the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Sep 2025
Transported African Dust in the Lower Marine Atmospheric Boundary Layer is Internally Mixed with Sea Salt Contributing to Increased Hygroscopicity and a Lower Lidar Depolarization Ratio
Abstract. Saharan dust is transported across the Atlantic, yet the chemical, physical, and morphological transformations dust undergoes within the marine atmospheric boundary layer (MABL) remain poorly understood. These transformations are critical for understanding dust's radiative and geochemical impacts, representation in atmospheric models, and detection via lidar remote sensing. Here, we present coordinated observations from the Office of Naval Research's Moisture and Aerosol Gradients/Physics of Inversion Evolution (MAGPIE) August 2023 campaign at Ragged Point, Barbados. These include vertically resolved single-particle analyses, mass concentrations of dust and sea spray, and High Spectral Resolution Lidar (HSRL) retrievals. Single-particle data show that dust within the Saharan Air Layer (SAL) remains externally mixed, with a corresponding high HSRL-derived linear depolarization ratio (LDR) of ~0.3. However, at lower altitudes, dust becomes internally mixed with sea spray, resulting in increased particle sphericity likely due to an increase in hygroscopicity, which suppresses the LDR signal to below 0.1 even in the presence of high dust loadings (e.g., ~120 µg/m3). The low depolarization in the presence of high dust in the MABL is likely due to a combination of the differences between the single scattering properties of dust and spherical particles, and the potential modification of the dust optical properties from an increased hygroscopicity of dust caused by the mixing with sea salt in the humid MABL. These results highlight the importance of the aerosol particle mixing state when interpreting LDR-derived dust retrievals and estimating surface dust concentrations in satellite products and atmospheric models.
-
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
(1754 KB)
-
Supplement
(497 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1754 KB) - Metadata XML
-
Supplement
(497 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
- RC1: 'Comment on egusphere-2025-4584', Anonymous Referee #1, 07 Oct 2025
-
RC2: 'Comment on egusphere-2025-4584', Konrad Kandler, 13 Oct 2025
Review of Sujan Shrestha et al., “Transported African Dust in the Lower Marine Atmospheric Boundary Layer is Internally Mixed with Sea Salt Contributing to Increased Hygroscopicity and a Lower Lidar Depolarization Ratio”
Content Summary
Saharan dust transported over the Atlantic to Barbados undergoes transformations in the marine atmospheric boundary layer (MBL) that are poorly constrained by remote sensing alone. During the August 2023 MAGPIE campaign, the authors combined high‐spectral resolution lidar (HSRL) profiles with vertically resolved single-particle chemical and morphological data. They find that while dust in the Saharan Air Layer (SAL) remains externally mixed and yields high lidar linear depolarization ratios (LDR ≈ 0.30), dust descending into the lower MBL becomes internally mixed with sea spray, acquiring more spherical, hygroscopic character that strongly suppresses the depolarization signal (LDR < 0.10) even at high dust loads. The discrepancy between expected and observed LDR is further compounded by differing backscatter efficiencies (lidar ratios) of dust vs marine aerosols, which bias the lidar’s sensitivity toward more spherical particles. The authors conclude that neglecting mixing state and morphological evolution can lead to underestimates of surface dust concentrations from depolarization-based retrievals, with implications for satellite retrievals, dust modelling, and air quality analyses.
I consider the topic as interesting and relevant, suitable for ACP. The work adds new detailed data to the pool. The paper written in a clear language and makes a nice reading.
I have some major points regarding details of the methodology (details further down)
- When making conclusions from the aerosol concentrations and the single particle data for the interpretation of the lidar measurements, the step of including ambient conditions (humidity) is missing. As key statements of this work depend on the comparison, this must be regarded.
- With respect to SEM: there is information missing for the classification schemes and how the shape is calculated. This is relevant for understanding the results and a comparison with similar works. This comparison is largely missing, too; it could however help a lot in assessing the relevance of the present work.
- The number of particles analyzed from the airborne sample is too low to allow for conclusions without a careful characterization of statistical significance.
I encourage the authors to work on the major issues, as I think the work should be published. I hope my comments are helpful.
Details
Line 141: As far as I know the BACO tower was constructed to have the top outside the lowermost marine boundary layer. Is it there justified to call the concentrations ‘surface’?
Wouldn’t it have made sense, if sea salt mixing is a major topic, to compare the top of the tower with measurements at its base or near sea level?
Line 156-157: How about sulfate and organics, that might come along?
Line 175: The selection of paper should be rethought. If you want to refer to general CCSEM papers, the first ones were probably in the 80s, like 10.1080/00022470.1983.10465674 and 10.1016/0048-9697(87)90438-4. If you want to refer to the use of CCSEM for mixing state of dust and marine particles, it could include also 10.1126/science.232.4758.1620, 10.1016/S1352-2310(03)00506-5, 10.1029/2005JD005810, 10.5194/acp-18-13429-2018, 10.5194/acp-25-5743-2025, which worked in comparable settings.
Line 185: Please comment on the quantification limits of N (and O). In many studies, these elements are excluded, if the substrate contains carbon, as the quantification is assumed to have a large error.
Line 197: When the TEM grids are carbon coated and the film contains carbon, why is only Cu excluded?
