the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jul 2026
Strengthening dust–cirrus relationships over Central Europe (2001–2025)
Abstract. Saharan dust is increasingly recognized as an important driver of aerosol–cloud interactions over Europe, yet its long-term influence on cloud formation remains poorly understood. This study investigates changes in the relationship between dust intrusions and upper-level cloud properties over the Carpathian Basin during the 2001–2025 period using a combination of MERRA-2 reanalysis, MODIS satellite observations, HIRS outgoing longwave radiation (OLR) data, and a database of 276 validated Saharan dust events.
The frequency of Saharan dust intrusions increased by nearly 70 % during the study period. Dust outbreaks were associated with enhanced high-cloud cover, increased cirrus reflectance, reduced effective ice particle radius, and lower OLR values. These relationships were particularly pronounced during spring, when thermodynamic conditions are most favourable for heterogeneous ice nucleation and cirrus development.
To quantify the temporal evolution of dust–cloud relationships, generalized additive models were applied to cloud and radiative variables. Atmospheric dust loading was a highly significant predictor in all seasonal models, while significant dust–time interaction terms demonstrated that cloud responses to dust have systematically strengthened over time. The strongest signals were observed for cirrus reflectance, high-cloud fraction, and OLR, indicating increasingly pronounced upper-tropospheric cloud responses under comparable dust-loading conditions.
The results suggest that Saharan dust has become an increasingly important factor influencing cloud formation and radiative processes over Central Europe. Together with growing evidence for more frequent ice-saturated and ice-supersaturated environments, the findings support the hypothesis that evolving upper-tropospheric conditions increasingly favour dust-associated cloud formation.
- Preprint
(2084 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-3742', Anonymous Referee #1, 19 Aug 2026
-
AC1: 'Reply on RC1', György Varga, 16 Sep 2026
Response to Reviewer 1
We sincerely thank Reviewer 1 for the positive assessment and the detailed and constructive comments. Following these suggestions, we revised the Introduction, Methods, Results and Discussion, and added a vertically resolved MERRA-2 dust analysis to better assess the presence of Saharan dust at cirrus-relevant altitudes.
Comment: The abstract should be as compact as possible
Response: Thank you. The Abstract has been shortened and reorganized to focus on the motivation, methods, and principal findings.
Comment: A more appropriate introduction into the cirrus literature would be desirable.
Response: We agree. The Introduction has been expanded to provide a broader background on cirrus formation, radiative effects, homogeneous and heterogeneous ice nucleation, and mineral-dust INPs. The recommended studies have been incorporated.
Comment: Lines 46–51: References on ice nucleation, Saharan dust observations, and feldspar.
Response: The recommended references have been incorporated, including Ansmann et al. (2003) on Saharan dust transport and Atkinson et al. (2013) on the ice-nucleating efficiency of feldspar.
Comment: Are you sure that the basin-like topography favors accumulation of air masses after long -range transport and provides favorable conditions for the formation and persistence of high level clouds? Are there articles that confirm or support your hypothesis?
Response: We agree that this interpretation was too speculative. The text has been revised to avoid attributing upper-level cloud formation or persistence directly to the basin topography.
Comment: The study is based on hourly dust COLUMN mass density fields. In the case of a dust event, how sure can you be that that dust occurred at cirrus formation height level? Height information.
Response: We addressed this issue with a new vertically resolved MERRA-2 analysis. Dust exceeding 10-9 kg kg-1 was present at 7–12 km in 95.7% of the 276 SDEs, demonstrating frequent vertical overlap between Saharan dust and cirrus-relevant altitudes. This analysis has been added to the Methods and Results. The results of the vertical analyses were also plotted on a new figure.
Comment: How reliable are the small changes in MODIS effective ice-particle radius?
Response: We agree that differences of only a few micrometres should not be overinterpreted. The revised manuscript therefore treats the effective-radius response as weak and does not use it as evidence for a major long-term microphysical change.
Comment: A reference to the model GAM would be fine.
Response: The statistical methodology has been clarified and appropriate methodological description have been added. The revised analysis is performed on daily observations rather than cumulative threshold averages.
Comment: About the large tables
Response: We agree. The table has been simplified.
Comment: The relationship between synoptic transport patterns and dust at cirrus altitudes should be clarified.
Response: We now distinguish between the large-scale transport environment and the actual vertical location of dust. The latter is independently assessed using the new vertically resolved MERRA-2 analysis. The different synoptic types were discussed in detail in the cited papers by Varga et al..
