the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Mar 2025
Enhancing dust aerosols monitoring capabilities across North Africa and the Middle East using the A-Train satellite constellation
Abstract. North Africa and the Middle East encompass the most active dust sources on the planet. Due to the limited availability of ground-based aerosol observations across the deserts, spaceborne retrievals represent the most reliable source of information for monitoring dust particles over these vast areas. In the current study, we present a synergistic approach incorporating aerosol retrievals acquired by active (CALIOP) and passive (POLDER-3, MODIS) instruments mounted on satellites of the A-Train constellation. Our main objective is to dynamically (in terms of space and time) estimate the dust lidar ratio (LR) throughout the CALIPSO operational period (2006–2023) by collocating columnar aerosol optical depth observations (POLDER-3/GRASP, MIDAS) and vertically resolved dust aerosol profiles obtained by CALIPSO. According to our findings, the agreement among the satellite-based retrievals improves, when the default and constant CALIPSO dust LR (44 sr) is adjusted. Specifically, increasing and decreasing dust LR tendencies are recorded in North Africa and in the Middle East, respectively, whereas a narrow transition zone of neutral declinations appears between the two regions. Furthermore, compared to the standard dust LR, higher values are recorded east of the Caspian Sea, while negative departures are found in Iran and southern Pakistan. The proposed dust LR adjustments are maximized over/near dust sources and during the dry period of the year, indicating a dependence on dust activity. The evaluation against AERONET observations clearly demonstrates that the refined dust LRs improve CALIOP’s performance when mineral particles are probed, but further justification is needed. The integration of the refined dust LRs significantly modifies the standard CALIPSO dust optical depth (DOD) climatological patterns, particularly in the Western Sahara, the Bodélé depression (northern Chad Basin), the Libyan Desert and Saudi Arabia. Positive and negative shifts in the regional DOD timeseries are found in North Africa and in the Middle East, respectively, with no notable modifications on the inter-annual and intra-annual trends in either region. A key priority of the analysis carried out is to improve the efficiency of CALIPSO dust retrievals towards advancing their utility in a wide range of scientific applications. The advent of the EarthCARE satellite mission, along with the incorporation of new aerosol models into the forthcoming CALIPSO Version 5 aerosol retrieval algorithm, will serve as a reference for our calculations. In this context, our findings also highlight that a synergy of multisensor aerosol products with modelling tools can enhance the spatiotemporal representation of aerosol properties, such as the lidar ratio, further improving retrieval utility.
Competing interests: Co-author Vassilis Amiridis is in the editorial board of AMT and coordinating the special issue that the paper is submitted.
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.- Preprint
(3462 KB) - Metadata XML
-
Supplement
(2297 KB) - BibTeX
- EndNote
Status: open (until 28 Apr 2025)
-
RC1: 'Comment on egusphere-2025-888', Anonymous Referee #1, 08 Apr 2025
reply
General comments:
This manuscript presents an important contribution to the field of aerosol remote sensing by enhancing the estimation of dust lidar ratios (LRs) through a synergistic use of multiple satellite datasets (CALIOP, MODIS/MIDAS, and POLDER-3/GRASP). The research primarily addresses a critical limitation in the CALIPSO dust retrieval algorithm—the use of a fixed dust lidar ratio (LR) of 44 sr.
The effort to improve the accuracy of CALIOP-derived Dust Optical Depth (DOD) in North Africa and the Middle East is commendable and of clear relevance to both the scientific community and stakeholders.
A noteworthy contribution of the study is the clear demonstration that regional variability in dust properties necessitates the use of spatially and seasonally resolved LR values. The analysis reveals that higher LR values are required over North Africa (~53 sr), while lower values (~37 sr) are more appropriate for the Middle East. These adjustments lead to measurable improvements in DOD retrievals, particularly in regions like the Bodélé Depression and the western Sahara, which are known for intense dust activity.
The authors further validate their findings against AERONET sun-photometer measurements, reporting substantial reductions in both Mean Bias Error (MBE) and Root Mean Square Error (RMSE). Seasonal analysis also reveals strong correspondence between revised LR values and observed dust activity patterns, highlighting the method's robustness across varying climatic conditions.
While the methodology is generally robust and the findings are well-supported, I would like to offer several constructive observations that may help refine the study and guide future work as follows.
1. Although the improved dust LRs clearly lead to enhanced agreement with satellite and ground-based AOD measurements, the broader implications for climate modeling and radiative forcing are not fully explored. A quantitative analysis of the impact on surface energy budgets, direct radiative effects, or global dust forcing would elevate the study’s significance, therefore I suggest to write a statement for future work related to this.
2. The authors reference the relevance of their results for upcoming missions (e.g., EarthCARE) and future algorithm updates (CALIPSO v5). While promising, the manuscript does not provide a detailed roadmap for how the improved LRs will be operationally integrated. Outlining a pathway or collaboration with mission teams could enhance the practical value of the study.
