Corrigendum to “Systematics of azimuthal anisotropy harmonics in proton–nucleus collisions at the LHC from the Color Glass Condensate” (Physics Letters B (2019) 788 (161–165), (S0370269318308438), (10.1016/j.physletb.2018.09.064))

In [1], we argued that the multiplicity dependence of momentum integrated two particle Fourier harmonics [Formula presented] observed at the LHC in p+Pb collisions can be explained, at least qualitatively, based on the power counting of the projectile nucleus color charge density, [Formula presented], in the dilute-dense limit of the Color Glass Condensate (CGC) effective field theory framework. We concluded that the dilute-dense CGC multiplicity scaling is [Formula presented] To go beyond the scaling arguments in Eqn. (1) and determine the magnitude of this signal, we considered the numerical framework introduced in [2]. Here the momentum integrated Fourier harmonics can then be defined as [Formula presented] where the even and odd harmonics are respectively given in terms of [Formula presented] and [Formula presented] The even and odd single inclusive distribution, [Formula presented], are given explicitly in Eqns. (2) and (3) of [1], and are of order [Formula presented] respectively. In [1] we argued that the momentum integrated even harmonics [Formula presented], and their multiplicity dependence, estimated in our numerical approach, are in agreement with p+Pb data from the ATLAS collaboration at the LHC. However, we discovered a numerical error in our calculations, whereby a factor of ħc was missing in the unit conversion for the momentum. Therefore, the momenta [Formula presented] in Eqns. (3) and (4) were instead a factor of ħc smaller than expected. While this does not change the power counting and multiplicity scaling of [Formula presented] as argued in [1], it does affect the magnitude of [Formula presented] estimated in the numerical calculations, presented in Fig. 3 of [1]. Revised results, with comparison to recent ATLAS results [3], are presented in Fig. 1. We find that the multiplicity independence of both [Formula presented], as argued in [1] based on the scaling arguments given in Eqn. (1), is unaffected. However the quantitative agreement for [Formula presented] is no longer evident. The results presented here used lattices with [Formula presented], [Formula presented]; as noted in [1], the convergence of [Formula presented] is especially slow, so while our studies indicate things have converged with this discretization, definitively ensuring the continuum limit is reached is challenging given current resources.