Line 202-203: What were the criteria used for that? E.g., how were organics and aged sea spray distinguished (Fig. 3)? For example, if I look at Fig. 5a, the sulfate particle (no. 6) or the aged sea spray (no. 2) have a much higher C signal than the organic shell of no. 5. In fact, C is visible in all spectra.
Some more detailed information is given in supplement, but it does not get clear for example, what were the criteria to separate dust/sea salt mixtures from pure dust or pure sea salt.
This section needs to be a bit more elaborated, so I suggest that a part of the supplement S2 is moved here or at least mentioned and summarized and amended with the missing information.Line 211: If only 113 particles were analyzed in total, please comment on the statistical significance. How large are the confidence intervals for the percentages given?
Consider this also for the statement made in line 355-357 and 367-369.
I don’t doubt the general statement, as it is known from previous studies you refer to that dust over Barbados is not strongly internally mixed, but the low numbers here limit the data applicability for such a statement.Line 217: Figure 3 seems to be the first figure reference.
Line 238: Why is dry quoted?
Line 256: It seems like the dust AOD precedes the BACO concentrations a bit. Which would make sense regarding the downmixing of the dust from above.
Line 291-294: “Conducive” seems a bit a misleading word here. Most of the common atmospheric compounds with any hygroscopicity should be expected to be droplets at these humidities. As the soundings were done in the afternoon, it can be expected that the humidity has been higher during other times of the day and, as a result, the particles are on the high branch of the growth vs. humidity hysteresis.
Line 293: Why refer to publications on longwave radiation in the Arctic in this context? Or to a work on Ca and Mg salts (without any direct reference to optical properties)? Remove and replace by suitable ones, if Titos et al. is not deemed to be sufficient.
Line 296: Indicate in a, when the soundings were done (e.g., with an arrow or line).
Line 296: In Fig. 2a, what is the red feature on Aug 16 at 5 km? Does it coincide with the high humidity?
Line 329: The expected LDR should be strongly depending on the humidity, as the sea salt particles would grow by factors of 4 in mass/volume at 80 % and beyond 8 at 95 % RH (vs. dry state), e.g. 10.1038/ncomms15883. As a result, the LDR would be expected to be shifted towards low values at high humidities.
Nevertheless, in the manuscript only the dry sea salt concentrations are compared (line 325-326), and eqns. 7 and 8 in the supplement only seem to take the dry concentration into account (line 164-165).
As a result, it seems that this estimate doesn’t have much relevance for the humid layers. I suggest including a growth model into the estimate and rethinking the conclusions made (e.g., lines 345-347, lines 365-366).
This also affects the derived statements, e.g. lines 397-399.Line 338-341: How much contribution would we expect from the particles > 80 µm?
Line 369-374: this seems to refer to the ground-based measurements, but in line 375 you jump back to the airborne ones. Please order more clearly.
Line 394-396: How do we learn about the variability of particle composition aloft to compare with? As far as I can see, there is no series of samples available.
Line 402: … either … ?
Line 403-405: The aspect ratio measured in SEM is representative for SEM conditions, e.g. very dry (vacuum), under which NaCl and other compounds are crystallized. In contrast, in the humid atmosphere we can expect that the sea salt fraction of the particle is in droplet shape. Therefore, it cannot be expected that the aspect ratio for hygroscopic compounds can be transferred from SEM into the atmosphere. This applies also to line 432.
Lines 405 is this the average (i.e. daily) size distribution or simply the integral of all particles? In the latter case, was the number of particles analyzed for each sample similar?
Line 407: ‘… only available from …, …focuses on …’ ?
Line 417-419: As there are different methods to obtain an aspect ratio from a 2D image, specify the one used, because they yield different results (e.g., doi: 10.1029/2019GL086592 and references therein). In particular, for a square like a cube projection, they do not necessarily come up with 1.
Line 419: Probably not ‘cuboid’ here, but ‘cubic’. Note that a cube is also aspheric by definition.
Line 422: If the number after the +/- is the standard deviation, it does not make much sense in the context of a distribution, which is probably rather log-normal than normal. I.e. in this case, the lower end of the standard deviation range would be 0.9, which is an impossible value for the 2D aspect ratio. Either use parameters of a suitable distribution (e.g. 10.1016/j.atmosenv.2007.06.047, 10.1016/j.atmosenv.2015.07.020) or use percentiles.
Line 434: Fig. 5d ‘Aged …’
Line 445: There have been some investigations with CCSEM in and around the Caribbean before (10.5194/acp-22-9663-2022; 10.5194/acp-18-13429-2018; 10.1029/2002JD002935; 10.1098/rsos.231433; 10.5194/acp-5-3331-2005; Roldan, Lizette: Characterization of microphysical properties of Saharan dust aerosols during trans-Atlantic transport. Howard University, 2006; 10.1002/2015GL065693; 10.5194/acp-25-5743-2025). How do these results compare?
