Measurement of the $p\overline{p}\to t\overline{t}+X$ production cross section at $\sqrt{s}=1.96\mathrm{TeV}$ in the fully hadronic decay channel

A measurement of the top quark pair production cross section in proton antiproton collisions at an interaction energy of $\sqrt{s}=1.96\mathrm{TeV}$ is presented. This analysis uses $405\pm 25{\mathrm{pb}}^{-1}$ of data collected with the D0 detector at the Fermilab Tevatron Collider. Fully hadronic $t\overline{t}$ decays with final states of six or more jets are separated from the multijet background using secondary vertex tagging and a neural network. The $t\overline{t}$ cross section is measured as ${\sigma }_{t\overline{t}}={4.5}_{-1.9}^{+2.0}(\mathrm{stat}{)}_{-1.1}^{+1.4}(\mathrm{syst})\pm 0.3(\mathrm{lumi})\mathrm{pb}$ for a top quark mass of $m_t=175\mathrm{GeV}/c^2$.

Figure 1

The $H_T$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line), and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 2

The $E_T^{56}$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line), and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 3

The $\mathcal{A}$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line) and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 4

The $⟨{\eta }^2⟩$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line), and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 5

The $\mathcal{M}$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line), and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 6

The $M_{\min }^{34}$ distribution for single-tag events (a) and double-tag events (b). Shown are the data (points), the background (solid line), and the expected $t\overline{t}$ distribution (filled histogram) multiplied by 140 (60) for the single (double)-tag analysis.

Figure 7

The output discriminant of an artificial neural network ($NN$) with six input nodes. All distributions are normalized to area. $NN$ is optimized to distinguish between fully hadronic $t\overline{t}$ Monte Carlo events (signal) and the background from multijet production (background) as predicted by the tag rate functions.

Figure 8

The performance of the TRF prediction on double-tag events (points), without including the correlation factor $C_{ij}$ (dashed histogram), and including $C_{ij}$ for two different functional parameterizations (solid histograms).

Figure 9

The distribution of the $NN$ output variable for single-tag events. Shown are the data (points), background (hashed band), signal (filled histogram), and $\mathrm{\text{signal}}+\mathrm{\text{background}}$ (dashed histogram). The vertical line represents the used cut of $NN>0.81$.

Figure 10

The distribution of the $NN$ output variable for double-tag events. Shown are the data (points), background (hashed band), signal (filled histogram), and $\mathrm{\text{signal}}+\mathrm{\text{background}}$ (dashed histogram). The vertical line represents the used cut of $NN>0.78$.