Characterisation of dust aerosols from ALADIN and CALIOP measurements

Song, Rui; Povey, Adam; Grainger, Roy G.

doi:https://doi.org/10.5194/egusphere-2023-2252

Preprints

https://doi.org/10.5194/egusphere-2023-2252

Preprints

11 Oct 2023

| 11 Oct 2023

Characterisation of dust aerosols from ALADIN and CALIOP measurements

Rui Song, Adam Povey, and Roy G. Grainger

Abstract. Atmospheric aerosols have a pronounced effect on climate dynamics at both regional and global scales, but the magnitude of these effects is subject to considerable uncertainties. A major contributor to these uncertainties is the incomplete understanding of aerosol's vertical structure, largely due to observational limitations. Spaceborne lidars can directly observe the vertical distribution of aerosols globally, and are increasingly used in atmospheric aerosol remote sensing. As the first spaceborne High Spectral Resolution Lidar (HSRL), the ALADIN instrument onboard the Aeolus satellite was operational from 2018 to 2023. With its sophisticated design, ALADIN can retrieve aerosol backscatter and extinction coefficients separately without an assumption of the lidar ratio. This study is dedicated to assessing the performance of ALADIN's aerosol retrieval capabilities by comparing them with CALIOP measurements. A statistical analysis of retrievals from both instruments during the June 2020 Saharan dust event indicates good consistency between the observed backscatter and extinction coefficients. A detailed comparison of extinction coefficients for dust layers reveals that ALADIN is more susceptible to signal attenuation than CALIOP. During this extreme dust event, CALIOP-derived aerosol optical depth (AOD) exhibited large discrepancies with MODIS Aqua measurements. Using collocated ALADIN observations to revise the dust lidar ratio to 63.5 sr, AODs retrieved from CALIOP are increased by 46 %, improving the comparison with MODIS data. Further, the combination of measurements from ALADIN and CALIOP can enhance the tracking of aerosols' vertical transport. This study demonstrates the potential for spaceborne HSRL to retrieve aerosol optical properties. It highlights the benefits of spaceborne HSRL in directly obtaining the lidar ratio, significantly reducing uncertainties in extinction retrievals. This work paves the way for forthcoming spaceborne HSRL missions, particularly the ESA ATLID space lidar (set for a 2024 launch) and Aeolus-2.

Received: 03 Oct 2023 – Discussion started: 11 Oct 2023

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 27479 KB)

Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Preprint (27479 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

26 Apr 2024

Characterization of dust aerosols from ALADIN and CALIOP measurements

Rui Song, Adam Povey, and Roy G. Grainger

Atmos. Meas. Tech., 17, 2521–2538, https://doi.org/10.5194/amt-17-2521-2024,https://doi.org/10.5194/amt-17-2521-2024, 2024

Short summary

Rui Song, Adam Povey, and Roy G. Grainger

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2023-2252', Anonymous Referee #1, 01 Nov 2023
This manuscript presents the meaningful demonstration of the capabilities of ALADIN in retrieving aerosol optical properties, specifically the backscatter coefficient, extinction coefficient, and lidar ratio. The dust layers’ lidar ratios used for CALIOP is also revised according the simultaneous measurements of ALADIN. The manuscript is well written and its contents are of high quality and scientific interest. The benefits of this study would be great for the accurate estimation of Aeolus and CALIOP aerosol data products. Hence, I recommend the acceptance of this manuscript after the necessary revisions.

The specific comments are listed below:
My main concern about the extinction coefficient/backscatter coefficient comparisons between CALIOP (at 532nm) and ALADIN (at 355nm) is their wavelength dependence. In this manuscript, the authors compare the aerosol products directly without any wavelength convert, even you mentioned it in line 244.

Line 66: could you please give some more detailed comments on why the extinction coefficient is not affected by the misdetection of the cross-polar component?

Line 104-105: The authors should be aware that the horizontal resolution for Rayleigh channel and Mie channel is different.

Line 120: Have the authors ever try to estimate the performance of the products from MLE? You mentioned the MLE method has positive effect on the products retrieve, however, why the Level-2 SCAmb products are applied in your study?

Figure 1: why the temporal disparity of 9 hours and the maximum spatial difference of 200km are set as thresholds? Is there any physical basis for these selections? For example, wind direction? Air mass transport?

The color bar in Figure 1 somehow misleads me. I suggest the authors may use the color bar oppositely, be like Figure 2.

Line 244: the spectral difference between 532 nm and 355 nm could be corrected somehow with the use of typical Angstrom exponent of dust. Have you ever tried to do this work?

Line 312: what is the time difference between the measurements from MODIS and CALIOP?

The wavelength band that MODIS applied should be pointed out. Hence, we can figure it out whether the wavelength convert should be carried out. From this point of view, the underestimation may be solved.

The technical corrections:
Line 193: “the blue dots in (d) represent the footprint…” should be changed to “the blue dots in (d) represent the footprints…”

Please provide the color bars’ label for the green/blue gradients in Figure 4 and 5.

Why there is only one red profile in Figure 4(a) and 5(a) between 12.5 km and 17.5 km? Is it because there is only one measurement case reach that height? Then I would suggest the authors provide the total numbers of measurements at different heights.

It should be “Comparison of aerosol extinction coefficients…” instead of “Comparison of aerosol backscatter coefficients” in the caption of Figure 5. Please correct it.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC1
- AC4: 'Reply on RC1', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC4-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC4
RC2:
'Comment on egusphere-2023-2252', Anonymous Referee #2, 02 Nov 2023
The authors present the combination of CALIOP and ALADIN measurements of a dust plume over the Atlantic ocean. They show that ALADIN measurements align well with CALIOP observations as long as proper data filtering as applied. They also re-iterate known limitations of the analysis of CALIOP observations of mineral dust by using ALADIN measurements of the dust lidar ratio to show that the dust lidar ratio in the CALIOP data retrieval is still set too low. The paper would benefit from re-organisation and shortening. Hence, minor revisions are needed:
I suggest to organise the work in a more conventional way with fewer sections. Please put methods and results into the corresponding sections rather then mixing them up as in the case study. The introduction is also bit lengthy and could be sharpened towards what's relevant for the presented work.

The text includes plenty of figure descriptions that should be covered solely in the figure captions (e.g. lines 193-196, 233-234, 270-275, 306-309, 3043-346). Please omit from the main text.

Please unify the colour axes in Figs. 1 and 2. The inversion is not very intuitive. Figure 2 might be expanded by a panel that shows all latitudes.

Why are the authors discussion cloud masks that are not being used for this work? I suggest to stick to what has been used (the MSG-SEVIRI dust mask) and to provide a statement that other cloud masks have either not yet been available or less useful for your purpose. Figure 3 would need to be revised accordingly.

Figures 4, 5, and 10 could be improved by adding the number of profiles that contribute at the different height bins. Also, colour scales for the gradients in Figs. 4 and 5 are missing.

The discussion of Figure 4 - particularly of the particle depolarisation ratio - would benefit from comparisons to findings of SAMUM-2 at Cape Verde.

line 251: shouldn't it be feature type identification?

