EGUsphere Paper Review

Airborne observations of cloud properties during their evolution from organized streets to isotropic cloud structures along an Arctic cold air outbreak

This paper investigates micro- and macrophysical properties and transition of marine boundary layer cloud in an Arctic cold-air outbreak over the Norwegian Sea. It also investigates the role of surface heat fluxes, buoyancy forces, and vertical wind shear.  

Major comments

1. The Introduction does not develop the raison d’être of this study. It provides a textbook summary of cloud streets, then summarizes two papers (one from 1996 with airborne observations and one from 2023 with satellite observations). Then the reader is left with the question why, where does this lead? Nowhere. What follows is the standard paper section summary, then Section 2. What is the objective? Literature should only be cited as part of a rationale for the study.
2. Section 5 explores the dynamic reason for the observed change in cloud depth and cloud macro/microphysical changes during repeat passes over the same location. The changes over the course of a few hours are relatively minor, and a variety of synoptic to mesoscale factors may have contributed to this, e.g. changes in surface wind speed or surface heat fluxes, or large scale subsidence. As an example, Fig. 1 reveals a remarkable mesoscale convergence band emerging from the apex of ice-free ocean north of Svalbard. Clearly this feature is east of the area of interest (although it may well have been included in passes #2 and 4, see Fig. 4b), but other leads (polynyas) could generate mesoscale circulation stretching far downwind. Alternatively, a slight change in flow trajectory and thus fetch from the ice edge or MIZ could bring about the observed change. I recommend examining multiple satellite images, as well as Arome Arctic model output, including Lagrangian trajectories.
3. If I understand correctly, the estimation of -H/L in Section 5.1 only depends on the wind speed. This is an oversimplification. The wind speed supposedly is measured at a level of 90 m, and dropsonde data are used for this. Are the 4 times mentioned in Table 1 the 4 dropsonde times, associated with the passes shown in Fig. 5a? The bulk aerodynamic approach has its limitations (e.g., assumptions about the exchange coefficients), and normally is applied to 10 m wind and 2 m T/q, not 90 m wind (Brümmer’s soundings may have lacked vertical resolution?). One cannot assume that the 90 m temperature is constant, in fact Fig. 7e shows a variation of ~1K, and H varies too (as mention in the Abstract). The eddy correlation fluxes approach (eqn 5) is much preferred, even at a FL above 90 m within the convective BL, and it appears that the data are available. This approach requires averaging over a substantial distance, and this integral approach results in a more representative outcome than a point measurement. Spaceborne Synthetic Aperture Radar (SAR) surface wind speed imagery shows significant small-scale variations in wind speed in the cloud street regime, presumably associated with coherent roll circulations, in other words, the wind speed variations are not representative of a larger area and do not demonstrate a trend. I do not place much credence in the -H/L values listed in Table 1, and I would not classify regimes based on the apparent small variations across the -H/L threshold of 15. I cannot agree with the conclusion arising from Sections 5.1 and 5.2. CARRA does show a weakening of the BL wind over a 9 hr period. In short, the paper does not prove its basic conclusion stated in Section 6: “For −H/L < 15, turbulence generated by wind shear dominates, favoring the formation of organized cloud streets. However, as −H/L increases beyond this threshold, buoyancy becomes increasingly significant, and the organized cloud streets collapse into more isotropic cloud patterns. In our case study, the reduction in wind speed is the dominant factor increasing −H/L, disrupting the organized convection which is required to sustain cloud streets.”
4. Finally, the writing style can be improved. The text is very figure-centric. Rather, the figures should merely serve to confirm statements in the text, e.g., “A similar trend, but with higher uncertainty, is also visible in Fig. 6b, which shows the correlation between A and B” is better stated as: Consistent with this, A and B are negatively correlated, although the correlation is weaker (Fig. 6b). There are many other examples.
5. In general, caution is warranted with the interpretation. This paper is based on a remarkably small dataset, and slim evidence.

Minor comments

L26: “The inflection point in the vertical cross-roll wind profile turned out to be too weak for that kind of dynamic instability”. The inflection point cannot be too weak. Is it that the cross-roll wind shear is too weak?

L36: “… had in common that a moderate CAO was simulated …” Do you mean that helical roll circulations were simulated? The CAO simply is a large-scale condition of cold air advection and strong surface heat fluxes.

L38: “the more TKE production by wind shear and the less by buoyancy plays a role in the entire ABL” better: … the more TKE production in the ABL is dominated by shear rather than by buoyancy.

L38: define the Monin-Obukhov length here, rather than on line 252. Or at least mention the definition in words here.