E.g. 10.1098/rsos.231433 shows in Fig 5a dust mixing, which is in respect to sea salt similar to the results of the present work (dust mixing on arrival into the boundary layer) but differs for sulfate (already found on the African side). 10.1002/2015GL065693 shows that dust can remain relatively unaltered after trans-Atlantic transport. 10.5194/acp-25-5743-2025 again show considerable mixing at BACO.Line 449-450: Combining data from Fig. 3 (assuming that the data is representative) and from Fig. 5: if the dust becomes gradually more mixed during the downward-mixing in the boundary layer, why do we see less dust/sea salt mixtures at BACO compared to the flight measurements?
Like 452-455: Check if the statement can be kept up.
‘Peak’ refers to conditions at BACO?Supplement, S2 reference to SEM images is not correct (now 5a or missing?)
Fig S2: Commonly kernel density estimators are used to smooth a histogram. But these curves are not smooth. Why? What estimator was used? Or is that a size distribution with a density on y? Please check.
In general: check the capitalization of the labels in the plots. E.g. 5a, image no. 1: capital at the start. No 2.: capital at the second word. No. 5: All words capitalized.
Citation: https://doi.org/10.5194/egusphere-2025-4584-RC2 -
RC3: 'Comment on egusphere-2025-4584', Anonymous Referee #3, 15 Oct 2025
This paper presents measurements of dust and sea salt over Barbados acquired during August 2023 and describes the impact of these aerosols on the ground-based HSRL measurements of depolarization and backscatter. The major focus of the paper is in assessing how these aerosols mix to reduce the lidar linear depolarization ratio (LDR) measured within the marine boundary layer (MBL), thereby frustrating efforts to use the LDR as an indicator of dust. This is an important topic as lidar measurements of LDR are often used to quantify dust amounts and evaluate model predictions of dust transport. The authors use tower and airborne measurements of particle size and composition to show that large amounts of dust were present even when the LDR was low (<0.1). The authors conclude that using LDR will often result in underestimates of surface dust concentration and argue that in situ measurements must be combined with such lidar measurements to correctly determine dust concentration.
The topic is suitable for ACP. The paper is generally easy to read and publication is recommended after the authors address the comments below.
- Line 34 (and elsewhere). “linear depolarization ratio” It’s not clear here (and elsewhere) whether this means volume (or total) linear depolarization ratio or particle depolarization ratio. (see description in the Burton et al. 2015 reference). From the values provided, it appears to be particle depolarization, but this should be clearly indicated here and elsewhere.
- Line 35 The wavelength of the lidar measurement should be indicated (532 nm).
- Section 2.1 Did the authors ever consult the MPI Raman lidar measurement images that are available on-line at https://barbados.mpimet.mpg.de/? These show extensive measurements of aerosol backscatter and depolarization over Barbados that also tend to support the HSRL measurements presented in the paper. These images exist for several years and include August 2023.
- Line 212. This section discusses airborne particulate samples for single particle analysis. If the CTO inlet has a cut point of 3.5 mm, how were samples as large as 25 mm sampled and manually analyzed?
- Line 214. Can the authors be more specific about how the first maximum in the relative humidity profile was assigned to be the CBH? How large did this maximum in RH have to be? Were these CBH values compared with those that can be readily determined from the HSRL measurements?
- Line 225. Readers looking for the details of the SSEC HSRL are supposed to consult the Razenkov and Eloranta references. What are the uncertainties in the HSRL measurements of aerosol backscatter, depolarization, and lidar ratio? I could not find these in the Razenkov reference. I also wonder whether this reference is still relevant for measurements acquired 15 years after the thesis was written (i.e. has the instrument and analyses remained the same during this period?) The Eloranta reference discusses the design and construction of the NCAR airborne HSRL; note that this reference is not readily available at my institution. I also wonder if this reference provides such uncertainty estimates and if so, whether they apply to the ground based lidar in the same way to the airborne lidar. Given that the HSRL measurements of LDR (and to a lesser extent lidar ratio) are a major topic of this paper, there should be at least a brief discussion of the uncertainties in these measurements; such discussion is absent from the paper. How large are the uncertainties in the volume and particulate depolarization and lidar ratio?
- Line 228. When operating pointing vertically, how close to the surface are profiles of aerosol backscatter and depolarization obtained?
- Figure 1c shows that the largest LDR (105 m) occurred on 08-23-23, yet the dust concentration at the top of the tower (Fig. 1a) was very small (negligible?) Why? This seems to indicate that there can be substantial differences between 50 m (top of tower above sea level) and the lowest lidar measurement height (105 m). Were there measurements made at/near the base of the tower or closer to sea level to study the vertical variations close to the surface?
- Figure 1. It would be interesting to see wind speed also during this period to see if/how the depolarization, backscatter, and lidar ratio varied with wind speed and also the amount of sea spray.
- Line 284. “…depolarization measurement responds to the 180-degree backscatter efficiency of the particulates (lidar ratio).” This is why there needs to be a better description of the LDR that is referred to. The particulate depolarization depends on the total (volume) depolarization as well as the particulate scattering ratio. The sentence currently is confusing since it mentions (lidar ratio) at the end of the sentence. I could easily see how a reader not familiar with lidar (and HSRL) can become confused.
- Line 288. Following on the last comment, “…the measured depolarization will be weighted lower due to the backscatter efficiency difference between the aerosol (i.e., lidar ratio)”. Lower than what? Do the author mean instead (or also) “…the measured backscatter will be weighted lower due to the backscatter efficiency difference between the aerosol (i.e., lidar ratio)”.
