Cold pools mediate mesoscale adjustments of trade-cumulus fields to changes in cloud droplet number concentration

Alinaghi, Pouriya; Jansson, Fredrik; Blázquez, Daniel A.; Glassmeier, Franziska

doi:https://doi.org/10.5194/acp-25-6121-2025

Articles | Volume 25, issue 12

https://doi.org/10.5194/acp-25-6121-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-25-6121-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 12

Research article

|

20 Jun 2025

Research article |

| 20 Jun 2025

Cold pools mediate mesoscale adjustments of trade-cumulus fields to changes in cloud droplet number concentration

Pouriya Alinaghi, Fredrik Jansson, Daniel A. Blázquez, and Franziska Glassmeier

Download

Final revised paper (published on 20 Jun 2025)
Preprint (discussion started on 20 Nov 2024)

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-3501', Anonymous Referee #1, 05 Jan 2025
This manuscript, "Cold pools mediate mesoscale adjustments of trade-cumulus fields to changes in cloud-droplet number concentration," presents an excellent study on the interplay between cloud-droplet number concentration (Nc) and trade-cumulus cloud fields, with a particular focus on cold pools. The study is timely, well-executed, and provides valuable insights into a topic essential for understanding cloud-climate interactions. The study identifies two distinct cold-pool regimes driven by Nc: one characterized by long-lived, organized structures formed through frequent cold-pool interactions at low Nc and another dominated by sparse, short-lived cold pools at high Nc. The study concludes that mesoscale dynamics, driven by Nc variability, significantly contribute to trade-cumulus feedback comparable to large-scale cloud-controlling factors.
In my opinion, this is a very well-written and clear manuscript that will contribute meaningfully to the field. The study is carefully designed, and the methods and results are robust. I have a few minor questions to further enhance the clarity of some concepts and results. I would like to recommend this manuscript for publication with only minor revisions.

L118-119:

The sentence “The lower boundary of hmix is determined by subtracting the difference between the upper boundary and the mode from the mode” is somewhat unclear to me.
Would it be more correct for me to understand it as: “The lower boundary of hmix is calculated by first determining the difference between the upper boundary and the mode, and then subtracting this difference from the mode”?

Figure 4, cold pool numbers and sizes

Figure 4. Caption:
“Note that the reason for the cold pools in simulation Nc = 70 /cm³ forming earlier than Nc = 20 /cm³ is that cold pools featuring a diameter of <5 km are not considered to be part of the cold-pool mask.”
Could you clarify why an additional filtering condition could not be applied to exclude such small cold pools from the threshold? Additionally, why do such small but significantly varying cold pools appear specifically in the Nc = 70 /cm³ simulation? Understanding this would provide further insight into the dynamics of cold pools in this regime.
Overall, I think the manuscript does an excellent job explaining the differences between Nc = 20 /cm³ and Nc = 1000 /cm³. However, Figure 3 suggests that the cold pools in the Nc = 50, 70, and 100 /cm³ cases exhibit the most pronounced structures and counts. Yet, these cases do not appear to receive the same level of synthesis or discussion as the extreme cases (e.g., Nc = 20 and 1000 /cm³) shown in Figure 5. Could you elaborate on the reasoning for this or include more analysis of these intermediate cases?

Figure 4. Caption:
“Time series of cold-pool number and size are slightly smoothed using Gaussian filtering.”
Could you provide more details about the Gaussian filtering method used here? Specifically, was the same filtering applied to both cold-pool number (Figure 4c) and size (Figure 4d), and how might this smoothing influence the interpretation of the results?

L155 & Figure 4d:
The term “domain-mean size of cold pools, Scp” lacks clarity regarding its units and definition. In Figure 4d, the unit is shown as kilometers (km). Should this instead be km² if it represents the area, or does “size” specifically refer to the radius or diameter of the cold pool? Adding this clarification would help ensure consistent understanding.

