Dijet photoproduction at HERA and the structure of the photon

TL;DR

This ZEUS study measures dijet photoproduction at HERA to probe the photon's parton densities and test perturbative QCD. By comparing high-precision dijet cross sections to NLO QCD predictions with different photon PDFs, the work shows good agreement in the high-x_gamma_obs and high-transverse-energy regime, while low-x_gamma_obs data provide strong constraints on the photon's parton content. The analysis demonstrates that current photon PDFs are not fully adequate to describe all features across the kinematic range, highlighting the photon structure as a critical area for refinement with higher-order or resummed calculations. Overall, the results reinforce the utility of jet photoproduction as a sensitive probe of photon structure and QCD dynamics, with implications for global PDF fits.

Abstract

The dijet cross section in photoproduction has been measured with the ZEUS detector at HERA using an integrated luminosity of 38.6 pb$^{-1}$. The events were required to have a virtuality of the incoming photon, $Q^2$, of less than 1 GeV$^2$ and a photon-proton centre-of-mass energy in the range $134 < W_{γp} < 277$ GeV. Each event contains at least two jets satisfying transverse-energy requirements of $E_{T}^{\rm jet1}>14$ GeV and $E_{T}^{\rm jet2}>11$ GeV and pseudorapidity requirements of $-1<η^{\rm jet1,2}<2.4$. The measurements are compared to next-to-leading-order QCD predictions. The data show particular sensitivity to the density of partons in the photon, allowing the validity of the current parameterisations to be tested.

Figures (13)

• Figure 1: Examples of (a) direct and (b) resolved dijet photoproduction diagrams in LO QCD.
• Figure 2: Comparison of data and MC simulation for (a) $x_{\gamma}^{\rm obs}$, (b) $y_{\rm JB}$, (c) $E_{T}^{\rm jet1}$, (d) $E_{T}^{\rm jet2}$, (e) $\eta^{\rm jet1}$ and (f) $\eta^{\rm jet2}$. The data are shown as points compared to Herwig (solid line) and Pythia (dashed line). Also shown is the LO component of direct photon processes in Herwig (hatched area). The simulated sample is normalised to the data and the fraction of direct and resolved photon processes combined according to a $\chi^2$-fit to the $x_{\gamma}^{\rm obs}$ distribution in (a).
• Figure 3: Measured cross sections as a function of $|\hbox{$\cos\theta^*$}|$ for (a) $x_{\gamma}^{\rm obs}$$<0.75$ and (b) $x_{\gamma}^{\rm obs}$$>0.75$, compared to NLO predictions. The data are shown with statistical errors (inner bars) and statistical and systematic uncertainties added in quadrature (outer bars). The uncertainty due to that of the jet energy scale is shown as the shaded band. The NLO prediction corrected for hadronisation effects is shown calculated using the GRV-HO and CTEQ5M1 PDFs for the photon and proton, respectively, and the scale set to $E_T/2$ (solid line). The hatched band represents the quadratic sum of the theoretical uncertainties as discussed in Section \ref{['sec:theo_uncert']}. The prediction using the AFG-HO photon PDF is also shown (dashed line). In (c) the cross sections are area-normalised and the data shown for $\hbox{$x_{\gamma}^{\rm obs}$} \ <0.75$ (solid points) and $\hbox{$x_{\gamma}^{\rm obs}$} \ >0.75$ (open circles).
• Figure 4: Measured cross section as a function of $E_{T}^{\rm jet1}$ for events with $\hbox{$x_{\gamma}^{\rm obs}$} \ > 0.75$. The measurement is divided into six regions of the pseudorapidities of the jets. The cross sections are multiplied by the scale factor indicated in brackets so that all regions can be displayed in the same figure. For further details, see the caption to Fig. \ref{['fig:cos']}.
• Figure 5: Ratio of cross sections to the central theoretical prediction as a function of $E_{T}^{\rm jet1}$ for events with $\hbox{$x_{\gamma}^{\rm obs}$} \ > 0.75$. The measurement is divided into six regions of the pseudorapidities of the jets. For further details, see the caption to Fig. \ref{['fig:cos']}.