Seeing in Color: Jet Superstructure

TL;DR

This paper introduces a new class of observables called superstructure variables, like pull, which capture inter-jet features such as color flow, providing new tools for event analysis in high-energy physics.

Contribution

The paper presents the concept of superstructure observables, exemplified by the pull variable, to analyze color flow and inter-jet correlations beyond traditional jet properties.

Findings

Pull can distinguish color singlet decays from background

Superstructure variables improve background discrimination

Applicable to Tevatron and LHC data

Abstract

A new class of observables is introduced which aims to characterize the superstructure of an event, that is, features, such as color flow, which are not determined by the jet four-momenta alone. Traditionally, an event is described as having jets which are independent objects; each jet has some energy, size, and possible substructure such as subjets or heavy flavor content. This description discards information connecting the jets to each other, which can be used to determine if the jets came from decay of a color singlet object, or if they were initiated by quarks or gluons. An example superstructure variable, pull, is presented as a simple handle on color flow. It can be used on an event-by-event basis as a tool for distinguishing previously irreducible backgrounds at the Tevatron and the LHC.