Figure 4:
In addition to the mean size shown in these panels, would it be possible to include the time evolution of the 90th and 10th percentiles, or perhaps the 75th and 25th percentiles, of cold-pool sizes? This could provide a clearer picture of the size distribution over time, especially since the maximum and minimum sizes—particularly for Nc = 70 /cm³—are both extremely large and highly variable across different Nc values. Including this information might help better illustrate the variability and outlier behavior in the size distributions.

Would Nc affect initial cold-pool field?

Figure 5:
This is a general question. The manuscript does an excellent job of explaining how, at large Nc values, the fewer cold pools and their greater distances reduce the subsequent collisions that could generate new convective events (e.g., L238, L349). However, my question concerns the initial cold-pool field: why do simulations with large Nc values already exhibit fewer and more widely spaced cold pools from the very beginning of the convective mode? Are there specific physical mechanisms at play here? For instance, could Nc directly affect precipitation evaporation rates, thereby influencing the initial cold-pool field?

Definition of Self-Organization

This manuscript presents an intriguing conclusion: when cold pools are suppressed, self-aggregation of cloud fields is enhanced by increased Nc (e.g., L333). However, the calculation of convective organization metrics appears challenging and nuanced. For instance, the evolution of the Iorg index shown in Figure A1 (g) and (h) does not align with this conclusion. On the other hand, I appreciate the use of the spatial standard deviation of the liquid-water path (σL) in Figure 6(a), which provides a unique perspective on cloud organization.
That said, the definition of "organization" used in this paper is somewhat unconventional, particularly since cold pools seem to play a limited role under this definition. I would suggest adding more discussion about how "organization" is defined and the physical meaning behind this choice. Such an explanation would clarify why the conclusion here—that cold pools are suppressed but the organization is enhanced—differs from the commonly understood relationship where cold pools generally increase convective organization.
Additionally, I would like to request more explanation for the Nc = 70 /cm³ case. As seen in Figure 3, the clouds under this condition exhibit pronounced spatial heterogeneity (and apparent organization) from hour 30 to 50. Providing further analysis of the medium Nc case could help me—and other readers—better understand its unique characteristics and their implications.
Citation: https://doi.org/10.5194/egusphere-2024-3501-RC1
RC2: 'Comment on egusphere-2024-3501', Anonymous Referee #2, 06 Feb 2025

Summary
Alinaghi et al. investigate the sensitivity of trade cumulus clouds and their cold pools to changes in droplet number concentration. They expand on a previously developed ensemble of high resolution, large eddy simulations (Cloud Botany) to include simulations experiencing varied, fixed droplet number concentrations (Nc= 20,50,70,100,200,1000/cc). This novel, expanded ensemble allows them to investigate how cloud behavior changes under fixed Nc and how the Nc influence compares to various cloud controlling factors (CCFs, including the diurnal cycle) that were explored in the Botany ensemble. Specifically, the analysis focuses on how cold pools, which control the organization of these clouds through their dynamics, are influenced. In addition to cold pool specifics, they analyze rain and liquid water amount and radiative properties expected to change due to the Twomey effect. Their results for rain, liquid water, and cloud amount are consistent with limited cases in the literature that had more complex microphysics. They find cold pool behaviors are different at low and high Nc concentrations, the latter leading to more spread-out cold pools and less dynamic interaction. They further find that the influence on cloud albedo is comparable to other CCFs. Overall, this work makes a strong and compelling case for needing to understand aerosol forcing in trade cumulus as well as cumulus feedback, a cutting edge concept. The manuscript is extremely well done with careful analysis that has been conducted skillfully and presented succinctly. I have a few clarifying comments and suggest that the limitations/behavioral influence due to fixed Nc should be made clearer. I recommend prompt publication after these issues are resolved as this manuscript will be a valuable contribution to the field.

