - `For other measured VOCs`, provide details in SI.

Results and Discussion:

Line 206 - How many extra species were measured by PTR?

Technical Corrections:

Line 53 – “The measurement of total OH reactivity (ROH) provides an effective approach to quantify the total amount of reactive gases in terms of reacting with OH radicals.” I think this sentence could be reworded to be clearer and more concise.

Line 166 – “The multiple linear regression (MLR) have been successfully applied to quantify the sources of air pollutants (Li et al., 2019;Yang et al., 2016a).” Grammar – `have` should be `has`.

Line 176 – “Calculated from observed isoprene and its photochemical products MVK and MACR”. Define acronyms.

Line 261 - ` Given the larger missing VOCR level during the high missing- VOCR days, we focus on the high missing- VOCR days in the following analysis.` This line seems unnecessary and repetitive.

Line 290 – Compensate rather than compensates.

Define units for all equations and variables where appropriate throughout.

Some of the details of how the box model was run from section 3.3 might be more appropriate in the methods section than in the results.

General Comments

This paper presents an interesting analysis for understanding missing VOC sources in urban areas. The science and the methodology are sound; however, the paper itself lacks depth. It seems as if the authors were hesitant to list out too much information within the manuscript.  It also assumes any reader has the exact background knowledge to follow all arguments so the authors don’t elaborate. While it is true readers can check the references themselves, and should, it would be nice if the authors could share tangible evidence from the sources that support their work. A paper should not only introduce a new idea or method but also be written in a way that subsequent papers can test the method for themselves and apply it to other data sets. As it is currently written that is not possible. This work definitely should be published but the authors need to go over the manuscript and hone the message. I suggest this paper is accepted with minor revisions because no additional analysis needs to be done but share a clearer message for what has already been done.

Specific Comments

Lines 90-91: You mention that missing VOC_R is a large part of total VOC_R but you don’t give numbers or examples for the reader. You mention suburban sources in reference to R_OH (Line 76) but not for VOC_R

Line 127: Could you show how a NO_x correction is applied? Perhaps in the supplement. What does “carefully applied” mean?

Line 177 – 178: Why don’t use you include values for a, b, c and C_background in your paper? A main point in your conclusion is how this is new method to use the CO with VOC_R to get at anthropogenic missing fraction but then you don’t show any concrete numerical examples using this new method.

Lines 211-213: Did you ever have negative VOC_R? In other words, observations higher than the calculated missing VOC_R? It looks like you have some periods no visible gray in the figure for missing VOC, if negligible but still some 'missing' also good to point out when that happens as well as the general 20% since that is impressive given the issues you introduce in the beginning of the paper from previous work.

Line 256: Include the reaction rate constants used

Line 293: By what factor did you increase all the NMHC? Were they all increased the same amount? What was the process here?

Section 3.3: The writing is unclear about the sensitivity studies. Were the individual VOC species (represented by the 3 examples) and the “all measured NMHC” done together or 4 different sensitivity studies? I’m assuming it was 4 different model runs but as written that isn’t apparent and it sounds like they were done all together. It wasn’t until looking at the figure it seemed like 4 runs. In particular, line 293 “and the one considering” suggests one model run which means you didn’t look at the impact of each species. Were the individual species taken from the measured results? Where did those numbers come from?

Lines 347-348: in regards to the parametric equation “developed here” do you mean being able to separate them all out? It isn’t just “versus CO” according to equation 4 so this is misleading.

Lines 348-350: “are also expected” doesn’t make sense in the context

Figure Comments:

Figure 1: Why are a and b blue but c red? If for primary vs secondary that isn’t referenced in the caption so is irrelevant since it doesn’t match the color scheme in c or d. For example in d, missing VOC is red but in c it would be secondary sources. In e, what are the green squares? You don’t reference them anywhere. Why are the circles magenta/pink? Not necessary and detracting.

Figure 2: Again the different colors for a and b seem unnecessary and then don’t match c and are in fact opposite. CO = anthropogenic but it is blue and red in a and c, respectively.

Figures 4 and 5: Nice use of colors here that tie the idea together.

Figure 5: there is no blue bar a but that is referenced in the caption. Also for a, it would be nice to note the dashed line represents the NO_x vs VOC limited regimes. 

Technical Corrections

Line 57: Indented but shouldn’t be

Line 68 and elsewhere: Perhaps a personal preference but the oxford comma can be very useful with complicated lists in sentences

Line 120: “thank to the” should be “thanks to the”

Line 131 and elsewhere: Be sure to have a space between a ) and the next word

Line 140: “cannot” is the more common spelling for this

Line 166: “The multiple linear regression (MLR) have” is awkward and incorrect tense. Perhaps something like “Multiple Linear Regression (MLR) has been”

Line 228: “As the” should be “As a”

Line 247: no period between 22:00 and )

Line 256: The wording of “higher aging degree of air masses” is awkward. Perhaps "higher degree of aging air masses" or "higher degree of air mass aging" based on what you write below at 259

Line 287: and elsewhere: Be consistent with – and spacing, sometimes an extra space and sometimes not