Dear authors,

I have now received three reviews of your revised submission. All reviewers feel that additional, and in some cases significant, further revisions are required before the paper can be considered for publication in AMT. Two reviewers have particular queries about the calculations for the cirrus cloud parts of the manuscript. I would encourage you to consider submitting a further revised version, with attention to the below points, which I would then further discuss with referees. Please let me know if you have any questions.

-----
Report #1:
General

The revised version is not ready for publication. The quality is rather low, and if this would be the first round of review, I would vote for rejection. I am a bit disappointed about the answers of the first author (I speculate the co-authors were never involved in the review). I have the feeling after reading the responses, the first author is not willing to invest more work, needed to become a good science article. This is disappointing.

Zenith pointing of the laser beam (this is obviously the case for the Italian lidar, rather than the required off-zenith pointing) and multiple scattering can introduce severe uncertainties in cirrus extinction profiling with lidar so that all the subsequent radiative transfer computations are useless when the multiple scattering effect is not considered and, independent of that, very questionable when the lidar is pointing to the zenith.

As a consequence, the cirrus studies in this paper are practically useless. More details are given below.

The paper is clearly not state-of the-art from the basic and fundamental lidar point of view.

At least, major revisions are required.

Details ( I refer to the manuscript in the review letter, with all the red sections…):

L27: … with respect to cirrus… must be added because in the case of liquid-water clouds the single-scattering lidar ratio is 18 sr, and when considering multiple scattering the apparent one is around 8-12 sr.

L45-47: strange argumentation…. to link geometrical properties of aerosols and clouds with passive remote sensing…. which have practically no profiling capability (only the model behind all the column integrated measurements may allow the retrieval of geometrical properties, but then with large error bars).

L68-70: I would remove these lines and Madonna references. Almost all experts in this field do not believe that the measurements and conclusions are ok.

L90: I do not know why we have equations in the introduction, I would move them to section 2 (2a: lidar methods). And why do we need the simple lidar equation? And why is there the overlap term missing (O(r))? The overlap profile affects MPL observations up to 4-5 km height. That should be mentioned.

In such a Section 2a, I would present the Fernald equation and the Raman lidar equation. They are used later on, and in this new Section 2a, all problems and differences could be explained in detail…, which is not possible (or should not be given) in the introduction.

L99: There are meanwhile so many lidar ratio papers (real measurements!)…. Please provide some references, e.g. Muller 2007, Gross, Tellus 2009, ACP 2013, 2015? , Ferrare, JGR…, Sakai, JGR… Veselovskii 2016 and many many others . Ackermann (1998) is just a very simple simulation study.

L105: Raman roto-vibrational … bad wording…

L133: The impact of these differences on end-user applications have never been evaluated…
Please be more precise, what do you mean with end-user applications?

L195: Now the RFOV is mentioned. The RFOV is 1.5 mrad. This means the multiple scattering effect is rather large in case of clouds. The measured apparent cirrus extinction coefficient is roughly a factor of 2 lower than the desired single-scattering extinction coefficient (throughout the cirrus layer). The single-scattering extinction coefficient is the basic input in radiative transfer equations. Multiple scattering depends on cirrus crystal microphysics (size, shape, amount…).

And such a huge multiple scattering effect is not considered in the paper? … the authors tell us: they ignore it!

This is not acceptable! These effects must be considered. There is no way to circumvent this problem.

L202: I complained last time with good reasons: There is no information on the laser beam pointing (zenith or off zenith).

Even in the revised version there is no hint on beam pointing. I speculate the laser beams were pointed to the zenith. This in turn means that the cirrus backscatter profiles are probably strongly affected by specular reflection. In case of zenith pointing, the backscatter coefficients (from which the extinction coefficients are estimated when using the Fernald method) are an order of magnitude larger in the case of aligned, falling crystals than the true backscatter values (obtained at off zenith pointing, 3-5 degrees off the zenith is sufficient). The alignment effect depends on particle size (only >100 mikrometer particles are able to be horizontally oriented), thus the effect can vary from height to height. So, cirrus backscatter and extinction profiles from zenith pointing elastic backscatter lidars are highly corrupt and uncertain, and to my opinion these Fernald extinction profiles are useless for further use in radiative transfer computations.

Figure 1: The aerosol measurements are ok. Multiple scattering effects can be neglected, as well as beam pointing effects.

Why are there different descriptions (1b, left: Iterative, Iterative, Raman, 1b, right: Raman, Elastic, Elastic) ?

X-axis and y-axis text and numbers of Fig 1a are much too small!

All plot frames should have exactly the same size (at least for 1a and for 1b)!

Figure 2: Same concerning x-axis and y-axis numbers and text, sizes of the frames of the different plots as in Fig.1.

Figure 2b, left: extinction values of 0.1, 0.2 …., 0.4 m-1 are wrong…

A clear sign of a strong multiple scattering effect is the vertical shift of the Raman extinction profile (by 500m) compared to the Fernald extinction profiles. The true cirrus top is at 13 km, and the increasing and decreasing multiple scattering effects at cloud base and cloud top, respectively, leads to the wrong Raman extinction profile, shifted towards greater heights. This was frequently discussed on lidar conferences. Dave Whiteman (NASA) often pointed to this feature.

To be clear, all the three profiles suffer (in addition) from the overall multiple scattering effect reducing the measured COD by about 40-60% compared to the true cirrus optical depth.