Major comment
The authors are careful to focus on the Twomey effect and not fully discuss adjustments in their aci analysis. Adjustments are not possible with fixed droplet number so this focus is reasonable. However, it would be helpful to be clearer about the simplifications in your treatment of aci and how that influences the behavior of cumulus, their cold pools, and their dynamic interactions that ultimately drive their organization. Specifically, these will all be different from their behaviors in reality. The brief discussion at the end about these simplifications is good but it is worth expanding on these limitations in Section 2 when presenting the simulations and discussing their potential impact when there are key points sensitive to these assumptions. This would be particularly helpful in Section 3.4 and the discussions of cumulus lifetimes, cold pool cycling, and cold pool interactions. This added context would also help strengthen the connections between this work and current observational analysis (e.g., from EUREC4A, the basis of the Botany ensemble).

Detail comments
Intro: Compelling and very well written, synthesizing the research in the field while motivating this work well.
Intro: typo, should be “the Twomey” and “the Albrecht” effects.
Line 30-32: Consider providing a reference length and time for the “mesoscale” and “meso-timescales” mentioned.
Line 75-76, throughout: This is an admirable goal. However, because not all clouds have cold pools isn’t this a subset of the trade cumulus behaviors that you are looking at? Might be worth mentioning that somewhere.
Line 118-119: typo, repeated “the mode”
Figure 4, throughout: It is great to see the similarity between your results and the cases from the literature that have used more complex microphysics, really strengthens your analysis discussion.
Figure 4, 6, and 7, A1 and A2: I understand your choice of colors here when you show all the Nc’s. However, if you are focusing on the three (20, 70, 1000) it is a little confusing that 1000 is also red (line weight is great though, very clear, thanks). If not time consuming, would it be possible to change that line purple or something so that the colors gradate from warm to cool? Otherwise, beautiful and clear figures throughout, nice work.
Line 200-202: the moisture memory hypothesis is very interesting and would be worth testing at some future point. Is their literature that supports this idea as well?
Line 202-203, 240, throughout: I don’t totally understand what you mean by compensating here. This seems like a key point, could you expand on what is being compensated and clarify your thinking?
Figure 4, 6, 7, 8, A1: Could you clarify how you are defining “sensitive to Nc”? Is there a statistical test? The error bars seem to overlap for many of the connected points… so it seems more like whether there is a trend? I found myself wondering if you could have non-monotonic sensitivity to Nc, e.g., due to some sort of buffering in cloud behavior at higher Nc concentrations…
Figure 7 caption: worth mentioning the hour ranges (or marking them on the timeseries?) for quick reference of the transient and equilibrium periods.
Line 304-310: intriguing discussion, these future studies will be very interesting.
Line 358-59: it would be useful to mention here, and throughout the paper when you focus on these variables, that the liquid water path and cloud fraction responses do not consider adjustments. Generally, I think you are careful about this (focusing on the Twomey effect) but worth clarifying that this is not the whole story because of the fixed Nc strategy and the behavior could differ (would you be able to generally indicate how it might differ based on what we know about adjustments?).
Line 375-386: Strong ending summary and discussion of some of the limitations while still highlighting the value of this paper. I would recommend discussing more about the limitations earlier on, both when you introduce the set up and when you are discussing the aci, cold pool interactions, and time evolution parts of the analysis (see other comments).

Citation: https://doi.org/10.5194/egusphere-2024-3501-RC2
AC1: 'Comment on egusphere-2024-3501', Pouriya Alinaghi, 21 Mar 2025

Our responses to both reviewers are included in the document attached.

Citation: https://doi.org/10.5194/egusphere-2024-3501-AC1

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Pouriya Alinaghi on behalf of the Authors (21 Mar 2025) Author's response Author's tracked changes Manuscript

ED: Publish as is (28 Mar 2025) by Raphaela Vogel

AR by Pouriya Alinaghi on behalf of the Authors (28 Mar 2025)

Short summary

Shallow clouds in the trades are a major source of uncertainty in climate projections. These clouds organize into striking mesoscale patterns that are exactly what climate models lack. This study explores the origin of such patterns and investigates how variations in microscale properties control them. The importance of microscale effects is compared to that of large-scale forcing on the mesoscale organization of trade-cumulus fields.