In conclusion this work represents a valuable contribution to aerosol remote sensing, offering practical improvements to satellite dust retrievals in one of the most important source regions globally. Addressing the aforementioned limitations in future research would not only strengthen the current findings but also broaden the applicability of the proposed methodology across other regions and satellite missions.
Specific comments
Page 7 in paragraph 220-please provide a reference for the interpolation procedure to derive AOD at 532nm.
Figure 3-page 11- the shades of red/purple are very close and not easily distinguishable-please choose contrasting colors.
Figure 4 page 14 – in caption for the correlation factor, better use “R”.
Page 15 line 398 -I think could be useful to write in here when you refer to Figure S1 and S2 to specify that they could be found in the supplement material.
Figure 8 page 21- Very difficult to follow all information in panels a and b; too much information; For me it could be much better if you put on separate panels the graphs related to annual/seasonal.The shades of red and brown for me are very confusing- difficult to distinguish; I would change brown with a different color-maybe black or grey.
Line 500- when you refer to depolarisation thresholds are you referring that the value is set for 0.28 for dust? In general when reading the word “thresholds” I would look for an interval between max and min values. Please clarify.
Line 510 Add respectively after “derived from the CALIOP-POLDER-3/GRASP and CALIOP-MIDAS synergies”.
Line 514- I have noticed you have used the word “departure” several times when you refer to the difference between two values- not sure if it is correctly used- please check with an English speaker.
Lines 528-529- “Within the zone of high DODs, the two peaks are more distinguishable when the new LRs have been calculated from the collocated CALIOP-MIDAS sample on a seasonal basis.” I do not agree with this statement. That area looks very similar in the 2 cases. At least is not \'more distinguishable\' for me.
Figure 11- Again, I have difficulties to distinguish red and brown curves.Citation: https://doi.org/10.5194/egusphere-2025-888-RC1 -
RC2: 'Review of egusphere-2025-888', Anonymous Referee #2, 16 Apr 2025
reply
The paper by Moustaka et al. describes the synergistic use of long-term CALIOP vertical aerosol profiles and two passive satellites to adapt the CALIOP lidar ratio choice for aeolian mineral dust. They replace the standard lidar ratio selection algorithm by a variable synergistic approach which scientifically improves the aerosol extinction retrievals obtained from CALIOP in the vicinity of dust and thus provide a data set which can be used to estimate radiative forcing more accurately.
The authors adapt the lidar ratio in a way that the total AOD (as for dust dominated scenes only, equivalent to the Dust Optical Depth) fits to the one retrieved by the satellite-based imagers (POLDER and MODIS). By doing so, they retrieve a map of dust lidar ratio across two research areas (North Africa and Middle East). The results are validated against selected AERONET stations in the region of interest. Furthermore, they apply these new results (lidar ratios) to CALIOP backscatter profiles to obtain enhanced dust optical depth climatologies in the research regions and compare them to the ones obtained with the standard CALIPSO retrieval.
The paper is highly relevant and of great interest for the research community and very well written. Usually, I would have only minor comments, but as I am struggling with Equation 6, which is essential for the whole publication, I have to propose “major revisions” (see major comment below). I suspect that the “Equation 6 issue” can be solved fast or is just a misunderstanding. In this case, the paper can be published after addressing the other comments. If there is a substantial mistake in the retrieval (what I doubt given the presented results), the results need to be newly calculated and interpreted.
Given that the “equation 6 issue” is minor and/or just a typo, the paper is suitable as a highlight paper, as it gives a very important improvement in the global data set of dust load in the atmosphere. However, for that reason, it would be great if the authors could make the newly derived products available for the research community, i.e., the CALIOP results with adapted lidar ratio retrieval.
Major /General comments:
- Equation 6 cannot be correct: According to Eq. 2, the absolute difference between the DODs is calculated and then used in Eq. 6. However, this would lead to negative adapted lidar ratios in case the DOD differences are less than 1.
If one assumes a let’s say DOD difference of 0.1, Equation 6 would lead to: 44sr * (0.1 -1) = -39.6 sr as updated lidar ratio which cannot be possible because it is negative.
A DOD difference of 2 would lead to an updated lidar ratio of 44 , i.e. the same as already used. And no difference would lead to a negative updated LR of -44.
Please check and comment. I am confident that it can be clarified fast, because this formula actually cannot lead to the results presented later on. - With respect to the correction of this issue, the methodology section should also be updated with some formulas for non-lidar experts on how to retrieve the DOD (AOD) from CALIOP. I.e., lines 265 to 270 are the most important part of the methodology but were yet just briefly described (especially when comparing to all the other extensive descriptions). For non-lidar experts this might not be obvious. Thus, these very important 5 lines should be more highlighted and even extended, i.e. by a sketch and showing the formulas (I.e., the (D)AOD is the integrated (dust)bsc times the Lidar ratio written as formula could help and you later can always refer to it.)