I suggest to add lidar curtain plots to Figs. 6 and 7 as those would clearly demonstrate the effect of signal attenuation. It would also be nice if the comparison of extinction coefficients had some quantitative element, such as a correlation plot.

I don't think that Figure 8 is needed.

It is not clear to my what the investigation related to Fig. 11 and Tab. 2 is supposed to tell the readers. Okay, the mean profile shape is different for cases with AOD below or above an arbitrary threshold. But can this be used somehow? If anything, I would expect that type (b) with the higher extinction peak would correspond to the attenuated lower AOD profiles - but it doesn't. This part of the paper left me puzzled and I suggest to omit it.

I don't think that Section 7 provides any information about the vertical transport of dust aerosol. Fig. 12 is certainly a nice plot that combines the observations of the two platforms. However, it would me more informative if it was to provide information on the longitudinal and height distribution as well. It seems to me that a similar plot could already be produced using much more data from MODIS observations.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC2
- AC1: 'Reply on RC2', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC1
RC3:
'Comment on egusphere-2023-2252', Anonymous Referee #3, 04 Nov 2023
Review report for amt-2023-2252
The paper "Characterisation of dust aerosols from ALADIN and CALIOP measurements'' aims to assess the performance of ALADIN by comparing with CALIOP, and uses the synergy of both sensors to improve and increase information over an extreme dust transport episode. The manuscript falls within the scope of the journal. However, the presentation and discussion of the paper is not clear and the submitted study is subject to deficiencies. I would recommend publishing considering the major revisions and addressing the specific comments that follow. Furthermore, the text needs rearrangements, especially in terms of the structure in the presentation of the methodology, figures, and discussion in specific sub-sections (see comments below).
Major comments:
I don’t see any direct evidence presented by the authors that supports the statement that “ALADIN is more susceptible to signal attenuation from CALIOP”. I suggest removing that statement from the abstract and elsewhere.

The authors state that their work paves the way for forthcoming spaceborne HSRL missions, particularly the ESA ATLID space lidar (set for a 2024 launch) and Aeolus-2. How does this paper do that?

In abstract and elsewhere, please mention the limitation of ALADIN/Aeolus on retrieving the total particle backscatter coefficient, otherwise the reader will get into confusion (for example, while the term “co-polar backscatter” is not a perfectly valid one, it could be used to distinguish from total backscatter).

QC flags: The paper extensively utilizes CALIOP and ALADIN, however pre-processing and Quality-Assurance criteria applied on the data used for the comparisons of backscatter and extinction coefficients are not sufficiently presented nor discussed.

Spectral conversions: Furthermore, assumptions on 532-355 nm spectral dependencies on depolarization, lidar ratio, extinction and backscatter should be supported (preferably using as reference the paper of Floutsi et al., (2023) where averages for these properties using long-term ground-based lidar measurements are given).

For 4 & 5 (aforementioned comments) to be tackled, I would suggest that the authors extensively elaborate on the datasets and methodology sections, to clearly present the processing chains and assumptions leading to the study conclusions.

Collocation: Section 2.3 “Collocation of Aeolus and CALIPSO”: The paper presents and discusses the following collocation criteria: (1) “3◦×3◦ grid, sets the maximum temporal disparity at 9 hours and the maximum spatial difference at 200 km”, (2) “a spatial distance under 1◦ and a temporal discrepancy not exceeding 24 hours, based on data between 30th June 2019 and 28th September 2021”, and (3) “between 30◦ N and 30◦ S, most collocated observations are within 4 hours and 100 km”. It is not clear at all the selected collocation criteria that are applied in the framework of the study. More important is the authors to discuss atmospheric homogeneity in terms of aerosols and clouds. How is it ensured that the two satellite sensors probe the same air masses? For example, the authors could provide a study on spatiotemporal representativeness in terms of the selected and applied criteria including literature - discussion on NA meteorology. Ensuring that the two systems probe the same air masses is fundamental for the follow-up intercomparison, non assessing it makes the outcome conclusions questionable.

Cloud contamination: The dust transport event examined was extreme, however the extensive presence of clouds may affect the scenes examined. Here, with respect to Aeolus Cloud Filtering, three methods are presented and discussed, the (1) SEVIRI CLM cloud mask, (2) the CM SAF cloud mask, and (3) AEL-FM. However, it is not clear which - if not all of the aforementioned cloud-screening datasets are applied. Please elaborate on this aspect, including discussion of the quality of the applied procedures, assumptions, and uncertainties. With respect to CALIOP, which cloud filtering criteria are applied?

In the conclusions section, the authors draw generic conclusions on Aeolus and CALIPSO, however their work is based on a single event, which is an extreme one, over a specific domain, and for a time period not exceeding a few weeks. Thus, the outcomes should not be treated as generic since the statistical study lacks the depth to support the argument.

Specific Comments:
A CALIPSO-based mean depolarization ratio profile is provided, reporting also mean particulate depolarization of 0.32. However, this depolarization is accompanied by a standard deviation ~ ±0.15 which translates to non-pure dust layers apparent in the atmosphere, resulting in particle depolarization values lower than 0.3. How do the authors treat those layers? Treating them as pure-dust layers and applying dust-related conversion factors contaminate the outcomes, so the authors have to address the aerosol mixtures accordingly. Moreover, please mention the pre-processing chains in terms of particle depolarization ratio profiles leading to the non-noisy profile in Fig.4. How do the authors treat larger than 1 and lower than 0 CALIPSO V4 L2 5km depolarization values?

You can modify Figures 4a, 4b, 5b, and, 5d, to linear instead of logarithmic x-axis scales, add a colorbar for the gradient values, and a second axis reporting on the sample size of profiles resulting in the mean profiles.

Results presented in Figures 4a, 4b, 5b, and, 5d: you may provide statistical metrics reporting on the intercomparison of backscatter and extinction coefficient profiles (e.g., σ, r2, mean/relative biases, …). Prior to doing this analysis the authors should elaborate how they get CALIPSO to the same horizontal and vertical resolution to Aeolus.

You can apply linear scales also to figures 6(b,c,e,f )and 7(b,c,e,f).

Line 198 and Figure 3: According to the authors the method of Ashpole and Washington (2012) is applied. Since this is a crucial section, please provide discussion on the method, assumption, performance, and uncertainties. Since CALIPSO is used, which is the reason for not implementing CALIPSO aerosol subtype classification as dust identified aerosol layers?

Lines 242-248: This sentence actually is generic to a degree that it doesn’t provide any information, since none of the aforementioned source of discrepancies is assessed and no effort in quantifying the effect of each factor is provided in the manuscript. Please elaborate more on this.

Please describe clearly in Section 6 how you apply corrections to CALIPSO and provide the formulas.

Please take care of the units in the manuscript, in some places they are missing.

Editorial corrections:
Lines 12-13: Provide the wavelength

Lines 59-75: Here additional references on the technical documents from ESA and the published papers could be used to support the technical description on the operation principle of ALADIN and the retrieval of the aerosol products, as already listed in related studies using ALADIN/Aeolus aerosol data.

