the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Jun 2025
Measurement report: Lidar observations of cirrus cloud properties with CALIPSO from midlatitudes towards high-latitudes
Abstract. Cirrus clouds play a crucial role in the Earth’s radiation budget. However, direct observations and model simulations of cirrus at high-latitudes are still sparse. In this study, we present the occurrence rate (OR) and geometrical thickness as well as extinction and particle linear depolarization ratio (PLDR) of cirrus at midlatitudes (35–60° N; 30° W–30° E) and high-latitudes (60–80° N; 30° W–30° E) using lidar measurements of CALIPSO in the years 2014 and 2018–2021. The results indicate a distinct seasonal cycle in the cirrus occurrence and optical properties. The seasonality in ORs and geometrical thicknesses generally becomes more pronounced with increasing latitude, while the altitude ranges of cirrus decrease with increasing latitude. The extinction coefficients decrease with increasing altitude at both high- and midlatitudes and are, in addition, larger at midlatitudes than at high-latitudes in all seasons. The calculated effective optical depths also show larger values at midlatitudes than at high-latitudes, while the differences across latitudes in winter are negligible. The distributions of PLDR in each 5-degree latitude bin show a general decrease with increasing latitude, leading to a remarkable latitudinal difference with larger values at midlatitudes than at high-latitudes. Finally, we compare the aerosol concentrations at different latitudes acting as ice-nucleating particles (INPs) to trigger heterogeneous freezing, as reported in previous studies. It turns out that aerosols such as mineral dust and soot (including aviation-induced soot) indicate much larger concentrations at midlatitudes than at high-latitudes.
- Preprint
(3158 KB) - Metadata XML
-
Supplement
(2810 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2052', Anonymous Referee #1, 25 Jul 2025
The authors present spaceborne (CALIOP) lidar observations of cirrus clouds and compare the measurements performed at midlatitudes and high latitudes. This is a good contribution to cirrus research at mid to high northern latitudes. However, several parameters are not well defined. The discussion of the results needs to be improved. Uncertainty ranges need to be given.
Minor revisions are needed.
P1, line 10: We need a clear definition of the ‘effective optical depth’. Why do you introduce ‘effective’?
P2, line 57: Here, you could add the recent MOSAiC publication on smoke and cirrus (Ansmann et al., ACP; 2025).
P3-P4: Most of the information on page 3 and page 4 are not needed and could be left out. The shorter the introduction the better. Come to your point of research in 1.5-2 pages!
P4, Eq.(1): Why do you use ‘effective’? It is simply the cirrus optical depth! In this context you may already explain how you get the single-scattering extinction coefficient sigma-ci. How did you correct for multiple scattering? Maybe you use the backscatter coefficient, obtained from the CALIOP observation with a lidar ratio that considers multiple scattering? What lidar ratios does the CALIOP team assume in their retrieval of the cirrus backscatter coefficient? And what lidar ratio do they assume in the multiplication of the backscatter coefficient to obtain the single scattering extinction coefficient? This is important information that is missing in this lidar paper on cirrus optical properties.
More general, how is the extinction coefficient profile determined? How large are the uncertainties when using the Klett technique? I speculate, probably much larger than 50%!
P4, line 123: CALIOP is nadir pointing? It measures at an off-nadir angle of 3°! So, it is off-nadir pointing! One could briefly explain why an off-nadir pointing is selected.
P5, line 149: ‘Occurrence rate’? You probably mean: Frequency of occurrence! The word ‘rate‘ points to occurrence per second. More common is to use ‘Frequency of occurrence’, i.e., number of cirrus layers occurring within a given latitudinal belt within a given season.
P5, line 158: You write: The altitude ranges in which cirrus formed… can be seen in Figure S1. How do you know where cirrus formed? I speculate that you often detect just virga segments far below the height where ice crystal nucleation (in situ cirrus formation) took place. Please, be more clear in this respect. You may also use ‘height interval’ as an alternative to ‘height range’. In conclusion, you mean the height interval in which cirrus segments were found… or cirrus clouds occurred. Please state that clearly!
P6, lines 166-168: The extent, size, or depth of the height interval in which you detected cirrus layers mainly depends on the size or vertical extent of the virga zones, i.e., by the base heights of the detected virga layers. That means: ‘Height range of cirrus formation’ is definitely misleading wording. This hold for the entire page 6.
P7, l 197: The geometrical thickness of a cirrus cloud is obtained from the knowledge of the base and top height of a given cirrus layer. Your definition is confusing (line 197): the geometrical thickness of cirrus cloud is defined as the vertical distribution of cirrus clouds at different latitudes. What do you mean here? What do you want to tell us?
