Dear authors,

This is a very interesting particulate matter (PM) source apportionment study performed in two high-altitude cities in South America. This study is relevant, since there is little published data on this subject about  La Paz and several other South American metropolises, as most of the studies are concentrated in cities such as Santiago or São Paulo in this subcontinent. Furthermore, these results are very important since these sites are located at a very high altitude and this may influence vehicle emissions, being helpful for studies in cities with similar topography. PMF treatments are described in a very didactic way and that can be useful for further source apportionment studies. 

Minor revisions should be performed to have it published, most related to technical corrections, although there are some specific comments, as well. As a non-native English speaker myself I recommend that the authors send the paper for an extra grammar check.

Specific comments:

1. Lines 31-32 - Some studies as Serafeim et al., 2023 (Atmospheric Environment v. 298,  119593) do include PAHs, and WSOC together with inorganic species for São Paulo (Brazil) particulate matter, however, this may be the first including hopanes and alkanes.
2. Line 183 - It is always important to mention that non-sea salt potassium is not exclusively emitted by biomass burning. This species may be also very linked with soil resuspension and is also related to fertilizers (Urban et al., 2012 - Atmospheric Environment, v. 61, p. 562-569), especially in dusty sites such as El Alto (which depends on the local soil composition, of course).
3. Lines 271-276 - Are these studies comparable?  Were they also performed year-round (or specific polluted period)? This should be discussed.
4. Lines 280-284 - Some authors attribute those differences to adsorbed water in the aerosol and to the presence of non-measured species (i.e. carbonates ), which depends on the way PM was reconstructed (Pio et al., 2013 - Atmospheric Environment, v. 71, p 15-25)
5. Section 3.2 - The authors may consider using Latin American review studies for the source apportionment interpretation and discussions, such as La Colla et al., (2020 - Environmental Reviews v. 29, n. 3), or emission studies such as Brito et al (2013 - ACP, v. 13, p.12199-12213) and more recently Pereira et al. (2023; v. 856, 159006) performed for the biofuel-impacted fleet in São Paulo (Brazil), they can be helpful. The use of biofuels in Bolivia could be discussed, as well.
6. Lines 350-354 - Were these mass concentration values given by those authors or calculated here for this manuscript? If these values were calculated by you, this should be mentioned somewhere in the manuscript.
7. Lines 430-432 - These Lev/Man ratios were related to sugarcane burning, some ratios can be associated with specific types of biomass. That should be considered in the discussions. Check the excerpt from Zhang et al. (2015 - Atmospheric Environment v. 102, p. 290-301):  “[...] An overview of biomass burning tracer ratios derived from various source profiles is provided in Table 4. In general, low LG/MN ratios (typically 3–7) are generated by softwood combustion. In contrast, burning of hardwoods and crop residues is typically associated with higher LG/MN ratios with a rather broad range of 10–83. Similar characteristics were also observed from the burning experiments of typical biofuels in South China.[...]”
8. Line 432 - The contribution difference seems to be quite small (8 and 9 %).
9. Line 441 - Sn is duplicated here. Cu was associated with gasohol and ethanol-powered vehicles’ exhaust in Brazil (which may also be related to the corrosion of engine parts with the increase in the use of ethanol fuel) by Brito et al (2013) and Pereira et al., (2023).
10. Line 453 -  The authors could perform correlations of PMF factor contributions and chloride concentrations to confirm that.
11. Line 469 - Do these factors increase with the same wind direction? Could they be from vehicular sources from different locations?
12. Line 598 - What do the authors mean by  "natural anthropogenic sources of PM "?

