NNLO predictions for dijet production in diffractive DIS

R. Žlebčík, D. Britzger, Jan Niehues, T. Gehrmann, James D. Currie, Alexander HussarXiv 2017 · arXiv

TL;DR

Researchers have achieved the first-ever Next-to-Next-to-Leading Order (NNLO) predictions for dijet production in diffractive deep-inelastic scattering (DIS). While these high-precision calculations significantly improve the description of the shape of kinematic distributions, they consistently overshoot the data in normalization, signaling an urgent need for a new generation of proton structure maps (DPDFs) calculated at the same level of accuracy.

What Problem Does This Paper Address?

In the world of subatomic physics, "diffractive" processes are a bit of a mystery. In these events, an electron hits a proton, but instead of the proton shattering completely, it remains intact (or nearly so), leaving a "large rapidity gap"—a void in the detector where no particles are found. This process is mediated by the exchange of a "Pomeron," an object with the quantum numbers of a vacuum.

To predict how often these events produce "dijets" (two sprays of particles), physicists use perturbative Quantum Chromodynamics (pQCD). Until now, calculations were stuck at Next-to-Leading Order (NLO). The problem? NLO predictions had massive theoretical uncertainties—sometimes as large as the measurements themselves—making it hard to tell if our models were actually correct. This paper pushes the boundary to NNLO to provide the precision needed to truly test our understanding of diffractive physics.

Methods

The authors utilized a sophisticated computational toolkit to tackle the "infrared divergences" that plague high-order QCD calculations:

  • NNLOJET: A program based on the antenna subtraction method, which allows for the correct handling of real and virtual particle emissions at NNLO accuracy.
  • fastNLO framework: Because NNLO calculations are incredibly "expensive" (requiring over 100,000 CPU hours), the team used fastNLO. This allowed them to calculate the computationally intense part of the interaction once and then quickly "convolute" it with different proton structure models and coupling constants.
  • Scale Choice: They tested various functional forms for the renormalization and factorization scales (like photon virtuality $Q^2$ and jet transverse momentum $p_T$) to see how they impacted the stability of the prediction.

Datasets

The study compared its new predictions against six distinct experimental measurements from the HERA collider (the world's only electron-proton synchrotron, which operated in Hamburg, Germany):

  • H1 Collaboration: Data from HERA I and HERA II periods, using different methods to identify the intact proton (FPS, VFPS, and LRG methods).
  • ZEUS Collaboration: Data using the Large Rapidity Gap (LRG) method. In total, they analyzed 49 single-differential and five double-differential distributions.

Key Insights and Ideas

  • The "Positive" Surprise: The NNLO corrections are large and positive, increasing the predicted cross sections by about 30% compared to NLO.
  • Shape vs. Normalization: This is the "aha moment." The NNLO calculations describe the shape of the data (how particles are distributed across different angles and energies) much better than NLO. However, the total number of predicted events is now significantly higher than what was measured.
  • The DPDF Bottleneck: The authors argue that this normalization mismatch isn't a failure of the theory, but a sign of "inconsistent ingredients." Currently, we only have Diffractive Parton Distribution Functions (DPDFs)—the maps of the Pomeron's internal structure—extracted at NLO. Using an NLO map with an NNLO calculation is like using a low-resolution map to navigate a high-precision drone; the results won't align perfectly.

Key Results

  • Reduced Uncertainty: NNLO predictions showed a marked reduction in scale dependence compared to NLO, meaning the theoretical "guesswork" is finally shrinking.
  • Better Kinematics: For variables like inelasticity ($y$) and photon virtuality ($Q^2$), the NNLO shape provides a superior description of the HERA data.
  • Quantifiable Mismatch: The NNLO predictions typically exceed the H1 and ZEUS data by roughly 30%.
  • Gluon Sensitivity: The results show that dijet production is highly sensitive to the gluon content of the Pomeron, confirming that dijet data must be included in future fits to accurately map out the "glue" that holds these diffractive exchanges together.

Limitations and Future Work

The primary limitation is the lack of NNLO DPDFs. Because the authors had to use NLO-level DPDFs as inputs, the normalization of their results is technically "mixed-order."

The clear path forward is a "consistent determination": the physics community needs to perform a global fit that extracts DPDFs from scratch using NNLO theory for both inclusive and dijet data. Once these high-resolution "maps" are available, the normalization discrepancy observed in this paper is expected to vanish, providing a definitive high-precision picture of diffractive scattering.

Summary updated

Abstract

Cross sections for inclusive dijet production in diffractive deep-inelastic scattering are calculated for the first time in next-to-next-to-leading order (NNLO) accuracy. These cross sections are compared to several HERA measurements published by the H1 and ZEUS collaborations. We computed the total cross sections, 49 single-differential and five double-differential distributions for six HERA measurements. The NNLO corrections are found to be large and positive. The normalization of the resulting predictions typically exceeds the data, while the kinematical shape of the data is described better at NNLO than at next-to-leading order (NLO). Our results use the currently available NLO diffractive parton distributions, and the discrepancy in normalization highlights the need for a consistent determination of these distributions at NNLO accuracy.
BibTeX
@misc{zlebcik2017nnlo,
  author = {R. Žlebčík and D. Britzger and Jan Niehues and T. Gehrmann and James D. Currie and Alexander Huss},
  title = {{NNLO predictions for dijet production in diffractive DIS}},
  year = {2017},
  eprint = {1710.03521},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1710.03521},
}
Figure