the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Jan 2026
Interacting effects of droplet number and ice formation processes on mixed-phase cold-air outbreak clouds
Abstract. Shallow, mixed-phase clouds associated with cold-air outbreak (CAO) events are natural laboratories for studying mixed-phase cloud processes which are important for estimating cloud-phase feedback in the warming climate. Recent studies show that CAO clouds are sensitive to aerosol-cloud interactions and ice formation processes, but many modelling studies perturbed model parameters individually, limiting the investigation of joint effects from multiple processes on cloud properties. Here we investigated how six cloud microphysics parameters jointly affect CAO cloud properties by building model emulators trained on output from perturbed parameter ensembles of a high-resolution regional model. The parameters are cloud droplet number concentration (Nd), ice-nucleating particle concentration (NINP), efficiencies of three secondary ice production processes, as well as the mixed-phase overlap factor (mpof). For the CAO case studied, Nd and NINP most strongly control the cloud radiative properties in the stratocumulus region; whereas in the cumulus region, effects from varying Nd and mpof are the strongest. Our results show that these parameters have non-linear joint effects such that the magnitude and even sign of cloud responses to a parameter are highly dependent on the values of other parameters. For example, the sensitivity of cloud albedo to increases in NINP varies between near zero to strongly negative across the sampled parameter space. Therefore, perturbing parameters individually is an inadequate method for determining the cloud responses to model parameters. This work demonstrates the power of model emulation and the importance of a full exploration of parameter space in order to understand the factors controlling cloud properties.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics.
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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(17673 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 06 Mar 2026)
- RC1: 'Comment on egusphere-2026-311', Anonymous Referee #1, 21 Feb 2026 reply
-
RC2: 'Comment on egusphere-2026-311', Anonymous Referee #2, 24 Feb 2026
reply
This paper develops a cloud emulator based on perturbed parameter ensembles of a regional model to investigate how six cloud microphysical parameters affect mixed phase CAO clouds over the Labrador Sea. The results show that the sensitivity of cloud properties to any single microphysical parameter are highly dependent on, and interconnected with, the values of the other parameters. Overall, the manuscript is generally well written and clearly structured. However, I have several major concerns regarding the construction of the emulator and the selection of the sensitivity parameters, particularly the omission of CCN-related processes. I therefore believe that major revisions are required to clarify several key methodological and physical aspects prior to publication.
Major comments:
One of my main concerns is the author’s selection of cloud droplet number concentration (Nd) and ice-nucleating particles (INPs) as the perturbed parameters. The selection of Nd instead of CCN, while simultaneously selecting INP concentrations, is physically difficult to justify. INPs are a subset of aerosols with strong ice-nucleating ability and are generally associated with way larger uncertainties than CCN. The authors state that aerosol-derived Nd is not used to avoid additional uncertainties, but this reasoning appears inconsistent given that the study includes the much more uncertain INP parameter. If INP variability is considered acceptable within the emulator framework, then CCN should be also included rather than fixed-Nd profile.  The omission of CCN effects through a prescribed Nd should be more carefully justified. I am particularly concerned because, in real CAO environments, Nd is strongly controlled by aerosol aging processes and activation to CCNs. Also, CCN will have competition of water vapors with INPs in mixed-phase cloud, hence affects INPs response as well. Treating Nd as an independent tunable parameter, rather than linking it to aerosol activation, may bias the INP response and the interpretation of the relative importance of microphysical sensitivities.
My second concern comes from the sparsity of the sampling in the 6-D parameter space. The PPE uses 66 ensemble members, which is reasonable, but 66 samples in a six-dimensional nonlinear parameter space may still be sparse, especially given the strong parameter interactions and equifinality demonstrated in the manuscript. This raises questions about the robustness of the emulator, particularly near the boundaries of the parameter space where extrapolation risks may be larger. I therefore suggest that the authors conduct a convergence test to evaluate the stability of the emulator results. In addition, a more explicit discussion of emulator uncertainty quantification and potential extrapolation errors in the manuscript would be helpful.
Finally, while the authors acknowledge case specificity, many conclusions are framed in broader terms regarding mixed-phase CAO clouds and aerosol–cloud interactions. Given the strong dependence of mixed-phase cloud processes on thermodynamic structure, aerosol regime, and synoptic conditions, I suggest that the authors more explicitly frame their conclusions as case-specific and discuss how different aerosol forcing and environmental conditions (such as continental versus marine environments), could alter the inferred parameter sensitivities.
