Some thoughts on how to match Leading Log Parton Showers with NLO Matrix Elements

TL;DR

The paper addresses the challenge of incorporating NLO QCD corrections into Monte Carlo event generation by aligning NLO matrix elements with parton showers to capture both fixed-order accuracy and soft/collinear resummation. It proposes a subtraction-inspired scheme that decomposes the NLO 2→2 cross section into LO, virtual, and reclassified 2→3 contributions, splitting the problematic 2→3 piece into an approximate singular piece (3a) and a finite remainder (3b) to achieve positive weights. A new interface to parton showers is described, where showers are restricted by jet-clustering criteria to avoid double counting, with different handling for 2-jet and 3-jet configurations, enabling consistent ME/PS matching. The work is conceptual, with no implementation yet, but outlines a road for future development to improve NLO QCD tests and to facilitate realistic comparisons with data through hadronization.

Abstract

We propose a scheme that could offer a convenient Monte Carlo sampling of next-to-leading-order matrix elements and, at the same time, allow the interfacing of such parton configurations with a parton-shower approach for the estimation of higher-order effects. No actual implementation exists so far, so this note should only be viewed as the outline of a possible road for the future, submitted for discussion.