The paper makes an important contribution to the literature and can provide input to the modeling community regarding the sea-salt emissions.

Some minor comments before publication:

Using ECMWF model constraints for RH would not necessarily remove clouds from L2A measurements (it is well-known that clouds are not well-represented in models)

It is better to use marine particle depolarization at 355nm from the Delian model (Floutsi et al., 2023). Gross' paper reports depol values at 532nm (even though the difference is not large, 1.3 vs 2%).

CALIPSO cannot verify the presence of a specific aerosol type, since the aerosol type is inferred based on assumptions on the surface type. Even though the regions selected are dominated by marine particles, it is better to rephrase as it concerns CALIPSO and further validate through a global model that there are no other types present (e.g. from ship emissions).

In 4.2 the comparison of Aeolus with CALIPSO on extensive properties (a, b), should be restricted for backscatter only (CALIPSO cannot deliver extinction). Extinction could be evaluated against passive sensors such as MODIS AODs over the region.

1. Original Submission

1.1. Recommendation

Major Revision

1. Comments to Author:

Correlation between marine aerosol optical properties and wind fields over remote oceans with use of spaceborne lidar observations

Overall opinion: The paper elucidates the relationship between marine aerosol optical depth (AOD) and near surface wind speeds using Aeolus. In short, you demonstrated that Aeolus can unveil positive relationship between aerosol optical properties and wind speed over oceans efficiently. Aeolus can resolve gradient of the aerosol optical variations depending on the wind speed as well. The differences between marine boundary layer and the layer above can be unveiled nearly exclusively by relying on Aeolus, which is a promising finding considering low signal to noise ratio of Aeolus. However, your current effort should be carefully revised because of poorly justified methodological choices, ambiguities (and, even more critically, inconsistency) in the research aim, introduction gaps, and text that is unfriendly to general readers. Most notably, some results are not persuasive; the relationship between marine AOD and wind speed has not been quantified in some cases, and only displayed in figures and discussed in the text in others without trends being quantified or correlations/causal connections being reported. Conclusions are just short and plain version of results (no summary, no overview, no holistic opinion presented in conclusions). Further details are outlined below.

