Automating the POWHEG method in Sherpa

TL;DR

The paper presents a fully automated implementation of the POWHEG method within the SHERPA event generator, enabling NLO-accurate, positive-weight simulations for processes with simple colour structure by integrating Catani-Seymour dipole subtraction and interfacing with external virtual corrections. It establishes a coherent theoretical framework linking POWHEG to standard parton showers, including real-emission decomposition, ME corrections, and approximate NLO seeds, and then details SHERPA’s concrete realization with automatic Born-zero handling. The authors validate the approach across a broad set of processes (e+e- jets, DIS, Drell-Yan, W, Higgs, ZZ, WW) and demonstrate internal consistency, ME+PS comparisons, and favorable agreement with experimental data, highlighting the method’s robustness and practicality. The work demonstrates that automated POWHEG within SHERPA provides accurate radiation patterns and cross sections, setting the stage for more extensive NLO merging and multi-jet applications in a single framework.

Abstract

A new implementation of the POWHEG method into the Monte-Carlo event generator Sherpa is presented, focusing on processes with a simple colour structure. Results for a variety of processes, namely e+e- to hadrons, deep-inelastic lepton-nucleon scattering, hadroproduction of single vector bosons and of vector boson pairs as well as the production of Higgs bosons in gluon fusion serve as test cases for the successful realisation. The algorithm is fully automated such that for further processes only virtual matrix elements need to be included.

Figures (20)

• Figure 1: Effective diagram for the splitting of (1) a final-state parton connected to a final-state spectator, (2) a final-state parton connected to an initial-state spectator, (3) an initial-state parton connected to a final-state spectator and (4) an initial-state parton connected to an initial-state spectator in the standard Catani-Seymour notation. The blob denotes the colour correlated leading order matrix element, and the incoming and outgoing lines label the initial-state and final-state partons participating in the splitting.
• Figure 2: $0\to 1$ jet resolution in $k_T$ clustered jets and transverse momentum of the $e^+e^-\!$-pair in $Z/\gamma^*$ boson production at the Tevatron. The standard parton shower effected on the leading order matrix elements (red) is compared to the POWHEG formulation (blue) and to POWHEG with the real emission matrix element $\mathcal{R}$ replaced by its parton-shower approximation $\mathcal{R}^\text{(PS)}$ (green).
• Figure 3: Dependence of the parton shower correction factor $w_{ij,k}$ on the $Z$-$H$-splitting parameter $\kappa_\text{res}$ for $W^-$ production at the Tevatron.
• Figure 4: Predictions for $0\to 1$ jet resolution in $k_T$ clustered jets and transverse momentum of the $W$ boson and in $W$ boson production at the Tevatron for different settings of the $Z$-$H$-splitting parameters $\kappa_\text{res}$ and $w_{ij,k}^\text{th}$.
• Figure 5: Predictions for the $0\to 1$ jet resolution in $k_T$ clustered jets in $Z/\gamma^*$ (left) and $W$ (right) boson production at the Tevatron.