| 14 Nov 2025
Exploring the processes of liquid water path sensitivity to aerosol-cloud interactions using output from a high-resolution large-eddy simulation
Abstract. Diagnostics from high-resolution Large-Eddy Simulations (LES) are used to investigate aerosol impacts on the liquid water path (LWP) sensitivity in a non-precipitating, single-layer liquid cloud regime. In two LES simulations, the 2013 conditions represent a low aerosol scenario, while the 1985 conditions represent a high aerosol scenario. Joint histograms of cloud droplet number concentration (Nd) and LWP reveal a non-linear relationship, with positive LWP sensitivity (increasing LWP with Nd) at low Nd and negative sensitivity at high Nd. The transition from positive to negative LWP sensitivity occurs at higher Nd values in the 1985 simulation (≈ 300 cm−3) compared to the 2013 simulation (≈ 100 cm−3), indicating that enhanced aerosol loading shifts the transition point. This shift reflects stronger droplet activation and sustained LWP growth under high CCN conditions. Diagnostics of the cloud dilution ratio indicate that negative LWP sensitivity is linked to enhanced cloud-top entrainment. The temporal evolution of the Nd–LWP relationship confirms increasing dominance of negative sensitivity in the 2013 case, while the 1985 case exhibits weaker LWP depletion. Additionally, aerosol perturbations also influence thermodynamic properties such as the apparent heating/cooling (Q1) and the moisture sink (Q2). Specifically, stronger cloud-top heating and moisture sinks are simulated during negative LWP sensitivity phases, particularly for high Nd in 2013, consistent with enhanced evaporation and entrainment. Aerosol perturbations thus modulate both microphysical and thermodynamic processes, producing distinct LWP sensitivity regimes with important implications for understanding aerosol–cloud–climate interactions.
Competing interests: I declare that neither I nor my co-authors have any competing interests.
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.
In this paper, diagnostics from high-resolution Large-Eddy Simulations (LES) are presented and analyzed to investigate aerosol impacts on LWP sensitivity in a non-precipitating, single-layer liquid cloud regime. The analysis focuses on the non-linear relationship between cloud droplet number concentration and LWP under contrasting aerosol conditions, highlighting distinct regimes of positive and negative LWP sensitivity. The results demonstrate how aerosol loading modulates droplet activation, cloud-top entrainment, and associated thermodynamic responses, thereby shifting the transition between LWP sensitivity regimes. These processes are shown to be highly relevant for understanding aerosol–cloud interactions and for improving the representation of cloud microphysical–dynamical coupling in climate models.
The topic of the manuscript is highly relevant, and the analysis is based on a unique modeling data set. However, I cannot recommend the manuscript for acceptance in its current form, as several aspects of the analysis require clarification. In addition, numerous minor errors and inconsistencies are present throughout the manuscript. I therefore strongly recommend careful proofreading and thorough checking of equations, units, and figures. Some of these issues are outlined below, but the list is by no means exhaustive.
I found it highly interesting that the transition from positive to negative LWP sensitivity occurs at different droplet number concentrations under different aerosol loadings. How does this behavior align with the commonly assumed entrainment feedback? This question is particularly relevant given that the transition occurs at Nd ≈ 300 cm⁻³ in the high-aerosol scenario. At such high concentrations, the mean droplet size is far from having a sedimentation velocity that would be dynamically significant and system shouldn’t be sensitive to changes in Nd.
More detail on the modeling approach is needed to better understand how methodological choices may influence the analysis. Specifically, how is Nd diagnosed from CCN? Is saturation adjustment applied in the cloud microphysics scheme? How is Nd treated during mixing and entrainment processes? What is the vertical resolution near the cloud top? Do updraft velocities differ between the simulations as a result of aerosol perturbations?
Line 35–36: “The numerous small droplets also lead to droplet sedimentation.”
Isn’t the opposite expected, with smaller droplets having reduced sedimentation velocities?
Line 36: “Enhance radiative cooling at the cloud top.”
This statement could be more precise, as radiative cooling occurs within a relatively thin layer near the cloud top.
Line 37: “Further, the entrainment of warm, dry air into the cloud leads to evaporation of the smaller droplets, resulting in decreases in LWP or negative LWP adjustment.”
Entrainment leads to LWP reduction even if all droplets lose mass homogeneously, not only through preferential evaporation of smaller droplets.
Line 49: “The Nd–LWP relationship is non-linear, and the co-variability between LWP and Nd primarily drives it.” While this is true, is it the co-variability between aerosol and meteorological conditions that fundamentally drives both LWP and Nd?
Line 53: Please clarify that the modeling referred to here concerns global, low-resolution models.
Line 95: Could the averaging of data to a coarser grid resolution affect the results? The original LES output likely resolves cloud-scale structures, but aggregation to 1.2 km places the analysis in the “gray zone” for cloud dynamical processes. Was the analysis repeated over smaller spatial scales to confirm that averaging does not affect?
Line 99–100: Why is this criterion used instead of one based on liquid water content? Such a layer may have very low LWC and therefore may not represent the actual cloud top, but rather a layer above it.
Line 102–104: This assumption is unlikely to hold over a one-day simulation. However, I did not find an analysis in the paper that could be affected.
Equation (2): Is the last term correct, given that vertical pressure velocity is used?
Equation (3): This equation appears to describe the mean volume diameter. If so, the subsequent calculation of the cloud dilution factor does not seem to be meaningful. Also the units appear incorrect.
Line 122: “Binned with a bin size of 1000.”
I assume this refers to the number of bins rather than the bin width; please clarify.
Line 143: The results are qualitatively consistent but far from being quantitatively consistent.
Line 148: “Which sustained droplet activation.”
The term sustained does not seem appropriate in this context.
Line 164: “Initially, both simulations exhibit positive LWP sensitivity.”
What is the reason for this behavior? Is it related to assumptions about the initial conditions? Are the initial thermodynamic profiles identical, differing only in aerosol loading?
Line 175: How are the different critical values determined? Are they derived across different spatial domains or over different time intervals?
Line 188: Related to the earlier comment on the definition of cloud top, how representative is Nd at cloud top of Nd throughout the cloud volume?
Line 192: This statement holds only if saturation adjustment is not used and condensation is explicitly calculated.
Line 195 onward: The effects of radiative cooling and heating are not discussed, despite being speculated as key mechanisms driving the negative Nd–LWP relationship. Why are these processes not analyzed in more detail?
Line 198: When discussing high-resolution simulations and processes occurring at the cloud edge, is advection the appropriate term, or would mixing be more accurate? More generally, modeling and also the analysis details may significantly affect the results, as numerical diffusion is unavoidable when Lagrangian processes are simulated in Eulerian framework. This may influence the diagnosed apparent heating and moisture sink terms.
Figure 1: Mean values could be added to both panels.
Figure 2: “the 1985 simulation uses the pre-industrial (1985) CCN”. This is a new definition for pre-industrial era.
Figure 6: left axis is cut of from the figure
Figure 7: ylabel “aparent”
Figure A1: Check the units.