Small Uncrewed Aircraft Based Microphysical Measurements of Polar Stratus Cloud During The Pallas Cloud Experiment 2022

Girdwood, Jessica; Brus, David; Doulgeris, Konstantinos-Matthaios; Böhmländer, Alexander

doi:10.5194/essd-2025-257

Preprints

https://doi.org/10.5194/essd-2025-257

Preprints

11 Aug 2025

| 11 Aug 2025

Status: a revised version of this preprint is currently under review for the journal ESSD.

Small Uncrewed Aircraft Based Microphysical Measurements of Polar Stratus Cloud During The Pallas Cloud Experiment 2022

Jessica Girdwood, David Brus, Konstantinos-Matthaios Doulgeris, and Alexander Böhmländer

Abstract. A dataset of in-situ observations of stratus cloud microphysics was created from measurements performed at the Pallas atmosphere-ecosystem super site. The data were collected using a small uncrewed aircraft (SUA) and the low-cost, lightweight Universal Cloud and Aerosol Sounding System (UCASS, Smith et al., 2019). Data from the instrument – platform combination was previously validated in Girdwood et al. (2022b) during a similar field campaign at the same site. The dataset contains cloud droplet size distribution, number concentration, and mass concentration, in addition to geolocation data, and meteorological variables. The flight pattern of the SUA was planned to provide a quasi-vertical profile. A total of 84 of these profiles across 39 flights were performed during the campaign period. The data from the SUA flights are available from 10.5281/zenodo.14756233 (Girdwood et al. 2022a).

Received: 02 May 2025 – Discussion started: 11 Aug 2025

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.

Download & links

Jessica Girdwood, David Brus, Konstantinos-Matthaios Doulgeris, and Alexander Böhmländer

Status: final response (author comments only)

RC1:
'Comment on essd-2025-257', Anonymous Referee #1, 05 Sep 2025
The authors present cloud micro-physical measurements taken by the UCASS optical particle spectrometer aboard a small uncrewed aircraft (SUA) during the Pallas Cloud Experiment 2022. The manuscript provides a good overview on the used instrument and data products, and illustrates an example of measurements. I am lacking some information on the retrievals used and a discussion of their uncertainties and limitations. I recommend publication of the manuscript after the following comments have been addressed by the authors.
General comments
Sec 3.2 “Data Analysis and QA”: The data analysis part of this Section is very short. I would suggest to add a separate Section 4 titled e.g. “Overview of Cloud Properties” summarizing the measured properties that the title of the manuscript advertises. Within this Section, more general information on the cloudy conditions should be presented, e.g. by summarizing where (geographically; also with respect to cloud base and top) clouds were profiled; if the whole cloud layer or only parts of it were measured, and where profiling relative to cloud base and top was performed; how does the data set fits in with other cloud data sets obtained during the campaign? How did profiles of the same cloud vary within one flight? What are the errors of the retrieved variables?

I am missing an overview and discussion of the retrievals applied to the measurements. How are the variables

mass and number concentration and effective radius (Tab 1) calculated and what errors and limitations arise from the retrievals and assumptions therein? How is phase information taken into account when retrieving in mixed-phase or ice cloud conditions? I would suggest to add a sub-section to either Sec 2 or 3 illustrating this currently missing information on the dataset.

Specific comments
Abstract
Line 2: add […] super site during the Pallas Cloud Experiment (PACE) in Autumn 2022.

L 4: How did measurements differ compared to previous data sets taken? (also see next comment)

Sec 1: Introduction
An additional paragraph is necessary to introduce the PACE experiment in more detail to the reader. Information should be added on other cloud measurements, previous in-situ observations within the PACE experiments, and how this dataset differs to previous ones.

L10: rather: Earth system models

L 11, ff: references given should be more specific and target gaps in the knowledge of clouds specifically observed at the PACE site.

L 13: also see comment above; which cloud processes are targeted with the measurements taken?

L 14: „One of the reasons for this“: unclear to what this is referring to

L 22: A short paragraph on the UCASS system should be added here to introduce it to the reader, including framing the system within other similar SUA systems. Some of the information given L 30 ff could be moved to an earlier stage of the manuscript.

Sec 2: Methods
Section title might be misleading. I would rather propose “Instrument and Flight strategy” as a title.

L 41, 56, 116: abbreviation (noa, b) needs to be explained

See GC2: a sub-section explaining the retrievals in more detail should be added here or in Sec 3.

