General comments:

The manuscript describes a method of using buoyancy in the ABL (in particular relatively close to the surface) as diagnostic parameter for process studies in ABL research and applies this method on rather limited data sets from RPAS and radio-soundings gained during two field campaigns (Flux-Capacitor and LAPSE-RATE). Although I clearly see the potential of the proposed method, I would like to be convinced a bit more on the usefulness by also providing some of the classical temperature/potential temperature plots as direct comparison with the more common way of looking into boundary layer development (that could e.g. be done by adding plots with the contours of buoyancy over the temperature/potential temperature field for figures 3, 8 and 9). Having in mind that there might be a radiation error in the temperature (and potentially also the humidity readings), in particular in the beginning of a flight during daytime, it might also be considered to use the observations of temperature and humidity from a fast installed and ventilated sensor at the ground to determine T_par in equation 1, instead of the value taken from the lowest level of the CopterSonde.

The language is not always scientifically concise and leaves some room for mis-interpretation/mis-understanding. I have stated some of the occurrences in the specific comment section, but I also suggest another round on that by one of the more senior co-authors of the manuscript. The figures are also not yet in a high-quality state for final publishing. Here I recommend (in addition to further comments in the specific section below) to consider the following changes:

• I suggest to remove the figure titles that are given in many of the sub-plots; this information is partly given in the axis text and in the figure caption
• I am not sure if the numbering/charactering of the figures in capital letters is the correct format for BLM
• The axis numbers/labels and legends/legend labels are often too small and hard to read

I clearly see a potential of the proposed method, the manuscript will, however, require certain major revisions described above before qualifying for publication in AMT.

Specific comments:

Line 16: sounds like only warm air is transported vertically, I suggest removing “warm” here

Line 20: I feel a bit complicated sentence structure at the end; could “are based off of decades old observation methods” be replaced by “are based on observation methods that are decades old”

Line 58: “Buoyancy is a fundamental principle in fluids caused by density or temperature differences” one example for those scientific inconcise/incomplete statements; buoyancy is related to density, period! It’s temperature and humidity that effect density! So mixing density and temperature asdone in this  sentence should be avoided.

Line 75: I feel there is a verb missing: “…but few studies have yet to substantiate…”

Line 121: CASS has already be used as abbreviation above, so no need to introduce it here again; but please check if it was introduced properly on first appearance

Line 131: the more appropriate reference here might be Kral et al., 2020:

Kral, S. T., J. Reuder, T. Vihma, I. Suomi, K. Flacké Haualand, G. H. Urbancic, B. Greene, G.-J. Steeneveld, T. Lorenz, B. Maronga, M. O. Jonassen, H. Ajosenpää, L. Båserud, P. B. Chilson, A. A. M. Holtslag, A. Jenkins, R. Kouznetsov, S. Mayer, E. A. Pillar-Little, A. Rautenberg, J. Schwenkel, A. W. Seidl, and B. Wrenger, Innovative Strategies for Observations in the Arctic Atmospheric Boundary Layer Project (ISOBAR) — Unique fine-scale observations under stable and very stable conditions. Bulletin of the American Meteorological Society, Early Online Release, doi:10.1175/BAMS-D-19-0212.1, 2020

Line 149/150: “with the lowest observed temperature and dew point used”, this might be misleading and is at least unclear; do you mean (as I assume) the temperature and dew point at the lowest elevation of your flight, or do you mean the lowest absolute value of temperature and dew point of your profile?

Caption figure 1: for a) and c) are the colors messed up! CopterSonde should be blue and Parcel red!

Figure 1, profiles for 15:15 UTC (lower panels): I assume the lowest 200 m indicate the mixed layer of the developing CBL after sunrise. I am a bit puzzled over the super-adiabatic slope of the environmental curve over the whole BL/ML depth. I can’t explain this other than by a systematic bias from your measurements, e.g due to sensor time lag!