Line 120: The version of the SCA algorithm (e.g. Baseline XX) used for retrieving the ALADIN profiles that are presented in the study could be mentioned here.

Lines 210-223: For the conversion between the total backscatter coefficient and the backscatter coefficient that ALADIN measures (co-polar backscatter) Paschou et al., AMT, 2022 could be also cited here, since it provides an extended discussion on the physical background along with the equations for converting the backscatter coefficient and the lidar ratio.

Lines 234-236: The ALADIN backscatter is depicted in Fig.4 (b) while the CALIOP backscatter in Fig.4 (a). Please correct the typos.

Lines 237-238: How much is the standard deviation of the mean value of particle linear depolarization ratio at 532 nm that is used (δlinear, 532part=0.32) for correcting the ALADIN backscatter coefficient from the missing cross-polar backscattered signal?

Line 237: It would be nice if “particle” could be used to define the type of depolarization ratio that is discussed. Here most probably is the particle depolarization ratio. Please elaborate through the manuscript.

Figure 5: Do you mean “Comparison of the extinction coefficients between CALIOP and ALADIN etc ”?

Lines 275-276: “Collocated CALIOP retrievals are upscaled from a resolution of 0.03 km to match this resolution." => 0.06 km in CALIPSO L2/L3.

Line 287: The authors mention that the case of 19th June 2020 is characterized by high AOD values. How high are the AOD values for this case, and from which source did they have been obtained?

Lines 313-315: It is not clear if the CALIOP corrected AOD (i.e. after exclusion of the fully attenuated bins) or the CALIOP AOD is used in the comparison. Please be more specific.

Figure 9: “The CALIOP AODs have excluded profiles containing fully attenuated bins at any altitudes”. Here it is not clear if the authors refer to the AOD values calculated from the CALIOP corrected data. Please make the description more specific. For example you could change to “The CALIOP AODs (red) are calculated using all the available profiles and the CALIOP corrected AODs (green) are calculated after excluding profiles containing fully attenuated bins at any altitudes” or similar.

Line 319: Regarding the dust lidar ratio variability you can also refer to the latest work of Floutsi et al., AMT, 2023.

Line 328: According to figure 10, the presented ALADIN LR has been corrected for the missing cross-polar backscattered signals. However, the reported LR value (63.5 sr; how much is its standard deviation?) is very high, when ground-based measurements at 355 nm show a variability of 53 ± 7 sr (Floutsi et al., AMT, 2023), so the value is on the upper limit of the distribution.

Lines 329-331 and Table 1: The authors could also refer to the recent work of Floutsi et al., AMT, 2023 where an extended data collection of aerosol-type-dependent optical properties are presented from a large number of archived observations and experimental campaigns.

Lines 337 – 342: Only now, after Fig. 10 and its discussion, it is explained by the authors what the “corrected CALIOP” represents in Fig. 9. Moreover, Fig. 9 is further discussed after the first discussion that took place in lines 311–315. This structure is very confusing, so it would be good if you could re-organize the text in order to discuss the figure outcomes just before or after you first introduce the figure and also to explain how new products/parameters that are depicted in the figure are obtained.

Line 340: What do the authors mean with a “subset of CALIOP AOD” and where do they refer to (e.g. to “CALIOP corrected” AOD)?

Lines 343 – 358: In Fig. 11, the authors are using the CALIOP extinction profiles after correction with the correction factor for lidar ratio (LR_ALADIN/LR_CALIOP), in which the fully attenuated bins are filtered out. Then they group the extinction profiles based on the total AOD from CALIOP (AOD < 1.8 and AOD > 1.8). It is not clear the reason for selecting a threshold of 1.8 AOD and not some other threshold. Moreover, why do the authors state that for the group of extinction profiles with AOD < 1.8 the strongly attenuated bins are included in the profiles, but on the other hand, for the group of extinction profiles with AOD > 1.8 the strongly attenuated bins are properly excluded?

Line 354: The filtering strategy is the exclusion of the fully attenuated bins from the CALIOP observations? Please be more specific here.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC3
- AC2: 'Reply on RC3', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC2
RC4:
'Comment on egusphere-2023-2252', Anonymous Referee #4, 09 Nov 2023
The paper of Song et al exploits the aerosol spin-off products for the European wind lidar mission Aeolus for a specific extreme event, namely a heavy Sharan dust outbreak observed over the Atlantic. To compensate the drawbacks of the wind lidar Aladin in the vicinity of non-spherical particles, the authors use the polarization observations from NASA CALIPSO mission to correct Aeolus‘ co-polar backscatter coefficient and SEVIRI dust mask as a cloud screening proxy.
The paper is of interest for the scientific community, exploits the synergy between different space-born profiles and describes an intense extreme event in a changing climate based on vertically resolved optical properties. It furthermore shows, how different sensors could be used in a synergistic way to retrieve optimized aerosol profiles. It is this worth publishing, however, only after addressing the issues listed below.

Major/General comments:
Most of the comparisons, especially of extinction coefficient are plotted on logarithmic scale in separated plots. However, by doing so, it is not possible to see major differences in case of strong backscatter and extinction as it is the case for in this paper. Thus, comparisons should be shown in linear scale an, maybe divided by low and high values, to allow the reader to see, how well the results agree. Best, also in the same Figure. Later you state, that „Assessing the accuracy of ALADIN’s aerosol retrievals within the upper atmospheric region exceeding the dust layer is beyond the scope of this work.“ Thus, there is no need to use a log scale.

At least, I cannot follow many conclusions you have drawn based on the log-based figures you provided.

In my opinion, the first part of the section 4, case study is a methodological part and should be put in a respective section. This section should be expanded with respect to CALIPSO observation which have been used: E.g., the quality controls are not clearly described. I can't figure out which CALIOP cloud screening is applied.

Furthermore, have you used mean Calipso depol profiles for correction or did you make a case by case correction? It is not clearly stated.

In Section 6, important information is missing, e.g. on how the columnar AOD is calculated from Calipso profiles which obviously are not available down to the ground. Currently, the section is really misleading.

As the authors focus on a specific atmospheric scene at a specific time of the Aeolus mission, conclusions drawn should not be too general.

Specific comments:
1: Please invert color scale, in all other plots of this color map high values are dark and low ones light.

Fig. 4b: Please use a different color for the mean, hardly seen. And please use contours instead of gradients as wording.

Please explain all abbreviations (e.g. HSRL) and reference if appropriate (e.g. for A-Train).

Lines 104-105: "corresponding to an along-track horizontal resolution of approximately 87 km". Here it should be mentioned that this nominal along-track horizontal resolution of ~87km corresponds to one Basic Repeat Cycle (BRC) also referred as Observation, and pointing to the L2A Algorithm Technical Basis Document (ATBD).

Line 105: "Each observation is comprised of 24 vertical bins". This is only valid for SCA, the SCAmb used within the study being aligned with only 23 vertical range-bin.

