This study uses ERA5 data and a multi-model ensemble of APARC QBOi models to investigate how QBO teleconnections are modulated by ENSO.  To separate the QBO and ENSO signals, simulations were conducted with annually-repeating prescribed SSTs corresponding to idealized El Niño or La Niña conditions.   Models are unable to represent the observed (ERA5) enhanced Holton-Tan effect during La Nina, where QBO W favors a stronger NH winter polar vortex.  Models are also unable to represent the observed increase in SSWs during El Nino.  Overall, the polar vortex responses to the QBO are much weaker than to ENSO in the models.   In addition, the equatorward shift of the boreal winter Pacific subtropical jet (APJ) observed during QBO W in not seen in the models.   In the tropics, the model experiments do not show a robust

or coherent QBO influence on precipitation.   It was further found that QBO effects on the Walker circulation exhibit a complex dependence on season, longitude, and phase of ENSO.  They that suggested that weakness of the QBO polar vortex coupling in the models might arise from systematically weak QBO amplitudes at lower levels in the equatorial stratosphere, polar vortex biases in winter, and inadequate representation of stratospheric-troposphere coupling, while an inadequate representation of QBO effects in the tropical troposphere might arise from the systematically weak QBO amplitudes at lower levels, precipitation bias, and inadequate representation of the Walker circulation in these models.   This paper documents the results of a considerable effort in the QBOi community,  with well-organization presentation and choice of figures.   The narrative provides an authoritative interpretation of the detail and status of observed and modeled QBO/ENSO influences on the extratropics.   I recommend publishing with minor revision.

1. Idealized time mean La Nina and El Nino states. Would the model results be noticeably different for a time-varying ENSO (then binned by ENSO phase), versus two perpetual ENSO phases?   It seems possible that the two-state method represents an upper bound on possible effects.

1. l216-217, Fig. 13: This is a kind of discretized time-height section. It is similar to Reed et al.’s original 1961 figure which shows a time-height section of zonal wind.  The Hovmoller diagram was originally defined to be the variation of geopotential height or another quantity near 60N as a function of longitude and time.   It was generalized to mean a longitude-time diagram, which is usually used to indicate wave propagation.  You have a table with dependence on season and altitude and you are not discussing wave propagation in longitude.  Please use the phrase “season-altitude variation” instead of Hovmoller diagram to indicate what you are showing.

1. l221-226: “when the QBO phase is not defined by the preferred 70 hPa level” – does this mean that there are other ways to define it or that sometimes the 70 hPa level index isn’t well defined?  In this discussion of how multiple indices affect significance calculations, please give a sense of the meaning and outcome.  For example, If you use more than one index definition at different levels, perhaps one might ascribe reduced significance to a result, but in your method it appears that alpha is reduced, therefore implying greater significance.  A little more information would be helpful for understanding this paragraph.

1. Fig.1: It looks like only ECCAM5, WACCM  and MRI are reasonably correct for neutral ENSO, but none get El Nino right.  Maybe ECCAM5 and MRI get LaNina right (relative to ERA5).

1. Fig. 2: Only MRI seems to represent the basic sense of the ERA5 signal.

1. Fig. 4 caption: suggest adding information to the effect of “La Nina, CTL, and El Nino, from left to right”, to orient the reader about the order of the triplets, and maybe move to near the beginning of the caption.

1. l356: suggest refer to (Fig. 4c). In this paragraph, and at times elsewhere, it might be beneficial to include more in-text references to figure panels being discussed.

8.  l387,  150W-150E: How sensitive are results in Figs. 6 and 7 to the choice of longitude band?

review of "QBOi El Niño Southern Oscillation experiments : Teleconnections of the QBO" by Naoe et al 

This study aims to examine how QBO teleconnections are modulated by ENSO using a multi-model ensemble of QBOi models. The specific simulations examined are simulations in which SSTs are either climatological, El Nino, or La Nina, which allows for examining potential nonlinearities between QBO teleconnections and ENSO teleconnections. The use of ~10 models allows for assessment of model sensitivity and robustness. The authors examine four different QBO teleconnections  - polar vortex response, subtropical jet, tropical precip, and Walker Cell. They conclude that the QBOi models generally fail to simulate the first three of these teleconnections, and hence it is difficult to conclude anything as to the possibility of ENSO and QBO teleconnections interacting. They find a robust effect of the QBO on the Walker Cell, however, the specifics of the QBO phase and season with maximum impact differ across the models.  