Line 167: There should be an “and” before “then”. Or did I mis-understand the meaning of this sentence

Line 175: replace “greater” by “higher”

Line 176: replace “radiosondes” by “radiosonde ascents” or “radiosonde releases”

Figure 2: legend by far too small font size! I also suggest to plot the CopterSonde data also only as dots (as for the radiosondes) and not as line; the time labels should be limited to hours:minutes and the font should be increased; this last comment applies for all plots with corresponding time axis labelling

Line 180/181: “This is a consequence of the shell being heated by the sun before takeoff.” Could this be an explanation for the superadiabatic slope observed in figure 1; the enclosure leads to an overheating in the beginning of ca. 1 deg C that is then slowly reduced due to ventilation during ascent; would perfectly explain an artificial super-adiabatic slope.

Line 182: “Moreover, the warm bias reduces to near-zero at 2103 UTC.” For me this is already the case for 18:30! It would be nice to have the time series of global irradiance to see e.g., the effects of cloudiness in the afternoon reducing/avoiding this warm bias.

Line 199/200: this sentence is very hard to read, I suggest to replace “as potential temperature method, CopterSonde derived heights” by “as those derived from the CopterSonde data.

Line 203/204: “The CopterSonde is more likely to find these sharp gradients near the surface because of the increased data resolution at lower levels.” However, the CopterSonde could also pick up artificial gradients here, just due to the fact that the potentially overheated sensor compartment slowly adapts to real ambient temperatures during flight.

Line 259/260: the statement “the potential temperature is independent of the surface conditions” is at least highly disputeable! Maybe you should specify in more detail what you wnt to achieve with this sentence.

Lines 263 and 272: replace “Radiational” by “Radiative”

Line 267/268: “Figure 9 shows how deep the stable layer is until 1400 UTC.”; How deep would you diagnose the SBL at that time? At around 600 m where the buoyancy gradient seems to disappear, or somewhere between 200 and 400 m, where the gradient becomes distinctly weaker? I miss a bit more explanation how the buoyancy concept can be used as diagnostic tool for SBL height.

Line 269/270: “Not until 1530 does the wind direction shift to southerly, generating some positive buoyancy (Fig. 9, 10).”; This sounds like the wind direction shift is causing the change in buoyancy, I would argue that the change in wind direction is a result of the processes (e.g., differential surface heating) that affect buoyancy; maybe worth of rephrasing this sentence.

Line 279: replace “more” by “a more detailed” or “an in depth”

Line 280/281: “below the ABL”, that should either read “in the ABL” or even in the “mixed layer of the ABL” or “below the capping inversion”

Line 306/307: “There are countless applications for RPASs to sample previously difficult aspects of the atmosphere.” This is a very oversimplified formulation, please write a bit more detailed what you mean, e.g. something along the lines “to collect atmospheric measurements for processes that are still not adequately understood”

General comment for the references; check citation of pages, sometimes with pp.! mostly without; so please homogenize

Line 330: journal name should be upper case

Line 333: reference Båserud is incomplete (journal name, volume, pages etc)

Line 345: journal name should be upper case

Line 354: journal name should be upper case

Line 360: journal name should be upper case

Line 373: journal name should be upper case

Line 396: journal name should be upper case

Line 406: journal name should be upper case

Line 407: journal name should be upper case

Line 419: journal name shouldn’t been abbreviated

Line 429: journal name should be upper case

Line 434: journal name should be upper case

Line 439: journal name should be upper case

Line 457/458: journal name should be upper case

Line 464: journal name should be upper case

General Comments:

I find the manuscript to be well-written. The authors demonstrate how observations of buoyancy from remotely piloted aircraft systems operating in the ABL can have practical applications in the diagnosis of ABL height, convective initiation and low-level jets. The authors utilize two field studies for evaluation, both of which are appropriate for the kind of (free-convection-driven) environments where buoyancy measurements are likely to be useful.  Specifically, the authors demonstrate how the proposed buoyancy-based method for ABL height diagnosis mitigates misdiagnosis from noise in the UAS-measured fields and is a viable method for both UAS and radiosonde-based profiles. One could argue that this article is less about RPAS observations and more about buoyancy as an ABL diagnostic, which would make it better suited for a journal other than AMT. I think the authors’ well-written introduction and conclusion sections provide enough context for this manuscript to stand as an RPAS application, however.

