NLO corrections to hard process in QCD shower -- proof of concept

TL;DR

The paper proposes a new method to incorporate QCD NLO corrections into the hard process of the initial-state Monte Carlo parton shower, tested via heavy-boson production at hadron colliders. It replaces traditional subtraction-based schemes with a monolithic weight $W^{NLO}_{MC}$ applied to LO ladder configurations within an angular-ordered evolution framework, aiming for complete phase-space coverage. Numerical results validate the approach: LO distributions reproduce collinear-factorization predictions with high precision, and the NLO corrections agree with analytic expectations while contributing modest changes (≈1–2%) for the tested observables. The work discusses differences with MC@NLO and POWHEG and outlines refinements needed for practical, broader application.

Abstract

The concept of new methodology of adding QCD NLO corrections in the initial state Monte Carlo parton shower (hard process part) is tested numerically using, as an example, the process of the heavy boson production at hadron--hadron colliders such as LHC. In spite of the use of a simplified model of the process, all presented numerical results prove convincingly that the basic concept of the new methodology works correctly in practice, that is in the numerical environment of the Monte Carlo parton shower event generator. The differences with the other well established methods, like MC@NLO and POWHEG, are briefly discussed and future refinements of the implementation of the new method are also outlined.

Figures (8)

• Figure 1: In the upper plot the LO distribution of $\eta_W^*=\frac{1}{2}\ln(x_F/x_B)$ from the CMC LO parton shower (purple) and from the strictly collinear formula (green) are shown. The lower plot shows the ratio of the two, the agreement of $<0.5\%$ is obtained.
• Figure 2: The pure $(-)$ NLO correction to the distribution of $\eta_W^*=\frac{1}{2}\ln(x_F/x_B)$ in CMC LO parton shower in $W$ boson production (purple). It agrees with the strictly collinear formula (green) to within $<1\%$ of the NLO correction itself.
• Figure 3: The comparison of the LO (purple) and of the pure $(-)$ NLO corrections (green) to the distribution of $\eta_W^*=\frac{1}{2}\ln(x_F/x_B)$; the overall normalization is in GeV$^{-2}$.
• Figure 4: The distribution of the NLO weight $W^{NLO}_{MC}$ of eq. (\ref{['eq:NLODYMCwt']}).
• Figure 5: (a) The inclusive distribution of gluons on the log Sudakov plane of rapidity $t=\xi_{\max}$ and $v=\ln(1-z)$. (b) Contributions from all gluons weighted with the component weight $W^{NLO}_j$.