This manuscript is appropriate for ACP, however I do have some concerns regarding several points.

The latitude ranges used are confusing.  According to the text, the zonal wind in Figure 1 is at 60N but the omega* anomalies and temperatures are 70-90N.  Wouldn’t it therefore be more sensible to show the zonal wind at 70N.  Can the authors can provide a reason why they did not do this?  

Ideally I would have thought that his study would be done using some type of coordinate relative to the vortex edge, but perhaps this is difficult in the mesosphere.  However, given that the authors have chosen 70N, it would be useful to have some indication of what fraction of the 70-90N area is inside-the-vortex (at least at levels where the vortex can be defined) at the time of the FSW and SSW events.  Perhaps the answer is “almost all of it”.  If this is the case please state this.

Line 194 – “The main benefit of the ground-based observations is the much higher temporal resolution of two hours, which permits to estimate of the sampling bias from the satellite MLS taking data only at two local times.”  There is no discussion anywhere else in the study suggesting that MLS sampling bias is a problem for this study, so please either delete this sentence or explain why it is relevant.  

Line 198 says “The results indicate a good agreement between MERRA-2 and MLS with GROMOS-C observations.”, yet in figure 3 – MERRA-2 ozone at altitudes above 0.1 hPa is clearly not in agreement with MLS and GROMOS data.  

Figure 4 – Given the MERRA-2 ozone values shown in Figure 3, it does not seem sensible to show these ozone anomalies from 0.1 to 0.01 hPa in Figure 4.

Line 360 - The authors claim an increased occurrence of SSW events during midwinter in the NH.  This is not shown or referenced anywhere else in the paper.  The statement regarding early FSW events is similarly problematic.

Line 395 – It is not clear what point this sentence is trying to make.  The claim that “ozone chemistry has become increasingly important in governing climate variability” certainly needs some justification that is not to be found here.

In this paper, ozone changes in the middle atmosphere around sudden stratospheric warming (SSW) and early final stratospheric warming (FSW) events are analyzed. MERRA-2 data are used to disentangle the different dynamical and chemical processes affecting ozone in the stratosphere and mesosphere. MERRA-2 data are evaluated against observations of GROMOS-C and MLS, and the good qualitative agreement in the ozone anomalies between GROMOS-C, MLS, and MERRA-2 below 0.1 hPa justify the use of MERRA-2 for the analysis of dynamical versus chemical processes. This is an interesting analysis, and the approach used to disentangle advection, turbulence, and chemistry is to my knowledge novel. In this sense, the paper brings new aspects to our understanding of these large disturbances of middle atmosphere dynamics; as SSWs are know to affect the whole atmosphere from the troposphere to the ionosphere, understanding its drivers and impacts is of course very important. However, I found the paper in parts difficult to follow. In particular how ozone is implemented in MERRA-2, and how it depends on the different chemical and dynamical terms is unclear (Section 2.3). As this is a prerequisite to understand the analysis and interpret the results, this should be clarified, and I have listed more specific comments to this issue as Major points below. I also have a rather long list of minor comments mostly regarding unclear wording listed below.

Major points:

Lines 111-112: in the stratosphere, odd oxygen is dominated by ozone, but in the upper mesosphere, it is dominated by atomic oxygen. This has to be taken into account when comparing MERRA-2 ozone to GROMOS-C and MLS, and presumably can explain the much higher values of MERRA-2 in the upper mesosphere.

Lines 111-114: I don’t understand the relationship between the different models here. As I understand your description, MERRA-2 uses tendencies of Ox (not ozone) from the GEOS CTM as ozone tendencies. This would explain the very high ozone values of ozone in the upper mesosphere, as Ox (and Ox tendencies) is significantly higher there, than ozone. But why are the assimilated meteorological data from GMAO used here, not from MERRA-2? I presume that you mean that these are used within the GEOS-CTM, not within MERRA-2, but this is not clear. Also, if the GMAO meteorological data are used to derive the ozone tendencies than used in MERRA-2, that would imply that the resulting ozone fields in MERRA-2 are inconsistent with dynamics of MERRA-2. I don’t think this is the case, but can you please clarify? 

Line 115: what is vertically integrated here – ozone, or the ozone tendency? Vertically integrated ozone would be total ozone; as you are discussing ozone profiles here, is it possible that you mean “vertically resolved”, not “vertically integrated”?

Line 117: shouldn’t the left-hand side be the total derivative, not a partial derivative?

Line 121: can you explain in a bit more detail how ANA refers to assimilation of ozone? Also, is this done within MERRA-2, or within the GEOS CTM?

Line 189 and Section 4: in my view, the value of the GROMOS-C (and MLS) data is that the evaluation of the performance of MERRA-2 ozone in relation to the FSW/SSW events; the good qualitative agreement is the justification to use MERRA-2 to analyze the dynamics versus chemistry in a later step. This should be made very clear here.

Line 198: If I understood correctly how ozone is implemented in MERRA-2 (description in Sec. 2.3), this is Ox, not O3; in the stratosphere and lowermost mesosphere, the difference is negligible, but in the upper mesosphere, there is significantly more Ox than O3 – this presumably explains the very high values of MERRA-2 “O3” shown in Figure 3 c) and f). If this is correct, MERRA-2 ozone can not be used above 0.1 hPa. Please clarify.

