Dear authors,

Thanks for the revised manuscript and addressing the reviewer's points. Reviewer 2 provided a re-review as attached below. Please consider point 1 from reviewer 2 about the hygroscopic growth of sea salt, revise your manuscript accordingly and resubmit it.

Best regards,
Ulrike

----

Report from reviewer 2:

I thank the authors for their careful and detailed responses to my comments. I am happy to see you added a short discussion of background CDNC, the importance of which often tends to get overlooked.

1) The addition of Figure 1 and the surface-area discussion is useful, but if the distributions shown in Fig. 1 represent dry sea salt, then hygroscopic growth will alter the light scattering surface areas, possibly increasing the GISS surface area more than the others because of the differences in the mass distributions. Your discussion of surface area on pages 20-21 may need some modification.

I offer a couple of additional comments for future consideration:

2) Considering the large masses, the water uptake in the GISS model will be dominated by particles in the range of 1-10 um diameter. Salt particles that are more than one micron in size are unlikely to activate due to the large amounts of water required, and the parametrizations (even in a “fully prognostic” scheme) are unlikely to treat that properly. The bottom line is that you may end up with something that may appear to be a cloud, but strictly-speaking it will be just a strong haze (direct effect). Whether cloud droplets form on the smaller particles will depend on the strength of the water uptake by the larger salt particles, but I suspect that most of what would be created from the GISS distribution is just haze. What does the considerable injection of large sea salt particles in the GISS case do to a cloud that is trying to form? A true parcel model could be helpful.

3) It would be nice to see the models include effects of the additional energy input necessary to produce and sustain these injections. As the other reviewer suggested, these injections may require substantial amounts of energy. On that note, the other reviewer suggested that the sizes of Aitken particles may be most efficient for modification by the indirect effect. That will be true if there are very few pre-existing larger particles (e.g. Leaitch et al., ACP, 2016), but such a scenario could be infrequent.

Dear Dr. Ahlm,

I'm happy to accept your revised manuscript as it is for publication in ACP.

Best regards,
Ulrike Lohmann