- Line 309. Maybe this will be discussed later in the paper, but what is the basis of the statement “…due to the predominance of larger but less backscattering mineral dust particles”. If the RH is higher in the MBL, and the particles are hygroscopic, wouldn’t the particles in the MBL be larger?
- Figure 3. It would be interesting to see profiles of the lidar ratio (532 nm) and backscatter color ratio (ratio of backscatter at 532 to 1064 nm) for this case also. These are two important aerosol parameters that can also help in interpreting the aerosol vertical distribution.
- Line 363. “underestimated” should be “overestimated”.
- Line 379. I don’t understand why the transition layer is described as “indifferentiable”. The LDR clearly shows this layer as different from the SAL and MBL.
- Figure 5. What do the black lines in 5b and 5c represent?
- Figure 5c. This seems to show only sea spray and not aged sea spray. Is that correct?
- Figure 5d. Are these measurements made at low RH? If so, wouldn’t the aspect ratios be very different at ambient RH?
- Line 449. Should be “internally” `
- Lines 467-477 This paragraph expresses the desire/need for combining lidar data with in-situ single particle analysis to improve the interpretation of lidar data in dust regions. While this may be true for the best interpretation of such data, this can’t be done routinely and globally on a continuous basis; that is why remote sensing techniques are pursued. The authors seem to indicate that there are no other alternatives and all remote sensing techniques are doomed to significantly underestimate dust impacts near the surface. Have the authors considered whether more advanced remote sensing measurements could provide additional data to help improve the interpretation of lidar data? For example, measurements of backscatter, extinction, depolarization at additional wavelengths? As a suggestion, the reviewers may want to examine the backscatter color ratio (or Angstrom exponent) using various wavelengths (ex. 355-532 nm) in such situations, especially when acquired by such HSRL systems. Examination of such data, in conjunction with depolarization data at multiple wavelengths suggests that, while the depolarization near the surface may be low suggesting that dust concentrations are low, the backscatter color ratio has similar values as observed in the SAL region and which are also different from values in other MBL regions where dust values are low, suggesting that backscatter color ratio may be an indicator of dust. The point is that the authors should not prematurely dismiss remote sensing techniques for providing accurate estimates of dust loading simply because of the limitations of the lidar measurements studied here.
Citation: https://doi.org/10.5194/egusphere-2025-4584-RC3 -
RC4: 'Comment on egusphere-2025-4584', Wenshuai Li & Yang Zhou (co-review team), 18 Oct 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-4584/egusphere-2025-4584-RC4-supplement.pdf
- AC1: 'Comment on egusphere-2025-4584', Sujan Shrestha, 23 Dec 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-4584', Anonymous Referee #1, 07 Oct 2025
Review of “Transported African Dust in the Lower Marine Atmospheric Boundary Layer is Internally Mixed with Sea Salt Contributing to Increased Hygroscopicity and a Lower Lidar Depolarization Ratio” by Shrestha et al.
This paper focuses on field measurements during ONR’s MAGPIE campaign in August 2023 at Ragged Point, Barbados. It focuses on how Saharan dust mixes with sea salt, yielding surprisingly higher hygroscopicity than what would normally be expected for dust. Of course this can be explained by the mixing with sea salt. This act of mixing appears to be localized in the bottom altitudes unlike higher up in the Saharan Air Layer. The enhanced hygroscopicity and mixing at lower altitudes coincides with reduced HSRL-derived depol ratios in contrast to higher altitudes where there is just dust. The topic is important and points to important nuances in the transport of dust and how we really must understand its mixing behavior with other aerosol types to know the chemical, physical, and optical properties of such air masses. I believe the paper will have interest among the readers of the journal.
The methods used are robust and include measurements of single particle properties, concentrations of sea spray and dust, and HSRL retrievals.
The presentation quality is good. In general, the paper was well done and publication is recommended. Some minor suggestions/questions are provided below.
Specific Comments:
Data Availability section: a bit odd to say “will be publicly available” to leave authors in suspense. Why not just make the concentration data available now? Also, it is uncertain about the format of single particle data, but is it common practice to archive those data publicly or not?
Lines 142-157: It would help readers if you explicitly say up front what part of the text is explaining the method of determining dust mass. It was a bit confusing.
Figure 3: nice visual depiction of results.
Line 128-130: The statement “BACO offers an optimal location for intercepting long-range transported Saharan dust with minimal interference from local anthropogenic emissions due to the prevalent Easterly trade winds” should be supported with appropriate references.
Line 208-214: The manual SEM/EDX analysis of a relatively small number of particles (40, 21, and 52) raises concerns regarding statistical representativeness and uncertainty. Since manual particle selection can introduce bias, it will strengthen the methodology if the authors clarify how particles were chosen (randomly or selectively) and whether any estimation of analytical or sampling errors was made.
Line 283-287: The explanation of lidar ratio differences between dust (~40 sr) and marine boundary layer aerosols (~20 sr) is unclear and could be misinterpreted. The phrase “a factor of two different with the marine sourced particles being twice as efficient per scattering cross-section compared to dust at backscattering energy” is confusing. Please clarify whether the statement refers to lidar ratio magnitude or backscatter efficiency.
