Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Uncertainties on parton distribution functions from the ZEUS NLO QCD fit to data on deep inelastic scattering

A. M. Cooper-Sarkar

TL;DR

The paper tackles the problem of quantifying experimental uncertainties on proton parton distribution functions (PDFs) and the strong coupling $α_s(M_Z^2)$ from inclusive deep inelastic scattering data. It adopts a global NLO QCD fit using DGLAP evolution to ZEUS $e^+p$ DIS and fixed-target data, carefully incorporating 71 correlated systematic sources and exploring two statistical treatments—offset and Hessian—to propagate systematics into PDF and $α_s$ uncertainties. The results show that the offset method provides robust, conservative error estimates, with $α_s(M_Z^2)=0.1166 \\pm 0.0008\,(uncorr) \\pm 0.0032\,(corr) \\pm 0.0036\,(norm)$ when $α_s$ is floated, and PDFs for the sea and gluon are consistent with MRST2001 and CTEQ6; model uncertainties are relatively small. The work underscores the importance of full correlated-systematic treatment in global DIS analyses and demonstrates that ZEUS data significantly constrain the gluon and sea distributions within the leading-twist NLO QCD framework.

Abstract

An NLO QCD analysis of the ZEUS data on $e^+ p$ deep inelastic scattering together with fixed-target data has been performed from which the gluon and quark densities of the proton and the value of the strong coupling parameter, $α_s(M_Z^2)$, have been extracted. The study includes a full treatment of the experimental systematic uncertainties, including point-to-point correlations. Different ways of incorporating correlated systematic uncertainties into the fit are discussed and compared.

Uncertainties on parton distribution functions from the ZEUS NLO QCD fit to data on deep inelastic scattering

TL;DR

The paper tackles the problem of quantifying experimental uncertainties on proton parton distribution functions (PDFs) and the strong coupling from inclusive deep inelastic scattering data. It adopts a global NLO QCD fit using DGLAP evolution to ZEUS DIS and fixed-target data, carefully incorporating 71 correlated systematic sources and exploring two statistical treatments—offset and Hessian—to propagate systematics into PDF and uncertainties. The results show that the offset method provides robust, conservative error estimates, with when is floated, and PDFs for the sea and gluon are consistent with MRST2001 and CTEQ6; model uncertainties are relatively small. The work underscores the importance of full correlated-systematic treatment in global DIS analyses and demonstrates that ZEUS data significantly constrain the gluon and sea distributions within the leading-twist NLO QCD framework.

Abstract

An NLO QCD analysis of the ZEUS data on deep inelastic scattering together with fixed-target data has been performed from which the gluon and quark densities of the proton and the value of the strong coupling parameter, , have been extracted. The study includes a full treatment of the experimental systematic uncertainties, including point-to-point correlations. Different ways of incorporating correlated systematic uncertainties into the fit are discussed and compared.

Paper Structure

This paper contains 11 sections, 12 equations, 1 figure, 1 table.

Table of Contents

  1. INTRODUCTION
  2. Description of NLOQCD fit
  3. Definition of $\chi^2$: treatment of correlated systematic uncertainties
  4. Offset methods
  5. Hessian methods
  6. Fit results: experimental and model uncertainties
  7. Experimental undertainties: offset method
  8. Model uncertainties
  9. Comparison of Offset and Hessian methods
  10. Parameter evaluation and Hypothesis Testing
  11. CONCLUSION

Figures (1)

  • Figure 1: Comparison of the gluon and sea distributions from the ZEUS-S NLO QCD fit for various $Q^2$ values. In this figure, the cross-hatched error bands show the statistical and uncorrelated systematic uncertainty, the grey error bands show the total experimental uncertainty including correlated systematic uncertainties (both evaluated from the standard fit with $\alpha_s(M_Z^2)=0.118$) and the hatched error bands show the additional uncertainty coming from variation of the strong coupling constant $\alpha_s(M_Z^2)$.