Following the sign correction the revised momentum budgets have a residual that is much smaller in magnitude and significantly smaller than that of the other forcing terms. This small remaining residual must be due to unaccounted-for contributions arising from differences between the momentum budget as we calculated it, see Outten et al. Eqs (1), (2) and (4), and the formulation used in the numerical model (the MetUM); for example, contributions from model diffusion. This was the explanation given for the residual in the original budgets (paragraph 3 on page 1945) and the discovery of a sign error does not make this incorrect. The large magnitude of the residual was a point of consternation in the original paper and was not fully explained but left to ‘further investigation’. It seems that further investigation is now not necessary.