Comment: You discuss effective-radius tendencies of 30.3, 29.9, and 28.2 micrometer…
Response: We agree that differences of only a few micrometres should not be overinterpreted. The revised manuscript therefore treats the effective-radius response as weak and does not use it as evidence for a major long-term microphysical change.
Comment: Line 251-255: You discuss Table 2, but you did not mention or introduce Table 2 in the main text body? Should be improved! Could be done in line 251.
Response: That table has been removed.
Comment: Are middle and high-level cloud fractions somewhere precisely defined? If yes, please indicate here where that was done, as a reminder. Please provide clear definitions somewhere.
Response: Definitions have been added to the Methods. MERRA-2 low-, middle-, and high-level cloud fractions correspond to the surface–700 hPa, 700–400 hPa, and above-400 hPa layers, respectively. The figure captions have also been clarified.
Comment: The interpretation should consider the competition between homogeneous and heterogeneous ice nucleation.
Response: We agree. The revised manuscript now explicitly discusses this competition based on Kärcher et al. (2022, 2023) and avoids assuming that increased mineral-dust INPs necessarily produce more ice crystals.
Comment: Figure 2 is the highlight of the paper! Again, I miss a clear definition of cloud cover and defined height regions for low, mid, and high clouds.
We need (a), (b), (c), etc.... within the six different panels.
Response: The definitions have been added, and table has been completed.
Comment: Can the direct dust effect on OLR be separated from the cirrus effect?
Response: No. We have clarified that our analysis cannot unambiguously separate these effects. OLR is therefore interpreted as an integrated radiative signature associated with dust-loading conditions rather than as an isolated cirrus radiative effect.
Comment: Atmospheric dynamics should be included in the Discussion.
Response: We agree. The Discussion now considers vertical velocity, wind shear, turbulence, and ice supersaturation, following Pieper et al. (2026) and Juurikkala et al. (2026). We emphasize that the observed relationships may reflect combined aerosol, thermodynamic, and dynamical controls.
Comment: Could the upper-tropospheric fraction of dust have increased over time?
Response: We examined this using vertically resolved MERRA-2 data. The 7–12 km dust mass mixing ratio showed no significant monotonic trend over 2001–2025, suggesting that the temporal changes cannot be explained simply by increasing upper-tropospheric dust abundance.
Comment: Are changes of only a few micrometres in effective radius physically meaningful?
Response: We agree that these differences are small. Their interpretation has therefore been substantially weakened, and the main conclusions are based on the more robust changes in cirrus reflectance, high-cloud fraction, and cloud-top properties.
Comment: The DIBS case should be attributed to the appropriate Ansmann et al. (2019) paper.
Response: Thank you. The citation has been corrected and the unnecessary Ansmann et al. (2019b) reference has been removed.
Comment: Sections 4.3 and 4.4 should be shortened.
Response: We agree. These sections have been condensed, while greater emphasis is now placed on unresolved questions concerning the combined effects of dust, atmospheric dynamics, and upper-tropospheric thermodynamic conditions.
We thank Reviewer 1 again for the constructive suggestions. In particular, the new vertically resolved dust analysis and the revised treatment of cirrus microphysics and atmospheric dynamics have substantially strengthened the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-3742-AC1 - AC3: 'Reply on RC1', György Varga, 21 Sep 2026
-
AC1: 'Reply on RC1', György Varga, 16 Sep 2026
-
RC2: 'Comment on egusphere-2026-3742', Anonymous Referee #2, 10 Sep 2026
This study examines associations between Saharan dust intrusions and upper-tropospheric cloud properties over the period 2001-2025. It makes the strong claim of a strengthening of the dust-cirrus relationship over that relatively short period. In the referee's opinion, the analysis is not reproducible and the manuscript does not demonstrate its main claim. The main concerns are detailed below; in particular, the GAM input may be redundant, and the manuscript repeatedly gives statistical links a causal interpretation which is not corroborated by the observational analysis presented. Furthermore, presenting the results in very large tables makes the demonstration very hard to follow. In light of these issues, I recommend that the paper be rejected at this stage.
Major issues
1) Presentation
The paper's tables contain roughly 500 populated entries that cannot be interpreted efficiently. Table 1 is particularly overloaded and split across pages, while Table 3 presents excessive numerical precision and many repetitive significance values. The authors should substantially simplify the main-text tables. In their present form, the tables obscure the findings.
