the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Dec 2024
Advances in CALIPSO (IIR) cirrus cloud property retrievals – Part 2: Global estimates of the fraction of cirrus clouds affected by homogeneous ice nucleation
Abstract. Cirrus clouds can form through two ice nucleation pathways (homo- and heterogeneous ice nucleation; henceforth hom and het) that result in very different cloud physical and radiative properties. While important to the climate system, they are poorly understood due to lack of knowledge on the relative roles of het and hom. This study differs from earlier relevant studies by estimating the relative radiative contribution of hom-affected cirrus clouds. Here, we employ new global retrievals (described in Part 1) of cirrus cloud ice particle number concentration, effective diameter (De), ice water content, shortwave extinction coefficient (αext), optical depth (τ), and cloud radiative temperature based on Imaging Infrared Radiometer (IIR) and CALIOP (Cloud and Aerosol Lidar with Orthogonal Polarization) co-located observations onboard CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation). Transition from het only to hom affected regimes are identified using αext and De. Over oceans outside the tropics in winter, the zonal fraction of hom affected cirrus generally ranges between 20 % and 35 %, with comparable contributions from in situ and liquid origin cirrus. Using τ distributions to establish a proxy for cloud net radiative effect (CRE), the τ-weighted fraction for hom affected cirrus over oceans outside the tropics during winter was > 50 %, indicating that hom cirrus play an important role in climate. Moreover, the climate intervention method known as cirrus cloud thinning could be an effective cooling method at high latitudes based on this τ-weighted hom fraction. A conceptual model of cirrus cloud characterization is proposed from these retrievals.
- Preprint
(11420 KB) - Metadata XML
-
Supplement
(1123 KB) - BibTeX
- EndNote
Status: open (until 31 Jan 2025)
-
RC1: 'Comment on egusphere-2024-3814', Blaž Gasparini, 06 Jan 2025
reply
This manuscript analyzes cirrus cloud properties using a satellite retrieval product. It introduces a criterion for distinguishing between microphysical formation mechanisms of cirrus, often based on simple thermodynamic theory. Notably, the study not only estimates the fraction of homogeneously formed cirrus but also assesses their optical depth-weighted contribution, providing a more accurate estimate of their radiative and climatic relevance.
This manuscript has the potential to make an important contribution to our understanding of cirrus clouds. However, several issues undermine its impact, mainly related with the manuscript length, logic, and figure presentation.
Key comments:
1. Excessive manuscript length (particularly figure number)
The manuscript includes 27 figures, many with 8-12 subplots. This abundance, coupled with occasional topic detours, dilutes the key messages and diminishes the paper’s relevance for the community. Reducing the number of figures (or at least figure panels) while focusing on the paper's key results is recommended. For instance, tropical data could often be excluded, as it is not the main focus of the study.
To improve readability/simplify the manuscript, the authors could limit the main text’s analysis to a cloud optical depth range of 0.3 to 3, moving additional analysis to an appendix.2. Colormap use
The manuscript frequently uses a broken colormap, which is (in my opinion) visually appealing but requires justification. If the switch between cold and warm colors aligns with a physically meaningful threshold, this should be explicitly stated. Otherwise, a perceptually uniform colormap should be used to ensure clarity and accessibility.3. Seasonality figures
The current plots make it challenging for readers to discern seasonality in cirrus cloud properties. Key figures could represent seasonality more intuitively, for example, by showing relative anomalies from annual means.4. Key science issue: Homogeneous freezing of solution droplets vs. homogeneous freezing of cloud droplets at homogeneous freezing temperature of water
The manuscript treats homogeneous ice nucleation in situ and homogeneous freezing of cloud droplets as equivalent, which they are not. This distinction is critical due to its implications for cirrus cloud thinning. While in situ nucleated homogeneous clouds are promising targets for seeding, current thinning methods cannot modify clouds forming at water’s freezing temperature. Although alternative modification strategies (e.g., convective invigoration or, possibly, the opposite, weakening; e.g. Varble et al., 2023, https://doi.org/10.5194/acp-23-13791-2023) could be developed, they fall outside the scope of this study.5. Additional key comment - data accessibility
To enhance the utility of the dataset for the community, the authors could consider publishing their post-processed data in a user-friendly format, such as NetCDF following CF conventions. This would facilitate its use by climate modelers.Specific comments:
I'm sorry if some comments repeat points already made in the key comments section.Line 117 and Appendix A:
Why should such cases be rare? This is hardly justified. It certainly cannot be the case for cirrus originating from deep convection, where anvils spread far from the deep convective core, see Gasparini et al. 2018, Fig. 5 (10.1175/JCLI-D-16-0608.1). In any case, it's hard to discriminate the origin with only snapshot data, without any means of tracking the cloud evolution.There are more studies that discriminate between liquid and in situ origin cirrus that could be mentioned, e.g. Wernli et al. 2016 (10.1002/2016GL068922), which uses a large statistical sample of clouds (relying on the imperfect reanalysis data, but at least with good statistics).