2.1. Comments:

1. Abstract: Although the paper is submitted to the special issue dedicated to Aeolus, I think you should foremost think about general readers. Please emphasize your research aim boldly, ensure this research aim agrees with what you state in the end of the introduction and in the end of the paper. Logically, there is little sense to start your abstract by introducing satellite-specific terms such as Level 2A product. At least, please mention Aeolus first. Ideally, state in the first sentence of the introduction why you think marine aerosol – wind speed relationship is important and Aeolus is a good choice to address it.
2. Introduction: Three problems here.
   • First, there is an extreme ambiguity on which aspect of AOD_mar and wind speed relationship you want to address. You include so-called highlights, which actually erode the clarity of your research aim. State directly: do you want to examine whether there is a relationship between AOD_mar and wind speed using Aeolus? If yes, say it boldly and underpin all highlights (i.e., objectives) to this research aim please.
   • Second, it looks like you placed everything what is related to AOD_mar – wind speed relationship, Aeolus and CALIPSO aerosol observations of aerosol together, but in barely comprehensible logic. To be more specific, your introduction is neither centered over your research aim, nor logically guides a reader to this research aim paragraph by paragraph though showing important milestones and gaps made in this research field. In other words, it is related to the study topic, but chaotically structured.
   • Third, you discuss marine aerosol optical properties – near wind relationship from lidar perspective, but omitted a large corpus of works dedicated to this issue. In particular, Josset et al. (2008) demonstrated that there is inverse relationship between wind speed and surface attenuated backscattering; while both these parameters have direct link to aerosol optical depth and this paper shows the formula how they are linked. Further works of Josset et al. (2008, 2010, 2018); Hu et al. (2008), Venkata and Reagan (2016) have elucidated this relationship in detail for CALIPSO and there were pre-launch Aeolus works on this topic such as Li et al. (2010). Moreover, some yet unfinalized studies, but ongoing efforts of Labzovskii et al. (2022) and Dionisi et al. (2022) addressing ocean surface-aerosol optical properties-wind interplay can be found as conference proceedings. By omitting all these CALIPSO and Aeolus-focused efforts, you hint that you are not aware about a hidden fundamental link between AOD and wind speed over oceans if your AOD is calculated using lidars. This can be a pitfall and a single point of failure for your methodology if this link does exist. Please check all the aforementioned works (I provided references in the end of this document) and incorporate their experience in your introduction, where the interpretation is up to you indeed.
3. Methodology: This part of the manuscript is a potential pitfall as well. In particular, you have introduced your own framework of marine aerosol domination, cloud screening and wind speed-aerosol optical property analysis. However, nearly every stage of the framework you showed on Figure 1 should be justified because you made numerous debatable assumptions. The assumptions about effective ability to classify marine aerosol using CALIPSO to be applied for Aeolus, assumption about efficiency of cloud screening based on Rayleigh channel information of Aeolus. I will raise the following issues
   • According to 2.1, it looks like you used the wind speed from official Aeolus observational product. However, Figure 2 implies that you used AUX_MET winds from NWP/ECMWF simulations and L2C data. Clarify this aspect in every section please.
   • First, you may describe the methodology and only then, introduce Figure 1. In most cases, the description of a figure comes before the figure itself. Moreover, think about a reader, without reading a text, he/she might be easily confused by reading terms and concepts that you not introduced, nor described yet.
   • You do not need to repeatedly use the term ‘aerosol optical depth’ as the acronym AOD is quite common. For instance, you keep using the term ‘aerosol optical depth’ without acronym even in Figure 1.
   • Once you state that SCA product is more robust than other algorithms, either mention these algorithms directly or just state that SCA product is robust.
   • Cloud screening. The presence of undetected clouds can severely plague your clear sky assumptions. How did you ensure that this screening strategy worked well? I did not notice any evaluation or statistical analysis of “cloud-free” and “cloud-containing” layers in your paper. Thus, the efficacy of this approach is questioned. Moreover, why one needs to use molecular optical depth and not Mie backscattering or extinction product (fundamentally more sensitive to thin clouds or liquid water clouds than Rayleigh product) directly to detect clouds at various heights? Let alone, you said that the SCA product is stable and you can confidently rely on these products.
   • Eliminating outliers. Okay, you referenced the paper from 1986 to justify statistical filtering of outliers. However, how did you ensure that in the particular case of Aeolus data, you have not filtered useful data using this particular statistical filter? I mean statistical filtering is helpful for sure, but only when you can understand when to apply this statistical filtering from physical point of view. Thus, please justify physical aspect of this choice here. The same is applied to outlier removal step introduced in Lines 368 – 375.
   • Why the title of 2.1 is “ALADIN/Aeolus” with right slash? Give more comprehensible name to the section please. Same applies for 2.2.