2) Reproducibility
The paper never states exactly what the GAM inputs are or over which geographic domain the averages are calculated: it is unclear whether the domain is Hungary, the Carpathian Basin, or another Central European region, and no spatial mask is defined. Under these conditions, the analysis cannot be reproduced. Given that the datasets have global coverage, the authors should also justify the regional restriction and demonstrate whether the reported relationship persists in a broader or global analysis, which would improve sampling and establish its generality.
3) Statistical significance and validity:
Section 2.4 appears to describe cumulative threshold curves: days above a high dust threshold also contribute to means at every lower threshold. If these
threshold means are used as GAM inputs, adjacent thresholds largely correspond to the same daily data and are not independent. This "replication" could be the reason for
the large R2 values and very small p-values. If cumulative means are indeed used, appropriate statistical tests should be performed; the inference should be
rebuilt at the daily or event level, accounting for temporal autocorrelation and using block-based uncertainty estimates.4) Causality versus correlation
Besides the significance issue, even with signficant correlations, dust transport and cirrus may well just share external controls such as circulation and upper-tropospheric humidity. Changes in synoptic weather regimes could therefore increase both transported dust and cirrus without any change in the causal effect of dust. This possibility is alluded to in the conclusion and in Sect. 4.2, but simple tests (e.g., including meteorological variables in the GAM) are not performed.
Moreover, the chosen dust predictor is questionable. Total-column dust mass does not show whether dust is present at cirrus altitude or spatially collocated with the cloud.
Citation: https://doi.org/10.5194/egusphere-2026-3742-RC2 -
AC2: 'Reply on RC2', György Varga, 16 Sep 2026
Response to Reviewer 2
We thank Reviewer 2 for the critical and constructive assessment. The comments prompted substantial revisions to the statistical framework, reproducibility of the analysis, vertical characterization of dust, and interpretation of the results. Most importantly, we no longer claim a uniform strengthening of the dust–cirrus relationship over 2001–2025. Instead, the revised manuscript describes a persistent but temporally and seasonally variable association, with the clearest temporal change occurring in cirrus optical properties under high springtime dust loading.
1. Presentation
Comment: The main-text tables are excessively large and obscure the key findings.
Response: We agree. The tables have been substantially simplified, and the principal results are now emphasized graphically. Excessive numerical detail has been removed from the main text.
2. Reproducibility
Comment on the GAM inputs and geographic domain (insufficiently described)
Response: We agree. The revised Methods now explicitly define the study domain (Hungary), spatial averaging, event windows, predictor and response variables, and statistical procedures, allowing the analysis to be reproduced.
Comment: Given the global coverage of the datasets, the regional restriction should be justified and a broader/global analysis considered.
Response: We have clarified that the study is intentionally regional and event-based, focusing on Hungary, and independently identified Saharan dust events affecting Hungary. A global analysis would address a substantially different research question and is beyond the scope of the present study. We now state this limitation more clearly.
3. Statistical significance and validity
Comment on the cumulative dust-threshold means, what are not independent and should not be used as GAM inputs. Statistical inference should instead be performed at the daily/event level and account for temporal dependence.
Response: We agree with this important criticism. The statistical analysis has been rebuilt using daily observations within independently defined SDE windows rather than cumulative threshold averages. Overlapping event windows were resolved to avoid duplicate observations, and the revised Methods now clearly distinguish descriptive threshold analyses from inferential statistics.
Comment on the high R² values (very small p-values may result from pseudo-replication among cumulative thresholds)
Response: We agree. Therefore, cumulative threshold means are no longer used for statistical inference. They are retained only as descriptive summaries, while significance testing is based on the underlying daily/event-level observations.
4. Causality versus correlation
Comment: Besides the significance issue, even with signficant correlations, dust transport and cirrus may well just share external controls such as circulation and upper- tropospheric humidity. Changes in synoptic weather regimes could therefore increase both transported dust and cirrus without any change in the causal effect of dust.
Response: We agree and have revised the manuscript throughout to avoid unsupported causal language. The results are now described as dust–cloud associations, and we explicitly state that the observational analysis does not establish causality.
The revised Discussion explicitly considers circulation, vertical motion, wind shear, turbulence, and ice supersaturation as potential common controls. We therefore interpret the observed relationships as potentially reflecting combined aerosol, thermodynamic, and dynamical influences rather than a dust effect alone.
Comment: This possibility is alluded to in the conclusion and in Sect. 4.2, but simple tests (e.g., including meteorological variables in the GAM) are not performed.
