GENERAL COMMENTS

The article aims at describing the improvements of the cloud and aerosol database of 

SCIATRAN software package. A precise description of the optical properties of clouds

and aerosol is an important part of a radiative transfer package. The article gives a good

overview of this database. The new features are well distinguished. The article presents a scientific interest and is worth being published in GMD. 

However, some points could be improved: 

- To avoid the need to refer to external articles, it would have been preferable to give some

more details on the components of CAS which have not been updated. Of course, if wouldn't 

be necessary to delve into the details, since they are available in the references.

- It would have been interesting to have more comparisons of modelizations done

with the new version of SCIATRAN/CAS with former versions and with other modelizations or measurements.

- Even if the acronyms are properly defined throughout the text, it seems to me that it would

be more easy for the reader if they  were gathered in an appendix.

SPECIFIC COMMENTS

Lines 41 to 52 about the use of SCIATRAN do not seem necessary to me, as well. The same applies

to figure 1.

Lines 94 and 95, it is said that the calculation can be performed 175.44 nm to 40 micrometer.

Lines 99 and 100, it is mentioned that aerosol and cloud scattering can be taken into account.

However, line 180 and 181, it is said that database can be used "at least"  between 225 nm

and 2.5 micrometer. Is seems to me that there is a contradiction there: is it possible

to model aerosols and clouds between 175.44 and 225 nm, and between 2.5 and 40 micrometer 

or not ? This point should be made clear in the article.

 

Line 111, it would be interesting to specify with jacobian can be computed (derivative

of radiances with  respect to which variables ?)

Line 118, I understand that the goal of the paper is not to delve into details

which have been extensively covered in other papers, but the list of the solvers

which can be used should be precised.

Line 187: by "optical characteristics" do you mean refractive indexes or (SSA, extinction coefficient and phase function ?) 

Line 205: it could be interesting to give references for OPAC version 3. This could also be

done in section 4.1.1, event if the details have been given in [Rozanov, 2014]

Section 4.1.1 : it could be interesting to give references for OPAC version 3. This could also be

done in line 205,  even if the details have been given in [Rozanov, 2014]

Line 205-212: I understand that the refractive indexes and size distributions from

OPAC are used to compute the topical properties of  an aerosol type. Why not use

directly optical properties (i.e. SSA, extinction coefficients and phase function) from OPAC ?

Line 218-225: it is mentioned above that the optical properties are computed from the refractive

indexes. Here, it is said that they can be set for different altitudes. Does it mean

that optical properties can either be directly specified or computed from refractive indexes ?

Even if OPAC v3 is not the focus of this paper, the way optical properties are defined

(specified or computed from refractive indexes) should be precised.

Line 243: the same applies to OPAC aerosols, where optical properties can be specified.

Line 265: logarithms/exponentials of dimensioned quantities units are used. The formula should

be modified so that no dimensioned quantity is used as an argument to  ln or exp.

Line 286 Same remark. Moreover, the dimension of both sides should be the same.

Line 296: it is said  that the user can define the shape of the aerosol number density vertical

  distribution. What is the default shape  (i.e. if the users does not define it)  ?

Paragraph  4.1.4 : why a "dust aerosol type" is mentioned, whereas, in this paragraph,

other aerosol types are mentioned (OPAC, XBAER-OC, MODIS-DT), which are considered

in other paragraphs ? This paragraph should focus on the Dubovik dust aerosol type,

which is not described elsewhere.  

Line 389: for the "size bins" tracer: is it an aerosol with a defined size ?

          This aerosol type should be precised.

Line 406: mixing ratio:  with respect to dry air ? 

Line 442: is soot accounted for by adding an extra aerosol type ? Or by 

          considering a layered sphere with both water (liquid or solid) and soot ?  

Line 490: in [Yang, 2013], there are 11 habits, not 9. 

Line 497:  It would be nice to have more details

          about the implementation, even if the full description is available in [Pohl 2020]

Line 565 and 580: without going into the details of Malinka's articles, it would be nice

    to recall briefly the principle of stereological approach and the definition of chord length.

Lines 588 and after: some details would be given about the RPV and Kernel-based Ross-Li.

    Not a full description, but something basic which would avoid the need to look in the references.

Line 775: "newly implemented in SCIATRAN aerosol types" => "aerosol types newly implemented

           in SCIATRAN". 

Figure 3b: it is difficult to see the difference between  the  forward peaks of the different aerosols, which are mentioned in lines 800-810. Perhaps adding another figure for the phase

  function, zooming on 0 degree scattering angle.

Caption of figure 3: "newly implemented in SCIATRAN aerosol types" => "aerosol types newly implemented in SCIATRAN". 

Figure 4: it would be nice to have another figure, zooming on 0 degrees, to better

          see the forward peak.

Line 888: for S+R+A, the agreement with MODIS is not so good above 900 nm , if we look

          at fig 5.

Line 904 and bottom of figure 5:  It would be more clear to plot the total TOA reflectance 

         and the contribution of each aerosol, instead of the difference.

Line 927: could you give some examples of such satellite instruments ?

Figure 6: it is not so easy to compare SCIATRAN BRDF to RTLSRS and FASMAR BRDF.

          Another figure, with the same kind of graphs, but with the difference

          SCIATRAN - RTLSRS and SCIATRAN - FASMAR would allow to see more clearly

          the differences. 

Lines 958 -> 968. 

