The manuscript presents a new set of POLIPHON conversion factors for dust aerosols at oceanic or coastal sites. The authors use a depolarization ratio-based metric to identify dust cases in AERONET measurements and further classify them into clusters of pure dust (PD) and dust-dominated mixture (DDM). They estimate the CCN- and INP-related conversion factors for these two clusters and also compare them with the already existing ones. In addition, they discuss the variations of these conversion factors along the transoceanic pathways. The manuscript presents considerable development in the field of lidar-based CCN and INP retrieval and has good potential for publication in AMT only after implementing and addressing the following comments.

Specific comments:

1. Since you mention the significance of how estimating the 3D distribution of CCN and INP can improve the climate models twice, in lines 52–54 and 71–72, could you please elaborate in the text with one or two sentences on how it may do that?
2. The microphysical properties of dust may change with ageing and mixing with other aerosol types (Kim and Park, 2012; Goel et al., 2020). This implies that the conversion factors estimated close to the main desert sources may not be the same as those estimated far away from the source region. This may be stated in the motivation.
3. Some validation studies that use POLIPHON conversion factors to estimate aerosol number concentrations and CCN concentrations from spaceborne lidar measurements highlight the need for improvement, especially for dust and marine aerosol mixtures (Choudhury et al., 2022; Choudhury and Tesche, 2022). Such studies can be included in the motivation for this study.
4. Line 136. Ansmann et al. (2019) use aerosol optical thickness (AOT) at 532 nm, not 500 nm. They estimate AOD at 532 nm using Ångström exponent and AOT at 500 nm. Since you compare your results with Ansmann et al. (2019), it is necessary to stay consistent. Also, 532 nm is the usual lidar wavelength. As you aim for application to spaceborne lidar, you should use this wavelength in your analysis. In general, the wavelengths mentioned in the paper for AOT and extinction coefficient up to this point are inconsistent. For instance, it is 532 nm in line 60 and 550 nm in line 139. Did you convert the AOT from 500 nm to 532 nm to estimate the conversion factors? If not, then how do you justify your calculations, as we cannot ignore the wavelength dependency of optical thickness?
5. Line 218. It seems like you did not convert the AOT to 532 nm when calculating the conversion factors. Please address the previous comment.
6. I found the transition from Section 2 to Section 3 to be very abrupt. I suggest adding a paragraph under Section 3 before going to Section 3.1, giving a brief introduction to the section and the analysis you present here so that readers have a basic idea of what to expect. This should have been included in the "Data and Methodology" section. However, looking at the organisation of the manuscript, I suggest adding it at the beginning of Section 3.
7. Lines 226-228. How can the number of data samples be linked to the level of agreement or difference between pure dust (PD) and PD + dust-dominated mixtures (DDM) clusters? Having fewer samples could only imply that the result is not significant or lacks confidence.
8. Line 231. Can you speculate on the local aerosol sources based on the sites or locations where you found the differences between PD and DDM clusters?
9. The comparisons presented in Section 3.1 are all qualitative. I would always prefer and recommend a quantitative analysis. Please quantify the differences in terms of percentages or absolute values.
10. Line 260. PD cases are already included in PD+DDM clusters, right? Please change "PD and PD+DDM" to "PD and DDM". Also, how do you define an adequate sample size? Please include it in the text.
11. Line 306. You haven’t mentioned the cluster you used in Figure 8. Looking at the values, I guess it is for PD cases. Here, I am assuming that the variations in conversion factors along the transoceanic pathways are for PD cases. From Table 2, I can see that the number of samples for stations at MI, ML, and TR are 26, 27, and 18, which are significantly less than other stations. Due to fewer samples, the variations that you report for these stations may not be realistic, as the long-range dust transports may vary seasonally and annually. One thing to look at is the distribution of the small number of sample points across different seasons and years. Are they limited to one season, or one year? In any case, you must mention these limitations in the paper. I would recommend using the DDM cluster, which has more than enough sample space to study the changes in the conversion factors along different transport pathways.
12. Why did you exclude the transoceanic variations of c100,d? I recommend adding a new panel to Figure 8 to show the variations of c100, d, and xd and discussing them in Section 3.4 of the manuscript.
13. As highlighted in the introduction of the manuscript, the ultimate goal is to apply the conversion factors to spaceborne lidar measurements to estimate global 3D CCN or INP data. How do you suggest applying these new conversion factors to spaceborne lidar? Should it be applied based on the geographical location? Or, as suggested by Ansmann et al. (2019), global average conversion factors should be used?
14. Figure 1 is missing some sites that are included in Table 2. Please modify.
15. Table 1 has an expression for nCCN but lacks the expression for n100,d. Please add.

