Nature 497, 478–481 (2013); doi:10.1038/nature12140

In this Letter, the QBO amplitudes were calculated incorrectly. The corrected values (consistent with the description of the methods in the paper) are all larger than those reported, generally by about 10–20%. The error has only a very small effect on the percentage trends of any of the time series examined, and the conclusions of the original paper are unaffected. The detailed claims throughout are unaffected, except that the significance threshold for the QBO amplitude trend reported for the MIROC-ESM-CHEM simulation during the 1953–2012 period should be P < 0.08 instead of P < 0.06 as reported originally. This error and Figs 1–3 and Supplementary Figs 3 and 9 have been corrected in the HTML and PDF versions of the original paper. We thank L. D. Oman for bringing this to our attention.