QCD corrections to electroweak l nu_l jj and l^+ l^- jj production

TL;DR

The paper computes next-to-leading order QCD corrections to electroweak Wjj and Zjj production in vector-boson fusion at the LHC, aimed at refining backgrounds to Higgs searches. Using a fully flexible parton-level Monte Carlo and Catani-Seymour subtraction, the authors separate vertex and box-line virtual corrections and validate their results via multiple cross-checks and width schemes. They find modest QCD corrections, with total cross sections increasing by about 10% and residual scale uncertainties below 2%, while PDFs contribute a few percent uncertainty; differential distributions show small shape changes and robust topology under NLO corrections. The study demonstrates strong perturbative control over EW VBF processes, enabling Wjj/Zjj production to serve as a calibration channel for Higgs VBF measurements and informing forward-jet tagging and central-jet veto strategies in experimental analyses.

Abstract

The production of W or Z bosons in association with two jets is an important background to the Higgs boson search in vector-boson fusion at the LHC. The purely electroweak component of this background is dominated by vector-boson fusion, which exhibits kinematic distributions very similar to the Higgs boson signal. We consider the next-to-leading order QCD corrections to the electroweak production of l nu_l jj and l^+ l^- jj events at the LHC, within typical vector-boson fusion cuts. We show that the QCD corrections are modest, increasing the total cross sections by about 10%. Remaining scale uncertainties are below 2%. A fully-flexible next-to-leading order partonic Monte Carlo program allows to demonstrate these features for cross sections within typical vector-boson-fusion acceptance cuts. Modest corrections are also found for distributions.

Figures (10)

• Figure 1: Feynman graphs contributing to the process $uc\,\hbox{$\rightarrow$}\, dc\ell^+\nu_\ell$ at tree level. For the generic VBF process discussed in this paper, seven Feynman-graph topologies contribute at tree level: the six topologies shown plus an additional bremsstrahlung graph, with the vector boson emitted off the final-state charm quark [mirror image of graph (b)].
• Figure 2: Examples of Feynman amplitudes with an initial gluon. Graphs like (a) and (b), with the gluon coupled to the initial quark line, correspond to vector-boson pair production and are eliminated. The two gauge-invariant subsets of graphs like (c) and (d), with the gluon coupled to the final-state quark pair, contain all $g\hbox{$\rightarrow$} q\bar{q}$ splitting contributions and are included in our calculation.
• Figure 3: Virtual corrections for a fermion line with two attached electroweak bosons, $V_1(q_1)$ and $V_2(q_2)$. The finite part of the sum of these graphs defines the reduced amplitude $\tilde{\cal M}_\tau(q_1,q_2)$ of Eq. (\ref{['eq:boxlinefig']}).
• Figure 4: Scale dependence of the total cross section at LO and NLO within the cuts of Eqs. (\ref{['eq:cuts1']})--(\ref{['eq:cuts4']}) for $W^-$ and $W^+$ production at the LHC. The decay branching ratio of the $W$ is included in the definition of the cross section, here and in all subsequent figures. The factorization scale $\mu_F$ and/or the renormalization scale $\mu_R$ have been taken as multiples of the vector-boson mass, $\xi\, m_W$, and $\xi$ is varied in the range $0.1 < \xi < 10$. The NLO curves are for $\mu_F=\mu_R=\xi m_W$ (solid red line), $\mu_F=m_W$ and $\mu_R=\xi\, m_W$ (dashed green line) and $\mu_R=m_W$ and $\mu_F$ variable (dot-dashed blue line). The dotted black curve shows the dependence of the LO cross section on the factorization scale. At this order, $\alpha_s(\mu_R)$ does not enter.
• Figure 5: Same as Fig. \ref{['fig:scale_depW']}, but for $Z$ production at the LHC, with the $Z\hbox{$\rightarrow$} \mu^+\mu^-$ branching ratio included in the definition of the cross section, here and in all subsequent figures.