Technical corrections:

1. Line 108 - The authors may consider rewriting this part “Although the characteristic tropical seasonal change between a dry and a wet season”
2. -Line 137 - PM₁₀ (“10” should be subscript).
3. Line 183 - “Biomass Burning”.
4. Lines 188, 194, 198, 202, 205, 218, 1038, and Table 1- The word “species” is the same in the singular and plural, the authors may revise that all over the text.
5. Line 251 - "Belis et al. 2019" - Check if the citation format is correct (in other parts of the text, as well).
6. Lines 263-269 - “10” should be subscript on PM₁₀ and “-3” superscript on m^-3.
7. Lines 236-238, 271, 288, 305, 311, and 323 - Check if figure and table citations follow the journal rules, there different formats throughout the manuscript.
8. Lines 285-287 - Consider rewriting this more clearly.
9. Lines 352, 432-433, and 458 - Martins Pereira, 2017 and Pereira 2017 are different articles from the same author. They may be cited as "Pereira et al., 2017a" and "Pereira et al., 2017b".
10. Line 360, 384... - Space between “μg” and “m^-3” throughout the text.
11. Line 442 - "Amato et al., 2011" - Check if the citation styles are correct (in other parts of the text, as well).
12. Lines 528 - "United States".
13. References section - Check if all the references are following the ACP guide for authors. Improve text details, as well (superscript and subscript).
14. Figure captions - Check superscript and subscript.

A review of the paper of authors:

Valeria Mardoñez, Marco Pandolfi, Lucille Joanna S. Borlaza, Jean-Luc Jaffrezo, Andrés Alastuey, Jean-Luc Besombes, Isabel Moreno R., Noemi Perez, Griša Močnik, Patrick Ginot, Radovan Krejci, Vladislav Chrastny, Alfred Wiedensohler, Paolo Laj, Marcos Andrade, and Gaëlle Uzu

Source apportionment study on particulate air pollution in two high-altitude Bolivian cities: La Paz and El Alto

 Submitted to ACP

The paper deals with source apportionment of PM10 concentrations and chemical composition in La Paz and El Alto, Bolivia. The comprehensive chemical composition database includes results of analysis of PM10 samples obtained at two urban background sites in parallel during one year long campaign on OC/EC, water soluble ions, sugar anhydrides, sugar alcohols,  elemental composition, PAHs, methyl PAHs, alkanes, thiophens, hopanes, and methoxyphenols. Such database is unique in South America and rare elsewhere.  

Thanks to presence of so many tracers, the authors were able to distinguish 11 factors. To overcome problems with too many variables and not so many (around 100 per site) samples, they limited number of variables to most specific tracers and major species together with grouping of highly correlated compounds. Moreover, they increased number of samples by combining both dataset into one after proving similarity between the two datasets in term of factors composition.

The PMF method is nicely described and the results seems reasonable. There is also nice discussion of PMF limitation and uncertainties. Nevertheless, there are several minor comments that should be solved before publishing it in ACP.

Minor comments

Fig. 1. The detail map of the sampling site has wrong color scale (opposite to the map of South America) and must be corrected.

Line 212: Typo - should be “hopanes” instead of “hopaes”

Line 225, Multisite PMF: the dimensions of the final matrix should be mentioned, nuber of variables used in final run is missing.

Line 240, fuel fingerprint: The units in results of chemical analysis of fuel in SI are apparently wrong, must be corrected. Moreover, the reviewer is surprised by high concentrations of arsenic in gasoline in comparison with lead and zinc. Is that really correct?

Line 250: Fig. number is missing in this line.

Line 278 -280, equation 4. The equation contents mistakes. Sulphates are corrected for sea salt sulphates, but no sea salt is added. In addition, crustal element mass calculation is a bit strange, why calcium is mentioned twice and only part of other elements are corrected for oxygen,  the same is valid for phosphor. 

Line 281 – 284: Water absorbed in aerosol and adsorbed on the filter can be also part of unresolved mass.

Line 319 - Dust factor in El Alto, the authors mentioned that sampling site was surrounded with dusty surface. Although the sampling site was on the roof of the building, it cannot be excluded that sampling site is influenced by the local dust. It should be mentioned here.

Line 320 – The factors associated to secondary aerosol … It should be mentioned which factors. Probably sulphates, nitrates and MSA rich factors, but it should be specified.

Line 441 – metals such as Cu, Sn, Sn and Pb. Not sure if one of Sn should be Zn or Sb (Zn seems more probable)

Line 473 – citation F. Amato et al., 2011 should be Amato et al. 2011

Line 598 – “natural anthropogenic sources” should be “natural and anthropogenic sources”