   • (1) You introduce a depolarization correction of backscattering based on the assumption that CALIPSO can ideally detect marine aerosol, but this is not the case. What if you just introduced positive bias in many cases of your analysis by assuming their marine nature, while their depolarization was not typical for marine aerosol cases?
   • Potential inaccuracies due to assumptions (not calculations or objective information) about MABL of ~1 km are not discussed.
   • LayerL and LayerH are counterintuitive terms. At least, LayerL could be referred as Layer_MABL for clarity throughout the text. Explain if I am wrong here and missing some intuitive links with L and H letters.
4. Results:
   • Some information you placed into the Results obviously fits the methodological description in more logical way (see the lines from the start of Results 4 to Line 260). Lines 368 – 370 as well, where you talk about elimination of statistical outliers; this should be explained in the methodology not in the middle of the results section.
   • Some terminological problems are visible. For instance, you say “it is considered that the Aeolus retrieved extinction and backscattering area reasonable”. First, considered by whom? Second, what is “reasonable” from physical point of view? (Line 305 and above). This aspect is better clarified in Line 315.
   • You dedicate considerable efforts to prove that your aerosol in ocean zones has marine/ocean origin. For instance, you speculate about winds in MABL and above. Why CALIPSO classification is not enough as methodological choice to determine once and for good that you have marine aerosol? You are eroding your research scope by devoting too much efforts to prove this point in the results, not in the methodology.
   • The style of reporting lacks references to certain figures, which hampers review process. You added three figures together in a row (4 – 6) and each consists of multiple panels. In this case, it is not helpful to refer to figures like “From Figures 4 – 6 you can see…” (line in the case of Line 316).
   • Line 316. I do not see this similarity qualitatively. Please try to use quantitative terms or more explicit qualitative description of similarity.
   • Line 325 Numerical reference to what is “evident high wind speed region” is missing. Use numbers or direct references to figures here and elsewhere.
   • Section 4.3 Use quantitative metrics while talking about such phenomena as increasing tendency. When I am looking at Figures 8 and 9, what we need here are rather: statistical agreement metrics (correlation or anything similar), metrics of statistical significance if this agreement exists, trend metrics. From first glance, I do not see any correlation for panels a, b, c. Perhaps, some correlation at panel d, but you did not articulate it.
   • Lines 403 – 405. You present an unsupported hypothesis here where it is not possible to establish whether you are facing the lack of wind data of > 15 m/s, some other physical phenomena such as response of ocean surface backscatter to stronger winds, which affects AOD in the end or even something else. Such unsupported surmises are not advisable for journals, focused on atmospheric physical phenomena.
   • Lines 411 – 429 In this paragraph, you quantify only wind speed values, not optical properties of aerosols. I do not see any value in these speculations of optical properties of aerosols becoming “larger” or “smaller” at certain wind speed intervals without (1) strong quantitative arguments about relationship of these optical properties to wind speeds, (2) consistent, centered narration around the pattern you identified. Moreover, it is unclear why you analyzed both extinction and backscattering, it does not seem that you make any difference between these parameters. Neither in the way treating them, nor in making conclusions, you just report that they change in some way with wind. Without quantitative arguments of relationship with wind, without explicit references to figures, where you noticed these patterns, why would it matter after all?
   • Line 468. Once again, Josset et al. [2008] have examined relationship between WS (AMSR-E from A-Train) and aerosol optical properties from collocated CALIPSO observations. Only over ocean. Please consider this point while writing your elaborations on WS-AOD_mar interplay here.
   • Line 508 Unsupported surmise about the presence of clouds. Arguments are needed here.
   • Line 510 Quite late to introduce lidar ratio as you already spoke about it before. Please address the consistency of terminology and your acronyms here and elsewhere
   • Figure 15. The deviations of LR from Aeolus are around 50%? For many types of data, such deviations make the quantification nearly meaningless. Your lidar ratio can jump from the values of 10 to >40. Moreover, some errorbars and even data points are omitted (see intersection of 40 LR and wind speed of 1 m/s).
   • Line 529. You cannot say that similar results have been shown in previous studies. Rather, your study being chronologically newer, reports similar results, not the other way round. Moreover, specify the study you meant here directly.
   • Line 531 You already used term AMSR-E, but explain it here once again.
   • Minor comments on this section you do not need to respond to, just consider this criticism while rewriting your results
     1. Avoid ambiguous terms like “explicit relationship”; stick to statistical, mathematical or physical terminology in such cases.
     2. Line 375 As mentioned in the comment about introduction, some previous studies have reported negative relationship between wind speed and aerosol optical properties given wind speed-AOD-surface backscattering fundamental relationship once we are dealing with water surface. The discussion on this aspect with the link to previous studies is missing in both introduction and results sections.