Response: We agree with the underlying concern. Rather than interpreting the statistical relationships as causal effects after meteorological adjustment, we have substantially narrowed the scope of our conclusions and explicitly discuss meteorological covariability as a limitation. A rigorous attribution of aerosol and dynamical effects would require additional upper-tropospheric dynamical and thermodynamic diagnostics and is beyond the scope of the present analysis.
Comment: Moreover, the chosen dust predictor is questionable. Total-column dust mass does not show whether dust is present at cirrus altitude or spatially collocated with the cloud.
Response: We fully agree and addressed this directly with a new vertically resolved MERRA-2 analysis. Dust exceeding 10⁻⁹ kg kg⁻¹ was present at 7–12 km during 95.7% of the 276 SDEs, demonstrating frequent vertical co-occurrence of Saharan dust and cirrus-relevant altitudes.
We additionally compared total-column and upper-tropospheric dust. Dust mass mixing ratios at 7–12 km showed stronger associations with several upper-level cloud properties than total-column dust loading, supporting the relevance of the vertically resolved dust metric.
We agree that the original wording of the manuscript on the strengthening response was too strong. The revised manuscript no longer claims a uniform or monotonic strengthening of the dust–cirrus relationship. Instead, our results indicate a persistent but temporally evolving relationship, with the clearest change found under extreme spring dust loading: median cirrus reflectance increased from 0.075 in 2001–2005 to 0.186 in 2021–2025, while the fraction of extreme dust days simultaneously exhibiting extreme cirrus reflectance increased from 15.6% to 47.6%. These changes were non-monotonic across the intermediate periods and are now described accordingly.
We thank Reviewer 2 for identifying these important limitations. The revised manuscript now uses daily/event-level inference, explicitly evaluates dust at cirrus-relevant altitudes, provides clearer methodological information, and substantially reduces causal interpretation. We believe these changes have resulted in a more reproducible analysis and a more appropriately constrained interpretation of the observed dust–cirrus relationships.
Citation: https://doi.org/10.5194/egusphere-2026-3742-AC2 - AC4: 'Reply on RC2', György Varga, 21 Sep 2026
-
AC2: 'Reply on RC2', György Varga, 16 Sep 2026
-
CC1: 'Comment on egusphere-2026-3742', Iarla Boyce, 28 Sep 2026
Find the review attached as a document. This review was completed as a joint exercise for the 2026 EGU Peer Reviewer Training.
- AC5: 'Reply on CC1', György Varga, 29 Sep 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 368 | 145 | 44 | 557 | 29 | 29 |
- HTML: 368
- PDF: 145
- XML: 44
- Total: 557
- BibTeX: 29
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General impression:
The manuscript is well written. It deals with a careful and very detailed analysis of long-term dust and cirrus data sets. The data base contains 276 Saharan dust events and covers the years from 2001-2025. A clear impact on cirrus properties by Saharan dust was found and thus a dust impact on ice cloud formation and development in the upper troposphere. This is a valuable contribution to the cirrus research branch.
Minor revisions may further improve the paper.
Detailed comments:
The abstract should be as compact as possible: goal/motivation, methods/data, main or key findings, all in all not more than 250 words.
Introduction: A more appropriate introduction into the cirrus literature would be desirable.
Please check the Meteorological Monographs, Special Issue, and here, e.g.,
Heymsfield et al., Met. Monographs, DOI: 10.1175/AMSMONOGRAPHS-D-16-0010.1
Please check the book: Lynch et al., Cirrus, 2002, Oxford University Press, see chapter 2: K. Sassen: Cirrus clouds, a modern perspective, …… and all the references in this chapter.
One of the pioneering cirrus articles is from 1986: Kuo-Nan Liou, Influence of Cirrus Clouds on Weather and Climate Processes: A Global Perspective, Review Article in Monthly Weather Review , 114, 1986, https://doi.org/10.1175/1520-0493(1986)114<1167:IOCCOW>2.0.CO;2
In addition, there are many modelling papers, please check papers of Kaercher et al., Jensen et al., Spichtinger et al.. in ACP and JGR, partly given in the references of the papers, listed next:
Important papers on dust INP parameterization:
DeMott, P. J. et al., Predicting global atmospheric ice nuclei distributions and their impacts on climate, P. Natl. Acad. Sci. USA, 107, 11217–11222, https://doi.org/10.1073/pnas.0910818107, 2010.
DeMott, P. J. et al., Integrating laboratory and field data to quantify the immersion freezing ice nucleation activity of mineral dust particles, Atmos. Chem. Phys., 15, 393–409, https://doi.org/10.5194/acp-15-393-2015, 2015.
