Review of “In-cloud characteristics observed in US Northeast and Midwest non-orographic winter storms with implications for ice particle mass growth and residence time”, by Allen and coauthors, egusphere-2024-3808

The authors address an important issue of ice particle generation and growth in cloud top generating cells found in winter storms sampled using observations from the IMPACTS and PLOWS field campaigns. The authors show that narrow updrafts of >0.5 m s^-1 associated with such generating cells are primarily responsible for ice particle generation, growth, and particle lofting within the cloud top layer while ice growth is likely constant or decreasing beneath in most cases. The authors also do a great job explaining their findings in the context of previous work on these winter storms and how previous understandings were either incorrect or limited. The overall manuscript is very well written with clear findings and well produced figures to support and detail the main findings which make a significant contribution to the scientific knowledge for ice particle formation and growth in such storms. I have only some minor comments for the authors to address:

Line 37: I would add that gravity waves can also yield vertical motions.

Lines 42 – 54: Maybe start with explaining residence time first in this paragraph (after the first sentence) then explain the changes or “microphysical pathway” the hydrometeor undergoes. I think you can fold the paragraph from lines 51-54 into the paragraph above. That way you can remove the “timescale of snow falling to the surface” and just use residence time. Just a suggestion not a must fix.

Line 173: Reword end of sentence. “within these types winter storms available” doesn’t make sense.

Line 174: Just say “warm and/or occluded fronts” and remove “air mass boundaries”.

Line 202: The reference to Fig. 5 makes it appear as if the figures are out of order. It would be good to add that this figure will be discussed later on in the text.

Lines 374 – 379 summarizes the section very well

The discussion section is very well done

Lines 450 – 456: Not a must fix but I’m not sure this information is really needed for this paper.

Lines 463 – 472: Good description of future work possibilities for the IMPACTS data.

Figure 11 caption: should be -22 C not 22

Overarching comment:

This paper presents analysis of aircraft vertical motions across two field campaigns, putting these observations in the context of the thermodynamic, kinematic, and structural features of the cyclone. The paper is well-written, has clear figures, and clearly lays out its findings. 

The definition of an updraft is fairly fundamental to the analysis of the paper, and something I find a bit confusing. The choice of 0.5 m s-1 is certainly defensible, but could use more explanation. What exactly is the uncertainty of the instrument? I’m not sure if the paper mentions it, but even if it does, putting it near Ln. 219 would help the reader understand why the value was chosen. Additionally, discussing these values in context of microphysics, particularly terminal velocity, could use a bit of clarity. Often, terminal velocity is discussed alongside vertical velocity due to the use of radar velocity measurements. Based on the rest of the paper, I think the point of the terminal velocity comparison is to note regions where the vertical air motion is sufficient to loft particles (that is, the particles themselves are moving upwards). I think making this clear in that section will make it more clear why the line is drawn where it is.



A related comment is that the paper discusses velocities approaching zero several times. Especially with the context of the discussion of vertical velocity scales earlier in the paper, it’s important to note the difference between approaching zero, but having a statistically significant difference from zero (that is, one can say that the mean is either up or downward motion), and approaching zero, where the mean could be zero. For instance, in Fig 1c, the caption notes that the maxima and minima approach zero, but values are still generally above zero (upwards motion). I suggest considering this distinction when the paper discusses velocities approaching zero. When averaged over larger scales, is the motion still upwards?

General comments:



Unless I missed something, it appears the manuscript references figures out of order. This makes it confusing for the reader to follow.

Ln 16: The trajectories are not the only control, but are an often-neglected control.

Ln 196: What does “handle” mean here - did they QC the data?

Ln 208-210: If QC was performed, was the PLOWS data handled the same way?

Ln 239-240: I think the spirit of this comment is correct, but another flight hazard - icing - will discourage flying in the largest updrafts (in certain temperature ranges) due to strong updrafts potentially reaching liquid saturation.

Ln. 273: I think there are good reasons to prefer a reanalysis vs a higher resolution NWP model here, but I would suggest the manuscript spells out for the reader why a reanalysis was chosen.

Ln. 295-297: Here’s an example of where choice of context may matter. Does ERA represent the environment sufficiently enough that the omega in Fig. 6 is on a scale that’s useful to the story - think back to the discussion of the scales of updrafts earlier.

Ln. 306-308: Could there be a sampling strategy difference, type of storm sampled difference (e.g. more cold Clipper type systems during PLOWS), or an altitude difference that explains this?

Ln. 344-346: What is the paper defining the “entire storm” to be? I think the general strategy for sampling these storms is to go from one edge of the precipitation echoes to the other (insofar as one can do so with the constraints of aviation navigation). These legs are likely not on the extremities of the precipitation region, so this statement is probably true to a certain extent, but how unrepresentative are these values for the comma-head? Stratiform precipitation in cyclones? Etc.

Discussion/conclusions section:



I generally agree with the points here. But I want to add a couple of thoughts that I had while reading that these sections don’t really address. Some of this may be “future work” rather than items that can be addressed with these results, but they are questions raised by the emphatically-worded conclusion in Ln 406-407.

1. Let’s say that the point here, that cloud top generating cells are where all the action is at, is correct. Why do operational NWP models - which neither resolve nor parameterize generating cells - manage to get usable forecasts for these storms? 
2. If upward vertical motions and cloud saturation happen near cloud top - 8 km up or more at times - does this imply conditions at the ground are essentially decided an hour plus in advance due to fall time (assuming negligible loss due to sublimation/virga processes)? This would have significant decision support implications!
3. Given that the processes around generating cells and their updrafts are of a small enough scale, what component of the findings would be easiest to apply to representing generating cells in models - adjusting microphysics parameterizations to grow particles closer to cloud top, or adjusting thermodynamic profiles to match the observations (highest RH near cloud top)? The paper touches on this at the end of the conclusions, but it might be interesting to compare observations vs the reanalysis data.