The referees, in particular the two who were more critical of the first submitted version of the paper, are all agreed that the paper has improved significantly in revision. Two now recommend publish as is, the third (Referee 2) requests further minor revision. I think that the comments of Referee 2 are very useful and therefore I request that you consider making further revision in response to them. I will consider the further revised version and your responses (which need only be brief) myself without consulting referees further and I expect at that stage to accept the paper for publication.

You have considered the remaining referee comments and made some changes in response and I am therefore pleased to accept your paper for publication in ACP.

I have accepted 'subject to technical corrections' to give you the chance to consider, and perhaps address, the two points I make below. You may also wish to make a final check of the paper to ensure that no minor typographical or similar errors have been missed.

l275: ‘Extratropical waves have vertical tilts in their temperature structures, and if the anomalies are large, they can effectively modulate tropopause height as explained next.’ — As noted by the referee, extratropical waves do not HAVE to have vertical tilts in their temperature structures (even if many waves do), and what you have written here might be interpreted (by a casual reader) as saying that waves affect tropopause height ONLY if they have vertical tilts in their temperature structures, which I don’t believe is true. So personally I would simply omit the reference to ‘vertical tilts’.

l710: ‘By tropopause-based zonal averaging of these
signals at certain latitude bands, we were able to quantify how much of the extratropical TIL at mid and polar latitudes is explained by the transient modulation of the tropopause region by the planetary and synoptic-scale waves.’ — my understanding is that you have constructed e.g. Fig 5b by removing what you define as the wave signal from all temperature/static stability profiles to generate a new set of ‘waveless’ profiles, then you apply tropopause-based averaging to those ‘waveless’ profiles. Your statement ‘tropopause-based zonal averaging of these signals’ does not seem to me to be a clear description of that, since you have just used ‘wave signal’ meaning what is removed, not what is retained. You may wish to consider changing this statement.