Measurement of the $t\overline{t}$ production cross section in $p\overline{p}$ collisions at $\sqrt{s}=1.96\mathrm{TeV}$ using events with large missing transverse energy and jets

In this paper we report a measurement of the $t\overline{t}$ production cross section in $p\overline{p}$ collisions at $\sqrt{s}=1.96\mathrm{TeV}$ using data corresponding to an integrated luminosity of $2.2{\mathrm{fb}}^{-1}$ collected with the CDF II detector at the Tevatron accelerator. We select events with significant missing transverse energy and high jet multiplicity. This measurement vetoes the presence of explicitly identified electrons and muons, thus enhancing the tau contribution of $t\overline{t}$ decays. Signal events are discriminated from the background using a neural network, and heavy flavor jets are identified by a secondary-vertex tagging algorithm. We measure a $t\overline{t}$ production cross section of $7.99\pm 0.55(\mathrm{stat})\pm 0.76(\mathrm{syst})\pm 0.46(\mathrm{lumi})$ pb, assuming a top mass $m_{\mathrm{top}}=172.5\mathrm{GeV}/c^2$, in agreement with previous measurements and standard model predictions.

Figure 1

$b$ tagging rates as a function of (a) jet $E_T$, (b) $N_{\mathrm{trk}}$ and (c) $\mathrm{E̸}_T{}^{\mathrm{prj}}$, for the data sample with exactly three jets in the event.

Figure 2

Tagging matrix check on data after preliminary requirements and before any additional kinematic selection. Observed and predicted number of $b$-tagged jets as a function of the jet multiplicity are shown in the figure, statistical errors only. The expected contribution coming from $t\overline{t}$ events based on the theoretical cross section of ${\sigma }_{t\overline{t}}=7.45\mathrm{pb}$ is also shown. Background error bands are centered on the inclusive top plus background prediction.

Figure 3

Distribution of neural network input variables for multijet data (background) and $t\overline{t}$ MC (signal) samples: (a) the transverse energy of the leading jet $E_T^1$, (b) $\Delta {\varphi }_{\min }({\mathrm{E̸}}_T,\mathrm{\text{jet}})$, (c) ${\mathrm{E̸}}_T^{\mathrm{sig}}$, (d) $\sum _{}^{}{E}_T$, (e) $\sum _3{E}_T$, (f) the centrality $C$, and the topology-related variables (g) sphericity $S$ and (h) aplanarity $A$.

Figure 4

Observed and matrix-predicted number of $b$-tagged jets versus neural network output in the multijet data for (a) all events with at least four jets, and for events with exactly (b) three, (c) four and (d) five jets, along with the expected contribution due to $t\overline{t}$ signal events. Background error bands are centered on the inclusive top plus background prediction.

Figure 5

Observed and matrix-predicted number of $b$-tagged jets by jet multiplicity in the multijet data after cut on neural network output ${\mathrm{NN}}_{\mathrm{out}}$ greater than 0.8, along with the expected contribution of inclusive $t\overline{t}$ signal normalized to the measured cross section. Background error bands are centered on the inclusive top plus background prediction.