Minor comments:
Page4 L109: in the manuscript, the author only considered dust as INPs. However, rmore and more studies have reported that sea-salt aerosols can also act as effective INPs, particularly in marine environments (e.g. Wagner et al., 2021; JV Trueblood et al., 2021). Given that this case is over the Labrador Sea, it would be helpful to discuss the potential contribution of marine (sea-salt) INPs and how excluding them might affect the inferred INP sensitivity. If feasible, I encourage the authors to include an additional sensitivity test (or at least an order-of-magnitude estimate) representing sea-salt INPs, or to justify why dust-only INPs are appropriate for this case.
Page5 Figure1: change the color blue for C323. The current lines are nearly invisible on printed paper
Page6 L134: please explicitly write the global mode name you used for dust concentration simulation
Page7 L165: please define the variable Niics,xy when first introduced
Page25 L483-484: It seems to me that you did not demonstrate where the dust source originates in your nested region. Therefore, I am not fully convinced that "purposefully omitting measurements from desert soil sources" is an appropriate choice. I would also suggest presenting the desert dust ns–T relationship in Figure A1 to show how those dust properties would differ from those examined here.
Page 27 L515: Page 27, L515: Why are only these three levels (surface, 1 km, and 2 km) selected? It would be helpful to clarify the typical upper bound of the cloud top height, how many model levels are located beneath it, and whether these selected levels adequately capture the vertical structure of the INP concentration variation.
Citation: https://doi.org/10.5194/egusphere-2026-311-RC2
Data sets
Model data from perturbed parameter ensemble simulations used in manuscript submitted to ACP "Interacting effects of droplet number and ice formation processes on mixed-phase cold-air outbreak clouds" Xinyi Huang https://doi.org/10.5281/zenodo.18267702
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 210 | 153 | 16 | 379 | 14 | 18 |
- HTML: 210
- PDF: 153
- XML: 16
- Total: 379
- BibTeX: 14
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript examines how mixed-phase cloud microphysical parameters influence macrophysical cloud properties during a marine cold-air outbreak over the Labrador Sea. Using two nested domains to represent upwind stratiform and downwind cumuliform cloud regimes, the authors conduct an extensive ensemble of simulations that systematically co-vary six microphysical input parameters. They employ Gaussian Process emulation to quantify relationships between input parameters and cloud-scale responses. The analysis identifies regime-dependent dominant parameters and demonstrates strong nonlinear interactions among them, underscoring the limitations of single-parameter sensitivity studies. The manuscript is clearly written, methodologically rigorous, and the analysis is comprehensive. I recommend publication following minor revisions.
Â
Minor concerns
l. 21 – I wonder if there are any other studies that could be cited here?
ll. 54-59 Please also include Elsaesser et al. (2025) here.
Sec. 2.1 I wonder if the author could list cloud-top temperature (CTT) here (e.g., from a typical configuration). Assuming that CTT varies across configurations (e.g., I would expect stronger ice processes to produce lower/warmer cloud-tops, thereby preventing themselves) and if not too much work, perhaps CTT could even be added to Fig. 6?
Sec. 2.2.2 If aerosol was considered here, it would affect both N_d and N_inp; it is unclear if the chosen dynamic ranges for N_d and N_inp correspond well? I wonder if the authors could briefly quantify and explain.
Fig. 6 I’m surprised by the large cloud cover in the Cu regime. How do the authors define cloud cover?
Fig. 7/8 Is the small sensitivity of mprof in Sc-dominated regimes connected to the large cloud fraction (as well as small IWP) ?
Fig. 9 How should the reader interpret a greater TWP in the Cu-dominated regime under increased E_HM? While a LWP reduction make sense to me, I’m surprised to see an even stronger IWP increase.
Â
Typos/Language
l. 342 “shown … show “
Â
References
Elsaesser, G., M. van Lier-Walqui, Q. Yang, M. Kelley, A.S. Ackerman, A. Fridlind, G. Cesana, G.A. Schmidt, J. Wu, A. Behrangi, S.J. Camargo, B. De, K. Inoue, N. Leitmann-Niimi, and J.D.O. Strong, 2025: Using machine learning to generate a GISS ModelE calibrated physics ensemble (CPE). J. Adv. Model. Earth Syst., 17, no. 4, e2024MS004713, doi:10.1029/2024MS004713.