Fig 3: it would be nice for the reader to get an understanding on where all flights were generally taking place. Is the selected plotted flight a random or representative example of the flight trajectory? It would be nice to see all flights on one map, e.g. color-coded by flight number; alternatively, if the plot gets too busy, a selection of representative or highlight flights could be added which could then be illustrated further in new Section 4 (see GC 1), including an illustration of how variable conditions were from profile to profile within the same flight. Additionally, an illustration of the vertical flight trajectories e.g. in a sub-figure (b) would add valuable information to the flight strategy.

Sec 3: Data
L 80: the global time step was set to 0.5s. What vertical and horizontal resolution, respectively, do the output variables have due to this averaging applied, and how does it change throughout the vertical profiles illustrated in Figs 5+6? This information would also be important to discuss in the Results Section (see GC 1)

L 87: “using one of a number of algorithms” should be specified in more detail; which ones were used? References should be added.

L 93: only discuss QA here (see general comment 1)

L 95: unspecific; constraints and limitations should be further discussed e.g. in previous Section

L 100: how were the threshold values chosen, and what uncertainty do they introduce to the masking?

L 101: unclear, sentence should be double-checked

L 103, 107: wind direction and speed were taken from Sammaltunturi station. From Fig 3, it seems like the flights were operated >3km away from the station. Were wind measurements only taken at ground? The authors should discuss in more detail how the wind information error due to the distance and orography affects processing, retrievals and data uncertainty.

Fig 5, 6: what is the minimal height that the instruments operate at? The plot seems to start around 400m; is this a system limitation or due to the observed cloud conditions or orography? Information should be added on where the particles were collected with respect to the cloud boundaries (see GC 2).

Minor xtick labels should be added to all respective flights sub-panels at the bottom. Are all panels showing the same xrange? It would be nice to add ‘10’ and ‘20’ to the top x-ticks.

It seems like lines are connecting points throughout clear air patches (eg Fig 6 profiles 25-28 between 750-1200m). Replacing lines with plotted dots would further illustrate the vertical resolution of the measurements.

Tab 2: Information on how long each flight lasted should be added. More information on the meaning of airspeed type ‘normal’ vs ‘corrected’ is needed in the text.

L 123 (also see GC 1): how variable are profiles change within each flight and how far spaced geographically?

Summary
A more concise outlook should be added to how this data set will be used in the future with respect to the questions and research areas raised in the introduction (L 12-14), and within other PACE measurements and modeling activities.

Technical Corrections
L 74: reference incomplete

L 94: spell out QA in Section title

L 109: Section title on new page

L 115, 116: what do ‘SHT’, ‘BME’, ‘AG’ stand for?

Fig 5, 6: top x-ticks don’t need 0 to match numbers given in Tab 2
Citation: https://doi.org/10.5194/essd-2025-257-RC1
- AC1: 'Reply on RC1', Jessica Girdwood, 24 Dec 2025
  
  ## Response to RC1
  
  ---
  # General Comments
  
  - The UAV profiles were all up to a set altitude of 2 km above ground level
  
  which was the legal limit. The cloud tops were beneath this limit for all but
  
  two of these profiles. The flight information section did contain this.
  
  Overall, instead of a separate section, the discussion of the cloudy conditions
  
  has been expanded upon in this section.
  
  - the flight information and constraints section was renamed to "flight
  
  information and conditions"
  
  - information, however it has been made more plain
  
  - The "Data analysis and qa" section was amended to explain all the derived cloud
  
  properties in the dataset.
  # Specific Comments
  
  - added "during the Pallas Cloud Experiment - 2022" on Line 2
  
  - "These measurements are intended to expand on the previous campaign since
  
  they form an extended dataset with the uncertainties already evaluated by
  
  previous experimental work" added to this line.
  
  - I have added a paragraph in the introduction to explain the difference
  
  between this and my previous work. It's worth stating that I am actually the
  
  same author as Girdwood et al 2022, I have just changed my name since then.
  
  - climate models changed to earth system models on L10.
  
  - Specific references to polar clouds added.
  
  - I added a clarification of the intended use case of the dataset at the end of
  
  the introduction since it made more sense to put it here after I introduce
  
  the SUA, instrument, location, etc.
  
  - rephrased L11 to clarify
  
  - UCASS explanation moved to L22 and more citations of previous use given. The
  
  UCASS is described in detail later in the paper, this is now cross
  
  referenced.
  
  - Altered the section headings to remove method section and move its
  
  subsections up a level, since there is already a more
  
  specific flight information section.
  
