LED spectrum stability: LED emission spectrum can be very sensitive to its temperature and as you described, the LED here is temperature controlled at 25 degrees. Can you show that this is also the case at high altitudes with environment temperatures well below zero? presumably, that would make life easier for the TEC unless the LED is not generating enough heat. 

optical alignment: I can see how fine alignment of the cell is not required. However, the device can experience a significant temperature gradient from ground level to a height of several km. In the altitude profile you presented this gradient can easily reach 40-50 degrees. With thermal expansion, especially when part of the device is kept at 25 C, how confident are you that the device is not too misaligned? can you provide data similar to Figure 4 from an actual flight? with temperature profile and LED temperature if you can.

section 3.2 is confusing to me. presumably, you are referring to the molecular absorption cross-section which should have cm2 or cm2 / molecule units. you used molecules / cm2. For water vapour, this should be orders of magnitude lower.

in the caption of Figure 8, you refer to the device as a "Sonde" although it was not flown here.

alongside a comparison to a state-of-the-art device at ground level I would like to see a comparison while performing under the intended instrument usage - i.e. during flight, under severe temperature and pressure gradients, and atmosphere composition changes. 

under Acknowledgements, you refer to drone flights. did I miss that in the main text? I didn't see a reference to any drone flights. 

I would like to see a discussion about the possible photolysis of NO2 at 405 nm considering its high quantum yield, or at least an explanation of why it is perhaps insignificant in this case. 

Review of “A Portable Nitrogen Dioxide Instrument Using Cavity-Enhanced Absorption Spectroscopy” by Bailey et al.

Bailey et al. present a description of a novel instrument for measuring NO2 in a small footprint for airborne or portable detection at high sensitivity. I recommend publication following the correction of the minor corrections listed below.

Line 16: 3 balloon and 1 UAV flight are listed in the abstract. At the end of the Introduction (line 56) it is stated that one flight is used. Please standardize across the paper.

Line 26: Check reference style for Cersosima et al. as shown in paper text with first initial.

Line 34: remove comma between optical and absorption

Lines 33-35: Check formatting of parentheticals with references. Also, list is missing any reference to electrochemical sensors.

Line 154: aerosol can also attenuate the light in the cavity, not just make the mirrors dirty.

Line 178: I₀ font consistency (should be italicized).

Line 221: Avoid personal pronouns. Replace “our pump” with “the pump”, same on line 220.

The authors describe a new lightweight instrument for measuring NO₂ from balloons or drones, using the IBBCEAS technique. They achieve 94 ppt precision in 1 second of averaging time, and demonstrate a test flight and comparison with an established NO2 instrument.

I think this paper is suitable for publication, but the authors should address the following concerns first:

• The authors compare PCAND to a recent lightweight LIF instrument, but should also compare its performance to other drone-based IBBCEAS NO2 instruments such as Zheng et al., 2024 and Womack et al., 2022.
• The sample cell is made of aluminum alloy, but the authors later say that FEP tubing was used in all plumbing, presumably to reduce NO2 losses. Have the authors tested losses of NO₂ on the aluminum alloy surface? Similarly, did the authors test for NO₂ losses on the Nafion dryer? And is the charcoal filter expected to completely scrub out the NO2 or will there be a small fraction remaining?
• There are inconsistencies in how the mirror reflectivity is reported. Line 68 says >99.9%, the Figure 1 says >99.98%, and Line 91 says 99.97%, which correspond to significantly different values when converted to effective pathlengths.
• The authors should discuss the 3 second flush time in the context of the speed of the drone or balloon. What kind of vertical resolution will be expected with this smearing? Is it sufficient for atmospheric chemistry studies?
• How frequently is the effective pathlength measured? Even if the cavity alignment is stable over months, is there any concern that mirror cleanliness will degrade faster than that?
• Section 4.1 is somewhat confusingly written. It’s not really clear how these two methods are derived from the equations. Has this method been used before?
• Additionally, I would recommend moving equation 8 to section 4.1, because it doesn’t follow from equation 7, but is rather derived in section 4.1
• Line 196: How are the “known” NO₂ concentrations provided? More detail is needed here.
• Line 201: More details should be included about how leaks and contamination could affect the data. How would they affect the data? Are leaks independently checked for?
• Figure 7: There are quite a few data points in this vertical profile with values close to -1 ppb. However, the reported uncertainty in the laboratory is 0.1 ppb. Do the authors expect that the precision degrades at higher altitudes? If not, how do they explain these negative values?
• Additionally, the profile shows NO₂ concentrations of >5 ppb at 7 km, which is unusually high. Did this occur in all the profiles? Was it possible the flight was affected by lofted biomass burning plumes? The authors should discuss this in detail, as accuracy at high altitudes will be critical if this instrument is to be used on balloon platforms.

References:

Zheng et al, 2024: https://doi.org/10.1016/j.atmosenv.2024.120361

Womack et al, 2022: https://doi.org/10.5194/amt-15-6643-2022