nNNPDF2.0: Quark Flavor Separation in Nuclei from LHC Data

TL;DR

This paper introduces nNNPDF2.0, a model-independent, machine learning-based determination of nuclear parton distribution functions using diverse LHC data, improving understanding of nuclear quark and antiquark modifications.

Contribution

The paper presents a novel nPDF extraction method incorporating new collider data and advanced statistical techniques, enhancing the accuracy and scope of nuclear parton distribution knowledge.

Findings

Good description of all datasets achieved

Quantification of nuclear modifications for individual quarks and antiquarks

Implications for strangeness and sum rules analyzed

Abstract

We present a model-independent determination of the nuclear parton distribution functions (nPDFs) using machine learning methods and Monte Carlo techniques based on the NNPDF framework. The neutral-current deep-inelastic nuclear structure functions used in our previous analysis, nNNPDF1.0, are complemented by inclusive and charm-tagged cross-sections from charged-current scattering. Furthermore, we include all available measurements of W and Z leptonic rapidity distributions in proton-lead collisions from ATLAS and CMS at $\sqrt{s}=5.02$ TeV and 8.16 TeV. The resulting nPDF determination, nNNPDF2.0, achieves a good description of all datasets. In addition to quantifying the nuclear modifications affecting individual quarks and antiquarks, we examine the implications for strangeness, assess the role that the momentum and valence sum rules play in nPDF extractions, and present predictions for representative phenomenological applications. Our results, made available via the LHAPDF library, highlight the potential of high-energy collider measurements to probe nuclear dynamics in a robust manner.