Finney et al. use observation-informed cloud-resolving modeling to investigate how both large-scale environmental conditions and microphysical properties influence anvil cloud albedo. Their cloud-resolving model simulations, constrained by observations from the DCMEX campaign of orographically forced convection in New Mexico, USA, reproduce the observed cloud and radiative properties in a reasonable way. The study reveals a substantial sensitivity of anvil cloud albedo to cloud droplet number concentration and, to a lesser degree, to ice-nucleating particle concentration.

The manuscript is clearly written, logically structured; it was easy to follow the line of thought and understand the key outcomes. However, I have several questions and suggestions that I believe should be addressed prior to publication. In particular, I encourage the authors to further explore the mechanisms underlying the reported changes in cloud properties.



General comments

1.) Mechanisms and physical interpretation

While the impacts of CCN and INP perturbations on anvil clouds are clearly described, the manuscript would benefit from further elaboration on the underlying mechanism leading to these changes. How exactly do changes in CCN or INP influence cloud albedo? How do the CCN propagate to changes in ice phase clouds? Are the effects primarily driven by direct microphysical modifications (e.g., changes in ice crystal size or number), or are they mediated indirectly through changes in updraft dynamics? How do specific microphysical process rates respond to the perturbations? Or, if we take a step back: What are the dominant processes that determine ice crystal number and mass in these simulations/this types of high clouds? How do they compare to those in more frequent type of anvil clouds, e.g. tropical anvils?



2.) Robustness of results

Are the key results robust? How sensitive are the findings to stochastic variability? What would happen if one were to run e.g. an ensemble of 5-10 simulations with perturbed CCN and INP conditions?

3.) Connection to observations

The manuscript would benefit from a stronger connection between the modeling results and observed high cloud albedo changes. Is there satellite evidence of similar anvil albedo changes over the Magdalena mountains under comparable dynamical but different aerosol conditions? Although the DCMEX campaign may not cover a long enough period to address this conclusively, long-term satellite records might offer additional context.

4.) Broader relevance

Although it may go beyond the scope of this study, the potential for generalizing these results is worth considering. Could e.g. long-term satellite retrievals combined with reanalysis data help assess the broader applicability of the findings? Additionally, is there an analogy between orographically driven convection and island-driven convection in the tropical Warm Pool?

5.) Selection of meteorological predictors

The choice of meteorological variables and cloud-controlling factors used in the analysis is not entirely clear to me. Why did e.g. the authors exclude some of the cloud controlling factors that are thought to be useful in explainig high clouds at climatological timescales, e.g. the mid-tropospheric updraft, upper tropospheric stability?

6.) Longwave cloud radiative effect and related quantities (e.g. cloud top temperature)

Although the study focuses on shortwave albedo effects, additional discussion of longwave  fluxes and related quantities such as cloud top temperature would be useful in bringin a more holistic view on anvil changes. For example Fig. 8 suggests possible changes in cloud top temperature. Moreover, given that cloud LW emissivity saturates at relatively low cloud optical depths (~2-3), LW fluxes are respond primarily to changes in thin anvils. Can the authors provide more insight on this aspect?

Specific comments

1.) 70 vertical layers are rather few for correctly representing thin anvils responses to any kind of forcing. Would the results hold with higher vertical resolution in the upper troposphere? Testing or at least discussing this would add credibility to the conclusions.

2.) The mechanisms by which CCN and INP perturbations affect anvil albedo appear relatively straightforward. Would similar sensitivities be found using a simpler, single-moment microphysics scheme? This question is especially relevant given that many global storm-resolving models use simplified microphysics. If such interactions are as robust as they appear, this would suggest that even models with basic microphysical representations might capture the essential response to aerosol perturbations. A comment on this would be helpful.

Section 4: How are cases with multiple cloud layers handled in the analysis? For instance, what if two high cloud layers are present? Does this occur frequently, and if so, how is it treated in the retrievals or model evaluation?

Data availability: I think the links don't have the model data uploaded, if I understand the website contents correctly.

Best regards,

Blaž Gasparini

Review of manuscript: “Microphysical fingerprints in anvil cloud albedo” by Finney et al.

General comment:

This manuscript addresses environmental and microphysical effects on SW and LW cloud radiative forcing via changes to cloud thickness and area via a modeling study of topographically forced convection over the southwest United States. Simulations with the UM model appear to reasonably represent the modes of convection and associated cloud cover. Sensitivity experiments suggest that variations in droplet number and INP have the greatest impact amongst their sensitivity tests. The paper is well written, and the explanations are easy to follow. There are, however, a number of comments below that should be addressed related to the model capabilities and the assessment of microphysical responses to changes in droplet number and INPs.

Specific comments:

1.Lines 99 and 105: How can you use a 75 second timestep with domain grid cell spacing of 1.5km and not encounter CFL errors, especially in the vertical? Also, given the rapid changes that can occur in clouds and their impact on radiation, a 15-minute time-step for radiation updates seems very long.

2.Lines 125-140: Are the droplet numbers constant over time? Or do the droplets undergo autoconversion, accretion, riming, homogeneous freezing, etc? This is rather critical since vertical transport of droplets to the anvil level and subsequent homogeneous freezing to generate high concentrations of small anvil ice can have a substantial impact on cloud top albedo.

3.Lines 219-220: Did you test cloud mass mixing ratio thresholds other than the one stated here? This threshold is rather low and may not constitute a visually apparent cloud. Does the satellite imagery use a similar sort of threshold for determining cloud presence?

4.Lines 285-286: While the focus may be on how SW CRE varies, this SW bias is often over 20% (from fig 4a). Some discussion should be included regarding how this bias could impact convective formation and diagnosis of SW albedo. Given that this study focuses on radiative effects, this bias is significant.

5.Line 291: The verbiage here is mixing the meanings of “lower” and “high”. Please refer to altitude using low and high. I assume that “lower” means less in this context. So perhaps use “less” and “more” or “increased” and “decreased” to refer to change in magnitude.

6.Lines 403-405: Why is the decrease in droplet number more impactful than the increase in droplet number toward change in cloud albedo?

7.Lines 453-455: The fact that the INP experiments lead to more ice hydrometeor mass than the increase in droplet experiment, combined with only a small increase in ice number for the INP experiments compared to the cloud droplet increase experiment, seems counter intuitive to me. Homogeneous freezing of droplets often dominates in adding the most mass and number to convective anvils. The results suggest that the ice hydrometeors are significantly larger in the INP experiments compared to the increased droplet experiment. Why would this be the case? Is there a default ice crystal size (from INP heterogeneous nucleation) in the microphysics scheme that could be influencing this outcome?