Line 446-477: The conclusion provides a strong interpretation linking microphysical evidence of dust-sea-spray mixing with the observed low LDR values and their implications for remote sensing. However. The authors may consider strengthening the conclusion by acknowledging additional contributors such as vertical heterogeneity within the MABL (e.g., overlapping marine and dust layers) and potential HSRL retrieval limitations could also contribute to the suppressed depolarization.
Citation: https://doi.org/10.5194/egusphere-2025-4584-RC1 -
RC2: 'Comment on egusphere-2025-4584', Konrad Kandler, 13 Oct 2025
Review of Sujan Shrestha et al., “Transported African Dust in the Lower Marine Atmospheric Boundary Layer is Internally Mixed with Sea Salt Contributing to Increased Hygroscopicity and a Lower Lidar Depolarization Ratio”
Content Summary
Saharan dust transported over the Atlantic to Barbados undergoes transformations in the marine atmospheric boundary layer (MBL) that are poorly constrained by remote sensing alone. During the August 2023 MAGPIE campaign, the authors combined high‐spectral resolution lidar (HSRL) profiles with vertically resolved single-particle chemical and morphological data. They find that while dust in the Saharan Air Layer (SAL) remains externally mixed and yields high lidar linear depolarization ratios (LDR ≈ 0.30), dust descending into the lower MBL becomes internally mixed with sea spray, acquiring more spherical, hygroscopic character that strongly suppresses the depolarization signal (LDR < 0.10) even at high dust loads. The discrepancy between expected and observed LDR is further compounded by differing backscatter efficiencies (lidar ratios) of dust vs marine aerosols, which bias the lidar’s sensitivity toward more spherical particles. The authors conclude that neglecting mixing state and morphological evolution can lead to underestimates of surface dust concentrations from depolarization-based retrievals, with implications for satellite retrievals, dust modelling, and air quality analyses.
I consider the topic as interesting and relevant, suitable for ACP. The work adds new detailed data to the pool. The paper written in a clear language and makes a nice reading.
I have some major points regarding details of the methodology (details further down)
- When making conclusions from the aerosol concentrations and the single particle data for the interpretation of the lidar measurements, the step of including ambient conditions (humidity) is missing. As key statements of this work depend on the comparison, this must be regarded.
- With respect to SEM: there is information missing for the classification schemes and how the shape is calculated. This is relevant for understanding the results and a comparison with similar works. This comparison is largely missing, too; it could however help a lot in assessing the relevance of the present work.
- The number of particles analyzed from the airborne sample is too low to allow for conclusions without a careful characterization of statistical significance.
I encourage the authors to work on the major issues, as I think the work should be published. I hope my comments are helpful.
Details
Line 141: As far as I know the BACO tower was constructed to have the top outside the lowermost marine boundary layer. Is it there justified to call the concentrations ‘surface’?
Wouldn’t it have made sense, if sea salt mixing is a major topic, to compare the top of the tower with measurements at its base or near sea level?
Line 156-157: How about sulfate and organics, that might come along?
Line 175: The selection of paper should be rethought. If you want to refer to general CCSEM papers, the first ones were probably in the 80s, like 10.1080/00022470.1983.10465674 and 10.1016/0048-9697(87)90438-4. If you want to refer to the use of CCSEM for mixing state of dust and marine particles, it could include also 10.1126/science.232.4758.1620, 10.1016/S1352-2310(03)00506-5, 10.1029/2005JD005810, 10.5194/acp-18-13429-2018, 10.5194/acp-25-5743-2025, which worked in comparable settings.
Line 185: Please comment on the quantification limits of N (and O). In many studies, these elements are excluded, if the substrate contains carbon, as the quantification is assumed to have a large error.
Line 197: When the TEM grids are carbon coated and the film contains carbon, why is only Cu excluded?
Line 202-203: What were the criteria used for that? E.g., how were organics and aged sea spray distinguished (Fig. 3)? For example, if I look at Fig. 5a, the sulfate particle (no. 6) or the aged sea spray (no. 2) have a much higher C signal than the organic shell of no. 5. In fact, C is visible in all spectra.
Some more detailed information is given in supplement, but it does not get clear for example, what were the criteria to separate dust/sea salt mixtures from pure dust or pure sea salt.
This section needs to be a bit more elaborated, so I suggest that a part of the supplement S2 is moved here or at least mentioned and summarized and amended with the missing information.Line 211: If only 113 particles were analyzed in total, please comment on the statistical significance. How large are the confidence intervals for the percentages given?
Consider this also for the statement made in line 355-357 and 367-369.
I don’t doubt the general statement, as it is known from previous studies you refer to that dust over Barbados is not strongly internally mixed, but the low numbers here limit the data applicability for such a statement.Line 217: Figure 3 seems to be the first figure reference.
Line 238: Why is dry quoted?
Line 256: It seems like the dust AOD precedes the BACO concentrations a bit. Which would make sense regarding the downmixing of the dust from above.
Line 291-294: “Conducive” seems a bit a misleading word here. Most of the common atmospheric compounds with any hygroscopicity should be expected to be droplets at these humidities. As the soundings were done in the afternoon, it can be expected that the humidity has been higher during other times of the day and, as a result, the particles are on the high branch of the growth vs. humidity hysteresis.
