Zheng et al. “Environmental Effects on Aerosol-Cloud Interaction in non-precipitating MBL Clouds over the Eastern North Atlantic”

Using a total of 20 non-precipitating single-layer marine boundary layer (MBL) stratus and stratocumulus cloud cases over the eastern north Atlantic (ENA) ocean, this study investigates the impacts of the environmental variables on the aerosol-cloud interaction (ACIr). Interesting results have been found with valuable discussions. For example, it shows that the ACIr values vary from -0.004 to 0.207 with increasing precipitable water vapor (PWV) conditions, indicating that re is more sensitive to the CCN loading under sufficient water vapor supply, owing to the combined effect of enhanced condensational growth and coalescence processes associated with higher cloud droplets and PWV. The paper is also well written. I would recommend its acceptance for publication after necessary minor revisions.

Detailed comments;

Line 41-44, two “verbs” exist for this sentence, which should be rephrased. Also, a few more studies are recommended here, particularly the longwave radiative property change of clouds by aerosols, such as Garrett and Zhao (2006, Doi:10.1038/nature04636).

Line 48-52, a few similar studies have also been carried out over the western pacific regions, which might be worthy to mention, such as Zhao et al. (2019, Doi:10.3390/atmos10010019), and Yang et al. (2019 , Doi:10.1016/j.atmosres.2019.01.027).

Line 66-69, Qiu et al. (2017, Doi:10.1016/j.atmosenv.2017.06.002) showed negative relationship between cloud re and aerosol amount for low precipitable water vapor condition in spring, fall and winter at southern great plain site, but positive relationship between cloud re and aerosol amount for high precipitable water vapor condition, which could be also cited here. Similar findings have also been found over other locations, such as western pacific region near Hebei province, China.

Line 281-283, similar height normalization method has been proposed and used by Zhao et al. (2018, Doi:10.1002/2017EA000346), which is worthy to mention here. Also, Similar findings (Line 283-287) have been found earlier in several studies, including the study mentioned here.

Line 319, Eq. (2). Earlier studies often define this for fixed LWC. How could the different definition affect the results?

Line 332-343, These are interesting findings and explainations. I wonder if this is related to the supersaturation adoped for CCN observed, or related to the true supersaturation status within clouds.

Line 358-376, The mechanism proposed here is valuable. If possible, I would suggest the authors illustrate the mechanism proposed here with a diagram.

Line 390, “that more close to adiabatic” shuold be “that are more close to adiabatic”

Line 432, “to narrows the DSD” should be “to narrow the DSD”

Review of Zheng et al., “Environmental Effects on Aerosol-Cloud-Interaction in non-precipitating MBL Clouds over the Eastern North Atlantic,” acp-2021-391

As the title suggests, this paper describes an aggregated analysis of aerosol-cloud interaction (ACI) in non-precipitating marine boundary layer clouds at the Eastern North Atlantic ARM remote sensing supersite. A relatively narrow view of ACI is taken in which the bivariate relationship between aerosol and cloud drop number concentration and the ACI index were calculated numerous times, compositing by various column-mean or column-integral quantities (e.g., water vapor path, cloud adiabaticity, lower tropospheric, turbulence). My main concern with the study is that each of these purported controlling factors is analyzed in isolation, which implicitly assumes no covariability among them. This assumption is not valid and no attempt to address this issue was given. As such, I find it difficult to accept many of the mechanistic arguments made by the authors. They cannot demonstrate cause and effect, and there are clearly confounding variables that limit their ability to draw stronger conclusions (for example, lines 243-244: “the coincidence of high NCCN and PWV does not necessarily imply a physical relationship”). I therefore recommend the manuscript be rejected and the authors encouraged to resubmit after broadening their analysis. The premise of evaluating ACI with the authors’ retrieval product is promising, but to understand the role of the controlling factors, they must be analyzed in a multi-dimensional framework (principal component analysis, k-means clustering, etc.) that allows the authors to identify and, more importantly, interpret co-variability among environmental factors. As it currently stands, the conclusions of this study point vaguely toward correlations with large-scale variables but give no clear guidance.

