Improving NLO-parton shower matched simulations with higher order matrix elements

TL;DR

This paper proposes a method to merge higher order matrix elements with NLO-parton shower simulations, enhancing the accuracy of multi-jet event modeling in hadronic collision simulations.

Contribution

It introduces a pragmatic approach to combine MEPS and NLOPS methods within the POWHEG framework, improving event sample quality without requiring modifications to existing simulations.

Findings

Distributions within NLO accuracy remain unchanged.

Multi-jet sensitive distributions match MEPS results.

Method applied successfully to W boson and top quark pair production.

Abstract

In recent times the algorithms for the simulation of hadronic collisions have been subject to two substantial improvements: the inclusion, within parton showering, of exact higher order tree level matrix elements (MEPS) and, separately, next-to-leading order corrections (NLOPS). In this work we examine the key criteria to be met in merging the two approaches in such a way that the accuracy of both is preserved, in the framework of the POWHEG approach to NLOPS. We then ask to what extent these requirements may be fulfilled using existing simulations, without modifications. The result of this study is a pragmatic proposal for merging MEPS and NLOPS events to yield much improved MENLOPS event samples. We apply this method to W boson and top quark pair production. In both cases results for distributions within the remit of the NLO calculations exhibit no discernible changes with respect to the pure NLOPS prediction; conversely, those sensitive to the distribution of multiple hard jets assume, exactly, the form of the corresponding MEPS results.