Line 293: Why refer to publications on longwave radiation in the Arctic in this context? Or to a work on Ca and Mg salts (without any direct reference to optical properties)? Remove and replace by suitable ones, if Titos et al. is not deemed to be sufficient.
Line 296: Indicate in a, when the soundings were done (e.g., with an arrow or line).
Line 296: In Fig. 2a, what is the red feature on Aug 16 at 5 km? Does it coincide with the high humidity?
Line 329: The expected LDR should be strongly depending on the humidity, as the sea salt particles would grow by factors of 4 in mass/volume at 80 % and beyond 8 at 95 % RH (vs. dry state), e.g. 10.1038/ncomms15883. As a result, the LDR would be expected to be shifted towards low values at high humidities.
Nevertheless, in the manuscript only the dry sea salt concentrations are compared (line 325-326), and eqns. 7 and 8 in the supplement only seem to take the dry concentration into account (line 164-165).
As a result, it seems that this estimate doesn’t have much relevance for the humid layers. I suggest including a growth model into the estimate and rethinking the conclusions made (e.g., lines 345-347, lines 365-366).
This also affects the derived statements, e.g. lines 397-399.Line 338-341: How much contribution would we expect from the particles > 80 µm?
Line 369-374: this seems to refer to the ground-based measurements, but in line 375 you jump back to the airborne ones. Please order more clearly.
Line 394-396: How do we learn about the variability of particle composition aloft to compare with? As far as I can see, there is no series of samples available.
Line 402: … either … ?
Line 403-405: The aspect ratio measured in SEM is representative for SEM conditions, e.g. very dry (vacuum), under which NaCl and other compounds are crystallized. In contrast, in the humid atmosphere we can expect that the sea salt fraction of the particle is in droplet shape. Therefore, it cannot be expected that the aspect ratio for hygroscopic compounds can be transferred from SEM into the atmosphere. This applies also to line 432.
Lines 405 is this the average (i.e. daily) size distribution or simply the integral of all particles? In the latter case, was the number of particles analyzed for each sample similar?
Line 407: ‘… only available from …, …focuses on …’ ?
Line 417-419: As there are different methods to obtain an aspect ratio from a 2D image, specify the one used, because they yield different results (e.g., doi: 10.1029/2019GL086592 and references therein). In particular, for a square like a cube projection, they do not necessarily come up with 1.
Line 419: Probably not ‘cuboid’ here, but ‘cubic’. Note that a cube is also aspheric by definition.
Line 422: If the number after the +/- is the standard deviation, it does not make much sense in the context of a distribution, which is probably rather log-normal than normal. I.e. in this case, the lower end of the standard deviation range would be 0.9, which is an impossible value for the 2D aspect ratio. Either use parameters of a suitable distribution (e.g. 10.1016/j.atmosenv.2007.06.047, 10.1016/j.atmosenv.2015.07.020) or use percentiles.
Line 434: Fig. 5d ‘Aged …’
Line 445: There have been some investigations with CCSEM in and around the Caribbean before (10.5194/acp-22-9663-2022; 10.5194/acp-18-13429-2018; 10.1029/2002JD002935; 10.1098/rsos.231433; 10.5194/acp-5-3331-2005; Roldan, Lizette: Characterization of microphysical properties of Saharan dust aerosols during trans-Atlantic transport. Howard University, 2006; 10.1002/2015GL065693; 10.5194/acp-25-5743-2025). How do these results compare?
E.g. 10.1098/rsos.231433 shows in Fig 5a dust mixing, which is in respect to sea salt similar to the results of the present work (dust mixing on arrival into the boundary layer) but differs for sulfate (already found on the African side). 10.1002/2015GL065693 shows that dust can remain relatively unaltered after trans-Atlantic transport. 10.5194/acp-25-5743-2025 again show considerable mixing at BACO.Line 449-450: Combining data from Fig. 3 (assuming that the data is representative) and from Fig. 5: if the dust becomes gradually more mixed during the downward-mixing in the boundary layer, why do we see less dust/sea salt mixtures at BACO compared to the flight measurements?
Like 452-455: Check if the statement can be kept up.
‘Peak’ refers to conditions at BACO?Supplement, S2 reference to SEM images is not correct (now 5a or missing?)
Fig S2: Commonly kernel density estimators are used to smooth a histogram. But these curves are not smooth. Why? What estimator was used? Or is that a size distribution with a density on y? Please check.
In general: check the capitalization of the labels in the plots. E.g. 5a, image no. 1: capital at the start. No 2.: capital at the second word. No. 5: All words capitalized.
Citation: https://doi.org/10.5194/egusphere-2025-4584-RC2 -
RC3: 'Comment on egusphere-2025-4584', Anonymous Referee #3, 15 Oct 2025
This paper presents measurements of dust and sea salt over Barbados acquired during August 2023 and describes the impact of these aerosols on the ground-based HSRL measurements of depolarization and backscatter. The major focus of the paper is in assessing how these aerosols mix to reduce the lidar linear depolarization ratio (LDR) measured within the marine boundary layer (MBL), thereby frustrating efforts to use the LDR as an indicator of dust. This is an important topic as lidar measurements of LDR are often used to quantify dust amounts and evaluate model predictions of dust transport. The authors use tower and airborne measurements of particle size and composition to show that large amounts of dust were present even when the LDR was low (<0.1). The authors conclude that using LDR will often result in underestimates of surface dust concentration and argue that in situ measurements must be combined with such lidar measurements to correctly determine dust concentration.
