General comments:

The authors of this study examined the depolarization ratio of pure black carbon fractal aggregates. Based on previous experimental data, the depolarization ratio (DPR) has been modeled and the model has been compared to the measured DPR obtained through SPIN measurements. It has been demonstrated that using random orientation in T-matrix models results in an underestimation of the DPR. The purpose of this article is to highlight the importance of using fixed orientation when modeling the optical properties of BC fractal aggregates.  The findings of this study are original in that they extend the previous work done by Sela and Haspel (2021), and there is a novelty to the analysis. I would like to provide some comments on the paper, and once these comments have been addressed, I recommend it for publication.

Specific comments:

1. 1. The authors begin their introduction by describing previous research. In my opinion, however, it is missing context about why we need to investigate the DPR of soot particles and how they might be relevant to remote sensing or radiative forcing estimates.
2. It would be better to move some information, in particular the mathematical equations in Section 2.2, to the Appendix about theoretical methods in order to make the publication easier to read.
3. The difference between eq. 8 and eq. 9 is not clear. The optical efficiency can be calculated as the ratio between the optical cross-section and the geometric cross-section. You have taken the ratio of two geometric cross-sections in equation 8.
4. In the results of the total scattered intensity over theta-sca (for example, Figs. 4, 6, 7 You may also so on), the authors have pointed out the differences in soot characteristics using terms such as SC_300_35, etc. There is an overwhelming amount of terminology used in the caption of these results, which makes it difficult for the reader to understand them. In each figure, it is recommended to indicate the point of difference within the panels. For examples in Fig. 4, the difference between the panels is the change in the mobility diameter. A description of the corresponding mobility diameter can be included in each panel by the authors. And specify the other common properties as a single line in the caption. Similarly, in the discussion, the authors can specify the relevant parameter that they are discussing in each section (rather than using long technical terms). The second part of this comment is a choice of style. The authors can choose to remain with their style.
5. The authors have discussed the difference between the modelled DPR and the measured DPR at the end of section 3.2. This was done by comparing the DPR from Table 2 with those from Table 4. It is an important result. Ideally, the differences should be visualized as a 1:1 corelation with a linear regression fit.
6. The results, particularly the tables about optical properties, should be moved to the Appendix or Supplementary sections.
7. In Table 9, the authors provide the range of optical properties for different realisations of pure BC aggregates. It would be preferable if the range was expressed as a percentage change. The study by Romshoo et al., 2021 (Fig. 4) has also demonstrated that there is variability in the optical properties of pure BCFAs at 660 nm for 30 realisations at different mobility diameters and fractal dimensions. What is the comparison between your results and theirs? Romshoo, B., Müller, T., Pfeifer, S., Saturno, J., Nowak, A., Ciupek, K., Quincey, P., and Wiedensohler, A.: Optical properties of coated black carbon aggregates: numerical simulations, radiative forcing estimates, and size-resolved parameterization scheme, Atmos. Chem. Phys., 21, 12989–13010, https://doi.org/10.5194/acp-21-12989-2021, 2021.
8. Similarly to comment 7, it would be more beneficial to have the range of optical properties and the DPR expressed as a percent change for random vs fixed orientation. It is easier for others to use or cite this information.
9. The authors mentioned the atmospheric implications in the conclusion, but no discussion/result has been provided. Would it be possible for the authors to explain how the DPR range (random vs fixed) affects the calculation of lidar backscattering? The sensitivity study can be demonstrated using a simple analytical equation, is there one?
10. The SPIN measures at 135 degrees, which is why the authors were able to compare the modelling and measurements of DPR for a single wavelength. Would it be possible for the authors to mention in the discussion whether there are other instruments that can be used to measure other angles or all angles? Is it possible to improve the understanding of this subject by using such instruments?

Technical corrections:

1. Line 53 – “vein” do the authors mean “way”
2. Line 92 – Section 2.1 instead of “section 1”
3. Line 41-44 – sometimes it is better to write two sentences instead of one long sentence
4. Captions of figures – the entire text does not need to be in Bold
5. At the end of the manuscript, in the Appendix, please provide the list of terms/abbreviations

Haspel et al. present experimental measurements and theoretical calculations of polarized scattered light for different types of BC aggregates. The measurements (2 polarization states, integrated over a scattering angle range from 115-155°) are performed for 2 types of 400 nm size-selected BC aggregates. The theoretical calculations (2 polarization states, angularly resolved plus integrated quantities) are performed using the multiple sphere T-matrix model for a large range of different BC aggregates (varying order/disorder, fractal dimension, number and diameter of primary spheres, aggregate orientation, refractive indices). The goal is not to perform a full quantitative intercomparison between the measurements and calculations, but rather to to see whether they follow the same qualitative tendencies. It is shown that they largely do. A limited quantitative comparison is also perfomed with respect to measured and calculated linear depolorization ratios.

