Given the scientific merit of this study lies in whether the proposed 'break' in the zonal mean wind trend is a real feature of the polar vortex evolution, I believe more rigorous statistical testing of its robustness is required before the manuscript can be considered for publication.

Please take into account the reviewers' suggestions as follows:

Reviewer 1, review 1
I suggest trying a bootstrap test as follows: resample with replacement from the 38 available years (giving say 1000 different 38-year composites). Within these composites then how often is there a zero (or near-zero) trend in late November? Is it more than 95% of the time?

Reviewer 1, review 2
I suggest that the authors test whether the trend, rather than the mean, is significantly lower in late November compared to early December.

Reviewer 2
Repeat the analysis taking into consideration the pre-satellite record in the JRA 55 reanalysis to increase sample size for the statistical testing. If the results are similar then this additional analysis need not be included in the manuscript but it should be noted in the text that this has been found. If the results are found to be different over this extended period, then the manuscript should account for this in the interpretation of the 'break' and its robustness.

Please delete the sentence on p.22 L7 "A global warming might delay the short break, but it is out of the scope of present study" as it is speculative.
p23 Two sample bootstrap test. The description of the test is rather short (the hypothesis is not clearly stated) and would benefit from editing by a native speaker.