     It is nice to see that SCIATRAN can well reproduce the measurements by choosing

     the parameters that minimize the residual. What would be interesting

     is to give the experimental values of  snow layer optical thickness

     and (if available) mean chord and DOM absorption. The comparison of these

     experimental values with the parameters would be interesting.

Lines 996 -> 995

     Same remark: It  would be interesting to give the experimental values of ice thickness, water depth and (if available) transport scattering coefficient.

TECHNICAL CORRECTIONS

Line 311: typo: wavelegths => wavelengths

Line 855: responce => response

Line 924 show => snow

Line 993: "the optical" => "an optical"

Line 994: "the geometrical" => "a geometrical", "the top height" => "a top height"

The paper describes the latest updates and additions to the already extensive suite of aerosol, cloud and surface reflectance databases implemented in SCIATRAN. The new databases are adopted from mostly recent work of, and published in the peer-reviewed literature by, a number of researchers. The paper is fairly well written and accessible. It should provide a useful reference to current and future users of SCIATRAN.

I understand the goal is to describe the updated databases, but I think a high-level description of the state of SCIATRAN would be useful in addition to the references that are already provided. This should include the radiative transfer solver(s) implemented, databases that existed before the recent update, as well as a reasoning for replacing them, if any. It isn’t entirely clear to me whether the new aerosol databases are in addition to the "old" aerosol types or the they replace the "old" ones.

The databases implemented in SCITRAN is now quite extensive. While reading the manuscript I was hoping to find some sort of guidance or recommendation as to which database should be used under various situations. I don’t mean the trivial scenarios like land aerosol models over land and oceanic aerosol models over ocean, or water and ice clouds depending on what the user want to simulate. I am thinking of especially of the bidirectional surface reflectance databases. Which one would be the most appropriate to use when all inputs required for the models are available?

SPECIFIC COMMENTS:

Lines 113-14: Are the fluxes only for selected wavelengths,  can the user request spectrally integrated quantities too?

Lines: 126-128: I assume this means the gas absorption calculations use the HITRAN 2020 database. I understand the discussion of it is out of scope, but please at least provide a reference.

Lines 259-262: It is said that the continental aerosol type is not included in the SCIATRAN database because  according to the authors’ investigation it is not used in the XBAER algorithm for MERIS. It may be fine since the continental aerosol type is generally regarded “old” by some researchers but I’d think there should be a more substantial justification for excluding it beyond the fact that the XBAER algorithm does  not use it for a specific instrument.

Lines: 445-449: In the previous version, the effective radius of water droplets was between 4 and 20 microns. So, do you mean the database was extended to include the range 2 - 4 microns, or that the "old" database was replaced with a new one having effective radius between 2 and 40 microns?

Table 3 caption: I think the correct URL is https://darktarget.gsfc.nasa.gov/atbd/ocean-algorithm

Line 459: Is there a reference for the relationship between mode radius and effective radius?

At places where implementation of databases is discussed the choice of the specific way a particular database  was implemented could be discussed. For example,  re the ice particle habit (lines 520-525), has the linear interpolation been suggested by the database developer? I am not saying that a linear interpolation is not adequate; I am only suggesting that when it is appropriate the choice should be justified as much as possible, especially in those cases when multiple choices are possible.

Line 577: Are "typical values" of the mean chord length and optical thickness of snow provided in the code or database for a broad category of snow, say fresh and old snow? I assume users could read the Malinka (2014) paper, but still, having such values listed in the paper would be useful for casual users who just want to see the sensitivity to these parameters without prescribing unrealistic values.

Line 808: The Dubovik and OPAC 4.0 dust models also have  relatively large differences in the backward scattering direction (scattering angle 150-180 degrees).

Lines 903-911: I am not sure I understand this part. What is being shown on the vertical axis in the lower panel of Fig. 5? Is it the reflectance from a single aerosol component divided by the reflectance from all 5 components, or the reflectance of 4 components (excluding one aerosol type) divided by the reflectance from all 5 aerosol types?

Line 2021: I assume the authors mean user-friendly interfaces. Is it really necessary to say it since it has not been shown in the paper?

TECHNICAL CORRECTIONS (could be more):

Line 145: “the explore of” should read “ the exploration of”.

Line 152: “of Young database” should read “the Young database”.

Lines 191-192: “... optical properties those six components” should read “... optical properties of those six components”.

Line 249: “Virible” should read “Visible”.

Line 311: “wavelegths” should read “wavelengths”.

Line 351: “Many investigations has …” should read “Many investigations have …”

Line 548: “Combing” should read “Combining”.

Line 588: Perhaps “build-in” should read “built-in”?

Lines 773-774: Would read better “... the new aerosol types implemented in SCIATRAN” or  “... the aerosol types newly implemented in SCIATRAN”.

Line 807: “paramterization” should read “parameterization”.

Figure 3 caption: See comment above for lines 773-774.

Line 824: “show layer” should read “snow layer”.

Lines 828-829: Instead of “the scattering processes plays very important role” the authors could write “the scattering process plays a very important role” or “scattering processes play a very important role”.

Lines 855-856: “response” should read “response”; “instrument” should read “instrument”.

Line 859-860: “to the time of writing” should read “at the time of writing”.

Line 927: “channels” should read “channels”.

Lines 983-984: “with the optical thickness of 20” should read “with an optical thickness of 20”; “the geometrical thickness” should read “a geometrical thickness”; “and the top height” should read “and a top height”.