Top-pair production and decay at NLO matched with parton showers

TL;DR

The paper presents an NLO+PS generator for top-quark pair production that includes NLO corrections in both production and decay within a narrow width approximation, augmented by an approximate finite-width and interference treatment. It extends the POWHEG BOX V2 framework to handle radiation from decaying resonances and interfaces with PYTHIA8 for realistic showering, offering multiple off-shell handling schemes (Breit-Wigner reweighting, double-resonant approximation, and full off-shell matrix elements). Through extensive comparisons with existing NNLO/ NLO results and internal cross-checks, the work demonstrates improved top-decay descriptions and clarifies the impact of off-shell effects on mass-sensitive observables. The approach provides a practical, flexible path toward more accurate top-quark mass measurements and sets the stage for fully consistent NLO+PS matching with intermediate resonances decaying into colored final states.

Abstract

We present a next-to-leading order (NLO) calculation of $t\bar{t}$ production in hadronic collisions interfaced to shower generators according to the POWHEG method. We start from an NLO result from previous work, obtained in the zero width limit, where radiative corrections to both production and decays are included. The POWHEG interface required an extension of the POWHEG BOX framework, in order to deal with radiation from the decay of resonances. This extension is fully general (i.e. it can be applied in principle to any process considered in the zero width limit), and is here applied for the first time. In order to perform a realistic simulation, we introduce finite width effects using different approximations, that we validated by comparing with published exact NLO results. We have interfaced our POWHEG code to the PYTHIA8 shower Monte Carlo generator. At this stage, we dealt with novel issues related to the treatment of resonances, especially with regard to the initial scale for the shower that needs to be set appropriately. This procedure affects, for example, the fragmentation function of the b quark, that we have studied with particular attention. We believe that the tool presented here improves over previous generators for all aspects that have to do with top decays, and especially for the study of issues related to top mass measurements that involve B hadrons or b jets. The work presented here also constitutes a first step towards a fully consistent matching of NLO calculations involving intermediate resonances decaying into coloured particles, with parton showers.

Figures (17)

• Figure 1: Example of real emission and underlying Born configuration in a simple model of neutral resonance production and decay. The thick solid arrow represents the incoming system. Solid lines represent coloured particles, dashed lines represent neutral ones. The particle labelled $c$ is the only Born level massless coloured parton in the final state. In the standard POWHEG BOX framework, the underlying Born is obtained by boosting the $n,p$ system of the real emission configuration in order to conserve the total energy in the rest frame of the incoming system. With this procedure, the resonance mass is not conserved.
• Figure 2: Comparison between our LO results and those of DDKP, obtained using dynamic scales. Plots are obtained with the cuts in eq. \ref{['eq:Dennercuts']}.
• Figure 3: Comparison between our NLO results and those of DDKP, obtained using dynamic scales. Plots are obtained with the cuts in eq. \ref{['eq:Dennercuts']}.
• Figure 4: Enhanced contributions in the region of small invariant mass of the $b\bar{b}$ system (left), and in the region of small $b$ transverse momentum (right).
• Figure 5: Comparison of our full, DR and BW results for the top and for the $t\bar{t}$ pair transverse momenta, without acceptance cuts.