Figure 3: All plot frames should have the same size (and should be well arranged above each other), all x-axis and y-axis text should use the same letter font, and PT

The caption should briefly explain what Raman, Full Res., and Iterative means. Full Res is obviously Iterative (60 m), and Iterative denotes Iterative (360 m)? Please harmonize this with the other figures…

Figure 4: The same concerning plot frames, and explanations of Raman, Full. Res. and Iterative. X-axis and y-axis text and numbers need to be harmonized.

To repeat: The cirrus results in Figure 4 are highly questionable. To my opinion, they are useless because of the lidar-related problems with zenith pointing and multiple scattering.

So, what is then at the end the message of the paper? I do not see a reason to support publication.

I leave it open to the editor (based on the other reviewer opinions) to accept it or not.

-----
Report #2:
General comments

The main objective of this revised paper, novelty entitled “Impact of varying lidar measurement and data processing techniques in evaluating cloud and aerosol direct radiative effects” is still to quantify inconsistences in aerosol (two case this time in the revised paper : dust aerosol and biomass burning event) and cirrus cloud (two cases in the revised paper : a thin and an opaque cirrus) radiative forcing at Top Of the Atmosphere and at surface due to two different ground lidar techniques (elastic and Raman lidar, i.e. the Multi-wavelengh System for Aerosols (MUSA) Lidar (Madonna et al., 2011)) and/or data processing (i.e. effect lidar measurement with different vertical resolution together with smoothing techniques).
The revised paper is based on the same approach and analysis technics as the initial paper, that is a good point.
Quality of references of the revised paper are now better and the title is now appropriate.
This revised paper still address relevant scientific topics within the scope of AMT.

Nevertheless, the revised paper shows dramatic drawback and inconsistency with the study based on cirrus cloud. Figures a not all clear. For a revised paper, it is quite annoying. This revised paper needs major revision.
Major remarks:
1) Line 689 (i.e. figure 2)
Right panels show the opaque cirrus. On the right top panel, the base and top altitude are between 6 km and 8 km, respectively, whereas the retrieved extinction coefficient profile is between 11 and 13.5 km.
Where is the problem?
2) Line 689((i.e. figure 2)
Optical depth is defined as extinction times distance. Bases on this simple formula and on information provided by the figure 2, optical depth tau of thin cirrus, is, roughly, tau=0.1(m-1)*100(m) = 10 and the optical depth of the opaque cirrus, is, tau=5e-5(m-1)*2000 (m)= 0.1 !
How can we qualify a thin cirrus with an optical depth of 10 and an opaque cirrus with an optical depth of 0.1 ?
3) Line 689((i.e. figure 2)
The comparison of the direct radiaitve effect of theses both cirrus is not suitable. Indeed, direct radiative effect is function of optical depth, effective radius (or asymmetric parameter and single scattering albedo) and of temperature (i.e. of the mean altitude of the cirrus but also of the cirrus cloud base and top altitude). In order to interpret differences of direct radiaitve effect between an optically thin and opaque cirrus, geometrical thickness must be the same, as well as the cloud base and cloud top altitude.
Moreover, why two different vertical resolution (420 m for the thin cirrus and 780 m for the opaque cirrus?
Based on my major comments, it is impossible to evaluate the conclusions of this revised work. This paper need major revision.


Specific comments
line 31-32 : This sentence is not clear. Please rephrase
line 48-52 : acronyms are missing
line 122 : please define lamba_0
line 144 : I think “…computed using…” is better than “… retrieved using…”
line 166 : the FLG model needs the phase function ? How is computed the phase function?
line 192 : please define a.s.l.
Figure 1 : - please define Tito(PZ)
- All the “no-information” (above 7 km, the blue color) is useless. Please rescale the figures.
- Please use the same color for the legends on right and left ( b)
- What is the difference between iterative and elastic (on the legend of figure 1b)?
- Why the range corrected signal is at 1065 nm whereas the retrieved extinction is at
Figure 2 : same global remarks as figure 1 but also :
- Why the Raman vertical profile is so different compared to the Klett profile for the opaque cirrus whereas it doesn’t for the thin cirrus ?
Figure 3 and 4 : please define Raman, Full Res. and iterative and put coherency with “elastic” and “Klett”

-----
Report #3:
The authors have addressed most of the concerns I had before recommending the article for its publication in Atmospheric Measurement Techniques. Personally, I think that the new title proposed is appropriate and that the new study cases for aerosol and cirrus clouds show more convincing results. But I have some minor concern I would like authors address before recommending the final publication:
- Although the authors responded well to my concern about the effects of aerosol microphysical properties in aersol radiative forcing computations, I miss such a paragraph in the revised manuscript.
- Authors refer in the abstract and in the text to aerosol and clouds optical and geometrical properties. Please, replace by 'aerosol and clouds optical and microphysical properties'.
- Although HSRL technique is not used in operational lidar networks such as EARLINET, their potential can not be ignored. Please refer this in the text and add appropriate references.
- The statement about the upcoming NASA Aerosol-Clouds-Ecosystems mission is not in the correct place. It is very important to reference such mission, and I recommend to move the sentence at the end of line 141. Also, please update reference to Whiteman et al., 2018: Whiteman, D.N., Pérez-Ramírez, D., Veselovskii, I., Colarco, P., Buchard, V. (2018) Simulations of spaceborne multiwavelength lidar measurements and retrievals of aerosol microphysics. Journal of Quantitative Spectroscopy and Radiative Transfer, 205, 27-39
- Please, define each term of equation 4.
- Axis of Figure 1 and Figure 2 are difficult to read

Best wishes,

Andrew

Dear authors,

Thank you for submitting your revision. After further consultation with one of the reviewers of previous versions, I am happy to accept your manuscript for publication in AMT.

Best wishes,

Andrew