Line 121 "Level-2 SCAmb products are used" and line 183 "the ALADIN L2A data from the study period". Here the L2A baseline reference (i.e. 2AXX) should be clearly mentioned as the exact date of downloading from the ESA ADDF.

Line 174: "official L2A Aeolus processor". The term official could be replaced by operational.

Line 235, „For the sake of comparison, the ALADIN aerosol retrievals 235 in Fig. 4 (a) have been converted from co-polar to total backscatter coefficients, aligning them with the CALIOP aerosol retrievals in Fig. 4 (b).“I think you mixed up here something. Please check!

Lines 237- 239: Did you use the mean depol value of Calipso or each single profile? At least stating that the depol ratio remains constant with a mean value of 0.32 is quite confusing.

Fig. 5. Caption wrong, its extinction not backscatter

Line 264 "CALIOP’s extinction retrieval relies on a predefined lidar ratio tailored for specific aerosol types". Here it might be interesting to point the lidar ratio value assigned to the tropospheric aerosol class highlighted in Figure 8.

280: „For Fig. 6(b), both measurements show an extinction of ∼15 km−1, except where ALADIN observations fail quality-control.“ How can I see that they fail quality control? Are these the non-existent data points? This is not clear. Please describe better and also which quality control was applied.

Fig. 6 and 7: Could you also plot the evolution of the Aeolus lidar ratio (after correction).

Line 293: "This example illuminates a common problem with ALADIN extinction retrieval: retrievals at the base of a thick aerosol layer are very likely to be significantly underestimated or excluded" by quality control due to low SNRs. What does this statement refer to ? Which test cases or analysis have been used to qualify it as a common issue ?

297: „A noteworthy observation is that ALADIN persistently records an extinction coefficient higher by ∼2 compared to CALIOP“ I do not see that in you plots.

Figure 9. Why do you not provide the Aeolus AOD as well?

Fig. 10. Please clearly indicate the wavelength in the plot (355 for Aeolus, 532 nm for CALIOP)

335: “This scaling method is an approximation, as a different lidar ratio can alter the lidar profile and subsequently affect the retrieval process.” Please describe a bit more. I guess you mean the lidar ratio choice already influences the backscatter retrieval during the Klett inversion?

Table 1.: I recommend to check the values and complete it with the recent publication of Floutsi et al. (DeLiAn).

Line: 343ff: According to Figure 11, there are no extinction profiles below ca. 1.8 km. Thus I was wondering how did you calculate the total AOD from Calipso? Did you interpolate? Did you just skip the lowermost altitudes? Please clearly describe.

Line 335: “The grouped extinction profile indicate a mean layer AOD 355 of 0.413 between the 0 and 2.4 km layer, accompanied by a considerable standard deviation due to the random distribution of strongly attenuated bins along the satellite track. Conversely, the alternative set of measurements devoid of strongly attenuated bins demonstrates a layer AOD of 1.015 between 0 and 2.4 km.” I do not understand this statement at all, please rephrase and describe more!

Conclusion: Please highlight a bit more the synergistic use of Calipso and Aeolus and Seviri for optimum aerosol profiles in this specific dust case.

Language:
Line 34: „Spaceborne lidars have the advantage of minimal aerosol loading between the instrument and the calibration region.“ .. I know what you mean with that, but a non-expert user will not understand what is mean there. Please rephrase.
- Lines 109-110: "Standard Correction Algorithm (SCA)" and "Standard Correction Algorithm middle bin (SCAmb)" should be replaced by "Standard Correct Algorithm (SCA)" and "Standard Correct Algorithm middle bin (SCAmb)"
References:

Floutsi, A. A., Baars, H., Engelmann, R., Althausen, D., Ansmann, A., Bohlmann, S., Heese, B., Hofer, J., Kanitz, T., Haarig, M., Ohneiser, K., Radenz, M., Seifert, P., Skupin, A., Yin, Z., Abdullaev, S. F., Komppula, M., Filioglou, M., Giannakaki, E., Stachlewska, I. S., Janicka, L., Bortoli, D., Marinou, E., Amiridis, V., Gialitaki, A., Mamouri, R.-E., Barja, B., and Wandinger, U.: DeLiAn – a growing collection of depolarization ratio, lidar ratio and Ångström exponent for different aerosol types and mixtures from ground-based lidar observations, Atmos. Meas. Tech., 16, 2353–2379, https://doi.org/10.5194/amt-16-2353-2023, 2023.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC4
- AC3: 'Reply on RC4', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC3

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2023-2252', Anonymous Referee #1, 01 Nov 2023
This manuscript presents the meaningful demonstration of the capabilities of ALADIN in retrieving aerosol optical properties, specifically the backscatter coefficient, extinction coefficient, and lidar ratio. The dust layers’ lidar ratios used for CALIOP is also revised according the simultaneous measurements of ALADIN. The manuscript is well written and its contents are of high quality and scientific interest. The benefits of this study would be great for the accurate estimation of Aeolus and CALIOP aerosol data products. Hence, I recommend the acceptance of this manuscript after the necessary revisions.

The specific comments are listed below:
My main concern about the extinction coefficient/backscatter coefficient comparisons between CALIOP (at 532nm) and ALADIN (at 355nm) is their wavelength dependence. In this manuscript, the authors compare the aerosol products directly without any wavelength convert, even you mentioned it in line 244.

Line 66: could you please give some more detailed comments on why the extinction coefficient is not affected by the misdetection of the cross-polar component?

Line 104-105: The authors should be aware that the horizontal resolution for Rayleigh channel and Mie channel is different.

Line 120: Have the authors ever try to estimate the performance of the products from MLE? You mentioned the MLE method has positive effect on the products retrieve, however, why the Level-2 SCAmb products are applied in your study?

Figure 1: why the temporal disparity of 9 hours and the maximum spatial difference of 200km are set as thresholds? Is there any physical basis for these selections? For example, wind direction? Air mass transport?

The color bar in Figure 1 somehow misleads me. I suggest the authors may use the color bar oppositely, be like Figure 2.

Line 244: the spectral difference between 532 nm and 355 nm could be corrected somehow with the use of typical Angstrom exponent of dust. Have you ever tried to do this work?

Line 312: what is the time difference between the measurements from MODIS and CALIOP?

The wavelength band that MODIS applied should be pointed out. Hence, we can figure it out whether the wavelength convert should be carried out. From this point of view, the underestimation may be solved.

The technical corrections:
Line 193: “the blue dots in (d) represent the footprint…” should be changed to “the blue dots in (d) represent the footprints…”

Please provide the color bars’ label for the green/blue gradients in Figure 4 and 5.

Why there is only one red profile in Figure 4(a) and 5(a) between 12.5 km and 17.5 km? Is it because there is only one measurement case reach that height? Then I would suggest the authors provide the total numbers of measurements at different heights.

