|I thank the authors for considering my comments and I feel now that the paper is more in-scope regarding the discussion of potential new insights into atmospheric chemistry. I’m more happy for this to be published in ACP, however I do have some comments that I would like to see addressed before publication.|
While the authors discuss in more detail the reasoning behind their method to exclude the larger peaks, I do not think they pay it enough attention to it in the revised manuscript. It is not sufficient to simply demonstrate that a factorisation looks unsatisfactory; the authors should discuss the reasons why it behaves the way it does, and better justify their chosen remedy in this context.
I thank the authors for taking on board my comment exploring rotational ambiguity and the use of FPEAK to explore this is certainly informative in some ways. However, I could question the following concerning the approach:
1. If little difference is found across the range of FPEAK values, then this can simply mean that FPEAK has not been varied across a sufficient range, noting that the effect of varying this parameter will vary dataset to dataset.
2. The authors seem to basing their choice of FPEAK on the level of separation of mass spectral profiles. I would note that such an approach will tend to favour the positive values of FPEAK, however this does not necessarily make the factors more ‘real’ because one would expect a certain amount of overlap between the factors in mass spectral space. The authors should explain this better (the assumption that siloxane should only feature in one factor could be considered reasonable, but the reasoning is left largely implicit).
3. There is an explicit recommendation in Paatero (2002) that nonzero FPEAK values should not normally be used for environmental data, so using this as the ‘final’ factorisation needs to be justified in this context. To be completely clear, my recommendation in the first review was to merely explore (and hopefully quantify) the rotational ambiguity, not necessarily to choose a more optimal FPEAK value.
Line 194: How are the authors defining ‘significant’?
Line 195: A more explicit description of the criteria used to choose the number of factors is needed.
Line 493: Regarding the two routes to nitrate formation, the caveat should be included that just because they are contained in the same factor, it does not mean that the same molecules are being produced. With the current wording, a reader could be mislead into thinking this to be the case.
Line 517: The end of this section is rather unsatisfactory, leaving the reader wondering whether the observation is linked to chemistry, time of day, insolation, temperature, etc. The authors make speculations regarding day vs night and sunny days, so can these be tested? As a bare minimum, the authors should (1) rule out temperature as a controlling factor because this is strongly anticorrelated with humidity and is known to influence reaction rates, branching ratios, etc., and (2) attempt to see if nocturnal nitrate chemistry could be responsible.
Line 33: The comment “These findings highlight the need for further studies to delve into gas-phase atmospheric processes of monoterpenes and sesquiterpenes” verges on pointing out the obvious, as terpene chemistry has been a major topic of atmospheric chemistry for decades.
Line 61: Replace ‘revealed’ with ‘showed’
Line 64: Remove ‘state-of-the-art’. Given that two authors have a financial interest in the instrument, it is important that the language describing it remains objective.
Line 468 (and elsewhere): Replace “oxidations” with “oxidation”