  - The abbreviation was actually a corrupted reference which is now fixed.
  
  - The flight in Fig. 3 was very characteristic, and the caption has been
  
  updated to reflect this. Any attempt at plotting multiple profiles resulted
  
  in an unreadable plot, though I am colourblind so this probably does not help.
  
  - The spatial resolution of the data depends on the ground speed and ascent speed
  
  of the aircraft. This changed throughout the campaign as a result of changes
  
  in wind speed. The dataset is in the time domain in order to comply with the
  
  data standards for the rest of the campaign outlined in Brus et al 2025. The
  
  manuscript has been amended to reflect this.
  
  - "one of a number of algorithms" is now expanded upon in the data analysis
  
  section, which now has a relevant cross reference.
  
  - data analysis and quality assurance sections split apart.
  
  - the constraining conditions for the UCASS being referred to here are the AoA
  
  and airspeed limits, which are discussed in the flight information section.
  
  This has now been cross referenced and expanded upon.
  
  - The reason for the choice of these limits was engineering constraints with
  
  the UCASS, which was previously stated in the flight information and
  
  apparatus sections.
  
  - Expanded the apparatus section to include information on UCASS uncertainty.
  
  - Line 101 was rephrased
  
  - Added paragraph discussing co-location error at the end of data QA.
  
  - The altitude is above sea level, this was clarified in the text. There is no
  
  minimum altitude, just the take off altitude.
  
  - Figures 5 and 6 have been altered to include the error for other reviewers
  
  comments.
  
  - I added references to the exact equations used to calculate corrected airspeed.
  
  - The take off site was the same each time, and the profiles were very similar
  
  and only changed in orientation with different wind directions. This was
  
  added to the text already in response to a different comment.
  
  - A paragraph was added to the summary with the intended uses of the data.
  # Technical
  
  - date now added to citation on L74.
  
  - L94 QA now spelled out.
  
  - L109 the manuscript is reformatted following publication anyway and the
  
  spacing of section titles is changed when the two column format is
  
  implemented.
  
  - SHT and BME are just the names of the sensors, and AG is a broken reference
  
  which is now fixed.
  
  Citation: https://doi.org/10.5194/essd-2025-257-AC1
RC2:
'Comment on essd-2025-257', Anonymous Referee #2, 19 Sep 2025
The manuscript presents a dataset of in situ data of polar stratus clouds collected from a fixed-wing small uncrewed aircraft (SUA) during the Pallas Cloud Experiment (PaCE) 2022 at the Pallas atmosphere-ecosystem super site. The dataset over 39 flights with 84 quasi-vertical profiles includes data on cloud droplet size distributions including derived products (number concentration, liquid water content, effective radius), meteorological variables, and SUA flight data. A quality check was performed on the data and different processing depths (“a1” and “c1”) were applied.
General comments:
I’m not sure if the dataset now only includes measurements of “arctic warm stratus” (line: 32), or also arctic cold stratus / mixed-phase clouds (line 104: “ice while flying”, line: 136 mixed-phase clouds).

If you were also measuring mixed-phase clouds and aerosols (line: 53) how did you handle the discrimination between cloud droplets, ice crystals, and aerosols? Also, due to their very different refractive indices that could affect the measurement. This could strongly influence the measure cloud droplet size distribution and derive variables.

If you measure cold cloud or mixed-phase conditions a flag in the data set would be helpful, and some short sentences on how you define these temperature regimes, i.e., using the BME temperature or other instruments.

You could also include a “QA masking impact”, i.e., how much of the respective flight/profile (e.g., the fraction) was masked.

What is largely missing in the manuscript is a compact uncertainty budget for (a) sizing (e.g., due to refractive-index / model dependence), (b) counting / concentration (e.g., sample volume), (c) LWC (e.g., density assumption, mixed-phase conditions), and (d) airspeed substitutions when the pitot tube froze (e.g., ground-station wind - based correction). Particularly the latter considering the wind speeds were presumably measured on ground level.

Please extend the analysis part of the data on how the variables are derived from the particle size distribution measurements. Which processes “proc” were actually used.

Summary/Conclusion: This part is very short and missing an outlook on what the dataset could be used for or how it could support other datasets if used in synergy. Mainly what is the fundamental value for future studies.

Specific comments:
Line 117-119: This should go into the data analysis part, where you introduce the masks.

Figure 5: Number concentration of cloud droplets, aerosols or both?

Do you always show the same x-range?

Are the concentration values saturated (e.g., profile 6) or just not displayed due to the x-axis range?

