aa r X i v : . [ h e p - e x ] J un TOP PAIR PRODUCTION CROSS-SECTION AT THE TEVATRON
URSULA BASSLER in behalf of the CDF and DØ collaborationLPNHE, 4 place Jussieu,75252 Paris Cedex 05, France
An overview of latest top quark pair production cross-sections measured at the Tevatron isgiven. These measurements have been carried out in the dilepton, lepton+jets and all-jetschannels with an integrated luminosity of about 1 fb − . The measurements are consistentwith NNLO calculations. Since the top quark discovery in 1995 by the CDF and DØ collaborations 1, top pair pro-duction cross-sections are one of the basic measurements to be carried out on each new datasample. During Tevatron Run I (1992-1996) an integrated luminosity of about 100 pb − at acenter of mass energy of √ s = 1 . . +1 . − . pb and 5 . ± . − of p ¯ p collisions with √ s = 1 .
96 TeVhave been produced and analyzed since. At this energy an increase of about 30% in the cross-section is expected. The most recent NNLO calculations predict a cross-section of 6 . +0 . − . pb 2or 6 . ± . top = 175 GeV.In the standard model (SM) | V tb | ∼ t → W b close to 100%.With a lifetime of τ ∼ − s top quarks decay before hadronization. Their decay channels areclassified according to the decay of the W bosons produced. The dilepton channel accounts forabout 6% of all decays, taking into account decays into ee , µµ and eµ and including leptonic τ decays. The l +jets channel represents about 34% of the cross-section with the leptonic τ decaysincluded in the e +jets and µ +jets channels. Decays into all jets occur in 46% of the events, theremaining 14% correspond to signatures with hadronic τ decays.Cross-section measurements are an important test of perturbative QCD at high p T as non-SM top production, for example resonant top production, may lead to higher a cross-section thanexpected. It is important to verify the consistency of different decay channels, as some non-SM able 1: Summary of the observed and expected number of events in the dilepton channels used for the DØcombined dilepton cross-section measurement and the CDF lepton + track measurement. ee µµ eµ ( ≥ jets ) eµ (1 jet ) l + track L in fb − .
04 1 .
05 1 .
05 1 .
05 1 . N bkgd . ± . . ± . . ± . . +1 . − . . ± . t ¯ t exp . ± . . ± . . +2 . − . . +0 . − . . ± . N obs
16 9 32 16 129 models for example t → H + or t → ˜ t , modify the contributions in different decay channels.Non- W top decays are probed from the comparison of the dilepton and l +jets measurements.Top quark event selections using b -jet tagging assume a branching ratio BR ( t → W b ) = 1. Theirconsistency with kinematic methods, free of this assumption, is an important check of the SMprediction.A cross-section is in most cases obtained from a counting experiment: σ ( p ¯ p → t ¯ t ) =( N obs − N bkgd ) /A tot L . N bkgd , the number of background events, estimated from Monte Carlosimulations and/or data samples, is subtracted from N obs , the number of observed events meet-ing the selection criteria of a top-event signature. This difference is normalized by the integratedluminosity L and the total acceptance A tot . A tot includes the geometric acceptance as well astrigger efficiency and event selection efficiency and is slightly dependent on m top . In all theMonte Carlo simulations m top = 175 GeV has been used. The signature of top dilepton events is two high p T , opposite sign leptons ( p T >
15 GeV), somemissing transverse energy ( E T >
35 GeV) and two or more, high p T jets ( p T >
20 GeV). Physicsbackground is due dominantly to
Z/γ ∗ +jets events, W W/W Z/ZZ +jets events, and estimatedfrom Monte Carlo simulations. Instrumental backgrounds occur due to fake isolated leptons,either as a mis-identified e or a µ in a non-reconstructed b -jet, as well as E T from detectorresolution, fake jets or noise in the calorimeter. These background are estimated from data.DØ measured the cross-section in the ee , µµ and eµ channels 4. Requiring 2 leptons and2 jets in the event selection yields to an acceptance of 8% and 5% in the ee and µµ channels,and 12% in the eµ channel. The acceptance in eµ channel could be further improved by anadditional 3% taking in account the events with only 1 jet in the final state. In total 73events are observed for 51 expected signal events and 24 expected background events. Detailsfor each channel are given in table 1. The combined result from the three measurements is σ t ¯ t = 6 . +1 . − . (stat) +0 . − . (syst) ± . .
3. The number of expected and observed events are also given in table 1,leading to a cross-section of σ t ¯ t = 9 . ± . ± . ± . − , explic-itly vetoing a fully reconstructed second lepton to allow for a combination with the dileptonmeasurements and using b -jet tagging to improve the purity of the sample. An update of thismeasurement with 1 fb − is in progress. ikelihood Discriminant0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 E ve n t s KS = 0.855DØ RunII Preliminary 900 pb -1 DATA 715 224ttW+Jets 403Multijet 88 E ve n t s ‡ ‡ ‡ DATA l+jets (8.3)pb fi tt ll (8.3)pb fi ttsingle topWccWbbWjj ttfi ZMulti-jet
DØ RunII Preliminary ‡ ‡ ‡ DØ RunII Preliminary
Figure 1: The Likelihood discriminant for the kinematic analysis (left) and the jet multiplicities in the 1-tagsample (middle) and 2-tag sample (right) for the b -jet tagging analysis with the DØ lepton+jets samples. For top decays into l +jets, the signature is a high p T lepton, large E T and 4 or more high p T jets. Dominating physics background is due to W +jets events. Instrumental background is dueto fake isolated leptons in multijet events. To separate the signal and background in the l +jetschannels either the kinematic properties of the events are used or b -jet tagging is required.For the first type of analysis, DØ constructs a likelihood discriminant based on six kinematicvariables without a b -jet tagging requirement 6. Its output is shown in figure 1(left) for the com-bined e +jets and µ +jets sample. For an integrated luminosity of L = 0 .
