We present , a Hessian ensemble of next-to-leading-order error parton distribution functions (PDFs) in a charged pion, constructed to enable precision comparisons with upcoming experimental data and QCD predictions. estimates several experimental, theoretical, and methodological uncertainties on pion PDFs. It follows a statistical procedure, grounded in de Finetti’s theorem, to systematically determine the component of the epistemic uncertainty associated with the choice of PDF functional forms. This uncertainty is estimated by fitting the data with multiple functional forms generated in a versatile format based on polynomial universal approximators (Bézier curves). The most distinct fits are converted into a joint Monte Carlo replica ensemble and then combined into one Hessian ensemble through a major update of the methodology, expanded in this article to combine PDFs for arbitrary hadrons and partonic flavor representations. This study also goes beyond the previous pion fits by estimating, for the first time, the uncertainty from nuclear PDFs for the tungsten target in the E615 pion-nucleus Drell-Yan data—a third-party input in our analysis, also included in the expanded combination. The upgraded framework, thus, altogether provides a principled algorithm to propagate parametrization dependence and other types of the epistemic uncertainty into conveniently represented PDFs. Notably, modifies the key physics conclusions from the earlier, more constrained analyses of the pion’s gluon and sea-quark content. It facilitates studies of PDF-induced correlations and comparisons to lattice QCD. We summarize the methodology, structure of the PDFs, and experimental implications.
Kotz et al. (Tue,) studied this question.