Table of Contents
Fetching ...

Pileup subtraction for jet shapes

Gregory Soyez, Gavin P. Salam, Jihun Kim, Souvik Dutta, Matteo Cacciari

TL;DR

The method rescues the possibility of using jet shapes in the high pileup environment of current and future LHC running, as it shows with examples of quark-gluon discrimination and top tagging.

Abstract

Jet shapes have the potential to play a role in many LHC analyses, for example in quark-gluon discrimination or jet substructure analyses for hadronic decays of boosted heavy objects. Most shapes, however, are significantly affected by pileup. We introduce a general method to correct for pileup effects in shapes, which acts event-by-event and jet-by-jet, and accounts also for hadron masses. It involves a numerical determination, for each jet, of a given shape's susceptibility to pileup. Together with existing techniques for determining the level of pileup, this then enables an extrapolation to zero pileup. The method can be used for a wide range of jet shapes and we show its successful application in the context of quark/gluon discrimination and top-tagging.

Pileup subtraction for jet shapes

TL;DR

The method rescues the possibility of using jet shapes in the high pileup environment of current and future LHC running, as it shows with examples of quark-gluon discrimination and top tagging.

Abstract

Jet shapes have the potential to play a role in many LHC analyses, for example in quark-gluon discrimination or jet substructure analyses for hadronic decays of boosted heavy objects. Most shapes, however, are significantly affected by pileup. We introduce a general method to correct for pileup effects in shapes, which acts event-by-event and jet-by-jet, and accounts also for hadron masses. It involves a numerical determination, for each jet, of a given shape's susceptibility to pileup. Together with existing techniques for determining the level of pileup, this then enables an extrapolation to zero pileup. The method can be used for a wide range of jet shapes and we show its successful application in the context of quark/gluon discrimination and top-tagging.

Paper Structure

This paper contains 7 equations, 2 figures.

Figures (2)

  • Figure 1: Impact of pileup and subtraction on various jet-shape distributions and their averages, in dijet, $WW$ and $t\bar{t}$ production processes. The distributions are shown for Poisson distributed pileup (with an average of $30$ pileup events) and the averages are shown as a function of the number of pileup events, $n_\text{PU}$. The shapes are calculated for jets with $p_t > 500\;\mathrm{GeV}$ (the cut is applied before adding pileup, as are the cuts on the jet mass $m_J$ and subjettiness ratio $\tau_{21}$ where relevant).
  • Figure 2: Left: rate for tagging quark and gluon jets using a fixed cut on the jet width, shown as a function of the number of pileup vertices. Middle: filtered jet-mass distribution for fat jets in $t\bar{t}$ events, showing the impact of the $\rho$ and $\rho_m$ components of the subtraction. Right: tagging rate of an $N$-subjettiness top tagger for $t\bar{t}$ signal and dijet background as a function of the number of pileup vertices. All cuts are applied after addition (and possible subtraction) of pileup. Subtraction acts on $\tau_1$, $\tau_2$ and $\tau_3$ individually. See text for further details.