Summary:

The manuscript by Kottmeier et al. describes the KITsonde dropsonde system, which can release 4 sondes simultaneously inside a dedicated launch container. This allows dropping either meteorological sondes of the same type, or sondes measuring different parameters such as cloud particles and gamma radiation. This system also allows direct transmission of data to the ground via a communication satellite.

This manuscript discusses the details of the dropsonde system and shows data from several field campaigns to demonstrate its capability.

The manuscript is overall well written. Some details require more discussions and clarifications, and I can recommend publication of this manuscript after some minor modifications following comments listed below.

Detailed comments:

Section 2.2: How long does the release of the individual sondes from the release container take and how much time or altitude below the aircraft is lost before measurements can be considered reliable as a result of this delay?

Lines 231ff: The heating feature of the DFM-17 reduces the risk for evaporative cooling after exiting clouds. The lack of this feature in the dropsonde variant is relevant and could extend the perceived cloud thickness to lower altitudes and could potentially lead to evaporative cooling at cloud base. This potential limitation needs to be discussed. Does the sonde measure and report the temperature of the humidity sensor?

Do you have access to raw humidity data without clipping? That would provide some feeling about the issues of measuring RH near saturation. This should be discussed to clarify the capabilities of the humidity sensor in this configuration. 

Section 2.4: This section should include the process how the meteorological sonde is released from its release container when satellite communication is used to transmit data to the ground directly. This is mentioned later in the manuscript.

Section 2.5: How can you be sure that the airflow through the particle detector is close to the ambient fall rate? Is the central channel straight without any obstructions? Can you account for any pendulum motion of the sonde, which would invariably affect the air flow through the instrument? Some discussion follows later but should be moved here. The meteorological sonde hanging below the OPC carries the risk of particle shedding, in particular after passing through clouds, which the particle detector may sense. This is probably not an issue for aerosol particles but should be mentioned. 

Line 253: The statement about size and cost is relative and should be deleted here.

Section 2.6: Can you show the profile of the test drop near Magdeburg? There is possibly a detectable increase in radioactivity due to cosmic ray activity as a function of altitude, which would be worthwhile showing. This would give a feeling of the baseline that the detector would see.

For temperature, humidity, pressure, and wind it would be good to know what the true uncertainties in flight are after all systematic and random errors are considered. The uncertainties listed in Table 1 are extremely optimistic and unrealistic for real world observations. This becomes relevant when the reader tries to interpret the differences of soundings discussed in section 3. Large scale structures shown in Figure 9 are certainly real, but when the differences reach those of the stated uncertainties, caution is warranted.

A particular challenge for dropsonde measurements is their validation. Are there any baseline measurements on the aircraft prior to launch? Are there any additional measurements on the ground prior to a mission? Have there been any attempts to compare the dropsonde observations with nearby balloon or remote sensing (e.g. lidar) measurements? Have there been any attempts to have four sondes from a release container to measure as close by as possible, i.e. using identical parachute size? If so, it would be good to show the scatter of all four sondes around their mean. Any additional information that could be used to quantify the accuracy of the measurements in flight would be helpful to get a better impression about the true capabilities of the meteorological sondes.

Line 379f: You correctly state that the atmosphere above 2 km is stably stratified. What then does the statement refer to that says “Two more stable shallow layers are present at about 6 and 8 km AMSL”, when the entire region is stable? Please clarify.

Line 381f: There seems to be a problem with the color coding of the profiles or with referencing the profiles in the text. The cloud top in profile 1d seems to be closer to 4.5 km (not 5.5 km). Profile 1c (not 1b) does not appear to fall through a cloud. Please check the description of the text of this section.

Is there a reason that all four sondes of drop 1 seem to be missing data in the region between 10 km and 11 km? The wind speed in that layer seems to be at or above 50 m/s (not 40 m/s as stated).

Lines 414ff: This explanation about the OPC sounding should be moved up to the instrument description in section 2.5.