Niemand et al., Parameterization of immersion freezing on mineral dust particles: an application in a regional scale model, J. Atmos. Sci., 69, 3077–3092, 2012
Ullrich et al., A new ice nucleation active site parameterization for desert dust and soot, J. Atmos. Sci., 74, 699–717, 2017.
Important INP review papers:
Kanji et al., Chapter 1: Overview of ice nucleating particles, Meteor Monogr. Am. Meteorol. Soc., 58, 1–33, https://doi.org/10.1175/amsmonographs-d-16-0006.1, 2017.
Knopf, D. A. and Alpert, P. A.: Atmospheric ice nucleation, Nat. Rev. Phys., 5, 203–217, https://doi.org/10.1038/s42254-023-00570-7, 2023.
Recently published simulation papers (the references contain many relevant papers since 2000):
Kärcher, B., DeMott, P. J., Jensen, E. J., and Harrington, J. Y.: Studies on the Competition Between Homogeneous and Heterogeneous Ice Nucleation in Cirrus Formation, J. Geophys. Res.-Atmos., 127, e2021JD035805, https://doi.org/10.1029/2021JD035805, 2022.
Kärcher, B., Jensen, E. J., Pokrifka, G. F., and Harrington, J. Y.: Ice Supersaturation Variability in Cirrus Clouds: Role of Vertical Wind Speeds and Deposition Coefficients, J. Geophys. Res.-Atmos., 128, e2023JD039324, https://doi.org/10.1029/2023JD039324, 2023.
Please check the papers of Krämer et al. , cirrus aircraft observations, overview papers!
Krämer et al., A microphysics guide to cirrus clouds – Part 1: Cirrus types, Atmos. Chem. Phys., 16, 3463–3483, https://doi.org/10.5194/acp-16-3463-2016, 2016.
Krämer et al., A microphysics guide to cirrus – Part 2: Climatologies of clouds and humidity from observations, Atmos. Chem. Phys., 20, 12569–12608, https://doi.org/10.5194/acp-20-12569-2020, 2020.
Krämer et al., A microphysics guide to cirrus – Part 3: Occurrence patterns of cloud particles, Atmos. Chem. Phys., 25, 13563–13583, https://doi.org/10.5194/acp-25-13563-2025, 2025.
Line 46: The used reference should be substituted by Hoose and Moehler 2012, Kanji et al. 2017, and Knopf and Alpert, 2023. With respect to dust, often cited papers are: DeMott et al., 2010, 2015, Niemand et al., 2012, Ullrich et al., 2017
Line 48: The first lidar network paper reporting a Saharan dust outbreak towards Central Europe (in October 2001) was published in 2003 (Ansmann et al., 2003). Could be mentioned!
Line 51: Here one should mention the paper of Atkinson et al., 2013, highlighting the feldspar impact.
Lines 86-88: Are you sure that the basin-like topography favors accumulation of air masses after long -range transport and provides favorable conditions for the formation and persistence of high level clouds? Are there articles that confirm or support your hypothesis?
Line 92: The study is based on hourly dust COLUMN mass density fields. In the case of a dust event, how sure can you be that that dust occurred at cirrus formation height level? To my opinion, Saharan dust loading is usually confined to the height range from the surface to 6 km height (over Europe), and the upper troposphere is often not affected by Saharan dust. However, cirrus preferably starts formation close to the tropopause (in situ cirrus). Any comment on this would be good!
Does MERRA-2 provide height information about dust loading? … so that on could check how often the upper troposphere is dust-loaded or dust-free when MERRA 2 indicates a dust event in terms of dust column mass concentration.
Lines 102-107: Later on, you show effective radius plots (Figure 2) in which the effective radius changes by a few micrometers within 20-25 years. How accurate are these MODIS retrieval products? I think some uncertainty bars in Figures 2d,e,f would solve the problem, whether the trends are trustworthy or not.
Line 115: A reference to the model GAM would be fine.
Lines 157-164: Here again, do you have the potential to check how often dust was present at heights above 7-8 km when the dust column mass density threshold was exceeded? Mattis et al., JGR, 2008 provide height resolved statistics on Saharan dust events over Germany? Please check that paper!
Mattis et al., 2008, JGR,113, D20202, doi:10.1029/2007JD009636, 2008
Result section:
Tables 1,2, and 3: I always ask myself in case of large and complex tables, what is the key information here? And this key information should be visualized! Readers are attracted by good section headlines and by figures, not by long tables. The tables could be moved into the Appendix section or in a Supplementary unit. I leave that open to the authors, what to do with the tables.