Figure 3 is misleading. You need to improve the figure, you have to write clearly: >0.1km, >0.3km, >1.0 km, >2.0km. Why do you not show histograms? If my interpretation of Figure 3 is correct, most detected cirrus layers have thicknesses from 0.3 to 2 km. I further speculate that thin cirrus layers with 100 to 200 m thickness are probably sublimating virga structures at all. Cirrus nucleation cells show immediately cirrus thicknesses of >300m within 20-30 minutes after nucleation of first ice crystals, resulting from ice growth and sedimentation processes.
P7, line 214-215: better write 39% , 13.5% and 15.5%.
P7, line 219: ‘rate of decrease’ is misleading! You simply mean ‘decrease’
P8, line 223-231: In such a cirrus paper, we need clear information on the CALIOP cirrus data analysis! As I already asked above: how is the extinction coefficient obtained from the CALIOP raw data. What lidar ratios did they use to obtain the Klett solutions for the cirrus backscatter coefficient? Please, provide numbers here. The lidar ratios, used in the Klett procedure, consider multiple scattering. Afterwards, the multiple-scattering-corrected cirrus backscatter coefficients is obtained. What lidar ratio did they use next to obtain the respective single-scattering extinction coefficient? This extinction profile can then finally be used to calculate the cirrus optical depth. It is simply not sufficient here to provide the reference Vaughan et al. (2009). At the end, we need to know how large the uncertainty in the used cirrus optical depths are! The use of ‘effective optical depth’ is confusing for all non-lidar scientists.
P8, line 232: The discussion of extinction coefficients shown in Figure 4 must include the uncertainty in the CALIOP retrieval products. The solutions for the extinction coefficients could be compared with the ones shown for the MOSAiC cirrus clouds (Ansmann et al., 2025) and also with other studies mentioned in that paper.
P9. Yes, there may be a daytime vs nighttime difference in the cirrus extinction coefficient. But daytime CALIOP data are much noisier than nighttime data… and the Klett solutions may be much more uncertain for daytime cirrus cases than for nighttime cirrus cases. Therefore, we need the CALIOP extinction uncertainty information! Trustworthy uncertainty numbers start at 50%! Uncertainties are larger for noisy daytime data than for less noisy nighttime data.
P10: In Fig.6 (and S6-S8), I do not see a clear change of the particle depolarization ratio with latitude, only a weak tendency. The variability indicated by the size of the boxes and bars is so large that a clear dependence is not obtained.
Also the differences (HL vs ML depolarization ratios) are quite small, and if day vs nighttime observations show strong differences one must be careful in the interpretation of the results because of the different background noise impact on nighttime vs daytime products.
Please, provide your hypothesis on the link between aviation (stronger at daytime), crystal size, and depolarization earlier, i.e., already on page 10. What about the impact of shape…. plates vs columns … on the depolarization ratio? Could be discussed as well.
P13: What about the impact of wildfire smoke on cirrus evolution? Should be included in the discussion. Furthermore, long-range transport of aerosol in the (upper) free troposphere occurs everywhere, at all latitudes. So, why should there be a decrease of the aerosol content in the upper troposphere with increasing latitude?
P14: Conclusions should be compact and short. Again you write: Cloud formation shows a clear decrease with latitudes. Maybe!, but you observed cirrus features from the top to the virga bottom. The height distribution of cirrus formation cannot be derived from CALIOP cirrus observations.
P14, lines 444-446: Again these temperatures are related to all cirrus observations, but not exclusively to locations of cirrus formation.
P15, l 452: You do not have measured aerosol concentrations! You did not report any aerosol observation (including typing) for the upper troposphere. So this last paragraph presents mainly your speculative ideas, rather than clear observations. It seems to be that wildfire smoke plays a major role in the upper troposphere over the mid latitudes as well as over the high latitudes, especially since the strong Canadian wildfires in the autumn of 2017. However, all these statements are just hypotheses. Please state that clearly!
Figure 1: dashed lines are shifted by 3%? What does that mean? a), b), c), d) is missing
Figure 2: You mean ‘Frequency of Occurrence’? a), b), c), d) is missing
Figure 3: Occurrence frequency! No longer OR. a), b), c), d) is missing
Figure 4: a) and b) is missing. Uncertainty bars (SD bars) are missing. They would show how trustworthy the observed differences are.
Figure 5: a), b), c), d), e), and f) is missing. The values could be compared with other studies (e.g., Ansmann et al., 2025).
Figure 6: a), b), c), d), e), and f) is missing. Very harmonic data, a tendency is visible…
Figure 8: a), b), c), d), e), and f) is missing. What can we learn from a distribution of temperatures linked to the detection of cirrus filaments and structures. The shown temperatures do not show cirrus formation temperatures.
