Review of “Vertical distribution of ice optical and microphysical properties in Arctic low-level mixed-phase clouds during ACLOUD” by Järvinen et al.

Recommendation: Accept with revisions.

This paper presents some interesting results on the vertical profiles of mixed-phase clouds measured during the ACLOUD campaign, finding results similar to those found from previous studies of this cloud genre in that liquid dominates the upper parts of clouds with ice phase observed in greater amounts in the lower parts of the cloud. This study is worth publication because it presents data from a new location and presents additional data about the measured asymmetry parameters that have not typically been measured during prior campaigns. They also present radiative transfer simulations to claim that the ice phase only has a minor effect on cloud transmissivity and albedo. The paper is well written, technically sound, and easy to follow. Thus, I am recommending that the paper be published, but there are some points that need to be clarified before the paper is published.

I have some uncertainty in how the flight profiles are used to construct the vertical profiles of cloud properties as a function of normalized altitude. It is stated that “vertical profiles were measured in-situ either by flying a double-triangle pattern, where altitude was changed at the outer vertices, or by flying stacked horizontal legs….[with sampling lasting] between 7 to 10 minutes.” However, it is unclear to me how many different altitudes were sampled during the stacked horizontal legs. Can you list the number of levels that were flown in Table 1? In Figure 3 and 4 can the specific altitudes used in the construction of these plots be indicated? Many other studies examining cloud profiles as a function of normalized altitude are based on measurements obtained during ramped ascents and descents so that 1) observations are obtained at multiple altitudes within the same cloud; and 2) there is clear knowledge of base and top due to when the cloud breaks out of cloud.  How many altitudes were sampled during a typical flight? How much did cloud base and top vary during the flights, because that would seem to be a major uncertainty in the calculation of the normalized altitude? How were the base and top determined? What is their uncertainty and how does that translate into an uncertainty in Zn?

A measure of uncertainty, the standard deviation, is provided in the plots of microphysical quantities versus normalized diameter. How does the uncertainty in the calculated microphysical quantities compare against this standard deviation? A good measure of uncertainty due to statistical counting is provided as proportional to the square root of the number of counts of particles that were counted in each of the size bins. Given that PHIPS has poorer sampling statistics than other microphysical probes, are there adequate statistics in 10 s to calculate N(D)? This should be quantified. Similarly, a better description of the uncertainty in the calculated LWC and IWC should be provided. Since the calculated IWC from the PHIPS and CIP is based on Brown and Francis (1995), there could be very large errors depending on the mixtures of habits that are actually present at a singe time. Were there any sensitivity studies conducted where either the Baker/Lawson (2004) technique or application of habit-dependent mass-dimensional relationships explored? And, finally, what are the uncertainties in the calculated vertically integrated water paths? The calculation of a water path from measurements that are collected on ramped ascents/descents is more straightforward as the Delta Z to multiple the measured IWC is straightforward. However, if the vertical profile is computed by combining data from several horizontal legs at specific altitudes, the computation is not quite as straightforward.  How much variability in the water paths were noted?  Finally, it is good that the specific definition to compute re is noted. It might be good to reference McFarquhar and Heymsfield (1998) to emphasize that there are several different definitions that can be used. Also, I think it is more typical to calculate the effective radii separately for the water droplets (rew) and ice crystals (rei).  Why was a single effective radius corresponding to both species used?

For the habit classification, can you add a figure that shows examples of each of the habits used in the classification scheme? That provides important observational evidence to assess how well the imaged crystals match these idealized crystals.

In general, there should be greater quantification of the uncertainties.

The section 6 on the case study of radiative transfer in the single-layer cloud system is a nice addition to the paper. However, its briefness and lack of detailed discussion makes it seem like this section was almost an afterthought to the paper. I think that this section should be given more prominence in the paper.

Please see attached file.