• Line 418 Growth rates become smaller is a dubious formulation. Did you mean growth rates slowed down? The same applies to “change points”. What are “change points” in line 420? Same for “wind speed distribution ranges are larger” (Line 413).

1. Line 423 24 – 28 m/s is not quite strong wind, it’s basically storm
2. Line 424 You basically said that extinction will sharply increase under the condition of increased extinction, right? Re-read the sentence please or explain what I understood wrongly here.
3. Line 425 Statistical significance should be supported by arguments here

• Line 440 Gradient change points? Where we can see that, which figure? Numbers are not mentioned here also to judge. Please update the entire paragraph with exact references to figures or numbers.
• Line 444 There is no section starting here. Please update this paragraph using the same guidelines as I gave above (more quantitative analysis).

1. Line 459. The language of this section has been visibly deteriorated compared to previous sections. Please revise it as well to make it more readable: the effort (line 461)?, grid? (465, maybe grid cell?)
2. Line 478 “Quite similar” is unscientific. Quantitative arguments please
3. Line 505 So how did you avoid enormously high errors due to wrongly fixed lidar ratios or at least quantified it?

1. Conclusions: Conclusions should be meticulously revisited after the revision. Most importantly, you missed the opportunity to give a holistic summary on what your analysis revealed in terms of physical behavior of aerosol. Did we learn something new about marine aerosols from physical standpoint? According to your conclusions – unlikely. Alternatively, you could fill up this summary by speculating on the value of your findings with regards to Aeolus capabilities (low signal-to-noise ratio, low resolution, presence of clouds, complex relationship between ocean properties, AOD and wind speed). In the current form, you just plainly repeated the results and methodological choices in briefer form. Other, more concrete problems are here:
   • You hinted that you identified marine aerosol in the results, but it is rather implied that you introduced some methodological tool to assume marine aerosol. Once again, it is unclear why CALIPSO classification was not enough for this purpose (Line 568)
   • Sensitivity analysis on distinguishing clouds from aerosols has not been shown, so I have doubts that you actually separated them quantitatively (Line 568). This statement definitely does not report your actual findings and therefore does not fit the conclusive tone it takes.
   • Vague methodological descriptions are redundant and uninformative for readers in the conclusions (Lines 568 – 570 about defining the areas of the study for example)
   • Repetitive formulations, partly reflecting your research aim/objectives are spotted (Line 574)
   • The information preceding the line 583 is redundant for conclusions from my point of view and should be either shortened or be more concrete in terms of reporting.
   • Line 600 – repetitive, you said it three times in the conclusions. Report conclusion directly without repeating research aim/question.
   • As mentioned, you do not shed the light on the difference between backscattering coefficient and extinction coefficient; they were treated identically and discussed as statistical parameters, not physical properties of aerosols.
   • Lines 595 – 599 You cannot judge about the size of the particle without either deriving microphysical properties of aerosols or at least by using Angstrom exponent. Lidar ratio does not fit such purpose, it just shows the ratio between light being extinct and light being scattered.
   • Line 603 You have not analyzed turbulence in your study.
   • Line 604 Wind speed bin is not scientific term known by a general reader.
   • Line 605 The statement about ‘not total similarity’ of aerosol variation tendencies does not bring any new knowledge and is, therefore, not useful for conclusions.
   • Lines 605 – 607 On the development of aerosols over ocean due to complex factors – you have not shown this in your study. As a suggestion, it is trivial because the complex relationship between aerosol evolution over ocean and processes on the ocean-atmosphere interface, as well as in the atmosphere; this relationship is obvious. Bring numbers and facts if you’d like to add some new value to this common knowledge.
2. Language and Format: Language should be revised either by seeking assistance of colleagues, more familiar with the standard of academic English or automated software for language check at minimum. Multiple stylistic (PDF-oriented font format is spotted at line 48 as example), and more critically, grammar errors are found (‘significate’ at line 45 for instance). Formatting caveats also include redundant spacing (line 131), double bracketing, inconsistent introduction of acronyms, etc. Please eliminate all these drawbacks, which currently emphasize the raw condition of your draft.

