See supplement.

This is a nice study of the reprocessing of the MIAWARA dataset, with a particular focus on the correction required to measurements taken with the AC240.  It is certainly appropriate for publication in AMT, but some changes and clarifications are required.

Line 18 – The references given here are fine, but generally this sentence would mention balloon borne measurements.  They are mentioned later, but only for the lower stratosphere (FPH balloons).  The next paragraph starts out with “Most obviously, radiosonde balloons measure in the troposphere”.  Perhaps the authors are specifically referring to remote measurements here, but they do not say this.  A general reorganization of these introductory paragraphs would be helpful.

Line 39 – This statement is too strong.  Even with just 2 measurements per day (one during the day and one at night), MLS is able to provide some information regarding diurnal variations.

Line 53 – This sentence regarding pressure broadening is very awkward.  The point is that the effect of pressure broadening is to increase the spectral width of the emission with increasing pressure.

Line 69 – What is being described here is a Dicke switching scheme.  The use of the word “calibration” here is very confusing.  

The phrase “calibrated spectra” is used repeatedly.  In almost all cases, replacing this phrase with simply “spectra” would reduce confusion.

Line 180 – “This improvement stems from the 14-bit analog-to-digital converter (ADC), a notable step up from the AC240’s 8-bit ADC, as well as advancements in digital signal processing that reduce numerical errors.” – No evidence is presented that any of the difference shown are caused by the 8-bit nature of the ADC. 

Line 222 – With regards to problems with the central channels on the AC240 a reference to Gomez et al. 2012 (RS1010, doi:10.1029/2011RS004778) would be appropriate here.

Line 223- “This was potentially due to the relatively large noise levels in these calibrations due to the short integration times of several hours, compared to the operational integration time of one day.”  Perhaps I am wrong, however I think that this has nothing to do with the integration times, but with the fact that on the scale shown on Figure 3 one cannot detect ~5-10% differences in the ozone line.

Figure 5 – Since these 2 panels are being compared please use the same ranges and ticks for the x-axes.

To what extent is the difference in the high altitude sensitivity in this figure is caused by the fact that the central 2 channels in the AC-240 are not being used?  Or have I misunderstood something here?

The bump in the AVK’s near 0.01 hPa in Figure 5 is very strange.  It shows up in the AC-240 plot as well, albeit not as clearly because retrievals from that spectrometer are not very sensitive near that pressure.  Is there perhaps a change in the thickness of the retrieved layers at this level?  Absent this, or some other a physically plausible explanation, it is difficult to believe that this is not indicative of an error in the retrieval code.

Figure 6 – A nice plot.

Figure 9 – A single contour plot here would probably be sufficient.  The pattern is the same for both spectra. 

Figures 10 – Given that the only difference in these two comparisons is a scaling of the bias, it is very surprising that there is a difference in the scaled and unscaled STD at this level in this Figure.  Presumably this occurs because there are differences in the profiles or spectra being used in the comparison.  If this is the case please state this.   If this is not the case then please provide another explanation.

Line 336 – The mesospheric H2O observed in these measurements is unrelated to the direct injection into the mesosphere noted in Proud et al. (2022).  I recommend a reference to Nedoluha et al. (2024)  https://doi.org/10.1029/2024JD040907