Fig. 3: the cloud fraction identification through separation across the red/blue channel ratio is not clear to me. What is the physical basis? An IR camera probably would have been better. The obvious limitation evident in Fig. 3a,b, that a slant view overestimates the nadir view albedo for any cloud of finite thickness, should be mentioned.

L221: pls provide more detail about how M is computed, limitations, and why M=0.01 can be used as threshold for (un)rimed particles. I suspect it uses in situ microphysics data only. Maherndl et al. (2024) describe two techniques.

L226: lack 4  track 4

On L228, it is stated that “Cloud streets show a stronger shear at cloud top with higher turbulence (higher TKE)” and “isotropic cloud patterns show a stronger buoyancy”. This is not demonstrated yet in the paper.

Fig. 5a: does it show the flight level of Polar 6 only? I see only 4 tracks, one for each pass. On L88, Polar 5 flight level is mentioned to be 1000 m above cloud top. How well synchronized were the two aircraft? Apparently not well in 2 of the 4 tracks.

L295: “because other parameters which might control the appearance of free rolls here, keeps constant”  because other parameters appear to vary less than wind speed?

Fig. 7: I suggest monotonically changing hues (color intensities) for the 4 different times. Easier to interpret

Fig. 8: “CARRA wind field on 4 April 2022 …” I suggest changing this caption : “… at 430 m ASL, which corresponds to cloud top height near point DS”

This study uses airborne observations to investigate the temporal evolution of cloud street structures and how changes in these structures affect both macrophysical and microphysical cloud properties, as well as the radiation budget. The authors introduce a new method called the "cloud street index" to describe the transition from organized cloud streets to a more isotropic cloud pattern.

Major comments:

1. Readers can easily become confused by the various flight paths and dropsondes from different aircraft. The terms “specific location”, “the same location”, “the before mentioned location” are repeated throughout this manuscript, leading to further confusion. It would be helpful to replace these phrases with “location DS” and provide latitude and longitude coordinates for different points. Additionally, consider modifying Figures 1, 5(a), and 7 to clearly illustrate which flight paths and dropsondes correspond to each aircraft.
2. One of the major findings of this study is that “decreasing wind speed drives the transition from cloud streets to isotropic cloud patterns.” This conclusion is primarily based on the vertical profiles from five dropsondes and the CARRA wind field at four different times. However, the sample size used to support this conclusion is quite small and insufficient to definitively claim that decreasing wind speed is the cause of this transition. While we did observe decreased wind speeds in isotropic clouds, this does not necessarily indicate that the decrease in wind speed is responsible for driving this transition.

Minor comments

line 29-30: It is unclear whether “this study” refers to this manuscript or Brown's 1972 study.

line 47-48: Since both the strength of the CAO and the surface heat flux play important roles in cloud structure and its transition, why doesn't this study explain the exclusion of the analysis of CAO strength and surface heat flux?

line 121: What does this mean “quantified by the cloud top altitude”. Are cloud top heights in Figure 4b retrieved from AMALi?

line 171-172: Why can the ratio of the red and blue channels be used to determine the cloud fraction? How is the 0.8 threshold determined?

line 182: It appears that the cloud fraction significantly depends on the size of the camera images. If two camera images overlap, could this impact the cloud fraction results?

line 212-213: Are passes 2 and 4 from Polar 6 or Polar 5?

Line 221-222: What airborne measurements are used in the calculation of M?

line 226: “…lack 4…”.  Did you mean pass 4 here?

line 228-229: More evidence is needed to support this hypothesis.

line 238-240: Based on Figure 1, there is a thick cloud band on the right side of the cloud streets. Are these clouds included in the analysis as well? Are they treated as isotropic cloud regimes?

line 266-267: How do authors calculate the SST using the values of H, theta_0, and delta_theta_as?

line 294-295: It is unclear how the authors reached this conclusion. Is this hypothesis based solely on Table 1? The sample size is too small to support this conclusion.

line 308: How do authors determine the strength of inversion?

Figure 1: The manuscript does not justify the selection of WP1 and WP2. Has any analysis been conducted using dropsonde measurements from these locations?  

Figure 3c: If Pass 1 was the earliest flight and Pass 4 occurred 120 minutes later, then Pass 4 is expected to have a smaller cloud street index compared to Pass 1 at the same location , as discussed later. However, this figure shows that Pass 1 and Pass 4 have similar cloud street index values when the distance to the sea ice edge is between 25 and 35 km, which is confusing. Additionally, it would be helpful to add a vertical line to Figure 3c to indicate the DS location.

Table 1: Why were the dropsonde results from HALO excluded from this table?