The authors reported measurement results of ambient particulate matter in Hong Kong using detailed chemical speciation methods. The results of the speciated organic portion by GC-MS suggested dominant contribution of secondary organic carbon (SOC) to total organic carbon (OC) measured. Focusing on the loading and sources of SOC, the authors used positive matrix factorization (PMF) and tracer-based method (TBM) to quantify the concentrations and source contributions of SOC in Hong Kong. Results showed that these two methods tracked well, but the latter method gave a much lower SOC loading compared to the former method (by a factor of 4). The authors went on to use linear correlation and multi-linear correlation to explore the relationships between different SOC types with other air pollutants or indicators, and found that nitrogen oxides (both NOx and NO3) and sulfate might play important roles in SOC formation in Hong Kong.

This is a rigorously performed field measurement study with very detailed chemical speciation information on ambient particulate matter in a typical urban environment in South China. The chemical characterization, data analysis, and discussion are scientifically sound. The manuscript is fairly well written. I therefore recommend Minor Revision with some comments provided below.

Major:

1. I was a bit confused about how different sub-types of SOC were calculated. On P11/L304, the authors showed that SOC_BB was calculated from OC_BB. But where was OC_BB from? The “BB/SOA” factor from PMF analysis? In addition, where were SOC_SOA and SOC_SS calculated from? The organics (since mostly polar and believed to be SOA species) in the “SOA” and “SS” factors from the PMF analysis? It would be better to have some description on how the concentrations of those sub-types of SOC were obtained.

2. Section 3.4: I am not fully convinced that a good correlation between daily concentrations of SOC factors and daily concentration of NO3 can support the statement that oxidation by NO3 radical plays a very important role in SOC formation (although this process might actually be important in Hong Kong, to be confirmed by other evidence). In the daily average concentrations, the diurnal profiles are not reflected, but the most likely case would be that most SOC factors would have peak concentrations during day time, while NO3 would have a peak concentration at night time (although one cannot completely exclude some presence of NO3 at daytime). Therefore, a good correlation between SOC factors and NO3 can at most suggest an indirect relationship between the two.

Minor:

P2/L16: monoterpenes to monoterpene

P2/L30&31: please add “and” before the last item of a list.

P8/L203: I do not see any relevance to HO2 channel from the discussion above. The paragraph is mainly about ring-opening reaction of IEPOX under conditions of high LWC and low pH. Am I missing something here?

P10/L269-289: better to put this detailed description of PMF in the Method section.

P13/L364: just “Offenberg et al. (2017)”.

P14/L390: mass to air mass?

P15/L410: better use proper scientific notation, instead of engineering notation, for the temperature-dependent rate constant.

The authors have carried out comprehensive chemical characterization of atmospheric aerosols and investigated the contribution of different sources to aerosol mass under various atmospheric conditions using positive matrix factorization (PMF) and tracer-based method (TBM). This work provide new valuable field data to better understand the sources, formation and composition of atmospheric aerosols in HKSAR. The paper is well written. However, I have one question about the data analysis.

Major comment

My major question is about the aerosol sample collection and some parts of the data analysis. Line 82, “A high-volume air sampler was used to collect PM2.5 onto a quartz fiber filter (20 cm × 25 cm) at a flow rate of 1.13 m3 min-1 for 24 h.”. Could the authors provide more details on how they obtain the 24-h or time averaged values for some parameters (such as NO3, particle acidity (HP+) and particle liquid water content (LWCP)) in their study? Further, the authors have conducted the correlation analysis of SOC terms with O3, NO2, SO2, OX, NO3, sulfate, particle acidity (HP+) and particle liquid water content (LWCP). Given 24-h samples were collected, could the authors comment how to quantify the effect of NO3 on the formation of atmospheric aerosols using 24-h samples? Same question for some species which have photochemical origins like O3. Could the authors also comment on this?

Minor comments

Line 99, “Air pollutants origin from the northern PRD region can travel together with air masses and transport into Hong Kong. Same as in our previous study (Hu et al., 2010; Ma et al., 2019), we carefully examined the air mass backward trajectories, the spatial distribution patterns of SO2, the concentration levels of both PM2.5 and O3, and the synoptic weather conditions during the sampling period.” Please provide this information in the supplement for reference.