Citation: https://doi.org/10.5194/egusphere-2025-2052-RC1 -
RC2: 'Comment on egusphere-2025-2052', Eleni Marinou, 26 Aug 2025
This article by Li and Gross presents an analysis on statistics of cirrus cloud properties at Mid and high latitudes, and discusses the different findings per season, altitude, temperature, and aerosol abundance. The results are new, and the study is relevant to the objectives of the journal. The work is complete, scientifically accurate, and significant, and the manuscript is well-written and well-structured. Overall, the study is suitable for publication. Certain sections could benefit from some additional clarifications, as described herein.
General comments:
Smoke layers from Canadian fires were frequently detected from CALIPSO during the years of the study. It would be interesting to include a discussion about the possible effect of the elevated smoke layers in the stratosphere on the ice in these altitudes in the 2 domains.
Consider including in the abstract the information that the cloud statistics of this work focus on temperatures <-38C, excluding ice observations above these temperatures.
In section 4.3, it is not clear what the contribution of this study is to this discussion. I suggest revising the text to make your results clearer in relation to or in addition to the past studies in this summary. The way it is written now could be part of the introduction of this paper.
Specific comments:
Page 1, line 9: “The distributions of PLDR in each 5-degree latitude bin show a general decrease with increasing latitude”: Suggestion to add the physical meaning of increased PLDR.
Page 4, line 124: “level 2 5-km cloud profile products”: It is useful to include the version of the product.
Page 4, line 125: “ all the atmospheric entities”: With this phrase, one may be confused whether the product also includes information on the aerosols. Consider revising.
Page 5, line 128: “ ..VFM..”: It is useful to include the version of the product.
Page 5, line 138: ”mid-latitudes (35–60◦N; 30◦W–30◦E) and high-latitudes (60–80◦N; 30◦W–30◦E)”: suggestion to include a figure with the map and the 2 domains (maybe in the appendix).
Page 5, line 142: “in 5 years of 2014 and 2018–2021 are analyzed”: Please include a short explanation why you exclude the other good CALIPSO years (2007-2017).
Page 5, lines 156-158, and Table 1: I suggest considering excluding the 2nd digit after the decimal point (statistically not significant).
Page 6, line 6: “variations in the altitudes with the maximum ORs along the latitudes are discernable, showing the largest values in summer”: suggestion to change as “showing the largest altitude values in summer”.
Page 6, line 192: “which is related to the larger variabilities in humidity at HL than at ML and is consistent with a recent model study showing larger INP effects on cirrus at higher latitude”: Isn’t the largest variability of temperature also a significant contribution for HL ORs also?
Page 7, line 202: “show that the thickest cirrus clouds formed in winter and the thinnest ones in summer”: I am not sure I see this in the plot, as overall Winter has the highest CR for thin clouds also. Can you rephrase this part to make it clearer?
Page 7, line 210: “presumed to be smaller at HL than ML”: Is this correct? It seems higher at HL.
Page 7-8, section 3.2. Consider including a sentence on why you concentrate in Spring among all the seasons, if there is a reason of interest to the reader.
Page 8, line 234: “indicating smaller and fewer ice crystals at higher altitudes”: suggestion to rephrase to and/or, and it could be one or the other also.
Page 8, line 237: “This is closely linked to the dominant formation processes of ice crystals depending on temperature and relative humidity over ice (RHi). “ This is also closely linked with the most frequent abundance of INP in these altitudes at ML. Consider revising this part to avoid confusion.
Page 13, line 400: “more irregular”: more in comparison to what? Please enhance this sentence for clarity.
Page 15, line 452: “we compare aerosol concentrations at different latitudes”: Where is this comparison shown? Consider adding a plot of the aerosol concentrations or revising the sentence by e.g., “we compare ice crystal concentrations in regions of different aerosol concentrations as reported in previous studies”.
Figure 1: If possible, add a scale indicating the magnitude of the OR[%] in these plots.
Figure 3: Based on these occurrences, ML has more clouds than HL in spring and summer. This is surprising. Can you include a comment in the manuscript on this, maybe backing up the findings with past studies on cloud abundance, or is this a new finding? Also, it would be useful to include a description or equation on how these occurrences are calculated.
Typos:
Page 5, line 138: midlatitudes: mid-latitudes
Page 6, line 160: discernable: discernible
Page 11, line 342: exam: examin
Figure 1: 2014 (2018): 2014 and 2018
Figure 3: occurrence frequency: occurrence rates?
Citation: https://doi.org/10.5194/egusphere-2025-2052-RC2
Data sets
Measurement report: Lidar observations of cirrus cloud properties with CALIPSO from midlatitudes towards high-latitudes Qiang Li and Silke Groß https://zenodo.org/records/15063832
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 330 | 42 | 17 | 389 | 24 | 12 | 29 |
- HTML: 330
- PDF: 42
- XML: 17
- Total: 389
- Supplement: 24
- BibTeX: 12
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1