Figure 6: The x-axis label is confusing looks like “510”, please at least include the ticks for every profile. Also here particularly evident is that you need to give some error estimates. Comparing profile 002 with 011 already indicates that due to counting statistics the uncertainty for 002 is higher compared to profile 011.

Line 120: Level “c1” means that the flight controller data was available and “a1” where it was not available? Please clarify.

Line 117: What do you mean by “The data level is also included here.” This needs more explanation, what is a data level?

Minor comments:
Line 44: what is (noa,a) after pitot tube?

Line 56, 116: what is (noa,b)?

Line 68: include space between 15° and angle

Lines 69,70, 99, 100: include small space between unit m\,s

Line 74: check reference year number

Line 75: Period missing

Line 101: remove first “time”

Table 1: kg\,m^-3

115: I thin the company is called Sensirion not Senserion

116: Change “Bosh” to “Bosch”

119: altitude not attitude :)

Table 2: define n/a in description (not available vs not applicable)

Throughout: change in-situ to in situ
Citation: https://doi.org/10.5194/essd-2025-257-RC2
- AC3: 'Reply on RC2', Jessica Girdwood, 24 Dec 2025
  
  ## Response to RC2
  
  ---
  # General Comments
  
  - Warm cloud changed to liquid cloud. There were no ice crystals in this cloud
  
  but there were supercooled droplets which froze on impact with the airframe.
  
  The manuscript has been changed to reflect this.
  
  - Droplets and ice crystals are discriminated against by size fraction
  
  separation. The scattering geometry of the UCASS means any ice encountered
  
  during the flights will have a scattering cross section beyond the range of
  
  detection. In addition, the probability of encountering an ice crystal during
  
  measurements at this time of year is negligible. The manuscript was changed
  
  to state this with the relevant references in the flight information section.
  
  - A section on data analysis was added for reviewer #1 and I think this also
  
  covers your comment here :)
  
  - A paragraph at the end of the summary was added to explain the intended uses
  
  of the dataset.
  
  - The data levels are consistent with the rest of the papers in this ESSD
  
  special edition for the PACE22 campaign. An explanation of this and the
  
  relevant reference was added to the manuscript.
  
  - I have added standard error to the effective radius plot in addition to a
  
  discussion in the manuscript.
  # Specific Comments (response in order)
  
  - broken reference which is now fixed
  
  - broken reference which is now fixed
  
  - L68 spacing now fixed
  
  - half space now added
  
  - Year number added
  
  - full stop added to end of sentence
  
  - This sentence was reworded for a different reviewer :)
  
  - spacing corrected
  
  - Senserion changed to Sensirion
  
  - bosh changed to bosch
  
  - attitude is an aviation term which refers to the angular position of an airframe in
  
  space, which is what i meant in this instance.
  
  - n/a defined in caption of Table 2.
  
  - changed in-situ to in situ
  
  Citation: https://doi.org/10.5194/essd-2025-257-AC3
RC3:
'Comment on essd-2025-257', Anonymous Referee #3, 01 Oct 2025

The manuscript describes a dataset collected in 2022 to understand cloud microphysics using a small UAS. The article contains useful information regarding the UCASS sensor and the available data. The platform would be beneficial for addressing the data gap in in-situ cloud observations. However, the manuscript lacks some key information about data processing.
General Comments:
I think it would benefit the reader if there was more description about how this article differs and expands upon the work from Girdwood et al. 2022b. As the introduction stands now, I do not see how this dataset is unique or what was learned from the previous campaign.
There should be more description on the data processing mentioned in Software. The authors mention that there is space for customization but do not describe the standard practice used on the dataset herein. Overall, the description of the QA process needs to be expanded. I do not understand the processes which went into creating the c1 level files.
The meteorological observations (T and RH) are mentioned but never described in their data quality or availability. I would imagine condensation on the sensors could pose an issue, but this is not mentioned, nor do we see any examples of meteorological data plotted in the manuscript. Were there any calibration offsets or post-processing to ensure accurate measurements?
The doi links in the Abstract and Data Availability statement do not work.

