This study characterized the diel dynamics of various organic markers in aerosols using a TAG system at urban Hong Kong during a summer period. The high time resolution observations of organic markers allowed to identify specific sources or processes that had crucial contributions to organic aerosols. This effort promoted our understandings of the sources of organic aerosols in urban HK, and the manuscript was well written. However, there have been some studies using TAG to apportion the sources of organic aerosols at urban/suburban areas. This study just looks like a supplement of TAG data in urban HK. Therefore, I recommend a major revision on the structure of writing/descriptions to point out some interesting findings. Detailed comments are as follows:

1. The title looks inappropriate. First, the study was only conducted in a summer period less than one month. I feel the results could not represent “An overview of organic aerosols” in urban HK. Second, as stated by the authors, the concentrations of organic markers were not quantified, and some important markers such as monoterpene derived SOA tracer were not identified. In addition, nitroaromatics, which are key brown carbon species, were not detected. Nitroaromatics may also help evaluate the contributions of oxidation products from biomass burning and vehicle emissions to organic aerosols. Third, the title did not carry impressive information.
2. The manuscript is mainly discussing the influence of different air masses (continental, coastal, and marine) on organic markers and OM sources. However, looking at Figure 3, there was only one day when air mass was from continental regions; only two days when air mass was from coastal regions and there was a one-day long maintenance of instruments when air mass was from coastal regions. One-day long observation is not representative. Therefore, grouping the data by continental, coastal, and marine air masses is not appropriate. For example, it is odd that cooking emissions had an insignificant contribution to OM when air mass was from continental region if the sampling site is surrounded by restaurants. This may be due to cooking plums did not significantly affect the sampling on just that day. It is feasible to regard the periods when continental and coastal winds were prevailing as cases to discuss the influences on organic markers but cannot regards the results as common situations when continental and coastal air mass dominated.
3. For the PMF analysis, it is not a good way to keep a factor which is unexplainable. The authors are encouraged to improve the PMF analysis by modifying inputs or others. As highlighted by the authors, cooking emissions should be a major source of OM at the sampling site, while the contribution from cooking oxidation products to OM could not be evaluated. A previous study (Huang et al., Comparative Assessment of Cooking Emission Contributions to Urban Organic Aerosol Using Online Molecular Tracers and Aerosol Mass Spectrometry Measurements, Environ. Sci. Technol. 2021, 55, 14526-14535) used azelaic acid, nonanoic acid, and 9-oxononanoic acid to indicate the oxidation of cooking emissions. I noticed azelaic acid has been detected in this study (No. 20 in Figure 2). The authors may try to conduct an analysis to evaluate how significant of cooking oxidation is in contributing to OM mass.
4. It was interesting to identify the periods lasting a few days when trough played a role. The meteorological parameters such as RH, Temperature, solar radiation, wind speed showed significant differences between “trough” days and “non-trough” days. I think it is valuable to focus on discussing the variations of organic tracers, especially SOA tracers, during the two distinct periods. It is good to see that in the manuscript the authors have pointed out the increases of phthalic acid and DCAs during “trough” days, indicating an aqueous formation of the species. The authors are encouraged to find more markers that had significant difference between the periods. Hopefully, the authors can evaluate how significant of aqueous formation is in contributing to OM mass. This would be a very interesting part.
5. Some inconsistencies may exist. For example, in Figure 7a, cooking emissions (dark yellow) constituted a significant part of OM sources during June 15th to 17th (continental and coastal air masses dominated according to Figure 3), while in Figure 7b and text, cooking emission was a very minor source of OM during the period. Another example is the X axis of Figure S8 shows a OM range of 0-22 μg/m3 while the Y axis of Figure 7a only reached 15 μg/m3. If X axis of Figure S8 shows the observation result, then there should be an unresolved percentage in Figure S8.
6. line 185-188: the explanation may need to be modified. Just like I mentioned above, one-day observation for continental air mass dominated period may not capture the influence from cooking emission. During fall-winter period, most air masses may come from the north, and I guess you may still find the contribution from cooking emissions if a long period of observation is available.
7. line 227: why not examine the correlation between pyrene and hopanes? NOx can also be emitted from biomass burning.
8. line 290: add “respectively” after PM1-OM.
9. line 318: please try to add azelaic acid in PMF analysis to see if cooking oxidation factor can be resolved.
10. Figure 1. How about show the locations of restaurants in the map?
11. Figure 3. Please add the time series of DCAs.