I have a number of other concerns the authors may also wish to consider:

• How good of a proxy is PWV for PBL relative humidity? Are there cases when non-drizzling stratocumulus occur with a relatively moister free troposphere? Perhaps you could estimate the fraction of PWV in the PBL using the interpolated sonde product or Raman lidar (note: Raman will only get you subcloud vapor)?
• Not enough information is given about how the vertical velocity variance TKEw is calculated. Is it a PBL average? A Doppler lidar column-deep average? Column max. value? And what Doppler lidar product are you using to get variance? The standard 10-minute integration? The median value seems low for surface-coupled stratocumulus cases. Are you evaluating any decoupled cases? There is also a diurnal and season cycle of turbulence at this site (at least, when sampling an undisturbed marine airmass; see more below), which may also be affecting your statistics.
• Have you controlled for wind direction in your analysis? It has been shown that there is an island effect when the surface wind is from the island (e.g., Zheng, Rosenfeld and Li 2016). Overland flow affects boundary layer turbulence and may also impact surface fluxes, PBL depth and CCN composition.
• How much does LTS tell us at a site like ENA, and what physical motivation do you have for including it as a sorting variable? I always envision LTS as having the most meaning in the subtropical eastern boundary current (EBC) areas, i.e., northeast/southeast Pacific and southeast Atlantic. The Azores are more of a mixed subtropical/midlatitude site that has much warmer SST than in the traditional EBC areas where MBL clouds are studied, and much of the cloud cover at ENA occurs in transient postfrontal subsidence vs. longer-lasting large-scale subsidence where the spatial gradient (of both subsidence and SST) matters more in defining cloud type transitions.
• For arguments you make about the relationship between entrainment, collision-coalescence and number concentration, it is problematic that your retrieval assumes constant Nc throughout the cloud layer. When entrainment-induced evaporation and/or collision-coalescence are active, this assumption is broken. In general, I don’t understand your argument that entrainment is a sink of Nc.
• High CCN events at ENA are not only from North America. They have also been traced to North Africa and Europe.

The authors analyze the impact of the environment on the aerosol-cloud interactions (ACI) from ground observations over the eastern north Atlantic. They find that both lower-trosospheric stability and turbulent kinetic energy influence the connection between water vapor, cloud-microphysics, and subsiquently ACI. For instance, they find that higher lower-tropospheric stability leads to higher cloud drop concentrations and ACI.

Overall, I think this paper is both well thoughout and written. However, I do have a number of issues that I would appreciate clarification on. Note that, even though I split my comments between major and minor, this is more of just a distinction between general and technical comments. Therefore, I recommend publication once these comments are addressed.

Major:

Line 147: Is LTS the most appropriate variable to use over the northeast Atlantic, considering the much larger influence of midlatitude cyclones compared to subtropical regions?

Line 171: How many potential non-precipitating cloud cases were there, and do your results suggest that most MBL clouds produce precip over the northeast Atlantic?

Line 193: You could highlight that the median LTS of 19.1 K is close to the value (18.55 K) used by prior studies to separate stratocumulus from shallow cumulus.

Line 226: You compare the logarithmic ratio that you find to other studies, but I don't understand what it actually means.

Figures 5 - 7: There doesn't appear to be much of a trend in the scatter plots, so what is the R² value for these regressions? Maybe this could be fixed by constraining your axes to closer to the limits of your datapoints?

Minor:

Line 77: "relatively shallower" should be "relatively shallow"

Line 78: I think "and is prone to" should be and "are prone to"

Line 80: "marine boundary layer maintained by" should be "marine boundary layer which is maintained by"

line 85: "regime of active coalescence process" should either be "regime of the active coalescence process" or "regime of active coalescence"

line 106: "particularly disentangling" should be "particularly by disentangling"

line 121: "operates at 910 nm laser beam" doesn't make sense, and maybe could be "operates at 910nm"

line 159: "from Doppler lidar" should either be "from a Dopplar lidar" or "from the Dopplar lidar"

line 183: "lay" should be "lie"

line 388: Unless I missed something why is Figure 5b discussed before Figure 5a, could you just flip those subpanels?

Figure 1: This may just be my printout, but the median dashed lines are difficult to see. Could you use a thicker line or a different color?