Line 424: Figures 13 and 14 are referred to before Figures 11 and 12. Please add a reference to Figure 11 and 12 before that.

Line 472f: Paragliders require active control or the parafoil. Are there any plans to implement this? Otherwise, maybe delete this statement.

Lines 494ff: Although I don’t doubt this capability, you only showed evidence for one Sahara dust layer, but not for any aerosol layer embedded in clouds. If such results exist, it would be good to include them, otherwise this statement should be revised.

Line 507f: What is meant by “aviation consulting processes”? Please clarify.

Line 508f: Similarly, what is meant by “a concept for validation based on measurement data”?

Line 62: The reference by Hartmann et al., 1996 does not mention any dropsonde releases. The Berichte zur Polarforschung Nr. 218 was published in 1997. If dropsondes were released during that project, a different reference would be helpful.

Table 1:

Table 1 lists uncertainties and reproducibilities. What is their significance level, i.e. one standard deviation (k=1) or two standard deviations (k=2) or something else?

The datasheet for the Bosch Sensortec pressure sensor lists its operating range as 300... 1250 hPa. Where does the measurement range to 1 hPa and the uncertainties at the lower pressures come from?

The measurement range for the wind speed is probably larger than 200 m/s and the sonde can probably sense the aircraft speed prior to launch.

Figures:

Figure 5: Where is the humidity sensor located and what is the airflow around it?

Figure 8, legend:  change “white circles” to “white dots”

Figure 11, legend: Please make sure that the unit µm is displayed properly (also in lines 438 and 440). Capitalize “Individual”.

Technical comments:

Lines 33-35: Rephrase to improve the clarity of that statement.

Line 48: A better reference to the NCAR dropsonde is:

   Hock, T. F., and J. L. Franklin, 1999: The NCAR GPS Dropwindsonde. Bull. Amer. Meteor. Soc., 80, 407–420, https://doi.org/10.1175/1520-0477(1999)080<0407:TNGD>2.0.CO;2.

Line 95: “Airborne” instead of “Advanced”

Line 135: Change “on” to “for”

Line 145: insert a hyphen into “system-control”

Line 189: replace “independently” with ““independent”

Line 319: Add “/” to “Upper Troposphere/Lower Stratosphere”

Line 338: At 200 m/s 1 to 2 min corresponds to 12 to 24 km. The distance given sounds a little short.

Line 339: What are typical dropsonde parachute cross sections?

Line 466: Replace “aöö” with “all”

Lines 355 and 470: Please specify the altitude for this fall rate. Or alternatively, specify the fall rate near the surface.

Line 484: Replace “exposition” with “exposure”

Table 1 lists a Tellit J-N3 GNSS module as well as a UBLOX MAX-MBC. I assume the text is correct and that the UBLOX MAX-M8C is used. Please correct Table 1.

General Comments

The paper describes a new dropsonde system – the KITsonde, which has novelties such as (1) multiple sounding profiles from a single launch, (2) a modular payload design thanks to a release-container concept and (3) reception from up to 30 channels simultaneously. The paper is well-written and coherent. The structure makes sense, starting with a background on dropsonde development, detailed description of the system itself, the different possible configurations and the tests done in the field. Overall, the measurement techniques described herein will improve the type of atmospheric soundings being taken currently as well as the aircraft strategies being employed. The KITsonde holds exciting potential for advancing atmospheric observations especially with the three aforementioned novelties.

However, I do have a few comments, which I wish the authors would clarify.

