The given manuscript discusses the importance of PMcoarse in formulating policies due to its growing relative contribution to PM10 loading in urban atmospheres. The paper uses the PM composition data derived from the measurements conducted in Hong Kong. The use of positive matrix factorization resulted in identifying four PMcoarse sources, including soil dust, Cu-rich dust, fresh sea salt, and aged sea salt mixed +secondary inorganic aerosols. Results also showed that these four sources can explain unidentified fraction of PMcoarse. Overall, this work presents a simple approach to understand PMcoarse composition/sources which can be applied to other locations with similar monitoring needs.

This work is nicely constructed, and my comments are listed below to consider:

• Please add reference to the sentence “For example, nitrate in the coarse mode is formed by the reaction between nitric acid (HNO3) from oxidation of NOx and pre-existing alkaline aerosols (e.g., sea salt and dust).”
• Please provide more details on the uncertainty matrix. I can’t find any information on the uncertainty of species those concentration is above the detection limit.
• Why was Deming regression applied?
• Line 155: What is the basis for using a ratio of 2 to calculate organics, author should provide clarification.
• Line 160: Add reference for using 1.6*OC.
• Add reference-“Coarse mode nitrate mainly forms by the uptake of HNO3 by pre-existing alkaline particles forming NaNO3 in reaction with sea salt and Ca(NO3)2 with soil dust.”
• Line 220: Add reference for Si estimation.
• I am not convinced with the factor 1, it is a mix of two sources. Apart from crustal elements, there is also a significant contribution of Pb, V, Mn and Zn which suggests this source is not properly resolved.
• The discussion about the seasonal contribution/variation of PMF factors should be enhanced. Currently, the given discussion is not sufficient to understand their origin. In addition, it would be great if author can also provide some insight on sources based on the previous receptor modeling results. Are the present results aligned with the previous observations?
• Is the contribution of Cu-rich dust and construction dust by Zhou et al., comparable? What is the similarity between these two sites? Are the air masses originates from construction active area?
• In the section 3.4, author mentioned that the aerosol samples were not corrected for sampling artifact of nitrate. Did author try to apply the correction and observed any change in the nitrate measurement?

This study performed chemical speciation for PM_2.5 and PM₁₀ samples collected in Hong Kong during 2020/01-2021/02. The results showed that the annual average concentration of PM_coarse (PM₁₀-PM_2.5 mass) accounted for ~50% of PM₁₀. Unlike PM_2.5, only ~75% of PM_coarse mass was explained by identified chemical components. The authors supposed that the unidentified part was dominated by geological components and aerosol liquid water. Moreover, several tools were utilized to apportion PM_coarse to specific sources and areas, particularly for the unidentified fraction. In general, this manuscript is well organized and written. But two major issues should be addressed before the consideration for publication.

1. In this work, the thermodynamic equilibrium model (ISORROPIA II) was adopted to estimate aerosol liquid water (ALW) in PM_coarse. After mass closure and PMF analysis, the authors concluded that the unidentified PM_coarse (4.1 μg m^-3, ~25%) was substantially contributed by ALW (1.2 μg m^-3).

Have the authors performed mass closure for PM_2.5 or PM_fine? Because the fine particles are more enriched with water soluble components (e.g., secondary inorganic ions), ALW should contribute more fractions to PM_2.5. According to section 3.1.1 (lines 158-160), it seems that PM_2.5 is mainly composed of NH₄⁺, NO₃^–, SO₄^2–, OC, and EC (~80%).

If ALW contributes a significant fraction of PM_coarse based on filter sampling, there’s no reason that it contributes less to PM_fine.

In fact, ALW is not stable on filters, and is subject to loss during long-term sampling and transportation.

So, the contribution of ALW to unidentified PM_coarse might not be estimated appropriately with the current study design.

2. When input PM_coarse mass for PMF analysis, it was presumed that the unidentified PM_coarse fraction have the same sources as identified components.

In this work, four factors linked with soil dust, copper-rich dust, fresh sea salt, and aged sea salt were identified using measured species data. Since understanding the sources and formation pathways of PM largely depends on how well they are identified, the sources of un-speciated coarse PM are unknown and might not be the same as measured species. If the unknown fraction of coarse PM was apportioned to the four identified factors, some factors contributions would be over-estimated. Because PMF may over-attributed PM_coarse to certain factors as it fits measured species (Shrivastava et al., 2007). This will occur if makers for unknown PM_coarse are not included in the PMF model (Shrivastava et al., 2007).

Therefore, the source apportionment method for unidentified PM_coarse mass is not appropriate. The authors should focus on sources of identified PM_coarse components.

References

Shrivastava, M. K., Subramanian, R., Rogge, W. F., and Robinson, A. L.: Sources of organic aerosol: Positive matrix factorization of molecular marker data and comparison of results from different source apportionment models, Atmospheric Environment, 41, 9353-9369, 10.1016/j.atmosenv.2007.09.016, 2007.