Lines 374 and following, Discussion and conclusions: Please be more precise which data you used, and what for. You used MERRA-2 temperatures and wind fields to identify SSWs. You probably did not use MLS wind fields for this as stated here, as MLS does not observe winds(?). You used MLS and GROMOS-C ozone to evaluate MERRA-2 ozone fields, and ozone anomalies, and justify its use for analyzing the dynamical and chemical drivers of the ozone changes.  Please clarify this here.

Minor points:

Lines 13-15: it is not clear what “this response pattern” refers to here. Maybe better “FSW events are associated with …”

Line 16: which chemistry-climate model? In the paper, model results are shown from MERRA-2 whose ozone product is based on a chemistry-transport model. No chemistry-climate model results are shown.

Line 15-17: this sentence should come after the sentence ending in line 10, to clarify where the results discussed in lines 7 and following come from. Else it is not clear where “the underlying dynamical and chemical mechanism” discussed in the sentence starting in line 10 come from.

Line 26: SSW events (plural)

Line 29: Observed FSW events …. depend (plural)

Line 31: atmospheric species, not atmosphere species

Lines 32-33: “…. ozone plays the most important role in the coupling between chemistry, radiation, and dynamical processes in the stratosphere and lower mesosphere. “ Ozone radiative heating and cooling peaks at the stratopause; in the upper mesosphere, heating by O2 becomes important as well.

Line 23-45: There are less studies of the impact of SSWs on mesospheric ozone, but there is some literature about this as well, e.g., Tweedy et al., JGR, 2023; Smith-Johnsen et al., JASTP, 2018. These should be discussed here as well.

Line 61: In addition, we show …

Line 61-63: this is stating the obvious – as total ozone is dominated by the amount of ozone in the lower stratosphere, anything affecting lower stratosphere ozone will have a correlating response in total ozone.

Line 75: using instead of leveraging

Line 80: and instead of which

Line 80-82: are you using the same retrieval and calibration version as in Fernandez et al (2015)?

Line 84: instrument, not instruments (?)

Line 88-89: please clarify what depends on the pressure here – the ozone profile, not the 240 GHz microwave band

Line 93: available instead of applicable

Line 135: X(dyn) consists of three terms, not four

Line 139: in this equation, one bracket is missing, probably at the end

Lines 164-165: I would say the westerly starts to weaken in 10-.1 hPa already a few days before the warming. During the warming, the wind reverses quickly to  easterly, and stays like this for at least 30 days below 1 hPa, but reverses back above that.

Line 165-166: I think you mean the westerly winds return after approximately 15 days? However, only above 1 hPa

Line 168: …. and cooling in the mesosphere > 0.1 hPa, which seems to onset a few days before the warming?

Line 168: In Fig. 1b) … please state here that now you are discussing FSWs, not SSWs. Note that the westerlies begin to weaken at lag -10 as well (similar to SSWs), and even reverse above 0.1 hPa before the event.

Line 170: at the stratopause. Temperatures in the lower stratosphere also increase strongly, but there is cooling in the upper mesosphere.

Line 173: the mean climatology of all years, or of all years without SSWs/FSWs? If SSW years are included in the climatology, that will diminish the anomalies somewhat.

Line 174: as shown in Fig. 2.

Line 181: in Fig. 2a?

Line 183: what does it mean that you have significant anomalies in w* extending to lag -30 before FSWs – those winters are significantly different to other winters much earlier?

Line 187-188: the lasting w* anomalies after the FSWs at and below 1 hPa are very small though

Line 196-197: in which altitude range?

Line 204: erase “with” before descending downward

Lines 205-206: … before the FSW onset, which is stronger than before onset of the SSW events.

Line 211: Climatological for all years, or for only those without SSW/FSW events? Please clarify.

Line 251-252: how does that explain the difference between 5 and 6?

Line 252-254: I agree that it is important to understand the interplay between dynamics and chemistry, which is particularly difficult during strong disturbances of the atmospheric dynamics like SSWs and FSWs. Still, I think the wording here “one of the keys to improving our understanding” is too strong. I would argue instead that SSWs are periods of known stratosphere-troposphere coupling, and that a better understanding of SSWs, and better representation in chemistry-climate models, therefore has the potential to improve medium-range weather forecasts during high-latitude winter.

Figure 7: I would say this figure shows essentially the same behavior as 3, though with less noise due to the better sampling; the figure and the discussion of it, are not really necessary, as they repeat things already discussed. In my view, the main use of Figures 3 and 4 is to justify the use of MERRA-2 data for the analysis of dynamics versus chemistry; it is not necessary to do this again.

Figure 8: here you use the absolute derivative for TOT, DYN and CHM. This is not consistent with the notation in equation 1, where all are given as partial derivatives. From the setup of equation 1, I think the correct use would be to denote TOT with total derivatives, DYN and CHM with partial derivatives; anyway this should be done in a consistent way throughout the manuscript.

Lines 332-334: the sentence is missing a verb.

Lines 336-337: as TCO is dominated by the lower stratosphere, changes in lower stratosphere ozone will map directly into TCO.

Line 341: 30-90°N, not 60-90°N.