The paper reports on the validation of ACE-FTS version 5.2 ozone profile data using ozonesondes. The underlying analysis is thorough and correct.

The main conclusions from the paper are that ACE-FTS ozone profile data agree with sonde data around 20 km altitude, are about 10% higher than sonde data at 30 km. At the extratropical tropopause, ACE-FTS ozone data tend to be 5 to 15%  higher than the sondes. Temporal ozone variations over the last 20 years are seen very similar by ACE-FTS and by the sondes. There are no significant drifts between the sonde and ACE-FTS time series. The paper is generally well written, although I find it somewhat lengthy and detailed. Nevertheless I think it is acceptable in its present form.

I have only two minor suggestions for changes: 

In Figure 10, I am wondering why the same data gaps appear for the ozone sondes and ACE-FTS. This is most notable at Irene where the gap goes from 2008 to 2013. Are really the same months missing in both data sets? I think this is probabably a plotting error and should be fixed.

In Figure 15, I find the green area misleading. In all the previous plots, the green area denoted the standard deviation range of the individual station results, whereas error bars denoted the standard error of the mean. Different from that, the green region in Fig. 15 denotes the standard error of the mean drift. For consistency with the previous plots, I suggest to change that and use error bars, instead of the green area.

This paper addresses the quality of ACE-FTS v5.2 ozone, and instrument drifts by comparing with correlative ozonesondes between upper troposphere/lower stratosphere (UTLS) and mid-stratosphere (~30 km). The overall scientific approaches and technical methods are sound. The research results are also useful to the community. There are, however, some issues/questions that need to be addressed or clarified before publication.

The following are specific comments and questions for this paper

1. Line 20. “The ACE-FTS ozone profiles exhibit in general high biases in the stratosphere, increasing with altitude up to ~10% at around 30 km, …”. The ACE-FTS ozone profile does not show systematic high bias in the stratosphere. The positive biases are mainly for altitudes above ~20 km. For altitudes below 20 km and a few kilometers above tropopause, it tends to show negative biases up to ~5 -10% (Figure 7). To make it clear, I suggest modifying the original text to “The ACE-FTS ozone profiles exhibit in general high biases in the stratosphere for altitudes above 20 km, increasing with altitude up to ~10% at around 30 km, …”, or other similar sentences.
2. Line 24, “…., and a small positive insignificant drift of 0 - 3 % dec^-1 in the tropics and southern mid-latitudes with overall uncertainties at 2 – 3 % dec^-1 (2σ level) in the low stratosphere”. The above sentence needs to be clarified. In Figure 15, it shows positive drifts of ~5% dec^-1 in the tropic and southern mid-latitude lower stratosphere (i.e. ~16 -19 km in the tropics, ~10-14 km in southern mid-latitude), with uncertainties of ~5-10% dec^-1. The results in Figure 15 are different from the above texts in line 24.
3. Line 121, “… and concluded that v4.1/4.2 showed slightly larger biases than v3.5/3.6 in the middle to upper stratosphere”. The ACE-FTS v4.1 and v3.6 ozone show the largest differences in the middle stratosphere (~30km), with positive biases of ~3% and 9% for v3.6 and v4.1 ozone, respectively (see Figure 2 in Sheese et al (2022)). The v4.1 ozone is not slightly larger (i.e. 6% larger) than v3.6 in the middle stratosphere. The above sentence needs to be revised or indicates up to ~6% differences between v4.1 and v3.6 ozone in the middle stratosphere.
4. Line 162, There is a typo in the Gaussian function, in the denominator square root of (2*pi*s) should be square root of (2*pi) *s  (i.e. “s” is outside of the square root). Another question is why s is set to be 1. The s value should correspond to the vertical resolution of ACE-FTS vertical resolution. Since the vertical resolution of ACE-FTS is ~3 – 4 km, the "s" value would be larger than 1 (see equation (2) in Sheese et al, 2017 (https://doi.org/10.1016/j.jqsrt.2016.06.026)). Using smaller s value would result in less smoothing ozonesond profiles (e.g. not comparable to ACE-FTS vertical resolution) and create larger differences (or artifacts) when comparing ACE-FTS ozone against ozonesondes in the regions with strong vertical gradient.  
5. Line 199, “Laeng et al. (2021)”. The published date is 2022 (i.e. Laeng et al. (2022)). Please double check the doi number of this paper (reference section in line 822). The doi number is “https://doi.org/10.5194/amt-15-2407-2022”
6. Line 454, “This is an interesting feature showing the differences appear to depend on the atmospheric temperature profiles”. Ozone in the UTLS is mainly controlled by dynamics. The larger differences between ACE-FTS and ozonesondes in the UTLS (or tropopause height) could result from larger dynamic variability and sampling biases such as using more relaxed coincident criteria. This can be seen in Figure 7, where differences between ACE-FTS and ozonesondes in the UTLS (and tropopause) are smaller than those in Figure 9 due to stringent coincident criteria. The ACE-FTS ozone retrieval in the UTLS regions could also be affected by aerosol or cloud. The above sentence (line 454) needs more clarification.  
7. Lines 601, “the aggregated mean instrument drifts show small insignificant drifts within ±1% dec^-1”. The altitude regions for ±1% dec^-1 drift needs to be specified. For example a few kilometer above the tropopause.