Erratum: A complete survey of texture zeros in the lepton mass matrices
Patrick Otto Ludl, Walter Grimus · Journal of High Energy Physics · 2014
Corrections in the numerical analysis: there was an error in the algorithm we had used to remove phases in the mass matrices by means of weak-basis transformations.As a result, in some cases too many phases had been removed.Furthermore, there was an error in the numerical computation of the mixing matrix, which affected the analysis of some of the models.We have corrected both errors in our programs and have repeated all numerical computations.The new results can be found in tables 9 to 13 which replace the corresponding tables in the paper.Consequences of the new analysis for the conclusions: though the set of textures compatible with experiment and the set of maximally restrictive textures have changed, the effect on the conclusions is small -see tables 9 to 13.In the case of Dirac neutrinos, there is only one relevant change, namely in the predictive power of the maximally restrictive textures with respect to the Dirac phase δ.Originally, since in general too many phases had been removed from the mass matrices, δ was forced to be exactly 0 or π.The correct rephasing leads to δ min ≈ 0 and δ max ≈ 2π for all but one of the textures -see the captions of tables 10 and 11.In the case of Majorana neutrinos, there is one notable change, namely in the predictivity of the texture (M ℓ , M L ) ∼ (6) with an inverted neutrino mass spectrum.The corrected analysis gives ∆(sin 2 θ 23 ) = 53.8< 100 and this case is thus predictive with respect to sin 2 θ 23 according to the criterion of equation (5.6) in the paper.However, tightening this criterion to ∆(O i ) ≤ 25, i.e. from 10σ to 5σ, there is still no texture which can predict any of the charged-lepton masses or the neutrino oscillation parameters.