|I acknowledge that the authors answered most of the questions raised in the review process. |
As you can see, most of the remainig issues are "minor". However, I really doubt your error spans on the intensive properties, see my remark below. Please come up with realistic values!
P 12, L 32: is it higher altitude or higher value?
P 15, L11 “h” in Athens missing
P 16, L 14 ff: I still think that the error of the IP are unrealistically small. For example, the authors state BEA(355 / 532) = 1.82+-0.0001
Let us assume that beta_aer at 532nm had zero error and beta_aer at 355nm had only 1% of error (and I think the authors agree that this is an absolute unrealistic low error estimation) this would translate into values of BAE between 1.7955 and 1.8445
As this is mainly a paper on the method the authors really should come up with a realistic error estimation.
P16, L 17 ff: you are using E NE etc for East, North East which is fine for me. But as you changed it elsewhere in the manuscript you may change it here
P16, L 22 please change: [the smallest particles] “are the most … backscattering” into “show the highest backscatter coefficient”
P19 L3-4: again SE Europe and NE Europe (which is fine for me but not consistent)
P21, L27 LRT event “mixed cases” (and the discussion before). Strictly speaking you had to show somehow that you really compare in your layers the mixture of 2 different BB events (but no additional source of aerosol!) As this is probably difficult to prove I recommend to be more cautious in your wording.
P22 L1 CE Europe is “Central East Europe”?
P22 L22 “with uncertainty analysis” cancel these 3 words? You took the data from Earlinet and rejected wrong entries. Which uncertainty analysis did you do beyond this?
P22, L25, Plural? Other possible sources (?)
Fig 9, sub-plot PDR: I don’t see the data
Fig 12: do the sub-plots belong to the same data set? I am wondering because in the oEUAF case in sub-plot LR532 vs LR355 the standard deviation in LR532 is from approx. 50 to 80 sr, while in sub-plot EAE vs LR532 the standard deviation is from 20 to 80 sr
p.4 L 10: I really recommend to cancel the sentence “Our approach …” You don’t need it. You describe, how you define a layer. Other authors may have done slightly differently. Honestly, I did not check all your quotes, but Veselovskii et al. 2018 did not describe, how they defined their layers. Even if they did you could only draw the conclusion that your layer detection is in line with Veselovskii if you had taken their beta profiles and derive layers with your code which are then almost identical to the region in which Veselovskii present their Angstroem exponents in their figure 8.