I have listed below some specific comments regarding some of the statements made and the figures. I think some non-trivial adjustments are needed in places, but the foundation of the investigation is sound and can be revised into a manuscript ready for publication.

Specific Comments:

L 48: The authors characterize “ABL transitions” on this line through its impact on ABL height and stability and proceed with a strong overview of ABL height estimation methods. I think the description of ABL transitions could be expanded a bit beyond this one line, given the history of this topic itself and its prominence in the manuscript title, to help frame the term for the reader. For example, when is the temporal evolution of an ABL a ‘transition’ and when is it not? I agree with the authors’ general characterization here; my only suggestion would be to include the influences from the 3-D flow on ABL transition that may not be diurnal, e.g., internal boundary layers, mesoscale circulations (such as the authors’ example later around L 270?), etc., and transitions in forced convection environments). It’s also worth noting that the word transition is used later (L 202) in a different context, seemingly referring to changes in the vertical temperature profile slope, rather than changes in ABL ‘regime’, which is the more common usage for the ‘transition’ term, I would argue. My impression after reading the full manuscript is that the authors are using low-level buoyancy from RPAS observations to diagnose boundary layer features and characteristics.  

L 128-132: Were the sensor bias corrections applied here identical a particular past campaign, or were new values used? If the latter, what were the new values and how were the corrections applied? 

L 173-174: Minor point, but I would disagree that radiosondes are Lagrangian tracers. The radiosondes here are balloon-launched, which provide lift that enforces upward motion (initially), regardless of the ambient vertical velocity (whether upward or downward). A tracer follows the ambient fluid velocity in three dimensions.

L 193-194: This sentence seems a bit unnecessary and I would suggest that it be omitted. Buoyancy is sensitive to temperature difference between a parcel and its environment, regardless of whether the parcel is in the lower troposphere.

L 195: The sentence reads: “ . . . buoyancy becomes convectively well-mixed .. . .” Do the authors mean to say that the ABL becomes convectively well-mixed? Buoyancy being a force, I’m unclear on what is meant by buoyancy being convectively well-mixed. Following the authors’ definition in (1), in a well-mixed unsaturated environment, the environmental temperature would follow the adiabatic lapse rate, which a buoyant unsaturated parcel would also follow. It’s arguably the components of the expression that are “well mixed” (or constant). Perhaps an alternative expression would be to call it invariant? The expression is also used around L 280 where the authors first introduce “isotropic buoyancy”. I think this term is closer to the conditions described, though isotropy does carry a connotation of direction.

L 202-203: As the authors insinuate, this example shown in Figure 4 is not appropriate for the potential temperature method applied to these specific CopterSonde data due the deep well-mixed layer. Despite that, have the authors experimented with applying some kind of low-pass filter to the CopterSonde measurements to truncate the high-frequency variability? For this example, I have to wonder if, imagining the CopterSonde flew higher or the mixed layer were shallow enough for the CopterSonde to capture it fully, would the method applied to the CopterSonde data in Figure 4 still yield erroneously low ABL heights?

L 210-212: The authors contend that local winds combined with differences in measurement time yield discrepancies between measurements on the two platforms. I think this hypothesis could use some further analysis, given that winds are much stronger later in this 24 hr period, where there appear to be non-trivial differences between measurement times between the two platforms as well. Is there some kind of forcing at the mesoscale or synoptic scale (or something else?) to explain this condition, where the inter-platform differences early on are much more pronounced than later?  To me it seems plausible that this difference could be partly related to the platform heating issue that was shown in Figure 2.