Technical corrections:

1. Line 41. Please cite the latest IPCC report. And since you quoted the IPCC report, I believe you mean effective radiative forcing and not radiation budget.
2. Line 60. A two-step dust separation technique for obtaining fine and coarse mode contributions separately, given by Mamouri and Ansmann (2014), has also been used in multiple studies and can be included here.
3. Line 74. Replace "regional-dependent and relevant to" with "regionally variable and dependent on".
4. Lines 81-82. The term "dust transport pathways" appears for the first time here without any prior explanation. I suggest explaining it briefly here.
5. Line 89. Do you mean "depolarization ratio"?
6. Lines 87-88. I suggest replacing "we plan to adopt another scheme to select data points that are representative of dust presence from AERONET databases" with "a different scheme to identify the presence of dust in AERONET measurements".
7. Lines 101-102. Replace "a previous study" with "the results from Ansmann et al. (2019)".
8. Line 138. Replace "AERONRT" with "AERONET".
9. Rephrase lines 201-202.
10. Line 208. The abbreviation FMF appears for the first time.
11. Line 209. A flow chart of what?
12. Line 241. AERONET.
13. Line 249. Remove "scatters regarding the".
14. Line 250. There's no need to mention the colours here. This should be included in the figure caption. Replace "situations" with "cases".
15. Line 253. Replace "participate" with "are considered".

There were many other obvious language-related errors that I have not included here. Some of them can be corrected during the manuscript's copyediting if it reaches that stage. However, I highly recommend the authors consult a professional language editor to improve the readability of the manuscript.

References

Ansmann, A., Mamouri, R.-E., Hofer, J., Baars, H., Althausen, D., and Abdullaev, S. F.: Dust mass, cloud condensation nuclei, and ice-nucleating particle profiling with polarization lidar: updated POLIPHON conversion factors from global AERONET analysis, Atmos. Meas. Tech., 12, 4849–4865, https://doi.org/10.5194/amt-12-4849-2019, 2019.

Choudhury, G., Ansmann, A., and Tesche, M.: Evaluation of aerosol number concentrations from CALIPSO with ATom airborne in situ measurements, Atmos. Chem. Phys., 22, 7143–7161, https://doi.org/10.5194/acp-22-7143-2022, 2022.

Choudhury, G., Tesche, M.: Assessment of CALIOP-Derived CCN Concentrations by In Situ Surface Measurements, Remote Sensing, 14(14), 3342. https://doi.org/10.3390/rs14143342, 2022.

Goel, V., Mishra, S. K., Pal, P., Ahlawat, A., Vijayan, N., Jain, S., and Sharma, C.: Influence of chemical aging on physico-chemical properties of mineral dust particles: a case study of 2016 dust storms over Delhi, Environ. Pollut., 267, 115338, https://doi.org/10.1016/j.envpol.2020.115338, 2020.

Kim, J. S. and Park, K.: Atmospheric aging of Asian dust particles during long range transport, Aerosol Sci. Technol., 46, 913–924, https://doi.org/10.1080/02786826.2012.680984, 2012.

Mamouri, R. E. and Ansmann, A.: Fine and coarse dust separation with polarization lidar, Atmos. Meas. Tech., 7, 3717–3735, https://doi.org/10.5194/amt-7-3717-2014, 2014.

Possibility to estimate the particle concentration (and so CCN and INP) from extinction/backscattering lidar measurements by using corresponding conversion factors is quite attractive. As suggested in publications of Ansmann with co-authors, the conversion factors can be estimated from AERONET measurements and this study improves the knowledge of these factors for dust, applying POLIPHON technique for different дщсфешщты. Paper is well written, presents new results and is suitable for publishing in AMT after minor revision.

Specific comments

Ln.89. ”The new scheme is based on the particle linear polarization ratio in AERONET Version 3 aerosol inversion product, which is considered a better indicator for non-spherical 90 dust particles (Shin et al., 2018, 2019)”

Depolarization ratio in AERONET is calculated with the model of randomly oriented spheroids. There are many challenges in application of this model to the dust particles and accuracy of depolarization calculation is the subject of discussions (e.g. Gasteiger, J., Wiegner, M., Groß, S., Freudenthaler, V., Toledano, C., Tesche, M., and Kandler, K.: Modeling lidar-relevant optical properties of complex mineral dust aerosols, Tellus B, 63, 725-741, 2011). I think corresponding comment should be added to the manuscript.

Table 1. Uncertainties of estimation is really important point. In estimation of dust backscattering depolarization ratio of smoke is assumed 0.05. But actually it varies in 0.04-0.09 range (Burton et al., 2013), though provided uncertainty 10-30% looks reasonable. However, lidar ratio of dust may vary in 30 sr-60 sr range, so uncertainty of extinction calculation should be higher, but authors provide 15-25% range. Uncertainty of mass concentration should be even higher, but in Table 1 the range 20-30% is given. I think these uncertainties should be clarified.

Ln 125. “a good proxy for dust-related CCNconcentration n_CCN,ss,d   “ I think should be explained, why “ss” used in subscript.

Eq.3. Notations look a bit strange for me. For example, authors use N_250,j and in the right part it becomes n_{250,d, j.} Why index “d” is absent in the left part?

Eq.6. What is c_d ?

Ln 195. Ok, here authors start discussion of spheroids model. But may be better to do it in introduction.

Eq.8 actually repeats Eq.1.