Merging Multi-leg NLO Matrix Elements with Parton Showers

TL;DR

This paper extends NLO matrix element and parton shower merging methods, CKKW-L and UMEPS, to achieve next-to-leading order accuracy in event simulations, improving theoretical consistency and practical implementation.

Contribution

It introduces the UNLOPS method, an extension of UMEPS, providing a more theoretically sound NLO merging approach, and implements these in Pythia8 for complex event generation.

Findings

Successful implementation in Pythia8 for W- and Higgs-production

Zero- and one-jet contributions corrected to NLO

Higher jet multiplicities described by tree-level matrix elements

Abstract

We discuss extensions the CKKW-L and UMEPS tree-level matrix element and parton shower merging approaches to next-to-leading order accuracy. The generalisation of CKKW-L is based on the NL3 scheme previously developed for e+e- -annihilation, which is extended to also handle hadronic collisions by a careful treatment of parton densities. NL3 is further augmented to allow for more readily accessible NLO input. To allow for a more careful handling of merging scale dependencies we introduce an extension of the UMEPS method. This approach, dubbed UNLOPS, does not inherit problematic features of CKKW-L, and thus allows for a theoretically more appealing definition of NLO merging. We have implemented both schemes in Pythia8, and present results for the merging of W- and Higgs-production events, where the zero- and one-jet contribution are corrected to next-to-leading order simultaneously, and higher jet multiplicities are described by tree-level matrix elements. The implementation of the procedure is completely general and can be used for higher jet multiplicities and other processes, subject to the availability of programs able to correctly generate the corresponding partonic states to leading and next-to-leading order accuracy.