L 230: Could the authors please clarify what is meant in this sentence beginning with ‘Contrary’? Shapiro et al. (2016) indicate in their paper (pg 3045) that buoyancy is specified to be at maximum a few hours before sunset, also depicted in Figure 7. From references to this elsewhere in the paper, surface buoyancy peaks at three hours prior to sunset. Given a local sunset time of about 0010 UTC, my visual estimate is roughly a four-hour difference between the peak of vertically integrated buoyancy and the time of sunset in the authors’ Figure 6. This seems reasonably close to the result in Shapiro et al. (2016), in contrast to the authors’ statement on L 230.  In addition, the authors are comparing maximum surface buoyancy of Shapiro et al. (2016) to their vertically integrated buoyancy expression. Could the authors please elaborate on why and under what circumstances these terms are directly comparable?

L 265: I’m unclear what the authors mean by “process continues throughout the night . . .”. The timeline of Figures 9 and 10 begins around 0530 LT (MDT) and continues until around 1100 LT. In the following sentence, are the authors referring to the inversion as shown on Figure 9, which increases after sunrise and then remains fairly constant in time?

Figure 3: Qualitatively, the two subplots appear to match-up well except in the first few hours, roughly before 1900 UTC. The CopterSonde wind measurements appear to noticeably underestimate those of the radiosonde at these times between about 400 m and 900 m AGL, but also in the region of roughly 100 – 300 m around 2000 UTC. Corresponding buoyancy values in this spatio-temporal region are also show non-trivial discrepancy in the initial hours at the upper levels. The authors provide some brief analysis later in the text (L 211-213). The authors contend that the discrepancy is due to a difference in location between the two platforms. I noticed that coherence between the two platforms is fairly high later on in the period, when wind speeds are considerably stronger. Is there a known physical reason to expect why the mixed layer temperature should be so different in an adjacent space 30 minutes later given the relatively flat and homogeneous nature of the domain early on? Can these buoyancy differences be explained by the platform solar heating problem (L 180-181)? Also, I would urge the authors to shift their analysis of Figure 3 earlier in the text, to where the Figure is first introduced.

Technical Corrections:

L 20: decades old -> decades-old

L 27: this type of data has –OR- these types of data have

L 82: “Flux-Capacitor” may not be immediately recognized by the reader as the name of a campaign at this point in the manuscript. Suggest: “The Flux-Capacitor campaign” or “Flux-Capacitor (2018)”, etc.

L 83: Suggest: “Southern Plains” or “southern Plains” as the authors see fit – somewhere in here there’s a proper noun and the authors are in a better position than the Reviewer to decide what that should be

L 99: Suggest: San Luis Valley -> San Luis Valley, Colorado, U.S.A. given that AMT is an EGU journal.

L 99, 111: Note that the same person (Gijs de Boer) is referenced as “de Boer et al.” on L 99 and “Boer et al.” on L 111. Suggest using the former naming format (and change to 2020a) and renaming the second reference as “de Boer et al. 2020b.”

L 103: Please define CASS here (first defined later on L 121)

L 116: “the CopterSonde” hasn’t yet been introduced at this point in the manuscript. Suggest replacing with “ . . . utilized a rotary wing quadcopter RPAS . . .” or perhaps “ . . .utilized a CopterSonde RPAS” for clarity.

L 124-125: “. . . pressure sensor within the autopilot.” Do the authors mean that the pressure sensor is affixed to the autopilot board?  Perhaps the authors could clarify here as the word autopilot is often used to refer to the software of the CopterSonde, too.

L 170: cross evaluated -> cross-evaluated

L 176: long established -> long-established

L 205-206: Doesn’t this sentence effectively repeat the statement on L 201-203? Likely can omit.

L 207-208: “the height of mixed potential temperature” -> “the height of the mixed layer”

L 210: replace comma with semi-colon;

L 226: Could the authors please include the equation used to compute vertically integrated buoyancy to verify the units as shown in Figure 6?