Theoretical calculations of light scattering by BC aggregates are only rarely compared with actual measurements. Therefore, I think this manuscript will be a useful contribution to the literature. However, there are a number of important issues that I believe should be addressed before the manuscript is accepted for publication.

L23-26: There are many potential reasons for the discrepancies between measured and calculated polarization ratios (as summarized nicely in Section 4). I am not convinced that this is the most likely explanation and therefore this part of the abstract reads to me as an over-interpretation of the results. 

L38: Throughout the paper there is conflation of the terms aggregate 'order/disorder' and 'fractalness'. For example, here is it is said that the authors will examine how the degree of aggregate disorder affects actual measurements. But the measurements are performed on aggregates of different fractal dimension but unkownn order, and ultimately they are then compared with the CCA aggregate calculations. It would be helpf ful if the differences (and similarities) between these two concepts were more clearly described and if more care was taken not to conflate the 2 terms throughout the paper.   

L41: The obvious atmospheric connection here is that the more fractal-like aggregates better represent fresh BC close to emission sources, while the more-spherical aggregates represent aged BC that has undergone restructuring. Perhaps this should be discussed here in order to better highlight the atmospheric relevance of the work. Can it also be said (with reference to past studies) that collapsed, spherical-like BC particles in the atmosphere are likely to be better represented by the IAS than the SC aggregates?  

Fig. 1: Are these SEM images of aggregates that have been size-selected? There appears to be quite some variability in the aggregate sizes. Also panel (a) shows aggregates that have coagulated into pairs. This will strongly affect their scattering properties. 

L74: A diagram would help grealy to visualize the positions of the different detectors. 

L88: These thesholds are high in terms of diameter (I assume optical diameter as measured by the side-scattering detector in the SPIN). It seems likely that some doubly-charged particles would still be included in the analysis? Is there a reason these thresholds were set so high? 

Table 1: The aggregate labels are difficult to take in and readibility later on becomes very laborious (e.g. comparing CCA_2.34_1.085_400_35 with CCA_1.92_1.085_400_35). I encourage the authors to think carefully about a better aggregate labelling system to improve the readibility of the text. 

Table 2: Perhaps my biggest overall criticism of the manuscript is that it is difficult to get a sense of the measurement uncertainties and how much can be interpreted from the differences observed. For example, there is clear systematic bias (~20%) between detectors P1 and P2. Why could this be? Could a similar bias be affecting detector S1 in relation to P1 or P2, and therefore the S/P ratios? Would it perhaps be better to use only P1 in the S/P ratio, since P1 and S1 are at least measuring the same angular position? In addition, measurements are only shown for BC aggregates of complex morphology. Did the authors also measure some simpler test aerosols that were either spherical or closer-to-spherical? It would be interesting to see the S/P statistics for such aerosols to get a sense of how significant the differences are between the COJ300 and R2500U sets. 

Table 3: The paper contains a lot of data tables that take substantial effort to comprehend. The presentation quality could be improved a lot, sometimes just with simple formatting. For example, in this table it would be useful to visually separate the different D_outer-envelope sets, even just with simple horizontal lines.

Figure 4 (and other similar figures): the figure captions are overly difficult to comprehend. Revision is required. A key piece of information is missing: what are the thick solid lines in the scattering region from 105 - 155? I had to dig into the text to realize these are purely for visualization purposes and they don't represent actual measurements. The figures themselves could also be improved so it is easy to differentiate the differences between the different rows. E.g. by adding labels D_outer-envelope = 300 nm, and so on. Same comment for all other similar figures and their captions. 

L285: The term 'side/back scatter' is rather imprecise and it shows its limitations in sentences like this. The precise quantity being considered is actually defined in Eq.7 and used in a number of the tables. I think the text would be improved by introducing and using a precise term for this quantity. 

Sections 3.4 and 3.5: It is good the sensitivities to these different parameters have been explored. However, the readibily and conciseness of the text would be improved greatly by moving some of the discussion, figures and tables in these sections to a supplementary information file. 

L656: In light of the many potential reasons for the discrepances between calculated and measured S/P ratios, many of which are nicely summarized in Section 4, I don't think it's reasonable to place particular emphasis on this one specific explanation here and in the abstract (as noted above).