Measurement of the top quark mass using the matrix element technique in dilepton final states

We present a measurement of the top quark mass in $p\overline{p}$ collisions at a center-of-mass energy of 1.96 TeV at the Fermilab Tevatron collider. The data were collected by the D0 experiment corresponding to an integrated luminosity of $9.7{\mathrm{fb}}^{-1}$. The matrix element technique is applied to $t\overline{t}$ events in the final state containing leptons (electrons or muons) with high transverse momenta and at least two jets. The calibration of the jet energy scale determined in the $\mathrm{lepton}+\text{jets}$ final state of $t\overline{t}$ decays is applied to jet energies. This correction provides a substantial reduction in systematic uncertainties. We obtain a top quark mass of $m_t=173.93\pm 1.84\mathrm{GeV}$.

Figure 1

The distributions in lepton $p_T$ and the ratio of data to predictions for the combined $ee$, $e\mu$, and $\mu \mu$ final states after applying requirements (i)–(vii).

Figure 2

The distributions in the number of jets and the ratio of data to the prediction for the combined $ee$, $e\mu$, and $\mu \mu$ final states after applying requirements (i)–(vii).

Figure 3

The distributions in jet $p_T$ after implementing the $k_{\mathrm{JES}}$ correction, and the ratio of data to the prediction for the combined $ee$, $e\mu$, and $\mu \mu$ final states after applying requirements (i)–(vii).

Figure 4

The distributions in $H_T$ after implementing the $k_{\mathrm{JES}}$ correction, and the ratio of data to the prediction for the combined $ee$, $e\mu$, and $\mu \mu$ final states after applying requirements (i)–(vii).

Figure 5

The response of the ME method in (a) $m_t$, (b) statistical uncertainty on the $m_t$, and (c) the pull width, shown as a function of the MC input $m_t$ for the combined $ee$, $e\mu$, and $\mu \mu$ channels. The error bars in (a) and (b) are invisibly small. The dashed line in (a) represents the case of ideal response.

Figure 6

The negative log-likelihood ratio for the combined $ee$, $e\mu$, and $\mu \mu$ data after calibration, as a function of the input MC $m_t$. The curve is the best fit to a parabola in the interval 168.4–179.7 GeV.

Figure 7

The distribution in the expected statistical uncertainty ${\sigma }_{\text{lhood}}^{\exp }$ for the combined $ee$, $e\mu$, and $\mu \mu$ channels after applying calibration, for the MC input $m_t=175\mathrm{GeV}$. The arrow indicates the statistical uncertainty for data after the calibration (${\sigma }_{\mathrm{meas}}$).