1. The current design is compatible with dispensers of RD-series dropsondes, which is great for integrating with existing aircraft systems. However, these dispensers are being phased out as NCAR and Vaisala transition to the NRD series (e.g., NRD-41). The newer, smaller sondes require research aircraft teams (incl. HALO) to modify or replace dispensers, often adopting the automated launcher systems. This is a crucial point because a smaller release container for the KITsonde would impact the first two novelties I list in the general comments. As the older dispensing systems are becoming obsolete, the KITsonde's compatibility with them is no longer an advantage. Could the authors address the KITsonde's compatibility with the new dispensers, either for HALO specifically or more broadly?
2. The variability in the parachute design is unclear to me and the paper will benefit from better description. For instance, phrases like “individually sized parachutes” (L97) and “different sizes of the parachutes” (L469) tell me that parachutes may vary in size to maintain separation in vertical space. If so, what are the possible sizes and configuration options? This information is critical for planning the launches because sounding strategies depend on the descent rate, which affects how closely the sounding trajectory approximates a vertical profile as well as how closely the sounding approximates to an “instantaneous” profile. For the KITsonde, in a 4-sonde release container, what parachute sizes are used and what are the expected descent rates for the 4 sondes, and is this configuration standard or could the user choose? Similarly, for a single-sonde release container (with or without other payloads), what is expected descent rate?
3. This third point relates slightly to the previous one in terms of descent rate. I am not completely convinced by the motivation for the satellite modem configuration. Currently (e.g. Ehrlich et al 2024, https://doi.org/10.5194/essd-2024-281) HALO dropsondes are sent to GTS in near real-time, i.e. as soon as the dropsonde makes a landing and therefore can also be sent to ground-support. Generally, the difference in real-time and near real-time is around 12-15 minutes and should not affect data assimilation too much. Of course this changes if the descent time is close to 45-60 minutes, which means that the aircraft telemetry could be out of range. But this advantage of the satellite telemetry then comes at the cost of a multi-sounding launch, which provides the novel spatio-temporal density I would argue is the best feature of the KITsonde. It does work as an example of different payload capabilities, but I struggle to find a practical use-case for it where it is advantageous over the conventional sondes, such as what the UCASS and radioactive payloads demonstrate. I would appreciate the authors’ clarification here.

Minor comments

Title: Why HALO specifically when the system has already been demonstrated with 2 other aircraft too?

L45-47 : Acronyms are not defined

L49 : Unclear why the horizontal spacing of 100 km between drops? Is there a reference for explanation?

L52 : NWS here is particularly the US National Weather Service as opposed to different countries’ weather services when defined in L44.

L81: What type of flexibility in operations?

L99-100: This is an excellent and very pragmatic advantage in favour of having the release container concept. :)

L157 : Is there a reference for 250 m/s? I believe 200 m/s might be closer to cruising speed during flight operations, but I am not familiar with all HALO payload configurations.

L194: Please mention the weights and CoG for the standard 4-sonde configuration?

L207: How many minimum satellite connections are needed for this?

L210: PTU 1.12 s (and wind 1 s)… Discrepancy with the abstract, where 1.2 s is mentioned for both.

L230: Humidity above 100% RH stated, but table shows range up to 100% RH

L232-233: It is unclear why the heated sensor is not used and why it is less relevant than to radiosondes. Could riming not be a case for when heated sensors prove to be useful?

L248: Are the interruptions often enough to justify an SD card buffer (thereby increasing waste and expense per sonde)?

L310-312: Have these comparisons been documented somewhere? Could references be provided?

L351: anomaly with respect to what?

L368: I believe “clear differences” is questionable phrasing except in the case of humidity.

L382: Shouldn't it be sonde 1c instead of 1b? For me, 1b looks like it went through the deepest cloud layer.

L403: meso-gamma… Is it per Orlanski (1975)? Please define spatial extent or provide suitable reference.

L423: Where is the “independent modelling” part in Figs 13 and 14?

L444: Is there a suitable reference for such coarse aerosol origins locally at 2 km altitude?

Figure 14: Why is there a 2-hour difference between the compared measurements? And why was the 3-5 km altitude window chosen?

L466: Typos in the first few words

At multiple places, figures are not numbered in the same sequence as their appearance in the text.