Specific Comments:

L4: It would be nice to add the year of the Girdwood 2022b campaign to distinguish the two datasets better.
L20: Have there been other UAS to gather cloud physics observations? If so, they should be mentioned in the introduction.
L41: What is (noa, a)?
L44-L47: All of the pointing to references to describe the data processing makes it hard to follow.
L54-56: Are there any differences between these two sensors? Or in their respected data quality?
Fig 3: You might consider adding a satellite image to understand the cloud types and coverage in the flight path.
L102: Are there differences in data quality between ascending profiles and descending? How is the concern for rotor-wash influence on the environment mediated?
L104: This seems like a fundamental issue given the goals of the platform. How often was the pitot tube unusable? Does that reduce the frequency of observations?
L128: Fix the citation

Citation: https://doi.org/10.5194/essd-2025-257-RC3
- AC2: 'Reply on RC3', Jessica Girdwood, 24 Dec 2025
  
  ## Response to RC3
  
  ### Jessica Girdwood
  
  ---
  ### General Comments
  
  - The purpose of the previous campaign was to validate the technique used to
  
  gather these data. During the 2020 experiment, we flew the aircraft --
  
  instrument combination in a number of configurations with a number of flight
  
  paths to experiment with how is best to gather these data. The introduction
  
  was expanded as part of another reviewers comments and I think that this is now
  
  clearer :)
  
  - The data analysis and data QA sections were split apart, and the data
  
  analysis was greatly expanded upon, I hope that this is now clear.
  
  - description of temperature sensor and housing added to the end of the
  
  apparatus section. Condensation on the sensor is an expansive issue with
  
  in-cloud temperature measurements and sparks a lot of debate in conventional
  
  aircraft measurement spaces. Solving this issue here is considered beyond the
  
  scope of this paper.
  
  - Links are now fixed.
  ### Specific comments (in order)
  
  - Date is now added for clarification.
  
  - The introduction was expanded upon with links to previous campaigns.
  
  - a broken reference which is now fixed
  
  - I'm not 100% sure what you mean by this, though the description of the UCASS
  
  was greatly expanded upon for another reviewer which I think might address
  
  this comment.
  
  - the only difference between the two UCASS variants is the size range, there
  
  is no difference in data quality.
  
  - unfortunately there was no satellite data. However I have added a more
  
  detailed description of the cloudy conditions and type of cloud encountered.
  
  - Yes. I discussed this in detail in Girdwood et al 2020, 2022. Rotor wash is not an
  
  issue here because this is a fixed wing SUA and it does not fly through its
  
  own wake at any point. I have altered the manuscript to state this.
  
  - you're right, this is an issue. I have recently worked on an upgrade to the
  
  UCASS which includes a hot wire airspeed sensor which does not have this
  
  issue, but it was unfortunately not ready in time for these experiments. The
  
  associated error with the corrected airspeed is now mentioned in the text, and
  
  the affected profiles can be found in table 2.
  
  - Fixed the citation.
  
  Citation: https://doi.org/10.5194/essd-2025-257-AC2

Jessica Girdwood, David Brus, Konstantinos-Matthaios Doulgeris, and Alexander Böhmländer

Data sets

Data From the Universal Cloud and Aerosol Sounding System Abord an Uncrewed Aircraft During the Pallas Cloud Experiment 2022 Jessica Girdwood et al. https://doi.org/10.5281/zenodo.14756233

Jessica Girdwood, David Brus, Konstantinos-Matthaios Doulgeris, and Alexander Böhmländer

Viewed

Total article views: 1,456 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,329	88	39	1,456	41	50

HTML: 1,329
PDF: 88
XML: 39
Total: 1,456
BibTeX: 41
EndNote: 50

Views and downloads (calculated since 11 Aug 2025)

Month	HTML	PDF	XML	Total
Aug 2025	237	16	4	257
Sep 2025	875	11	11	897
Oct 2025	101	19	10	130
Nov 2025	52	17	4	73
Dec 2025	50	25	10	85
Jan 2026	14	0	14

Cumulative views and downloads (calculated since 11 Aug 2025)

Month	HTML	PDF	XML	Total
Aug 2025	237	16	4	257
Sep 2025	875	11	11	897
Oct 2025	101	19	10	130
Nov 2025	52	17	4	73
Dec 2025	50	25	10	85
Jan 2026	14	0	14

Viewed (geographical distribution)

Total article views: 1,434 (including HTML, PDF, and XML) Thereof 1,434 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 09 Jan 2026

Short summary

In-situ data of cloud microphysics is essential for targeted studies of cloud processes, validating remote sensing, and both assessing and improving the accuracy of weather and climate models. In this work we adopt a novel technique using a small uncrewed aircraft (SUA) and a bespoke sensor to produce vertical profiles of cloud microphysical parameters. The data are publicly available from https://zenodo.org/records/14756233.


Total:	0
HTML:	0
PDF:	0
XML:	0