It should be “Comparison of aerosol extinction coefficients…” instead of “Comparison of aerosol backscatter coefficients” in the caption of Figure 5. Please correct it.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC1
- AC4: 'Reply on RC1', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC4-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC4
RC2:
'Comment on egusphere-2023-2252', Anonymous Referee #2, 02 Nov 2023
The authors present the combination of CALIOP and ALADIN measurements of a dust plume over the Atlantic ocean. They show that ALADIN measurements align well with CALIOP observations as long as proper data filtering as applied. They also re-iterate known limitations of the analysis of CALIOP observations of mineral dust by using ALADIN measurements of the dust lidar ratio to show that the dust lidar ratio in the CALIOP data retrieval is still set too low. The paper would benefit from re-organisation and shortening. Hence, minor revisions are needed:
I suggest to organise the work in a more conventional way with fewer sections. Please put methods and results into the corresponding sections rather then mixing them up as in the case study. The introduction is also bit lengthy and could be sharpened towards what's relevant for the presented work.

The text includes plenty of figure descriptions that should be covered solely in the figure captions (e.g. lines 193-196, 233-234, 270-275, 306-309, 3043-346). Please omit from the main text.

Please unify the colour axes in Figs. 1 and 2. The inversion is not very intuitive. Figure 2 might be expanded by a panel that shows all latitudes.

Why are the authors discussion cloud masks that are not being used for this work? I suggest to stick to what has been used (the MSG-SEVIRI dust mask) and to provide a statement that other cloud masks have either not yet been available or less useful for your purpose. Figure 3 would need to be revised accordingly.

Figures 4, 5, and 10 could be improved by adding the number of profiles that contribute at the different height bins. Also, colour scales for the gradients in Figs. 4 and 5 are missing.

The discussion of Figure 4 - particularly of the particle depolarisation ratio - would benefit from comparisons to findings of SAMUM-2 at Cape Verde.

line 251: shouldn't it be feature type identification?

I suggest to add lidar curtain plots to Figs. 6 and 7 as those would clearly demonstrate the effect of signal attenuation. It would also be nice if the comparison of extinction coefficients had some quantitative element, such as a correlation plot.

I don't think that Figure 8 is needed.

It is not clear to my what the investigation related to Fig. 11 and Tab. 2 is supposed to tell the readers. Okay, the mean profile shape is different for cases with AOD below or above an arbitrary threshold. But can this be used somehow? If anything, I would expect that type (b) with the higher extinction peak would correspond to the attenuated lower AOD profiles - but it doesn't. This part of the paper left me puzzled and I suggest to omit it.

I don't think that Section 7 provides any information about the vertical transport of dust aerosol. Fig. 12 is certainly a nice plot that combines the observations of the two platforms. However, it would me more informative if it was to provide information on the longitudinal and height distribution as well. It seems to me that a similar plot could already be produced using much more data from MODIS observations.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC2
- AC1: 'Reply on RC2', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC1
RC3:
'Comment on egusphere-2023-2252', Anonymous Referee #3, 04 Nov 2023
Review report for amt-2023-2252
The paper "Characterisation of dust aerosols from ALADIN and CALIOP measurements'' aims to assess the performance of ALADIN by comparing with CALIOP, and uses the synergy of both sensors to improve and increase information over an extreme dust transport episode. The manuscript falls within the scope of the journal. However, the presentation and discussion of the paper is not clear and the submitted study is subject to deficiencies. I would recommend publishing considering the major revisions and addressing the specific comments that follow. Furthermore, the text needs rearrangements, especially in terms of the structure in the presentation of the methodology, figures, and discussion in specific sub-sections (see comments below).
Major comments:
I don’t see any direct evidence presented by the authors that supports the statement that “ALADIN is more susceptible to signal attenuation from CALIOP”. I suggest removing that statement from the abstract and elsewhere.

The authors state that their work paves the way for forthcoming spaceborne HSRL missions, particularly the ESA ATLID space lidar (set for a 2024 launch) and Aeolus-2. How does this paper do that?

In abstract and elsewhere, please mention the limitation of ALADIN/Aeolus on retrieving the total particle backscatter coefficient, otherwise the reader will get into confusion (for example, while the term “co-polar backscatter” is not a perfectly valid one, it could be used to distinguish from total backscatter).

QC flags: The paper extensively utilizes CALIOP and ALADIN, however pre-processing and Quality-Assurance criteria applied on the data used for the comparisons of backscatter and extinction coefficients are not sufficiently presented nor discussed.

Spectral conversions: Furthermore, assumptions on 532-355 nm spectral dependencies on depolarization, lidar ratio, extinction and backscatter should be supported (preferably using as reference the paper of Floutsi et al., (2023) where averages for these properties using long-term ground-based lidar measurements are given).

For 4 & 5 (aforementioned comments) to be tackled, I would suggest that the authors extensively elaborate on the datasets and methodology sections, to clearly present the processing chains and assumptions leading to the study conclusions.

Collocation: Section 2.3 “Collocation of Aeolus and CALIPSO”: The paper presents and discusses the following collocation criteria: (1) “3◦×3◦ grid, sets the maximum temporal disparity at 9 hours and the maximum spatial difference at 200 km”, (2) “a spatial distance under 1◦ and a temporal discrepancy not exceeding 24 hours, based on data between 30th June 2019 and 28th September 2021”, and (3) “between 30◦ N and 30◦ S, most collocated observations are within 4 hours and 100 km”. It is not clear at all the selected collocation criteria that are applied in the framework of the study. More important is the authors to discuss atmospheric homogeneity in terms of aerosols and clouds. How is it ensured that the two satellite sensors probe the same air masses? For example, the authors could provide a study on spatiotemporal representativeness in terms of the selected and applied criteria including literature - discussion on NA meteorology. Ensuring that the two systems probe the same air masses is fundamental for the follow-up intercomparison, non assessing it makes the outcome conclusions questionable.

Cloud contamination: The dust transport event examined was extreme, however the extensive presence of clouds may affect the scenes examined. Here, with respect to Aeolus Cloud Filtering, three methods are presented and discussed, the (1) SEVIRI CLM cloud mask, (2) the CM SAF cloud mask, and (3) AEL-FM. However, it is not clear which - if not all of the aforementioned cloud-screening datasets are applied. Please elaborate on this aspect, including discussion of the quality of the applied procedures, assumptions, and uncertainties. With respect to CALIOP, which cloud filtering criteria are applied?

In the conclusions section, the authors draw generic conclusions on Aeolus and CALIPSO, however their work is based on a single event, which is an extreme one, over a specific domain, and for a time period not exceeding a few weeks. Thus, the outcomes should not be treated as generic since the statistical study lacks the depth to support the argument.

Specific Comments:
A CALIPSO-based mean depolarization ratio profile is provided, reporting also mean particulate depolarization of 0.32. However, this depolarization is accompanied by a standard deviation ~ ±0.15 which translates to non-pure dust layers apparent in the atmosphere, resulting in particle depolarization values lower than 0.3. How do the authors treat those layers? Treating them as pure-dust layers and applying dust-related conversion factors contaminate the outcomes, so the authors have to address the aerosol mixtures accordingly. Moreover, please mention the pre-processing chains in terms of particle depolarization ratio profiles leading to the non-noisy profile in Fig.4. How do the authors treat larger than 1 and lower than 0 CALIPSO V4 L2 5km depolarization values?

