Dear Evangelia and Spyros,

Many thanks for your revised manuscript and apologies for the slow processing. I wanted to go through the changes thoroughly as the manuscript had changed significantly from the first version.
In doing so I have identified a few additional minor points that should be addressed prior to publication, which I have listed below. Apart from this, the manuscript is in good shape for publication in ACP.

With best wishes,

Eiko


Scientific comments:

Line 221: Were the filters analysed for Na+ which could be used to estimate the seasalt sulphate that would not have been detected by the AMS?

Line 264: I am not convinced that the correlation coefficient would necessarily be poor if a significant proportion of the organics were contained in the submicron fraction. You should instead look at the quantitative comparison between organics in PM2.5 and those reported by the AMS, based on the C/OM ratio from the AMS.

Line 357: Please introduce the meaning and calculation of the angle theta or at least add reference Kostenidou et al. (2009). It’s only been used in a few papers and the reader may also not have a feel for what angle reflects close agreement. Has this ever been compared with the more typically applied comparison of spectra using the correlation coefficient? Maybe it is worth also adding R2 values to Table 2 also? Or at least re-iterating the relationship between theta and R? Given that R2 and theta are related, neither is a perfect descriptor for factor similarity as both are biased towards the main peaks. Section 3.3.4 may benefit from such caveat.

Factor names: I suggest you make it even clearer in the manuscript why you deviate from the standard description of the factors of LV-OOA/SV-OOA and HOA/COA. In particular, the diurnal pattern of HOA-2 looks very similar to that reported for (what was called) COA in other studies. The lunchtime peak in COA looks early in other studies also. You state that this is inconsistent with lunch time in Greece (Line 461). Could larger canteens and outside catering starting cooking earlier and dominate the lunchtime emissions? With the exception of COA from Barcelona, there is still marked similarity between your HOA-2 and any of the literature factors used in Table 2. Also, in the UK, COA has been attributed mainly to vegetable oils rather than meat cooking (Allan et al., 2010); not sure how this may relate to Greek conditions. There has always been a concern that what was reported as COA may include other (possibly traffic related) sources, but it is unclear why these would not be distinguishable and consistently lumped into one single factor given their different diurnal behaviour. In summary, I suggest you expand the discussion on HOA-2 and its relationship to COA factors (and temporal profiles) in the literature and review your decision not to call it COA.

HOA-1 / BC correlation: How important are emissions from two-stroke engines in Greek cities as a potential source of HOA-1 (without generating BC) and are there other possible sources of BC other than diesel trucks?
Line 377 and Table 1. Can you please indicate which correlations are significant? Maybe, in Table 1, you could show those R2 values that reflect statistically significant correlations (state P value) in bold.


Clarifications:

Line 145. Please provide a rationale. Was there another AMS operated in the city centre? If so why was this excluded?

Line 208. At this point you should spell out more explicitly that you are performing high resolution PMF as the references point to unit mass resolution analyses.

Line 216: Please clarify whether an average CE was used or whether this varied on a half-hourly basis.

Line 566: Could you make your reasoning more explicit here? Artifact for what? You are probably implying that organic S is not a good quantification of MSA aerosol, but you should spell this out.

Figure S11 and related text. If the disagreement between filter and AMS organics is due to uncertainties in the CE (as you suggest), why does this not affect the comparison for sulphate?


Technical corrections:

Line 102: characterization of OA sources

Line 115: Maybe this could be rewritten: “… in the dark in less than a day …” does not read quite right as it is only dark for less than half a day in summer.

Line 213: Suggestion: “was significantly higher compared with most other studies”

Line 227: “at the site.”

Line 315: redundant “that”.

Line 395: Add Allan et al. (2010) as the first study that identified COA in ambient air?

Figs. S13, S14, S25 and S26: please set offsets for vertical axes to zero.
Line 532. “Organic contributions in Patras and Athens were around”


References:

Allan, J. D., Williams, P. I., Morgan, W. T., Martin, C. L., Flynn, M. J., Lee, J., Nemitz, E., Phillips, G. J., Gallagher, M. W., and Coe, H.: Contributions from transport, solid fuel burning and cooking to primary organic aerosols in two UK cities, Atmospheric Chemistry and Physics, 10, 647-668, 2010.

Dear Evangelia and Spyros,

Many thanks for the additional amendments and clarifications.
The paper is now in good shape for publication in ACP.

With best regards,
Eiko.