Line 123, “ We then categorized all sampling days into three groups, i.e., days mainly influenced by the regional pollution from the PRD region (regional days), days influenced by long-regional transport of air mass from the northern and eastern China (LRT days), and days dominated by the locally generated pollutants (local days).” Could the authors provide the guidelines or justifications for this classification.

Line 180, “We applied the Kintecus  kinetic model to quantitatively evaluate the fractions of these two Isop_SOA intermediates that undergo gas-phase oxidation,  aerosol-phase acid-catalyzed ring-opening reaction, and dry deposition processes. Details of the model calculations were  provided in the appendices.” Do the authors run the model with the 24-h averaged values? With time revised data (if available), how would the simulated results vary with the hour of the days and day of the years (or different seasons) in this study?

Appendices, Line 476, “Simulation time was set to be 100 h to ensure the completion of reactions.” Why the simulation time was set to 100 h?  Also, why the completion of reactions was assumed in the simulations? How the variation of the parameters during the day and over the year would affect the simulations?

Line 212, “This seasonal trend of monoterpene SOA tracers may be partly due to the lower mixing height and temperature during autumn/winter, which favored the partition of Mono_SOA  tracers onto the aerosol phase.” Has this correction applied to all samples?

Line “353, “It has been well noted that results obtained from this tracer-based method are subject to potential uncertainties from various aspects, e.g., the larger variation of precursor concentrations and more complicated environmental conditions in the real atmosphere than in smog chamber experiments, the decay of some tracer compounds during transport, mismatch of ambient and smog chamber generated SOA compositions, using surrogates other than ketopinic acid for the quantification of tracer compounds, and so on (Ding et al., 2014; Hu et al., 2008; Kleindienst et al., 2012, 2007).” Could the authors comment more quantitively how these factors would affect the results in this study?

Overview:

Cheng et al. performed comprehensive chemical analyses on 49 air filters collected in Hong Kong using GC-MS. They used chemical tracers and standards to characterize the organic species from the filters and then applied both PMF and TBM methods to perform source apportionment. Their results show that the temporal trend of these two methods agree with each other however TBM factors only account for a small portion of the PMF factors. They also show that IEPOX pathway is the dominant pathway to form isoprene-derived SOA during their sampling period by combining chemical characterization with box modeling. Lastly, the team performed correlation analyses to examine the correlation between each SOC factors and common ambient chemicals such as ozone, nitrate, particle sulfate, and acidity and showed different compositions of the SOC may be affected by different chemical species. Overall the manuscript is well written and the chemical analyses are rigorous. The manuscript is recommended for minor revision. I have the following comments for the authors to consider before publication.

Major Comments:

The first major comment I have is mainly related to the description of the PMF method and how they are related to the conclusions. Did the author use the EPA PMF or the Igor based PMF? Which version of the software did the author use? Why the uncertainties of OC/EC and other tracers were set to 20% and 40%, respectively? In addition, any intermediate results on the PMF showing the optimal number of factors to use should be six instead of other values? The author could provide more information about PMF in the SI section to further validate the analysis.

Similarly, when the author characterized the filter samples into three different categories, i.e., the local sources, the long regional transport, regional source, the back trajectory data and the related information should be provided in the SI as well.

Another comment I have is about the correlation analyses. The author did both Pearson R analysis and the multivariant linear regression (MLR) analysis.  The correlation factors in these two analyses do not agree with each other for many categories based on the results in Table 3 and 4. Can the author provide some discussions to explain why the correlation factors from these two analyses do not agree with each other, or even showing opposite trends? And how would the author determine which values to trust?

My last major comment is about the sample collection time. The filter sample collection spans almost a year. The VOC emissions and atmospheric oxidation chemistry might be quite different for summer and winter seasons. Would combine all the samples from different seasons together miss any seasonal variabilities and lead to misinterpretation of the data?

Minor Comment:

L19: Should SOC be defined first?

L80: the sample collection

L223: “b” in beta should be lower case and “C” in Caryophyllinic should be upper case. Beta should be italic as well.

The categories in Table 3 and 4 should be in the same order for easy comparison.