Monte Carlo Top Quark Mass Calibration

TL;DR

This paper develops a calibration method linking Monte Carlo top quark mass parameters to field theory masses using 2-Jettiness in electron-positron collisions, enhancing the precision of top mass measurements.

Contribution

It introduces a calibration procedure using QCD predictions for 2-Jettiness to relate Monte Carlo top mass to MSR and pole masses, accounting for uncertainties.

Findings

$m_t^{ m MC}$ agrees with $m_{t,1\,\rm GeV}^{\rm MSR}$ within uncertainties.

$m_t^{\rm MC}$ can differ from $m_t^{\rm pole}$ by up to 600 MeV.

Converting MSR mass to pole mass yields larger, incompatible results.

Abstract

The most precise top quark mass measurements use kinematic reconstruction methods, determining the top mass parameter of a Monte Carlo event generator, $m_t^{\rm MC}$. Due to the complicated interplay of hadronization and parton shower dynamics in Monte Carlo event generators relevant for kinematic reconstruction, relating $m_t^{\rm MC}$ to field theory masses is a non-trivial task. In this talk we report on a calibration procedure to determine this relation using hadron level QCD predictions for 2-Jettiness in $e^+e^-$ annihilation, an observable which has kinematic top mass sensitivity and a close relation to the invariant mass of the particles coming from the top decay. The theoretical ingredients of the QCD prediction are reviewed. Fitting $e^+e^-$ 2-Jettiness calculations at NLL/NNLL order to PYTHIA 8.205, we find that $m_t^{\rm MC}$ agrees with the MSR mass $m_{t,1\,{\rm GeV}}^{\rm MSR}$ within uncertainties. At NNLL we find $m_t^{\rm MC} = m_{t,1\,{\rm GeV}}^{\rm MSR} + (0.18 \pm 0.22)\,{\rm GeV}$. $m_t^{\rm MC}$ can differ from the pole mass $m_t^{\rm pole}$ by up to $600\,{\rm MeV}$, and using the pole mass generally leads to larger uncertainties. At NNLL we find $m_t^{\rm MC} = m_t^{\rm pole} + (0.57 \pm 0.28)\,{\rm GeV}$ as the fit result. In contrast, converting $m_{t,1\,{\rm GeV}}^{\rm MSR}$ obtained at NNLL to the pole mass gives a result for $m_t^{\rm pole}$ that is substantially larger and incompatible with the fit result. We also explain some theoretical aspects relevant for employing the C-parameter as an alternative calibration observable.