The topic is suitable for ACP. The paper is generally easy to read and publication is recommended after the authors address the comments below.
- Line 34 (and elsewhere). “linear depolarization ratio” It’s not clear here (and elsewhere) whether this means volume (or total) linear depolarization ratio or particle depolarization ratio. (see description in the Burton et al. 2015 reference). From the values provided, it appears to be particle depolarization, but this should be clearly indicated here and elsewhere.
- Line 35 The wavelength of the lidar measurement should be indicated (532 nm).
- Section 2.1 Did the authors ever consult the MPI Raman lidar measurement images that are available on-line at https://barbados.mpimet.mpg.de/? These show extensive measurements of aerosol backscatter and depolarization over Barbados that also tend to support the HSRL measurements presented in the paper. These images exist for several years and include August 2023.
- Line 212. This section discusses airborne particulate samples for single particle analysis. If the CTO inlet has a cut point of 3.5 mm, how were samples as large as 25 mm sampled and manually analyzed?
- Line 214. Can the authors be more specific about how the first maximum in the relative humidity profile was assigned to be the CBH? How large did this maximum in RH have to be? Were these CBH values compared with those that can be readily determined from the HSRL measurements?
- Line 225. Readers looking for the details of the SSEC HSRL are supposed to consult the Razenkov and Eloranta references. What are the uncertainties in the HSRL measurements of aerosol backscatter, depolarization, and lidar ratio? I could not find these in the Razenkov reference. I also wonder whether this reference is still relevant for measurements acquired 15 years after the thesis was written (i.e. has the instrument and analyses remained the same during this period?) The Eloranta reference discusses the design and construction of the NCAR airborne HSRL; note that this reference is not readily available at my institution. I also wonder if this reference provides such uncertainty estimates and if so, whether they apply to the ground based lidar in the same way to the airborne lidar. Given that the HSRL measurements of LDR (and to a lesser extent lidar ratio) are a major topic of this paper, there should be at least a brief discussion of the uncertainties in these measurements; such discussion is absent from the paper. How large are the uncertainties in the volume and particulate depolarization and lidar ratio?
- Line 228. When operating pointing vertically, how close to the surface are profiles of aerosol backscatter and depolarization obtained?
- Figure 1c shows that the largest LDR (105 m) occurred on 08-23-23, yet the dust concentration at the top of the tower (Fig. 1a) was very small (negligible?) Why? This seems to indicate that there can be substantial differences between 50 m (top of tower above sea level) and the lowest lidar measurement height (105 m). Were there measurements made at/near the base of the tower or closer to sea level to study the vertical variations close to the surface?
- Figure 1. It would be interesting to see wind speed also during this period to see if/how the depolarization, backscatter, and lidar ratio varied with wind speed and also the amount of sea spray.
- Line 284. “…depolarization measurement responds to the 180-degree backscatter efficiency of the particulates (lidar ratio).” This is why there needs to be a better description of the LDR that is referred to. The particulate depolarization depends on the total (volume) depolarization as well as the particulate scattering ratio. The sentence currently is confusing since it mentions (lidar ratio) at the end of the sentence. I could easily see how a reader not familiar with lidar (and HSRL) can become confused.
- Line 288. Following on the last comment, “…the measured depolarization will be weighted lower due to the backscatter efficiency difference between the aerosol (i.e., lidar ratio)”. Lower than what? Do the author mean instead (or also) “…the measured backscatter will be weighted lower due to the backscatter efficiency difference between the aerosol (i.e., lidar ratio)”.
- Line 309. Maybe this will be discussed later in the paper, but what is the basis of the statement “…due to the predominance of larger but less backscattering mineral dust particles”. If the RH is higher in the MBL, and the particles are hygroscopic, wouldn’t the particles in the MBL be larger?
- Figure 3. It would be interesting to see profiles of the lidar ratio (532 nm) and backscatter color ratio (ratio of backscatter at 532 to 1064 nm) for this case also. These are two important aerosol parameters that can also help in interpreting the aerosol vertical distribution.
- Line 363. “underestimated” should be “overestimated”.
- Line 379. I don’t understand why the transition layer is described as “indifferentiable”. The LDR clearly shows this layer as different from the SAL and MBL.
- Figure 5. What do the black lines in 5b and 5c represent?
- Figure 5c. This seems to show only sea spray and not aged sea spray. Is that correct?
- Figure 5d. Are these measurements made at low RH? If so, wouldn’t the aspect ratios be very different at ambient RH?
