The manuscript by Cheng et al. investigated the characteristics of haze pollution during two sequential heating seasons in the central city of the Harbin-Changchun (HC) metropolitan area, with a focus on the major drivers responsible for the observed inter-annual variations of aerosol composition. Compared to traditional hotspots of air pollution studies in China (e.g., the North China Plain), HC is unique with respect to both meteorological conditions (e.g., the extremely cold winter) and anthropogenic emissions (e.g., those from the intensive energy use for central and space heating). Therefore, observational results from this region may provide additional insights into haze pollution in China. However, PM2.5 in HC remains largely unexplored with limit studies, which may explain the relatively slow improvement of air quality in HC compared to the North China Plain. In this context, the authors did a good job of unfolding the chemical compositions and sources of PM2.5 in HC. Many in the community of atmospheric science would be interested in the results presented. My overall assessment is that this manuscript could be considered for publication in ACP given the following concerns (mostly technical) could be properly addressed.

1. Line 54, suggest changing “aerosol” to “aerosols”.
2. Line 58, “its effects on SOA…” should be better.
3. Line 97, suggest removing the “the” after “with”; in addition, references should be provided for “limited studies”.
4. Line 177, as indicated by the PMF source profiles, it seems that zero EC was apportioned into the two factors representing secondary aerosols. Please confirm and revised the description if necessary.
5. Lines 207-208, I think the “during 2019-2020” here is redundant and unnecessary.
6. Lines 227-228, rephrase “the same OC average concentrations”.
7. Line 235, references should be provided for the ISORROPIA model.
8. Line 249, suggest adding a “usually” before “not”.
9. Line 261, suggest adding a “the” before “positive”.
10. Line 287, the meaning of “those” is not clear enough.
11. Line 291, references are necessary for this statement.
12. Line 307, suggest adding an “In addition” before “an obvious difference…”
13. Line 318, suggest changing “emission” to “emissions”.
14. Line 332, please note that the more rapid decrease of SO2 emissions compared to NO2 is valid only for recent years in China.
15. Line 637, change “contribution” to “contributions”.
16. Line 647, suggest adding an “an” before “ambient”, and changing “with” into “at”.
17. Annotation in Figure 8, suggest changing “Stronger impact of…” to “Stronger impacts of…”
18. Caption of Figure S4, suggest using “increasing strengths of biomass burning impact”.
19. Caption of Figure S6, I guess something was missing after “Relationship between OC/EC”.
20. Caption of Figure S7, rephrase the sentence “For the 2019-2020 campaign, (OC/EC)pri and OC* were determined as the slope…”
21. Captions of Figures S9 and S10, it is unnecessary to repeat the definition of “D” and “H” conditions.
22. Caption of Figure S11, suggest changing “temperature” to “temperatures”.

This work investigates the aerosol chemical composition in Harbin in the past 2 years based primarily on in-situ measurement. Generally, I find that this manuscript may be more like a measurement report since that it does not provide sufficient new insight into atmospheric chemistry but the dataset seems unique and presents some new perspectives. One main problem with this work is that it entirely attributed the variations of secondary aerosol to the chemical processes, but aerosol does have a lifetime of 1~2 weeks and regional transport could largely contribute to the temporal variation at one specific site. Overall, I think this work fits the scope of ACP but there are some issues that need to be addressed to improve this work. 

The method ought to be detailed in the main text.  Section 2 is too sketchy, in which some basic information like the geographic location of the campaign site, its representativeness, the main emission sources in the surrounding areas, the instrumentation and the QAQC should be included. Also, other analysis like AWC, Positive Matrix Factorization (PMF) results discussed in Section 3, which is not from the direct observation, should be described in this section.  This article itself missed a lot of necessary information on the method and analysis tools. I suggest moving the supplementary method to the main text. Another issue related to the method is that PMF analysis usually needs a large amount of data to support the factorizations since that it basically is statistical analysis. According to the Supporting Information, the sampling number is around 200~300, which is not enough to get a reliable result.

The discussion part is too descriptive and the majority of the main text is just describing the variation of different aerosol components. In some places, the authors jumped to some conclusions without careful investigation. For instance, the RH-dependent increase of OC was attributed to heterogeneous reactions without investigating the cloud-water chemistry and diffusion/dispersion analysis (Line 236-238). Additionally, the author concluded that the inter-annual (I would not call it inter-annual since it is just a two-year comparison) variation of OC was related to RH levels without other vital information of the regional transport pattern and other meteorological parameters. Besides, the higher threshold RH for the sharp increase of SOR in Harbin is not certainly indicative of the fact that the heterogeneous formation of sulfate was less efficient there. Many previous works have pointed out that the polluted air mass from the southern area of Beijing also brings about humidity, rather than heterogeneous chemical reactions at a local/city scale. Thus, some in-depth analysis and rigorous arguments need to be added to this work.

There are also some inconsistent arguments that needed to be checked. For instance, in the abstract, the authors claimed that "we found that open burning activities were actually not eliminated ……", but Line 168 stated that "indicating that agricultural fires were almost completely eliminated during the measurement period". It doesn't seem very clear.