Overall Summary:

This paper is within the scope of AMT and should be published. There are some confusing sections, discussed further below, but overall the paper does a nice job of explaining a new technique that will improve our understanding of stratospheric circulation and add important measurements for long-term climate relevant gases. They prove that this technique works well and appears easy to add on to already existing infrastructure, and is cost-effective. There is nice agreement with aircraft measurements and any discrepancies are reasonably explained. I look forward to seeing more StratoCore-GC-ECD data in the future. Given that only one section of this paper needs restructuring, it should be accepted with minor revisions.

General comments:

The main section of the paper that needs improvement is section 2.3 and relevant figures. Section 2.3 was the most confusing of the entire paper and also critical to the verification of this new system. It was hard to follow the steps of the different experiments. In part this was a challenge because of the different language used (test, experiment, etc) along with the two tests (but maybe more?) to study three scientific objectives. This section is important and a restructure would be helpful for the reader to follow the logic. For example, line 160 mentions ‘both experiments’ and then line 192 mentions ‘another set of tests’. Was that related to the first two tests mentioned at the start of the section? If so, set doesn’t make sense since was only two tests. Overall, this section is content heavy and the reader needs a clearer path to follow the important discussion.  

Related to Section 2.3, it is really hard to understand what figures 4 and 5 are trying to show. How does figure 4 show us that there is no contamination? What does the CO₂ and N₂O tell us in 5a? This figure and/or the relevant text to it needs to be reconsidered so the reader doesn’t need to spend the majority of their time trying to understand it.

Section 3 and Figure 6:  Is there a reason why the parachute is red? What is the ‘lightweight material’? Is the fact the balloon is off-white important?

Line 268: “data not shown in figure” – Could the comparison between model and observed be done in a supplement if no space in the paper? It is referenced in the conclusion that there is good agreement between model and observations but there is nothing that directly shows that for the reader. The RMSE is given but more information would be nice.

Lines 275-289: Vertical references are given as hPa but the plots are shown as km which is challenging to compare the figure to the text. Perhaps add a second y axis to Figure 9 with hPa?

Technical Corrections:

Line 20: We then launched

Line 44: chlorofluorocarbonsmolecules

Line 114: add degree symbol to 38C

Line 255: “after the descent” – during the descent? After the initial descent? The phrasing as written implies once it is on the ground

Line 299: in Kansas

Line 301: Are you saying the StratoCore data can get higher than the ER-2 or that it covers more of the stratosphere than the ER-2?

Figure 9: Is it possible to outline the symbols to make them easier to see? Or make them larger like in Figure 10.

This manuscript describes a neat and low-cost new development in characterising the trace gas composition of the stratosphere at an expanded range of species, while at the same time maintaining a high vertical resolution. The methods are well presented and have been thoroughly tested. My main concern is not with this new methodology itself, but with the detection method. It has been known for decades that ECDs are prone to interferences as they are sensitive to the dozens of chlorinated, brominated, and iodinated trace species in the atmosphere. This becomes very much a problem when compressing species with such a large range of boiling points into a 2-minute chromatogram - not so much for the low-boiling end (here: SF6, N2O) and CFC-12, but very much for H-1211, CFC-11, and CFC-113. This is completely ignored here and no evidence is provided that interferences have been identified (e.g., CFC-114 & -114a are know to elute at similar times as H-1211) or quantified. The Hintsa et al. (2021) paper that is cited does not disuss this problem either - but then the method is essentially the same (GC-ECD-based) and the paper also does not show plots of the problematic species. This interference problem might not be so pronounced in the troposphere, where none of these gases are photolysed. However, this changes very much in the stratosphere, especially as their interferents are photolysed at different rates. I therefore urge the authors to at least discuss these serious limitations, and ideally assess their influence on the results as well. Further specific and mostly minor comments can be found below.

l9-10 2 x "atmosphere in the first sentence.

l31 Throughout the manuscript: Please consider a consistent ordering when citing multiple references, e.g. alphabetically or by publication year.

l36-50 This discussion is not well balanced as it ignores recent work on the lifetime of SF6 (e.g. Ray et al., JGR, 2017, Leedham Elvidge et al., ACP, 2018, and Loeffel et al., ACP, 2022), which provides strong evidence that this compound might not be completely inert and therefore not a direct age tracer.

l51 Tans, 2009 is not in the reference list.

l81-82 It is not clear what the "ten stratospheric measurements" mean. Are these ten samples from one flight or ten flights? Also, why not state the actual published number instead of saying "around"?

l82 No 2 L AirCores were used for the Laube et al., 2020 paper.

l90 Please clarify that the top 25 % and the stratospheric portion are not the same thing as tropopause pressures can vary substantially with season and location.

l94-101 Please shorten to avoid the repetition of the (very valid) point on the advantages of "high resolution".

l127 Looking at Figure 9, SF6 (as expected) does not vary by 50-100%.

l128-130 When comparing Figure 3 and Figure 9, it is apparent that the observed stratospheric mole fraction range extends well below the range for which the GC-ECD response behaviour was characterised. The effects on the analytical uncertainties of such low mole fractions, and the resulting limitations on constraining circulation changes should at least be discussed in a qualitative manner.

l138-140 At what pressure is that cylinder gas being pushed through? Would this induce extra mixing?

l142-147 Please quantify "carefully controlled". What flow rates are being used, what are the related uncertainties, and how does this translate into sample volume uncertainties?

l149-159 This is a very nice experiment. However, two questions that are not addressed (and which might provide limitations to the conclusions drawn) are 1) According to Karion et al., 2010 the AirCore also usually contains a magnesium prechlorate drier. Was that also tested? And 2) Were these test carried out at the temperatures that AirCores cool down to during actual balloon flights?

L181 If the chromatogram, as indicated in Figure 2, is 2 mins long, the size of each sample is 4-5 ml (l92), and about 250 ml of air are analyzed (l136), this gives a minimum time of 100 minutes for the analysis of the upper part of an AirCore, not including any flushing, backflushing or calibration standard measurement times. This seems to be inconsistent with this experiment only taking 1 hour, unless it was carried out at higher flow rates (which would make it less representative of an actual flight).

l264 It looks like like "between" is missing after "imbalance".

l264 It is not clear how this pressure imbalance was measured.

l283 Please indicate the approximate altitude of the "650 K isentrope" or, alternatively, add an explanation of why this coordinate was used here.

l285 These values only "agree well" qualitatively. Also, please provide a reference for the expected photolysis rate order.

l288 It is not made clear to the reader, exactly how the authors derived that "variability on scales of days to weeks" was captured here.

l289-290 It seems like a missed opportunity not to show the CO2 and CH4 results as well. Why would you fly two AirCores alongside each other and then only display the profiles from the new, but not those from the established method? This is especially apparent here, where the latter results are discussed, but the reader left in the dark on how well these "similar structures actually" agree with each other.

l313-317 It is not clear to me, how the authors arrived at this conclusion. Looking at Figure 9, the lowest SF6 mole fraction also appears to have been measured at Flight 2.