You can modify Figures 4a, 4b, 5b, and, 5d, to linear instead of logarithmic x-axis scales, add a colorbar for the gradient values, and a second axis reporting on the sample size of profiles resulting in the mean profiles.

Results presented in Figures 4a, 4b, 5b, and, 5d: you may provide statistical metrics reporting on the intercomparison of backscatter and extinction coefficient profiles (e.g., σ, r2, mean/relative biases, …). Prior to doing this analysis the authors should elaborate how they get CALIPSO to the same horizontal and vertical resolution to Aeolus.

You can apply linear scales also to figures 6(b,c,e,f )and 7(b,c,e,f).

Line 198 and Figure 3: According to the authors the method of Ashpole and Washington (2012) is applied. Since this is a crucial section, please provide discussion on the method, assumption, performance, and uncertainties. Since CALIPSO is used, which is the reason for not implementing CALIPSO aerosol subtype classification as dust identified aerosol layers?

Lines 242-248: This sentence actually is generic to a degree that it doesn’t provide any information, since none of the aforementioned source of discrepancies is assessed and no effort in quantifying the effect of each factor is provided in the manuscript. Please elaborate more on this.

Please describe clearly in Section 6 how you apply corrections to CALIPSO and provide the formulas.

Please take care of the units in the manuscript, in some places they are missing.

Editorial corrections:
Lines 12-13: Provide the wavelength

Lines 59-75: Here additional references on the technical documents from ESA and the published papers could be used to support the technical description on the operation principle of ALADIN and the retrieval of the aerosol products, as already listed in related studies using ALADIN/Aeolus aerosol data.

Line 120: The version of the SCA algorithm (e.g. Baseline XX) used for retrieving the ALADIN profiles that are presented in the study could be mentioned here.

Lines 210-223: For the conversion between the total backscatter coefficient and the backscatter coefficient that ALADIN measures (co-polar backscatter) Paschou et al., AMT, 2022 could be also cited here, since it provides an extended discussion on the physical background along with the equations for converting the backscatter coefficient and the lidar ratio.

Lines 234-236: The ALADIN backscatter is depicted in Fig.4 (b) while the CALIOP backscatter in Fig.4 (a). Please correct the typos.

Lines 237-238: How much is the standard deviation of the mean value of particle linear depolarization ratio at 532 nm that is used (δlinear, 532part=0.32) for correcting the ALADIN backscatter coefficient from the missing cross-polar backscattered signal?

Line 237: It would be nice if “particle” could be used to define the type of depolarization ratio that is discussed. Here most probably is the particle depolarization ratio. Please elaborate through the manuscript.

Figure 5: Do you mean “Comparison of the extinction coefficients between CALIOP and ALADIN etc ”?

Lines 275-276: “Collocated CALIOP retrievals are upscaled from a resolution of 0.03 km to match this resolution." => 0.06 km in CALIPSO L2/L3.

Line 287: The authors mention that the case of 19th June 2020 is characterized by high AOD values. How high are the AOD values for this case, and from which source did they have been obtained?

Lines 313-315: It is not clear if the CALIOP corrected AOD (i.e. after exclusion of the fully attenuated bins) or the CALIOP AOD is used in the comparison. Please be more specific.

Figure 9: “The CALIOP AODs have excluded profiles containing fully attenuated bins at any altitudes”. Here it is not clear if the authors refer to the AOD values calculated from the CALIOP corrected data. Please make the description more specific. For example you could change to “The CALIOP AODs (red) are calculated using all the available profiles and the CALIOP corrected AODs (green) are calculated after excluding profiles containing fully attenuated bins at any altitudes” or similar.

Line 319: Regarding the dust lidar ratio variability you can also refer to the latest work of Floutsi et al., AMT, 2023.

Line 328: According to figure 10, the presented ALADIN LR has been corrected for the missing cross-polar backscattered signals. However, the reported LR value (63.5 sr; how much is its standard deviation?) is very high, when ground-based measurements at 355 nm show a variability of 53 ± 7 sr (Floutsi et al., AMT, 2023), so the value is on the upper limit of the distribution.

Lines 329-331 and Table 1: The authors could also refer to the recent work of Floutsi et al., AMT, 2023 where an extended data collection of aerosol-type-dependent optical properties are presented from a large number of archived observations and experimental campaigns.

Lines 337 – 342: Only now, after Fig. 10 and its discussion, it is explained by the authors what the “corrected CALIOP” represents in Fig. 9. Moreover, Fig. 9 is further discussed after the first discussion that took place in lines 311–315. This structure is very confusing, so it would be good if you could re-organize the text in order to discuss the figure outcomes just before or after you first introduce the figure and also to explain how new products/parameters that are depicted in the figure are obtained.

Line 340: What do the authors mean with a “subset of CALIOP AOD” and where do they refer to (e.g. to “CALIOP corrected” AOD)?

Lines 343 – 358: In Fig. 11, the authors are using the CALIOP extinction profiles after correction with the correction factor for lidar ratio (LR_ALADIN/LR_CALIOP), in which the fully attenuated bins are filtered out. Then they group the extinction profiles based on the total AOD from CALIOP (AOD < 1.8 and AOD > 1.8). It is not clear the reason for selecting a threshold of 1.8 AOD and not some other threshold. Moreover, why do the authors state that for the group of extinction profiles with AOD < 1.8 the strongly attenuated bins are included in the profiles, but on the other hand, for the group of extinction profiles with AOD > 1.8 the strongly attenuated bins are properly excluded?

Line 354: The filtering strategy is the exclusion of the fully attenuated bins from the CALIOP observations? Please be more specific here.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC3
- AC2: 'Reply on RC3', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC2
RC4:
'Comment on egusphere-2023-2252', Anonymous Referee #4, 09 Nov 2023
The paper of Song et al exploits the aerosol spin-off products for the European wind lidar mission Aeolus for a specific extreme event, namely a heavy Sharan dust outbreak observed over the Atlantic. To compensate the drawbacks of the wind lidar Aladin in the vicinity of non-spherical particles, the authors use the polarization observations from NASA CALIPSO mission to correct Aeolus‘ co-polar backscatter coefficient and SEVIRI dust mask as a cloud screening proxy.
The paper is of interest for the scientific community, exploits the synergy between different space-born profiles and describes an intense extreme event in a changing climate based on vertically resolved optical properties. It furthermore shows, how different sensors could be used in a synergistic way to retrieve optimized aerosol profiles. It is this worth publishing, however, only after addressing the issues listed below.

Major/General comments:
Most of the comparisons, especially of extinction coefficient are plotted on logarithmic scale in separated plots. However, by doing so, it is not possible to see major differences in case of strong backscatter and extinction as it is the case for in this paper. Thus, comparisons should be shown in linear scale an, maybe divided by low and high values, to allow the reader to see, how well the results agree. Best, also in the same Figure. Later you state, that „Assessing the accuracy of ALADIN’s aerosol retrievals within the upper atmospheric region exceeding the dust layer is beyond the scope of this work.“ Thus, there is no need to use a log scale.

