Measurement of the top quark mass in the dilepton channel using $m_{\mathrm{T}2}$ at CDF

We present measurements of the top quark mass using $m_{\mathrm{T}2}$, a variable related to the transverse mass in events with two missing particles. We use the template method applied to $t\overline{t}$ dilepton events produced in $p\overline{p}$ collisions at Fermilab’s Tevatron Collider and collected by the CDF detector. From a data sample corresponding to an integrated luminosity of $3.4{\mathrm{fb}}^{-1}$, we select 236 $t\overline{t}$ candidate events. Using the $m_{\mathrm{T}2}$ distribution, we measure the top quark mass to be ${\mathrm{M}}_{\mathrm{top}}={168.0}_{-4.0}^{+4.8}(\mathrm{stat})\pm 2.9(\mathrm{syst})\mathrm{GeV}/c^2$. By combining $m_{\mathrm{T}2}$ with the reconstructed top quark mass distributions based on a neutrino weighting method, we measure ${\mathrm{M}}_{\mathrm{top}}=169.3\pm 2.7(\mathrm{stat})\pm 3.2(\mathrm{syst})\mathrm{GeV}/c^2$. This is the first application of the $m_{\mathrm{T}2}$ variable in a mass measurement at a hadron collider.

Figure 1

The $m_{\mathrm{T}2}$ distributions from $t\overline{t}$ dilepton Monte Carlo events that pass the selection criteria for three input values of the top quark mass. Each distribution is normalized to have unit area.

Figure 2

Negative log-likelihood distributions.

Figure 3

Distributions of the three variables used to estimate the top quark mass, showing the $b$-tagged and non-$b$-tagged samples separately. The data are overlaid with the predictions from the KDE probability distributions using the top quark mass ${\mathrm{M}}_{\mathrm{top}}=169\mathrm{GeV}/c^2$ and full background model.