While the paper should eventually be publishable in WCD, major revisions are needed first.

major comments:

1. For the first three teleconnections where the models generally fail in the multi-model mean, there are still several models which are relatively more successful in capturing the observed response. There is no discussion  of  why there is spread across models for two of these teleconnections (vortex response and subtropical jet), while there is a very limited discussion of the third (namely Figure 10). The paper should include a detailed discussion for all three teleconnections as to possible causes of the intermodel spread in how well the models are doing. 

This could be similar to Figure 10, but instead of T100hPa, the authors could consider horizontal or vertical resolution, the mean state of the vortex or subtropical jet position, meridional width of the simulated QBO, strength of the QBO in each model in the lowermost stratosphere, strength of the QBO in the mid-stratosphere, etc. All of these factors could plausibly be linked to why some models are better than others, and by exploring all of them the paper might be able to provide some insights to model developers as to what needs to be improved.

On the topic of Figure 10, what is the correlation and slope of the best-fit line? Is the relationship statistically significant?

2. For the QBO signal in reanalysis, do you try to regress out a lingering signal of ENSO before plotting wqbo minus eqbo? Line 476-479 seems to indicate you don't do this, and it isn't clear whether this is done for the other teleconnections either.  If this is not done, then comparing the observed signal to the model signal isn't a fair comparison as there will still be a residual signal from SSTs.

3. For the fourth teleconnection examined, the Walker Cell one, the authors adopt a completely different methodology than for the first three. Why for this section only do you play with the season and pressure level, but for earlier sections you don't? For the first three the models did a poor job, and now for this teleconnection they appear to be doing ok. Is this success for the Walker Cell just because you are giving the models lots of opportunities to succeed? Why not use this methodology for earlier sections too? Either way, the fact that a single paper is using very different methodological approaches for different sections is confusing, and leads to the (in my opinion misleading) impression that the models are much better at the QBO-> Walker Cell connection than the others.

To be specific, previous work which allowed for different vertical levels to define the QBO can lead to very different conclusions as to whether models capture the HT effect of the polar vortex. See Rao et al 2020a. It could also be that the seasonality of the HT effect differs from one model to the next. It would be interesting to see if the QBOi models still struggle to represent the HT effect if the authors adopted Rao et al's methodology.

minor comments:

1. Somewhere in the paragraph from lines 109 to 116, and also near line 124,  Rao et al 2020b should be cited and discussed

2. Line 135: Trascasa-Castro et al 2019 and Weinberger et al 2019 should be cited and discussed

3. line 142-146: Ma et al should be cited and discussed

4. line 199: Pahlavan et al should be cited.

technical edits aren't included in this round, but will be provided after the major comments are addressed.

Ma, T., W. Chen, X. An, C.I. Garfinkel, and Q. Cai, Nonlinear effects of the stratospheric Quasi-Biennial Oscillation and ENSO on the North Atlantic winter atmospheric circulation, JGR, https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2023JD039537

Pahlavan, H. A., Fu, Q., Wallace, J. M., & Kiladis, G. N. (2021). Revisiting the quasi-biennial oscillation as seen in ERA5. Part I: Description and momentum budget. Journal of the Atmospheric Sciences, 78(3), 673-691.

Weinberger, I., Garfinkel, C. I., White, I. P., & Oman, L. D. (2019). The salience of nonlinearities in the boreal winter response to ENSO: Arctic stratosphere and Europe. Climate dynamics, 53, 4591-4610.

Trascasa-Castro, Paloma, Amanda C. Maycock, Yu Yeung Scott Yiu, and Jennifer K. Fletcher. "On the linearity of the stratospheric and Euro-Atlantic sector response to ENSO." Journal of Climate 32, no. 19 (2019): 6607-6626.

Rao, Jian, Chaim I. Garfinkel, and Ian P. White. "Impact of the quasi-biennial oscillation on the northern winter stratospheric polar vortex in CMIP5/6 models." Journal of Climate 33, no. 11 (2020a): 4787-4813.

Rao, J., Garfinkel, C. I., & White, I. P. (2020b). How does the quasi-biennial oscillation affect the boreal winter tropospheric circulation in CMIP5/6 models?. Journal of Climate, 33(20), 8975-8996.