A Cell Resampler study of Negative Weights in Multi-jet Merged Samples

TL;DR

This paper investigates the use of cell resampling to reduce negatively weighted events in Monte Carlo simulations for high-energy physics, focusing on diphoton background processes relevant to Higgs studies.

Contribution

It introduces the application of cell resampling to mitigate negative weights in multi-jet merged samples, a novel approach in this context.

Findings

Cell resampling effectively reduces negative weights.

Impact on kinematic distributions is analyzed.

Method applied to realistic LHC simulation samples.

Abstract

We study the use of cell resampling to reduce the fraction of negatively weighted Monte Carlo events in a generated sample typical of that used in experimental analyses. To this end, we apply the Cell Resampler to a set of $pp \rightarrow \gamma \gamma + \mathrm{jets}$ shower-merged NLO matched events, describing the diphoton background to Higgs boson production, generated using the FxFx and MEPS@NLO merging procedures and showered using the Pythia and Sherpa parton shower algorithms. We discuss the impact on various kinematic distributions.