At least, I cannot follow many conclusions you have drawn based on the log-based figures you provided.

In my opinion, the first part of the section 4, case study is a methodological part and should be put in a respective section. This section should be expanded with respect to CALIPSO observation which have been used: E.g., the quality controls are not clearly described. I can't figure out which CALIOP cloud screening is applied.

Furthermore, have you used mean Calipso depol profiles for correction or did you make a case by case correction? It is not clearly stated.

In Section 6, important information is missing, e.g. on how the columnar AOD is calculated from Calipso profiles which obviously are not available down to the ground. Currently, the section is really misleading.

As the authors focus on a specific atmospheric scene at a specific time of the Aeolus mission, conclusions drawn should not be too general.

Specific comments:
1: Please invert color scale, in all other plots of this color map high values are dark and low ones light.

Fig. 4b: Please use a different color for the mean, hardly seen. And please use contours instead of gradients as wording.

Please explain all abbreviations (e.g. HSRL) and reference if appropriate (e.g. for A-Train).

Lines 104-105: "corresponding to an along-track horizontal resolution of approximately 87 km". Here it should be mentioned that this nominal along-track horizontal resolution of ~87km corresponds to one Basic Repeat Cycle (BRC) also referred as Observation, and pointing to the L2A Algorithm Technical Basis Document (ATBD).

Line 105: "Each observation is comprised of 24 vertical bins". This is only valid for SCA, the SCAmb used within the study being aligned with only 23 vertical range-bin.

Line 121 "Level-2 SCAmb products are used" and line 183 "the ALADIN L2A data from the study period". Here the L2A baseline reference (i.e. 2AXX) should be clearly mentioned as the exact date of downloading from the ESA ADDF.

Line 174: "official L2A Aeolus processor". The term official could be replaced by operational.

Line 235, „For the sake of comparison, the ALADIN aerosol retrievals 235 in Fig. 4 (a) have been converted from co-polar to total backscatter coefficients, aligning them with the CALIOP aerosol retrievals in Fig. 4 (b).“I think you mixed up here something. Please check!

Lines 237- 239: Did you use the mean depol value of Calipso or each single profile? At least stating that the depol ratio remains constant with a mean value of 0.32 is quite confusing.

Fig. 5. Caption wrong, its extinction not backscatter

Line 264 "CALIOP’s extinction retrieval relies on a predefined lidar ratio tailored for specific aerosol types". Here it might be interesting to point the lidar ratio value assigned to the tropospheric aerosol class highlighted in Figure 8.

280: „For Fig. 6(b), both measurements show an extinction of ∼15 km−1, except where ALADIN observations fail quality-control.“ How can I see that they fail quality control? Are these the non-existent data points? This is not clear. Please describe better and also which quality control was applied.

Fig. 6 and 7: Could you also plot the evolution of the Aeolus lidar ratio (after correction).

Line 293: "This example illuminates a common problem with ALADIN extinction retrieval: retrievals at the base of a thick aerosol layer are very likely to be significantly underestimated or excluded" by quality control due to low SNRs. What does this statement refer to ? Which test cases or analysis have been used to qualify it as a common issue ?

297: „A noteworthy observation is that ALADIN persistently records an extinction coefficient higher by ∼2 compared to CALIOP“ I do not see that in you plots.

Figure 9. Why do you not provide the Aeolus AOD as well?

Fig. 10. Please clearly indicate the wavelength in the plot (355 for Aeolus, 532 nm for CALIOP)

335: “This scaling method is an approximation, as a different lidar ratio can alter the lidar profile and subsequently affect the retrieval process.” Please describe a bit more. I guess you mean the lidar ratio choice already influences the backscatter retrieval during the Klett inversion?

Table 1.: I recommend to check the values and complete it with the recent publication of Floutsi et al. (DeLiAn).

Line: 343ff: According to Figure 11, there are no extinction profiles below ca. 1.8 km. Thus I was wondering how did you calculate the total AOD from Calipso? Did you interpolate? Did you just skip the lowermost altitudes? Please clearly describe.

Line 335: “The grouped extinction profile indicate a mean layer AOD 355 of 0.413 between the 0 and 2.4 km layer, accompanied by a considerable standard deviation due to the random distribution of strongly attenuated bins along the satellite track. Conversely, the alternative set of measurements devoid of strongly attenuated bins demonstrates a layer AOD of 1.015 between 0 and 2.4 km.” I do not understand this statement at all, please rephrase and describe more!

Conclusion: Please highlight a bit more the synergistic use of Calipso and Aeolus and Seviri for optimum aerosol profiles in this specific dust case.

Language:
Line 34: „Spaceborne lidars have the advantage of minimal aerosol loading between the instrument and the calibration region.“ .. I know what you mean with that, but a non-expert user will not understand what is mean there. Please rephrase.
- Lines 109-110: "Standard Correction Algorithm (SCA)" and "Standard Correction Algorithm middle bin (SCAmb)" should be replaced by "Standard Correct Algorithm (SCA)" and "Standard Correct Algorithm middle bin (SCAmb)"
References:

Floutsi, A. A., Baars, H., Engelmann, R., Althausen, D., Ansmann, A., Bohlmann, S., Heese, B., Hofer, J., Kanitz, T., Haarig, M., Ohneiser, K., Radenz, M., Seifert, P., Skupin, A., Yin, Z., Abdullaev, S. F., Komppula, M., Filioglou, M., Giannakaki, E., Stachlewska, I. S., Janicka, L., Bortoli, D., Marinou, E., Amiridis, V., Gialitaki, A., Mamouri, R.-E., Barja, B., and Wandinger, U.: DeLiAn – a growing collection of depolarization ratio, lidar ratio and Ångström exponent for different aerosol types and mixtures from ground-based lidar observations, Atmos. Meas. Tech., 16, 2353–2379, https://doi.org/10.5194/amt-16-2353-2023, 2023.
Citation: https://doi.org/10.5194/egusphere-2023-2252-RC4
- AC3: 'Reply on RC4', Rui Song, 15 Jan 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2252/egusphere-2023-2252-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2252-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Rui Song on behalf of the Authors (15 Jan 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (19 Jan 2024) by Ulla Wandinger

RR by Anonymous Referee #3 (21 Jan 2024)

RR by Anonymous Referee #1 (25 Jan 2024)

RR by Anonymous Referee #2 (25 Jan 2024)

RR by Anonymous Referee #4 (05 Feb 2024)

Suggestions for revision or reasons for rejection

The authors have done a huge effort and tackled many aspects of the 4 reviewers.
The manuscript is now much clearer and better understandable and has significantly improved. However, it has not yet its full potential. Therefore, I have still some open, minor comments mostly related to responses to my previous concerns.