Major concerns

1. The study determines a series of organic markers in PM2.5 with two-hour resolution in Hong Kong for one month. However, the data management is inadequate. The mean and median values for most species are very different (at least for those shown in Figure 4), probably due to variability in compound concentrations. Therefore, it is required to know the distribution of the variables, to confirm the assumption of the normal distribution done by the authors.
2. The authors must show the distribution of the variables, thereby justifying the type of statistics to be used, whether parametric or non-parametric, or some other mathematical proposal for data management. They should double-check any statements or suggestions they made.
3. Throughout the document, no statistical evidence is shown to support the conclusions of the values comparisons between the different periods. Nor the use of Pearson correlations. Subjectivity should be avoided in the discussion. Appropriate statistical tests will verify the hypotheses proposed by the authors. The article cannot be published in its current state and must be resubmitted for consideration.

Minor comments

1. Lines 100-104. Place the mass of the compound isotopically labeled with carbon 13, added to the TAG.
2. Lines 105-109. Change below to "throughout the document".
3. Lines 110-114. This way of associating the variables with their respective concentrations is confusing. Each value must be associated with the corresponding contaminant. Apply to the entire document.
4. Lines 120 - 129. The description is very confusing. Explain more clearly why organic contaminants were not quantified with the TAG. Explain what the authors refer to as “internal standard scaled peak areas”. Check if it is an accepted term in the scientific community. Enter the corresponding equation for calculation.
5. Lines 150-154. Explain in detail and in the supplementary material the calculation of the method detection limit (MDL). Explain why it is only one value. In general, it should be one MDL for each compound.
6. What sense does the MDL in µg/m3 if the units of the compounds are shown in relative areas?
7. Lines 170-174. Avoid subjectivity throughout the document. It is not appropriate to simply say: slightly higher, slightly lower, etc., because nothing conclusive is got it. The level of significance associated with the statistical test must be set. In this case the averages are compared. As mentioned above, the use of the average must first be justified, and subsequently, the result of the comparison test and the significance level must be given. This applies to the rest of the comparisons made throughout the document.
8. In a separate section, the statistical tests must be described.
9. Lines 190-194. This assertion is not clear. Figure S4a shows that levoglucosan maximum indeed comes from the east, but between 7 and 18 km away from the site in the hour before its emission. From figure 1, it is not clear how the authors assume that the biomass combustion is located in the first kilometer east of the site. Clarify.
10. Lines 215-219. What is the evidence to associate these alkanes with sacrificial activities. Explain what the authors refer to as sacrificial activities.
11. Lines 220-224. Azelaic acid is not shown in Figure S5.
12. Lines 220-229. The pyrene is a clear example that the average is not the best metric to compare periods. In this case, the median of the hour 20:00 is like the rest of the hourly medians. This change the discussion and conclusions.
13. Line 227. It is well known that pyrene and NOx are emitted by vehicles that burn diesel. Could this correlation be associated with this source?
14. Line 228. Does the term funerary refer to crematoriums for humans and animals? since there are funeral homes that do not have this service. If so, the authors should cite studies showing the type of PAHs emitted by these sources found in this study. This information will be useful to strengthen their hypothesis.
15. Line 240. It's confusing. Is it sum or division?
16. Figure 3. Why does the time scale go from 6 to 26? What are the units? Eliminate the µg/m3 on the scale on the right. Change color between ammonium, nitrate and coastal air bar. It's confusing. Change the color between the UV and continental air bar. It's confusing.
17. Figure 7a. Why does the time scale go from 6 to 26? What are the units? Improve sharpness.
18. Figure S2. Why does the time scale go from 6 to 26? What are the units? Place the color bar for marine, continental and coastal air, as in figure 3.
19. Figure S5. Explain the calculation to normalize the variables.
20. Figure S6. There is inconsistency between the graphs and the legend.
21. Table S2. Verify the names of the compounds.