New constraints for nuclear parton distribution functions from hadron–nucleus collision processes

This work studies collinearly factorizable nuclear parton distribution functions (nPDFs) in perturbative Quantum Chromodynamics (QCD) at next-to-leading order in the light of hadron-nucleus collision data which have not been included in nPDF analyses previously. The aim is at setting new constraints on the nuclear modifications of the gluon distribution and on the flavour separation of quark nuclear modifications. The introductory part provides an outline of the theoretical framework of QCD collinear factorization and the used statistical methods and relates the work presented here to other similar contemporary analyses. As a result, a new set of nPDFs, EPPS16, is presented, including for the first time electroweak-boson and dijet production data from CERN-LHC proton-lead collisions and allowing a full flavour separation in the fit. The flavour separation is constrained with Drell-Yan dilepton-production data from fixed target pion-nucleus experiments and neutrino-nucleus deep-inelastic scattering data, which are shown to give evidence for the similarity of the u and d valence-quark nuclear modifications. For studying the gluon degrees of freedom, collider data are essential and in the EPPS16 analysis new constraints are derived from the dijet production at the LHC. Possible further constraints for the gluons are investigated in terms of the LHC data on nuclear modification ratios of dijet and D-meson production. Using a non-quadratically improved Hessian reweighting method, these measurements are found to put stringent constraints on the gluon modifications in the lead nucleus, reaching smaller values of the nucleon momentum fraction than previously accessible. A study on the future prospects of constraining nPDFs within a multi-observable approach with the BNL-RHIC is also given.