- Line 449. Should be “internally” `
- Lines 467-477 This paragraph expresses the desire/need for combining lidar data with in-situ single particle analysis to improve the interpretation of lidar data in dust regions. While this may be true for the best interpretation of such data, this can’t be done routinely and globally on a continuous basis; that is why remote sensing techniques are pursued. The authors seem to indicate that there are no other alternatives and all remote sensing techniques are doomed to significantly underestimate dust impacts near the surface. Have the authors considered whether more advanced remote sensing measurements could provide additional data to help improve the interpretation of lidar data? For example, measurements of backscatter, extinction, depolarization at additional wavelengths? As a suggestion, the reviewers may want to examine the backscatter color ratio (or Angstrom exponent) using various wavelengths (ex. 355-532 nm) in such situations, especially when acquired by such HSRL systems. Examination of such data, in conjunction with depolarization data at multiple wavelengths suggests that, while the depolarization near the surface may be low suggesting that dust concentrations are low, the backscatter color ratio has similar values as observed in the SAL region and which are also different from values in other MBL regions where dust values are low, suggesting that backscatter color ratio may be an indicator of dust. The point is that the authors should not prematurely dismiss remote sensing techniques for providing accurate estimates of dust loading simply because of the limitations of the lidar measurements studied here.
Citation: https://doi.org/10.5194/egusphere-2025-4584-RC3 -
RC4: 'Comment on egusphere-2025-4584', Wenshuai Li & Yang Zhou (co-review team), 18 Oct 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-4584/egusphere-2025-4584-RC4-supplement.pdf
- AC1: 'Comment on egusphere-2025-4584', Sujan Shrestha, 23 Dec 2025
Peer review completion
Post-review adjustments
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,324 | 142 | 35 | 1,501 | 47 | 26 | 23 |
- HTML: 1,324
- PDF: 142
- XML: 35
- Total: 1,501
- Supplement: 47
- BibTeX: 26
- EndNote: 23
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Sujan Shrestha
Robert E. Holz
Willem J. Marais
Zachary Buckholtz
Ilya Razenkov
Edwin Eloranta
Jeffrey S. Reid
Hope E. Elliott
Nurun Nahar Lata
Zezhen Cheng
Swarup China
Edmund Blades
Albert D. Ortiz
Rebecca Chewitt-Lucas
Alyson Allen
Devon Blades
Ria Agrawal
Elizabeth A. Reid
Jesus Ruiz-Plancarte
Anthony Bucholtz
Ryan Yamaguchi
Qing Wang
Thomas Eck
Elena Lind
Mira L. Pöhlker
Andrew P. Ault
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1754 KB) - Metadata XML
-
Supplement
(497 KB) - BibTeX
- EndNote
- Final revised paper
Review of “Transported African Dust in the Lower Marine Atmospheric Boundary Layer is Internally Mixed with Sea Salt Contributing to Increased Hygroscopicity and a Lower Lidar Depolarization Ratio” by Shrestha et al.
This paper focuses on field measurements during ONR’s MAGPIE campaign in August 2023 at Ragged Point, Barbados. It focuses on how Saharan dust mixes with sea salt, yielding surprisingly higher hygroscopicity than what would normally be expected for dust. Of course this can be explained by the mixing with sea salt. This act of mixing appears to be localized in the bottom altitudes unlike higher up in the Saharan Air Layer. The enhanced hygroscopicity and mixing at lower altitudes coincides with reduced HSRL-derived depol ratios in contrast to higher altitudes where there is just dust. The topic is important and points to important nuances in the transport of dust and how we really must understand its mixing behavior with other aerosol types to know the chemical, physical, and optical properties of such air masses. I believe the paper will have interest among the readers of the journal.
The methods used are robust and include measurements of single particle properties, concentrations of sea spray and dust, and HSRL retrievals.
The presentation quality is good. In general, the paper was well done and publication is recommended. Some minor suggestions/questions are provided below.
Specific Comments:
Data Availability section: a bit odd to say “will be publicly available” to leave authors in suspense. Why not just make the concentration data available now? Also, it is uncertain about the format of single particle data, but is it common practice to archive those data publicly or not?
Lines 142-157: It would help readers if you explicitly say up front what part of the text is explaining the method of determining dust mass. It was a bit confusing.
Figure 3: nice visual depiction of results.
Line 128-130: The statement “BACO offers an optimal location for intercepting long-range transported Saharan dust with minimal interference from local anthropogenic emissions due to the prevalent Easterly trade winds” should be supported with appropriate references.
Line 208-214: The manual SEM/EDX analysis of a relatively small number of particles (40, 21, and 52) raises concerns regarding statistical representativeness and uncertainty. Since manual particle selection can introduce bias, it will strengthen the methodology if the authors clarify how particles were chosen (randomly or selectively) and whether any estimation of analytical or sampling errors was made.
Line 283-287: The explanation of lidar ratio differences between dust (~40 sr) and marine boundary layer aerosols (~20 sr) is unclear and could be misinterpreted. The phrase “a factor of two different with the marine sourced particles being twice as efficient per scattering cross-section compared to dust at backscattering energy” is confusing. Please clarify whether the statement refers to lidar ratio magnitude or backscatter efficiency.
Line 446-477: The conclusion provides a strong interpretation linking microphysical evidence of dust-sea-spray mixing with the observed low LDR values and their implications for remote sensing. However. The authors may consider strengthening the conclusion by acknowledging additional contributors such as vertical heterogeneity within the MABL (e.g., overlapping marine and dust layers) and potential HSRL retrieval limitations could also contribute to the suppressed depolarization.