• Linear vs. logarithmic plotting
Thanks for making this approach and providing the linear plots to this response.
However, I still believe, as well as reviewer 3, that the linear scale would fit much better to your manuscript especially as you do not focus on the very small backscatter coefficients! Furthermore, it is anyhow questionable how trustable these values in regions with such low aerosol content are as they are beyond the detection threshold anyhow. Thus, please really consider again, which scale you prefer and if it is representative with respect to your argumentation and research focus ( e.g. signal attenuation in strong dust plumes)!
IF you stay with log scale, you should at least plot CALIOP and ALADIN mean profiles in the same plot for comparison, e.g. as Subfigure c.

• Me: In Section 6, important information is missing, e.g., on how the columnar AOD is calculated from Calipso profiles which obviously are not available down to the ground. Currently, the section is really misleading.
Authors: The CALIPSO column AOD is derived by integrating the 532 nm aerosol extinction profile from the 5 km Aerosol Profile Products. Importantly, we have excluded profiles containing fully attenuated bins. This detail has now been added to the last sentence in the first paragraph of Section 4 (updated Section number).
Me again: This is okay, but please state again that you exclude fully attenuated profiles and that in most time the marine boundary layer is not captured and thus the comparison to MODIS is per se biased to underestimation.

• Me: 297 „A noteworthy observation is that ALADIN persistently records an extinction coefficient higher by ∼2 compared to CALIOP“ I do not see that in you plots.
Authors: Thank you for your comment. This is illustrated in Fig.7(e), specifically within the latitude ranges of 8° N to 14° N, and 20° N to 22° N.
Me again: That's the problem of the log scales - it is hardly seen...
Nevertheless, I think the wording "persistently" is not valid in this context, as you also show under estimation ins the same figure. Please rephrase !

• Me: Figure 9. Why do you not provide the Aeolus AOD as well?
Authors: In this case of an extreme dust event, a significant number of Aeolus profiles contain bins with missing extinction retrievals. Given Aeolus's 1 km vertical resolution within the dust plume, attempting to integrate the extinction to compute columnar AOD would introduce a substantial bias. Consequently, our focus in this analysis was on utilising Aeolus's lidar ratio to correct CALIOP's extinction and AOD retrievals, rather than on the direct use of Aeolus AOD.
Me again: This is logical. Correct. But still, It would have be interesting to see the AOD profiles even though comparison results might not be well.

• Me: Line 343ff According to Figure 11, there are no extinction profiles below ca. 1.8 km. Thus, I was wondering how you did calculate the total AOD from Calipso? Did you interpolate? Did you just skip the lowermost altitudes? Please clearly describe.
Authors: The extinction is not zero below 1.8 km; merely < 1e-2 as shown in Fig.11. The integration of extinction coefficients to obtain AOD has considered all altitudes down to the surface.
Me again: Nevertheless, it is in contradiction with reality as there is a marine BL below. Thus, please make a statement in text. It is still a bit unclear, what conclusion one can draw out of it with these biased extinction profiles.

• Me: Line 34 „Space-borne lidars have the advantage of minimal aerosol loading between the
instrument and the calibration region. “ .. I know what you mean with that, but a non-expert
user will not understand what is mean there. Please rephrase.
Authors: This sentence has been rephrased as “Space-borne lidars often self-calibrate by assuming some section of the atmosphere lacks aerosol contamination, typically the stratosphere.”
Me again: well, this sentence is confusing again and has now a different meaning, what about “Space-borne lidars are often calibrated in atmospheric regions for which a very low aerosol content is assumed, i.e., typically the stratosphere. One further advantage is, that in contrast to ground-based lidars, no significant aerosol contribution is expected between the space-borne lidar and this calibration region.”

New comments
Line 197 of the tracked-changed-highlighted manuscript:
Please state here also the Aeolus Data Version (i.e., 2A11) when mentioning the CALIPSO version.

Figure 5 ff: As a compromise, could you plot the mean of both instruments in linear scale as Sub-Figure c?

Line 378 Please add „...AOD underestimation cannot be attributed to lost retrievals from the dust’s bottom layer AND MARINE BOUNDARY LAYER“. In general, a bit more discussion on the underestimation of CALIOP could be made. E.g. not only focussing on lidar ratio improvements but also discuss maybe on missing aerosol layers (marine BL etc.).

Line 397 of the tracked-changed-highlighted manuscript: The lidar ratio is actually not derived from CALIOP, maybe „ read out from each individual CALIOP profile“ is better?

Line 424 of the tracked-changed-highlighted manuscript:
You may add „…It is this latter set of profiles that tends to yield AOD values consistent with those derived from MODIS observations EVEN THOUGH THE MARINE BL IS EXCLUDED“

Hide

ED: Publish subject to minor revisions (review by editor) (05 Feb 2024) by Ulla Wandinger

AR by Rui Song on behalf of the Authors (15 Feb 2024) Author's response Author's tracked changes Manuscript

ED: Publish as is (22 Feb 2024) by Ulla Wandinger

AR by Rui Song on behalf of the Authors (04 Mar 2024)

Journal article(s) based on this preprint

26 Apr 2024

Characterization of dust aerosols from ALADIN and CALIOP measurements

Rui Song, Adam Povey, and Roy G. Grainger

Atmos. Meas. Tech., 17, 2521–2538, https://doi.org/10.5194/amt-17-2521-2024,https://doi.org/10.5194/amt-17-2521-2024, 2024

Short summary

Rui Song, Adam Povey, and Roy G. Grainger

Viewed

Total article views: 683 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
461	190	32	683	15	15

HTML: 461
PDF: 190
XML: 32
Total: 683
BibTeX: 15
EndNote: 15

Views and downloads (calculated since 11 Oct 2023)

Month	HTML	PDF	XML	Total
Oct 2023	180	80	6	266
Nov 2023	95	30	9	134
Dec 2023	25	15	3	43
Jan 2024	50	28	5	83
Feb 2024	39	13	4	56
Mar 2024	34	17	2	53
Apr 2024	38	7	3	48
May 2024	0
Jun 2024	0
Jul 2024	0
Aug 2024	0

Cumulative views and downloads (calculated since 11 Oct 2023)

Month	HTML	PDF	XML	Total
Oct 2023	180	80	6	266
Nov 2023	95	30	9	134
Dec 2023	25	15	3	43
Jan 2024	50	28	5	83
Feb 2024	39	13	4	56
Mar 2024	34	17	2	53
Apr 2024	38	7	3	48
May 2024	0
Jun 2024	0
Jul 2024	0
Aug 2024	0

Viewed (geographical distribution)

Total article views: 636 (including HTML, PDF, and XML) Thereof 636 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 29 Aug 2024

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (27479 KB)
Metadata XML

Short summary

In our study, we explored aerosols, tiny atmospheric particles affecting Earth's climate. Using data from two lidar-equipped satellites, we examined a 2020 Saharan dust event. The newer ALADIN satellite's results aligned with the CALIOP satellite. By merging their data, we corrected CALIOP's discrepancies, enhancing the dust event depiction. This underscores the significance of advanced satellite instruments in aerosol research. Our findings pave the way for upcoming satellite missions.


Total:	0
HTML:	0
PDF:	0
XML:	0