This study examines the generation and propagation of 10-day waves in the southern hemisphere upper atmosphere, using observations from meteor radar winds at two Antarctic stations, with supporting data from SD-WACCM simulations and MLS data. The main conclusion of this study is the identification of regions of barotropic/baroclinic instability in the upper mesosphere and lower thermosphere that can generate such oscillations.

I have several reservations about this study and at the moment I cannot recommend it for publication. I recommend the authors take a look at published works that compare meteor radar wind observations and model dynamics to become more familiar with state of the art. Start here: doi 10.1029/2021JA030177.

Major comments:

1. The main conclusion of the study is not new. The authors have ignored existing literature which offers corroborating or alternative investigations: doi 10.1002/grl.50373, 1016/j.asr.2022.10.054, 10.1029/2019JD031599 and 10.1016/j.jastp.2014.06.009, and many others published since Chandran’s study.
2. MERRA-2 is scientifically valid up to 60 km, the uppermost altitude where MLS data is used. Above 60 km and all the way to the lid (~75 km), it is just a sponge layer. Thus, I cannot associate any usefulness to the EXP75 presented in this study, where SD-WACCM is nudged with MERRA-2 fields all the way to MERRA-2 lid: it is like nudging a model with another model. Certainly, EXP60 is more useful but it is discussed only in terms of a difference from the EXP75 and briefly. If the goal of doing this study was to compare two simulations with higher and lower nudging fields, the authors should then be aware of these extant studies: doi 10.1002/2017JD027782, or 10.1002/2015GL065838.
3. MLS data is used in the data assimilation of MERRA-2. I find it odd to attempt to validate the model results that are nudged to MERRA-2 with a dataset that is part of the data assimilation cycle. In other words, an independent data set ought to be used. Why not SABER?
4. The authors identify the occurrence of a 10-day normal mode from the symmetric behavior of MLS geostrophic winds. That is not sufficient: the amplitude structure has nodes, and the phase is expected to be in a specific configuration in order to be 10-day free oscillation. See: DOI: 10.1175/JAS-D-11-0103.1, 10.1111/j.1600-0870.2007.00257.x, and the Salby’s papers already cited.

Minor comments:

Line 38-41: Normal modes or traveling PWs? If the former, normal modes by definition do not exchange momentum or energy with the background zonal flow.

Figure 1. I don’t understand how the averaged field (bottom panel) shows an overall minimum in April when 2015 has a pronounced amplitude peak. Are data in the shaded regions removed before averaging?

Comments on "Quasi-10-day wave activity in the southern high-latitude MLT region and its relation to the large-scale instability and gravity wave drag" by Lee et al.

Reviewed by Yosuke Yamazaki, Leibniz Institut of Atmospheric Physics, University of Rostock

This study focuses on the westward-propagating quasi-10-day wave (Q10DW) in the mesosphere and lower thermosphere of the Southern Hemisphere as observed by meteor radars and simulated by the WACCM model. Two WACCM simulations were used, in one of which the model was constrained by the MERRA-2 reanalysis from the surface up to 75 km (EXP75), and in the other simulation the model was constrained similarly but up to 60 km (EXP60). After showing qualitative agreement between the Q10DWs in the meteor radar observations and EXP75 simulation, the authors examined the cause of Q10DW for selected events, and demonstrated that the barotropic/baroclinic instability played a leading role. The authors also compared the Q10DWs in EXP75 and EXP60, and noted that the wave amplitude is generally greater in EXP60. They concluded that the difference resulted from the background atmosphere that is largely controlled by gravity-wave parameterization.

The new findings are the sensitivity of Q10DW to the nudging range adopted in model simulations and the importance of gravity-wave drag in it. I do not have an objection for this paper to be published in Atmospheric Chemistry and Physics. However, I feel that the paper could benefit from some revisions. My specific comments can be found below.

Major comments:

1. Model-data comparison

Figures 1 and 2 show the Q10DW from meteor radars and WACCM EXP75 simulation, respectively. The disagreement is profound. The patterns of intra-annual variations are different. Also, the wave amplitudes are different by a factor of 3 or so. This makes me wonder whether the radar analysis technique has been properly applied. If yes, and if the discrepancy between the data and simulation exceeds the uncertainties in the radar results, are the simulation results still a good representation of Q10DW? How do the WACCM results compare with the MLS results? It is easier to derive Q10DW using satellite data than radar data. If there is good agreement between the WACCM and MLS results, it would make more sense to present the MLS results instead of the radar results.

2. EXP60 vs EXP75

It is stated that the amplitude of Q10DW in EXP60 is excessive (in Abstract and Summary). However, I do not see evidence to support this statement. Looking at Figures 1, 2 and S2, the amplitude of Q10DW is too small in both EXP75 and EXP60. It is difficult to say which reproduces the observations (Figure 1) better. Again, comparisons with the MLS results might help.

3. Introduction

The paragraph starting at l. 65 highlights recent studies on Q10DW. It feels that there are many studies that the authors overlooked especially in the context of the Q10DW response to sudden stratospheric warmings (e.g., Matthias et al., 2012; Yamazaki & Matthias, 2019; He et al., 2020; Wang et al., 2021; Qin et al., 2022; Yin et al., 2023). It is important to address them, because some of them performed a similar analysis as the authors did in this paper and identified the importance of the barotropic/baroclinic instability.

Matthias et al. (2012) https://doi.org/10.1016/j.jastp.2012.04.004

He et al. (2020) https://doi.org/10.1029/2020GL091453

Qin et al. (2022) https://doi.org/10.1029/2022JD037730

Yin et al. (2023) https://doi.org/10.1016/j.asr.2022.10.054

--- only those which are currently not cited ---

Minor comments:

4. l. 62 "modulate the periods of tides"

Periods of tides would not change, as tides are defined by their periods (24h, 12h, 8h, etc.) The amplitude of tidal waves can undergo an apparent modulation due to the presence of secondary waves arising from the nonlinear interaction between planetary waves and tides.

Recommended reading

Miyoshi & Yamazaki (2020) https://doi.org/10.1029/2020JA028283

5. l. 72 "high-latitude mesosphere and lower thermosphere"

How did they find large Q10DW in the high-latitude region while their data are limited within +/-50 latitude?

6. l. 94 "the amplification mechanism of Q10DW-W1 still has not been investigated"

It may be clarified that this is about the seasonal amplification during equinoxes. As mentioned earlier, the amplification mechanism of Q10DW during sudden stratospheric warmings has been addressed in previous studies.

7. Figure 1

It would be informative if the geographical coordinates of the radars are mentioned in the figure caption.

8. l. 392 "F/sgn(A)" (also at l. 473)

What is "sgn"?

9. l. 427 "6 April 2015 case"

Unlike the other events, the wave propagation is poleward in the stratosphere. This is a Q10DW event during a final warming event (Yamazaki & Matthias, 2019; Qin et al., 2022). Qin et al. (2022) discussed that the inter-hemispheric propagation of the wave from the Northern Hemisphere was possible because of the phase of the quasi-biennial oscillation.

10. l. 432 "divergent"

It should be "divergence".

11. l. 463 "s>=20"

It is strange that only GWs with high zonal wavenumbers are resolved. Maybe "s<=20"?

12. l. 483 "February and November"

Which year?

13. l. 484 "February and November are chosen because ..."

The justification is weak. It would make more sense to select the events that are examined in Figure 4, where the Q10DWs are not only substantial in amplitude but also in qualitative agreement with observations.

14. Summary

I do not see the merit of listing items here (#1-#6). #2 and #5 are not the description of results but the description of what the authors did to get the results.