Line 175, but related to many of the figures:
Wouldn't it be enough to show only DJF and JJA in the main manuscript and the other seasons in the supplement?Line 177:
Should be predicted by the GCM, but only in the sampled range of cloud optical depths. The clouds at COD<0.3 are very common, and while not as radiatively important, will contribute significantly to the mean ice cloud properties.Line 188:
But also the agreement with ICNCFigures 2-7:
- If one does not pay attention and focus on some features, all subpanels generally look the same.
- In general, I like the choice of the colormap. However, it is a discontinuous colormap with a very sharp transition. This can only be used if there's a reason for such a choice. Otherwise, the reader will see patterns that aren't real, but just a result of the sharp colormap discontinuity.
Section 3.1 is, in my opinion, not important to the main story of the paper. It could be moved to the Appendix, along with the corresponding figures (or at least parts of some of the figures).
Section 3.2/Figure 12:
If the relevant threshold is 30 ICNC/liter, then the discontinuity in the color map should be set at that level.Section 3.2:
The use of extinction coefficient is not motivatedLine 320:
How is this equation derived? What is its source?Line 347:
Does "theoretical method" refer to the formula in Verlinde? I'm not sure where it comes from, but it doesn't seem like a theoretical relationship, more like an empirical fit to data?Lines 348-350:
The authors here assume that homogeneous freezing of cloud droplets is the dominant source of ice in convective plumes. It's hard to know how important it really is, but given that most convective clouds freeze at T of about -30°C, freezing of cloud droplets in the mixed-phase regime should be very important. There are quite a few references on this, although it's harder to find good information on deep convective clouds (e.g. Coopmann et al., 2020 https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2019JD032146; but at least in the tropics, most mixed-phase clouds are of convective origin).Moreover, while associated with a large uncertainty, secondary ice production is thought to contribute substantially to ICNC in deep convective clouds, see e.g. Hu et al., 2021 (https://doi.org/10.1175/JAS-D-21-0015.1)
Figure 14:
Not sure we need panels g-l, since they show the same as panels a-f.
Also, is there any physical significance to the discontinuity of the colormap?Line 400:
We cannot know how active hom is, and in any case it would be homogeneous freezing of cloud droplets.Figures 15 and 16:
Is there a physical meaning to the colormap boundary at 211 K?Lines 440-441:
What if updrafts are the key difference between land and ocean?Figure 17:
I see no difference between the dashed and solid lines, they seem to be parallel. Therefore, one of them could be removed from the plot for clarity.Figure 20:
Why is there a seasonality in the parameterization for the Northern Hemisphere (discontinuously moving from 30° to 60°N) but not for the Southern Hemisphere?Why is the word "corrected" in the title?
Figure 21:
Why do we need 4 temperature bins when the behavior seems to be pretty much the same in each of the bins (also, we are at figure 21 already, and it's starting to get harder to keep the focus).Figure 22 is, in my opinion, the key figure of the manuscript, but unfortunately it gets lost a bit due to the large amount of information presented.
Another comment, similar to the one above: I don't think it's fair to lump together in situ cirrus and cirrus of liquid origin. Only in situ cirrus can be modified. So the readers might want to have numbers of hom vs. het for in situ cirrus only.
Section 4:
This is a valuable section, but there's a lot of speculation. I am very intrigued by the results presented in Figure 24. These should be verified in the future with other observational datasets and model studies to confirm or reject the proposed interpretation of cirrus cloud properties relative to updraft.To increase the relevance of this section, the authors could compare their hypotheses with parcel model studies of cirrus, e.g. work by Bernd Kärcher, if possible.
Figure 25:
I assume that the INP number for Figure 25 and its description should be fixed. This should be stated explicitly.Does the model in Figure 25 hold for all cirrus or only for in situ cirrus?
A caveat to this interpretation is that cirrus data could also be explained as cirrus at different stages of cloud evolution.Section 5:
It seems a bit odd to have a separate section mentioning the otherwise very relevant study by Froyd et al., 2022.
I don't think the Froyd et al., 2022 study presented any ice residual measurements (unlike e.g. Cziczo et al., 2013).Figures 26 and 27:
Figures 26 and 27 are very useful for the purpose of scientific presentations, but lack some more content to qualify for a proper scientific publication. For example, can we be sure that the clouds shown are at temperatures below the homogeneous freezing temperature of water?In the best case, one could find a photo of a cirrus cloud with a coincident CALIPSO overpass and the analysis as done for the rest of the paper.
Lines 730-731:
Or, more physically, increase the vertical resolution of the model.
Or use some kind of subgrid cloud fraction in the vertical (similar to what is done in the horizontal dimension in coarse GCMs).
In any case, the pre-existing ice formulation tends to be less important when using high resolution models, and may become less important in future cirrus modeling studies.Lines 754-755:
The study by Froyd et al. 2022 used only an idealized model setup (based on dust measurements, but the trajectories were based on reanalysis data).Citation: https://doi.org/10.5194/egusphere-2024-3814-RC1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 140 | 46 | 8 | 194 | 33 | 3 | 3 |
- HTML: 140
- PDF: 46
- XML: 8
- Total: 194
- Supplement: 33
- BibTeX: 3
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1