Dear Colleagues:
I have reviewed your manuscript following revision. I am satisfied with your response to the more technical aspects of Reviewer 1’s comments but still have a number of concerns that need to be addressed before I can recommend publication. Some might be easier than others to clear up. I have checked the "minor revisions box" in an attempt to keep this moving efficiently and in the hope that you can clear these matters up quickly. I do, however regard them as more than minor and request careful consideration.

Graham

1) By using PCI, you now introduce a rather strange concept which is the interaction between CO and clouds, which to my knowledge is not a real interaction. What is real is that CO correlates with aerosol and so can be used as a proxy for aerosol or CCN, especially in the absence of wet scavenging. When wet scavenging exists, it is not clear what the CO-cloud relationship means.

In a number of places the text becomes very awkward.
For example, in the abstract:

“For a given set of meteorological conditions, we find that the liquid water path of arctic clouds does not respond strongly to pollution”

Why should LWP respond to CO? What is the physical mechanism?

or,
“If, however, aerosols are scavenged during long-range transport, then the cloud sensitivity to pollution is low even if the sensitivity to aerosols remains high.”

what do you mean by cloud sensitivity to pollution? I think you mean to the particles for which pollution (CO) is a proxy, but this sounds very odd. And if the aerosols are scavenged, how can the sensitivity of the clouds to aerosols be high.

2) I am very confused by why the PCI_re is often quite different from the PCI_tau. The authors seem to look for physical explanations, but if both are cover the same LWP range, and both use the same CO values, I do not see how large differences can follow, and any differences will likely be a result of methodology.

If you are not stratifying by LWP, then please make this crystal clear (see questions below on Eqns). But since the LWP range is the same for both tau and re, why should there be such large differences in the two PCI values? This should also be true when you stratify by LTS or q since you would still include the same LWP values.

3) This leads to my biggest concern about this study is that emphasis is placed on quantification whereas there exists a body of literature that points to ACI values being highly sensitive to methodology, data sources, aggregation scale, etc.

While I understand that the aggregation scale is consistent here, attempts to compare these values with other studies would be unwise. And I am concerned about sensitivity to methodology.

4) The question of representativeness of CO in FLEXPART is raised in the paper but one does have to be concerned about drawing conclusions. How confident are you that similar conclusions could be drawn using real CO measurements (granted, they are not available) or aerosol/CCN measurements in the non-precipitating cases. In my experience with these analyses, choice of data and methodology have a very large influence on the outcome.

5) On pg. 5, you state: “The goal of this study is to use satellite, tracer transport model, and meteorological data sets to determine the effects of
pollution on cloud microphysics due only to pollution itself and not to the meteorological state.”

Again, pollution is suggested as influencing clouds. In addition, the end of the sentence implies that you are not interested in the role of meteorology, but that can’t be the case since you bin by q and LTS.

6) Eq. (3) should be formulated in terms of aerosol or CCN for it to be of any interest. As it stands it is quite obvious.

7) Eqns 4) and 5): are these partial derivatives at constant LWP or full derivatives. See question 2.

8) Eq.6: What is being held constant for these partial derivatives?

9) Eq. 7 seems inconsistent with Eqns 4 and 5 (partial vs. full derivatives).

10) The methodology of placing less weight on outliers is an example of how methodology can influence results.

11) Pg. 9, line 12, I was looking for some explanation but never got one.

12) Pg. 9, line 19: is humidity a proxy of LWP? How well do they correlate?

13) Pg. 11, line 12: Higher LTS might mean clouds are more adiabatic, which would be consistent with Kim et al. 2008.

14) My personal opinion is that the Conclusions become too far reaching if one considers the uncertainties in quantification of the PCI or ACI values.

15) I assume black is the highest density in Fig. 3. How does one make the decision to place less weight on the outliers? If one places too little weight on them then the slope is no longer defined (e.g., a fit through the black and green points might give -ve PCI_re.)

16) Pg. 2, line 26: is orthogonal the correct word?

17) The tau, LWP, re relationship used is for a homogeneous cloud. Please state so.

18) “15% of total space”: Is this area, number of pts?

19) In addition to Tietze et al. the use of CO as a CCN proxy has also been applied by Berg et al. (2011, GRL) and Lance et al. (2012, ACP).

Dear Authors,
I probably have a philosophically different viewpoint on this manuscript but wont let that stand in the way of publication.
Best ,
Graham