General
The revised version is significantly better than the submitted one. However, there are still many points that need to be improved and clarified. The discussion and interpretation of the results is based on measured range-corrected signals (attenuated backscatter) and volume depolarization ratio at 355nm (VDR355), rather than on particle backscatter coefficients (BSC355) and particle depolarization ratios (PDR355). These quantities BSC355 and PDR355 would allow a much better and more clear interpretation of the observations.
The use of VDR355 means that any variation in VDR355 may be partly related to (a) changes in the particle-to-molecular backscatter ratio and (b) changes in the ratio of non-depolarizing droplets to depolarizing ice crystals. This ambiguity must be considered in the entire discussion. This ist not the case in the present version of the paper. This has to be improved.
As can be seen from my quite long list of remaining comments and questions, we need another round of revision.
Details:
P1 l20: the authors write: …the depolarization ratio of falling hydrometeors increases from liquid-water values to the ice/snow values… My question: How do you know that this is related to the changing ratio of droplet-to-crystal number concentration? As mentioned above, VDR355 is used and depends on Rayleigh depolarization (causes a depolarization ratio of about 0.01). The lower BSC355 (or attenuated backscatter), the lower VDR355. And vice versa, the higher BSC355, the higher VDR355. This can be simply caused by the decreasing impact of Rayleigh backscattering on VDR355 with increasing BSC355.
So please keep that in mind and improve the text accordingly.
P2, l54-55: The authors write: ..revealed the detailed vertical structures of falling mixed-phase virga…. Again, with VDR355, you are not able to give clear answers. Only with PDR355 you would be able to do that.
By the way, what is a mixed phase virga? A virga consisting of an external mixture of droplets and ice crystals? To my understanding, all virga above the height of the 0°C-level consist of ice crystals. There are no cloud droplets, they are too small to fall out of the cloud deck. Again, the VDR355 probably changes because of the changing Rayleigh backscattering impact.
P3, l83: please include after ….. is transmitted vertically into the atmosphere … (to the zenith).
P4, L107-110: How do you know about the mixture about droplets and ice crystals? The VDR355 cannot be used to clarify this due to the reasons discussed above. PDR355 would do a better job here.
‘Mixed phase’ hydrometeors … Again: What do you mean? … Internally mixed particles consisting of ice and liquid water? It is certainly an external mixture of droplets and crystals… Again, to my opinion there is only ice in these virga at temperatures below 0°C.
P8, l216-226. This part of the discussion is confusing and speculative and must therefore be changed. You do not know at what height the ice crystals were nucleated! You speculate about the role of contact freezing! There is no observational hint in the literature that contact freezing plays any relevant role! Furthermore, you mention again mixed-phase stratiform precipitation at heights more than 1 km above the 0°C height level. Do you have any measurement that supports all this? As mentioned, VDR355 does not help! So, please remove all the speculative statements.
Here, my most important suggestion to improve the manuscript: On P11, lines 315-321 you provide a perfect explanation of the vertical structure of the cloud-virga system. Therefore: Why not starting the discussion of the lidar observations with Figure 6 (28 Dec, 0005-0007 LT), before discussing the more complex and less clear cases shown in Figure 4 (28 Dec, 0112-0115 LT) and then Figure 5 (28 Dec, 0112-0115 LT)?
P9, l269-277: Please keep in mind: The decrease of VDR355 below the VDR355 maximum at 600 m height can also be the result of the increasing influence of non-depolarizing boundary-layer aerosol particles. The backscatter coefficient (in your case the range-corrected signal) strongly increases with decreasing height, probably caused by strong aerosol pollution.
P10, l285: One should clearly and more often emphasize that the radiosonde temperatures were most probably measured in virga-free air. This is needed to avoid the impression that temperatures were up to 5°C in the virga with ice crystals. The temperatures in the virga were most probably always close to 0°C.
P10, l295-315: As mentioned, VDR355 is a function of the particle-to-molecular backscatter, and thus a function of the particle backscatter coefficient BSC355. With increasing BSC355, the difference between VDR355 and PDR355 decreases. In lines 308-310, you mention the inverse relationship between backscatter and VDR355. Exactly that prevents you to make clear statements about the presence and number concentrations of droplets and ice crystals. The better parameter would be PDR355. To obtain PDR355 from VDR355 you need, however, to calculate the height profile of BSC355 first. BSC355 and PDR355 would allow a much better discussion, instead of using the basic and just qualitative lidar quantities, VDR355 and range-corrected signals. This is not just state-of-the-art.
P11, l315-335: This is the best part of the entire discussion! Nevertheless, the detection of the liquid-water cloud deck does not mean that the ice crystals formed in this layer. We do not have any information about potential cloud seeding effects …. from above so that ice nucleation may have taken place at -25°C. Nobody knows.
P11, l329-330: Again, do not forget that VDR355 changes with BSC355 (or the particle-to-molecular backscatter). VDR355 is thus influenced by ice crystals, droplets, and Rayleigh molecules. There is a mixture of information from molecules (1% depolarization) and particle (droplets 0-5%, crystals around 40%). And the lower BSC355, the lower the VDR355. In the case of PDR355, you would be able to make much more clear statements about the presence of droplets and crystals, because the Rayleigh impact is removed.
P11, l346: Again, to my opinion, the falling hydrometeors in the first 100-200m of their descent are ice crystals. Variations in VDR355 are probably caused by the variations in the particle-to-molecule backscatter ratio.
P12, l350: The Hallet-Mossop effect (secondary ice formation, SIF) causes a rather huge increase (orders of magnitude) in ice crystal number concentration at temperatures between -5 and -8°C. To my understanding, SIF is very efficient in the mixed-phase clouds (in the main cloud body with dense populations of crystals and droplets). So, my question is: Is SIF also very strong in (ice-dominating) virga? Can you provide references that SIF occurs in virga as well?
P12, l350-373: Again, it would be helpful to have PDR355 instead of VDR355 in the discussion.
P12, l375: Figure 7 shows again, that you can observe and describe virga properties. But there is no way to say anything about the ice nucleation processes higher up. The cloud deck in which ice crystals were initially nucleated remains undetected.
P12, l380-395: Again, the discussion is based on VDR355…. Low and high values are mainly controlled by the particle-to-molecule backscatter ratio (or BSC355), and only if particle backscattering is very high, VDR355 comes close to the particle-related PDR355.
The same statements about BSC355 vs VDR355 relationship hold for the final section 3.2.2.
P15, l470-475: You did not clearly observe any cloud layer in which ice nucleated. The focus in this manuscript is on the discussion on virga, and nothing else.
P16, l485: You do not have any observation that clearly indicates that you measured a mixture of droplets and ice crystals here….. VDR355 does not allow such conclusions. PDR355 would allow that. To my opinion, the virga purely consist of ice crystals. |