L 234 -245: The verb tense changes over the course of these two paragraphs, from past to present to future. Please be consistent with verb tense throughout.

L 260-261: Assuming that the authors mean that the buoyancy foretells of change to local convection strength (?), I would suggest ‘portend’, ‘herald’ or ‘precede’ as a replacement for ‘predate.’

L 269: Add ‘UTC’ after ‘1530’

L 281: “below the ABL” -> below the top of the ABL (or similar)

L 323: Note that AMT asks that authors include the DOI in the references where available (https://www.atmospheric-measurement-techniques.net/submission.html#references )

Figure 1: The caption indicates that the red lines are from the CopterSonde and the blue lines are from the parcel calculation, but the figure shows the opposite color assignment.

Figure 1: Could the authors please enlarge the text of the legend in subplots (a) and (c) slightly to improve legibility.

Figure 1: Would suggest indicating the UTC times of the examples in the caption to help the reader distinguish the top and bottom rows of the figure. 

Figure 2: In the caption, add a space between 2018 and Purple.

Figure 2: The time of sunset (sunrise) in Washington OK on 05 October 2018 was approximately 1910 CDT (0727 CDT), or 0010 UTC (1227 UTC) (e.g.,: https://sunrise-sunset.org/us/washington-ok/2018/10 ). The sunset (sunrise) graph on the figure indicates approximately 2020 UTC (0121 UTC), however. Could the authors please revise the figure? Also, the abscissa tick mark labels do not appear to cover the time range indicated by the caption (1500 UTC 05 October 2018 to 1430 UTC 06 October 2018).

Figure 2: The CopterSonde temperature graph (black) is drawn as a continuous line over this period (indicated to be approximately 24 hours according to the caption), but Table 1 indicates that these observations were taken during 46 flights averaging 30 minutes each. Why the black-shaded graph in this figure drawn this way? Could the authors please revise the figure to indicate where CopterSonde observations are present versus absent?

Figure 3: In the caption, some text is missing before the word filled: e.g. …”and wind speed (m s^-1) is shown in filled contours”

Figure 4: Could the authors please enlarge the text in the legend in the upper right corners of both subplots to improve legibility? Same for Figure 5 and other figures where this point size is used.

Figure 4: Please indicate in the caption the date and location of these measurements as done in Figure 3.

Figure 4: The graphs indicating time of sunset appears a bit off with respect to the abscissa tick mark labels in subplot (b). The time of sunset should be around 0010 UTC by my findings for 05 October 2018 in Washington, OK.

Figure 5: In the caption: buoyancy determined -> buoyancy-determined

Figure 7: Could the authors please add a ruler showing distance to these subplots? The authors make frequent reference in the adjacent text to distances from the stations and it would be helpful to have that reference on the subplots. Also, it is somewhat difficult to discern where the mountains are located with the radar reflectivity present and the presence of multiple colors shading the surface. Perhaps the authors could add a subplot showing relief? This is mostly in response to the statements on L 243-244 about issues with radar coverage and interference from mountains. I’m not entirely sure I follow this statement – are the authors referring to convective cell base height? I would anticipate convective cells within the valley would be recognized by radar if there were no obstructions between the cell and radar, no?

Figure 7: Please add the label ‘UTC’ to the times labelled on the subplots, for clarity.

Figure 8: Suggest some rewording for clarity: “Unfilled contours are buoyancy (m s^-2) where solid (dashed) are positive (negative) and filled contours are specific humidity (g kg^-1) from CopterSondes . . . “

Figure 8: Date format in the subplot titles differs from the format used elsewhere in the manuscript (DD MMM YYYY).

Figure 9 & 10: I would suggest merging these two figures as subplots of a single figure, as the authors did in Figure 8. The authors use both Figures 9 and 10 to describe the sequence of physical processes in the final paragraph of Section 5. I think it would be useful to have the figures side-by-side to facilitate comparison.