Mentioned references:

1. Dionisi et al., (https://meetingorganizer.copernicus.org/EGU23/EGU23-16196.html)
2. Hu, Y., Stamnes, K., Vaughan, M., Pelon, J., Weimer, C., Wu, D., Cisewski, M., Sun, W., Yang, P., Lin, B., Omar, A., Flittner, D., Hostetler, C., Trepte, C., Winker, D., Gibson, G., Santa-Maria, M., 2008. Sea surface wind speed estimation from space-based lidar measurements. Atmos. Chem. Phys. 10.
3. Josset, D., Pelon, J., Hu, Y., 2010 Multi-Instrument Calibration Method Based on a Multiwavelength Ocean Surface Model. IEEE Geosci. Remote Sensing Lett. 7, 195–199. https://doi.org/10.1109/LGRS.2009.2030906
4. Josset, D., Pelon, J., Pascal, N., Hu, Y., Hou, W., 2018. On the Use of CALIPSO Land Surface Returns to Retrieve Aerosol and Cloud Optical Depths. IEEE Trans. Geosci. Remote Sensing 56, 3256–3264. https://doi.org/10.1109/TGRS.2018.2796850
5. Josset, D., Pelon, J., Protat, A., Flamant, C., 2008. New approach to determine aerosol optical depth from combined CALIPSO and CloudSat ocean surface echoes: New approach to determine AOD. Geophys. Res. Lett. 35. https://doi.org/10.1029/2008GL033442
6. Labzovskii L., van Zadelhoff G-J, Donovan D, De Kloe J., Josset D., 2022. How sensitive are Aeolus Lidar Surface Returns (LSR) to the types of surface? Insights for LSR-based retrieval of AOD over ocean by using Aeolus; https://doi.org/10.5194/egusphere-egu22-12079
7. Li, Z., Lemmerz, C., Paffrath, U., Reitebuch, O., Witschas, B., 2010. Airborne Doppler Lidar Investigation of Sea Surface Reflectance at a 355-nm Ultraviolet Wavelength. Journal of Atmospheric and Oceanic Technology 27, 693–704. https://doi.org/10.1175/2009JTECHA1302.1
8. Venkata, S., Reagan, J., 2016. Aerosol Retrievals from CALIPSO Lidar Ocean Surface Returns. MDPI Remote Sensing 8, 1006. https://doi.org/10.3390/rs8121006

This manuscript combines Aeolus and CALIPSO data along with ECMWF wind products that assimilate Aeolus to understand the relationship between wind speed and marine aerosol optical properties in 3 remote oceanic basins. The manuscript is well written and results are clearly presented. I recommend publication with minor revision:

1) The authors have a hard cut-off at 1 km to mark the MABL top throughough the manuscript. I would suspect this would vary somewhat with region. Can the authors comment on any ramifications of this? Would using ECMWF MABL help?

2) Line 41. Please specify - by area, by mass.

3) Line 95. Change from "is" to "was" as ALADIN stopped science operations at the end of April 2023.

4) Line 174. Why was version 3.41 used for January 2022- July 2022? Was version 4 data not available? This has implications for continuity of the VFM algorithm.

5) The authors have not discussed the implication for the time offset of CALIPSO and Aeolus. While I don't think this is a major impact, the aerosol typing by CALIPSO will be after Aeolus observed winds and would be seeing the aftermath. Please consider discussing this point in the manuscript.

6) Equation 1. Consider labeling Bmar, Aeolus as Bmar, Aeolus co or Bmar, Aeolus II to indicate it is the co polarized return.

7) Line 555. The "upward" trend in the middle of the plot seems very faint. Is it statistically significant?