It should be also emphasized again that you assume that the backscatter profile of CALIPSO to be only representative for dust aerosol as I assume you do no separation of the backscatter in a dust and non-dust component (Later in Section 4.4. you explain it, but this is too late then). - Furthermore, the authors “just“ adapt the lidar ratio for the retrieval of the extinction coefficient from the given backscatter profile multiplied by the lidar ratio. However, for elastic lidars, the assumption of the lidar ratio is already used when calculating the backscatter coefficient (see, e.g., Klett-Fernald 1984). The authors should briefly discuss what effect an alternated lidar ratio (in the range they have retrieved it) would have on the retrieved backscatter profile to give evidence that there is no need to recalculate the backscatter profiles and the methodology they used is valid.
- For the result parts (Sec 4.4.), the authors always refer to both approaches using MODIS AND POLDER: However, it would be convenient for the reader if the final statements could be made based on only one of these data sources especially if the results are similar (MIDAS seems to be the choice for that as longer available).
- More a hint: The paper is very well written and of high quality, but partly a bit longish and repetitive. Partly there is an overlap and repetition between the data set and methodology section. This does not need to be changed here but should be considered for the next publication to be more concise.
- All data should be made available for the scientific community, also the LIVAS data set AND your novel Calipso dust extinction data set produced in this work. I.e., upload to zenodo or some dedicated database.
- References are partly missing and many typos in the reference section – please check carefully.
Minor/Specific comments:
- Please avoid the wording “departure”, i.e., neg. or pos. departures, it is not obvious what is meant with that phrase.
- Line 58: “…affects the atmospheric composition and radiative balance of the Middle East and western parts of Asia” – I guess the Eastern Mediterranean is missing here as well?
- Line 67: The references given for “passive and active remote sensing techniques” seem not to be appropriate especially with respect to the active instruments. Later on, the authors cite relevant publications with respect to active remote sensing - such should be also used here
- Line 77: Cannot find reference Hansen 1995, Mishenko 1997 in the list. Thus, the connection to MAP is not given.
- Line 94: “Past studies indicated…” Please give references for past studies.
- Line 109-111: “The adaptation of a more “realistic” dust LR, along with more reliable elastic lidar-derived AOD values, could contribute to a more robust aerosol typing.” It is not clear to me why this should be the case – please clarify.
- Line 138-139: “…lidar system with horizontal averaging and vertical resolution variable based on both wavelength and altitude…” I do not understand this sentence, please rephrase it.
- 157: LIVAS database: how can this be accessed?
- 226: Would be good to list the AERONET stations here already.
- Fig. 1b: Please show the same frame as Fig. 1a otherwise it is confusing (same lat lon box)
- Figure 2 is quite blurry, make sure to have high resolution plots when resubmitting
- Caption Fig. 2. Please try to add the CALIOP aerosol subtype description in the plot.
- Line 396: Why are the deserts in central Asia a blind spot and your approach cannot be used here?
- Line 396: Hofer et al 2020 not in reference list
- Line 431: “summer months” should be avoided in this context for regions close to the equator
- Figure 5: the distributions of the “LR boxes” seem a bit more extended (especially to the east) compared to the areas in Fig 1a, please comment!
- Line 473: “(via the integration of the vertical dust extinction coefficient profile)” here you could nicely refer to an equation once introduced in the methodology part.
- Line 490. I propose to make Section 4.4 a new chapter 5. It is a completely different topic and after the “validation" part the respective “application” part. Thus, it would fit better to have a dedicated chapter.
- Line 494-496: I do not understand this: Previously it seems that you just use pure dust cases. Now something different is stated. If a new approach is used here, it should have been explained in the methodology section already. Can you explain what is the difference here compared to the sections above?
- Line 639: “In elastic lidars (such as CALIOP), for the derivation of the aerosol extinction coefficient from the backscatter coefficient, it is required a factor known as lidar ratio”. Actually, it is also needed to retrieve the backscatter coefficient - see major comment 3.
- In the conclusion section, it would be good to emphasize that higher LR mean higher DOD and vice versa.
- Line 647: ”…which are less pronounced in the CALIOP-POLDER results.” Why?
- Line 653: “each synergy (combination) performing better or worse” Do they perform better or worse? Both is not possible.
- Line 655: “showed a notable decrease or increase” → notable deviation
- 691: Please make LIVAS available as well as your research data set.
- Line 706: I think it would be fair to also acknowledge the PI’s of the respective AERONET stations by name.
Citation: https://doi.org/10.5194/egusphere-2025-888-RC2 - Equation 6 cannot be correct: According to Eq. 2, the absolute difference between the DODs is calculated and then used in Eq. 6. However, this would lead to negative adapted lidar ratios in case the DOD differences are less than 1.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 129 | 38 | 6 | 173 | 16 | 6 | 5 |
- HTML: 129
- PDF: 38
- XML: 6
- Total: 173
- Supplement: 16
- BibTeX: 6
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 46 | 27 |
| Greece | 2 | 25 | 15 |
| China | 3 | 14 | 8 |
| Germany | 4 | 9 | 5 |
| Switzerland | 5 | 8 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 46