Editor's Review:

Note 1: This is the Editor's (Patrick Chuang's) review.

Note 2: I reviewed the manuscript up to Line 242/Fig 2 or thereabouts. There seems to be a major error (or misunderstanding on my part) in Fig. 2 that needs to be resolved before the rest of the manuscript can be evaluated since this figure provides the basis for much of the reasoning for the simulated effects. Upon receipt of a response to my review and revised manuscript, I will continue my review.

Lines 46-49: How do you define condensation or evaporation efficiency? Is it a column-integrated value? Later on (line 262) you reference Pinsky et al. 2013 and Seiki and Nakajima 2014, but in neither paper does the word “efficiency” actually appear. Do these papers use a different term? Are these really appropriate references if the term “condensation efficiency” does not appear? Koren et al. 2014 mentions efficiency once but does not define the term. Please define this term carefully at some point, as you use it in various parts of the manuscript.

154-157: This doesn't very well describe how this manuscript is different from the numerous previous studies. Please make this clear.

170-171: Only 2 hrs for spin-up? Lee et al. 2012 (in their Fig. 1; note: I was a co-author on this paper) shows 4 to 6 hrs was necessary for their case. Are you sure these simulations are properly spun-up? Also, do the simulations include a diurnal cycle? Or is it fixed sun zenith angle? Or is it night time?

175: So what is the fixed shaped of the aerosol size distribution? “based on...” is not particularly enlightening.

178: Some of these values are really very small. Are they really applicable to the atmosphere? (Also see comment regarding line 200 about adding Nd to Fig 1).

184-185: I'm confused by the fact that the aerosol “maintains constant mixing ratio with height” but simultaneously “a prognostic eqn is solved for the aerosol mass”. If aerosol mass can appear because of, say, cloud drop evaporation, how can aerosol mass mixing ratio be constant with height? If aerosol size distributions are held constant, this is of course not physically realistic... what are the consequences for interpreting these results in the context of the real atmosphere?

189: Is Khain et al. 2000 a reference to Kohler theory, or a scheme to calculate activation? If the former, that seems like a poor choice.

200: Analyzing the data in this way supposes that there's no overall tendency in time. Is this true? If not, what is the tendency relative to the de-trended variability? How would the results change if a different time window (say, 8 hrs instead of 14) were analyzed?

200: Please add a panel showing mean drop concentration.

200: What is the number of clouds in any given simulation? Does it change with aerosol number? Perhaps add this as a separate panel to Fig. 1.

218: Choosing the maximum value of any parameter seems very prone to outliers. Could you instead plot the, say, 90th or 95th percentile cloud top height to tell the same story?

242: I don't understand Fig 2. Lee et al. 2012 shows values between NET condensation of clean and polluted simulations to be on the order of 0.1 to 1 g/kg/day, which is about 1e-6 to 1e-5 g/kg/s. You show (gross) values that are as much as 4 orders of magnitude larger. These are not compatible. I think something is very wrong with your plot. Second, why is this value so much larger at high altitude than at low altitude, esp for the 2000 cm-3 case. Cloud base is somewhere around 500 m (Fig 4), but the mean condensation rate is something like 2 orders of magnitude higher at altitude (2000 to 3000 m). For an adiabatic parcel, the condensation rate generally goes DOWN (albeit not very quickly) with altitude because of the exponential dependence of the saturation mixing ratio on T with the possible exception of the tens of meters near cloud base where there's substantial supersaturation. Also, the difference between the cond and evap tendencies is huge in this area for the 2000 cm-3 case. This suggests that LWC should increase greatly at these altitudes (since precip is absent). But Fig. 4 shows that LWC is roughly constant with height in this simulation. I don't think these two are compatible. Lastly, the 1 to 2 orders of magnitude difference between aerosol cases seems wrong. There can be differences (see again Lee et al. 2012) but I don't think they can be that big. By my reasoning, something is very wrong with this plot.

Fig 1: Panel B does not have units of kg/m2. Please include some “uncertainty bars” to indicate the time variability in all panels.

Dear authors,

I have read your revised manuscript, together with your detailed responses to the referee comments. There are still some very minor issues that should be addressed; see below. After that, I will be happy to accept the paper for publication in ACP.

There are still some language and grammar mistakes; I didn't list them here as they will be taken care of during copy editing.

Best regards,
Barbara Ervens

--------------------------------------------
l. 10: Change ‘properties’ to ‘loading’ as you do not explore effects of any other aerosol properties (composition, size)

l. 191: If the aerosol size distribution is only scaled up/down, the shape should be identical, not similar

l. 341: How does the study by Dagan et al., 2016, differ from the current one?

l. 374: either ‘last’ or ‘third’ seems redundant here

l. 381: a) There is no Figure 1F
b) I am confused (but this might be due to the missing figure): The rain rate is given in mm/day (e.g. fig. 1F). How does this translate into percentages?