Lines 215-216: It would be nice to have a sketch to illustrate high and low pressure systems and resulting air mass transport features of types 1,2, and 3. But then the question arises, the meteorological pattern leading to types 1, 2 and 3 belong to the main Saharan dust layer? And the main Saharan dust layer is typically below 6 km above Central Europe. How is then the link between the air mass transport pattern and the cirrus formation event, mainly taking place above 8 km height?
It would be interesting to have even a class (some kind of type 4) without any dust, i.e., a dust-free class, i.e., MERRA indicates upper tropospheric (sulfate) aerosol background conditions and thus homogeneous freezing conditions?
Line240-242: You discuss effective-radius tendencies of 30.3, 29.9, and 28.2 micrometer. Is it really possible to derive effective radius information with an accuracy of better than 1 micrometer. For me, all the values just show effective radii around 30 micrometer and thus rather constant microphysical properties over the 25 years.
Line 251-255: You discuss Table 2, but you did not mention or introduce Table 2 in the main text body? Should be improved! Could be done in line 251.
Line 323: Are middle and high-level cloud fractions somewhere precisely defined? If yes, please indicate here where that was done, as a reminder. Please provide clear definitions somewhere.
Lines 336-338: Is that statement in line with simulations of Kaercher et al., 2022 (heterogeneous versus homogeneous ice nucleation, the paper is mentioned above)? Usually, simulation show that more ice crystals are produced in homogeneous freezing processes than in heterogeneous ice nucleation processes.
Figure 2 is the highlight of the paper! Again, I miss a clear definition of cloud cover and defined height regions for low, mid, and high clouds.
We need (a), (b), (c), etc…. within the six different panels.
Figure 2e (ice cloud particle radius). This effective radius information is probably related to the cirrus top region. In the virga zone, the effective radius is probably larger than 100 microns. Again, the range of values from 28 to 33 micrometer is so small and uncertainty bars are missing! Can we trust the trends in the different figures?
Figure 2f (outgoing longwave radiation): You were able to clearly separate the dust impact on the outgoing IR radiation fluxes from the respective cirrus impact on the outgoing IR radiation? Please comment on that within the discussions presented in lines 347-365 (Sect. 3.5).
Section 4 (Discussion):
Lines 373-379: Please check the papers of Pieper et al. (2026) (observations) and Juurikkala et al. (2026) (simulations) regarding the impact of atmospheric dynamics (wind shear, turbulence, and consequences for updraft strength and related supersaturation strength) on ice crystal nucleation! The fundamental question is: Is the enhanced INP number concentration responsible for all the trends in your results? Or is that related to changes in the dynamics when Saharan dust events occur. With other words, please include the dynamics aspect in your discussion!
Pieper, L., Seifert, P., Griesche, H., Ansmann, A., McCosh, G., Ewald, F., and Kneifel, S.: Turbulence structure of dusty cirrus and its relation to atmospheric conditions: A characterization using cloud-radar measurements and reanalysis data, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-3092, 2026.
Juurikkala, K., Raatikainen, T., Krämer, M., Williamson, C. J., and Laaksonen, A.: Mineral dust concentration controlling convective cirrus structure and persistence: a large-eddy simulation study with observational constraints, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-2427, 2026.
Line 382: You are using dust column values! Maybe the dust fraction (contribution to the column value) for the upper troposphere increased with time?
Line 388: Again, the effective radius changes are so small! Is it really possible to retrieve this property so precisely over 25 years and for different dust column mass values? Are all the trends trustworthy? Do changes of a few micrometers rather than by 20-50 micrometers or even more, indicate different formation conditions? To my opinion, changes by 2 micrometer with inceasing dust column mass and variations by 3 micrometer withing 25 years indicate more or less unchanged cirrus conditions from 2001 to 2025! Please defend your position and comment on this!
Line 468: DIBS was obviously presented in Ansmann et al. (2019a), not in the Ansmann et al. (2019b) conference paper. In the 2019a paper, the entire lidar/radar approach is described as well. So, Ansmann et al. (2019b) is not needed.
Sections 4.3 and 4.4 should be kept short. They are too long and not very interesting. Such a general and extended discussion is not needed in this paper. It would be more interesting to formulate some fundamental questions on future cirrus developments in a changing climate, and leave the answers open to the modelling community in order to find plausible solutions to these questions.