Dear Drs. Lindgren, Sheshadri,

your answers to the (mostly minor) reviewer comments look fine, but I did want to follow up on a few points and also wanted to bring up some other points that I noticed. Please make appropriate revisions, after which this paper should be acceptable for publication in WCD.

1. heating perturbation reaching 200 hPa: I don't find the figures comparing H2 and T2 in the response particularly convincing as they show the vertical EP flux in linear scale, which emphasises the troposphere (btw, is this total EP flux or just wave 2? the latter would be the appropriate quantity I think). The structure of vertical EP flux around the tropopause is almost impossible to tell from this scaling and because the strength of EP flux there is at the lowest contour level it is difficult to tell how big the difference between the two runs really is there (from a relative measure perspective the difference could be order 1). It does seem to me, as the reviewers both suggested, that with the setup of the heating perturbations you are putting some of the wave forcing into the (lowermost) stratosphere (this part therefore does not have to propagate through the tropopause). There is also evidence for this in Fig. 6b, where one can see a local minimum of wave 2 upward EP flux near 300 hPa (60 N). It'd be good to mention and discuss the implications of this setup a bit more.

2. meaning of EEIs (reviewer 1, specific comment 2): should you perhaps refer to this as wave-wave interactions (WWIs)? At least to me (and perhaps also to reviewer 1) the term "eddy-eddy interactions" provokes the connotation of smaller scale phenomena such as within a spectrum of scales in turbulence (more appropriate for tropospheric dynamics). In the stratospheric context here the interaction is mainly between two waves as you point out, so WWI may be a bit more to the point.

More importantly about the removal of EEIs/WWIs: the fact that you only remove these interactions in the zonal direction, and not in the meridional, would be good to emphasise more. Right now this is only mentioned in passing (beginning of section 2) and then briefly near line 370. Otherwise it's surprising why you should get high wavenumber structures in the meridional direction (those barotropically unstable vorticity "ripples"). I do wonder to what extent these high meridional wavenumbers, which occur much less pronounced in the full simulations (and the real atmosphere), may lead to artefacts?

3. barotropic instability: have you also checked for symmetric/inertial instability? In any case, one question I had on this is that if you find that the necessary condition for barotropic instability is met, why do you not observe barotropically unstable waves develop? Or do you? Perhaps because the non-linear breakdown of the instability requires wave-wave interactions it doesn't end up kicking off? Please discuss somewhere in paper.

4. tuning of GCMs to produce realistic SSW frequency: in the case of WACCM the Charlton et al. paper doesn't mention anything about tuning in WACCM, only that its SSW frequency is too low. So unless you have a reference where the tuning in WACCM to produce realistic SSW frequency is explicitly mentioned you should remove this part of the statement and only refer to idealised modelling studies.