91 fb − , 124 (100) t ¯ t events are expected in the e +jets ( µ +jets) channel for 168 (235) W +jets events and 62 (27) mul-tijet events, leading to a combined cross-section of σ t ¯ t = 6 . +0 . − . (stat) ± +0 . ± . b -jet tagging algorithm.The chosen operating point has a b -jet tagging efficiency of 55% and a fake tag rate of 1%. Forthe same fake tag rate this represents a 15% increase in efficiency with respect to the previous b -jet tagging algorithm. The cross-section result of σ t ¯ t = 8 . +0 . − . (stat) +0 . − . (syst) ± . e and µ final states,3-jet events, or 4 and more jet events, 1-tag or 2-tags separately. The jet-multiplicities for thecombined 1-tag and 2-tags samples are shown in figure 1(middle and right respectively). Withinthe errors, the results are consistent between the kinematic and the b-tagging analyses.CDF results in the l +jets channels have been presented at the previous Moriond QCDconference 8 on an integrated luminosity of L = 0 . − . For a method using only kinematicvariables the result is σ t ¯ t = 6 . ± . ± . ± . b -jettagging yields to a cross-section of σ t ¯ t = 8 . ± . ± . ± . Even though this channel has the highest branching ratio, it is largely dominated by multijetbackground. The preselection of t ¯ t → jets in the inclusive CDF multijet sample 9 requires 6 to 8jets with p T >
15 GeV separated by ∆
R > .
5. This preselection has a S/B ratio of 1/1300. Toreduce the background a NN is used with 11 kinematic input variables. A further improvement inthe S/B ratio is obtained by requiring a secondary vertex tag. The cross-section is then measuredfrom the number of observed tags with an expectation of n tag = 0 . ± .
07 per top eventdetermined from Monte Carlo. In total 387 signal and 846 background events have been foundfor L = 1 .
02 fb − . The cross-section obtained is σ t ¯ t = 8 . ± . +2 . − . (syst) ± . (pb)t t fi p(p s Cacciari et al. JHEP 0404:068 (2004)Kidonakis,Vogt PRD 68 114014 (2003) =175 GeV/c t Assume m * CDF PreliminaryCombined(old SLT,all-had) * – – – ) -1 (L= 760 pb (lumi) – (syst) – (stat) All-hadronic: Vertex Tag * – – – ) -1 (L=1020 pb MET+Jets: Vertex Tag * – – – ) -1 (L= 311 pb Lepton+Jets: Soft Muon Tag – – – ) -1 (L= 760 pb Lepton+Jets: Vertex Tag * – – – ) -1 (L=1120 pb Lepton+Jets: Kinematic ANN * – – – ) -1 (L= 760 pb Dilepton * – – – ) -1 (L=1200 pb Lepton+Track * – – – ) -1 (L=1070 pb dilepton/l+jets (topological) dilepton* (topological) ltrack/emu* (combined) tau+jets* (b-tagged) alljets (b-tagged) l+jets (b-tagged) l+jets* (NN b-tagged) l+jets* (topological)
230 pb –1 +1.2-1.2+1.4-1.1 pb –1 +1.2-1.1+0.9-0.8 pb
370 pb –1 +1.9-1.7+1.1-1.1 pb
350 pb –1 +4.3-3.5+0.7-0.7 pb
410 pb –1 +2.0-1.9+1.4-1.1 pb
420 pb –1 +0.9-0.9+0.0-0.0 pb
910 pb –1 +0.6-0.5+0.9-1.0 pb
910 pb –1 +0.9-0.8+0.7-0.7 pb
0 2 4 6 8 10 12
DØ Run II * = preliminary σ (pp → tt) [pb] Winter 2007
Kidonakis and Vogt, PRD 68, 114014 (2003) Cacciari et al., JHEP 0404, 068 (2004)m top = 175 GeV l+jets* ( μ -tagged)
420 pb –1 +2.0-1.8+0.0-0.0 pb expected error bandfrom combined result Figure 2: Top quark pair production cross-section measurements: summary of the current CDF (left) results andDØ results (middle). The right figure shows the m top dependence of the DØ dilepton measurement. Summaries of the CDF and DØ top quark pair production cross-section measurements are shownin figure 2(left and middle respectively). All the measurements are in good agreement with theNNLO-predictions. Results with an integrated luminosity of about 1 fb − are highlighted. Withthis luminosity the errors have been sizably reduced and reach in some of the decay channelsabout 15%. From a combination of all results an experimental error at the order of the theoreticalerror can be expected.During the conference the question was raised if m top could be determined from the cross-section measurements. A first answer concerning the precision that could be achieved can beinterfered from figure 2(right) showing the dependence of the DØ dilepton cross-section as afunction of m top . The shaded band shows the assumption of the total error for a combinedresult of the currently measured current cross-sections to be of the size of the theoretical errorand with the same m top dependence than the di-lepton cross-section. A determination of m top using the current production cross-section would lead to an error on m top of about ± Acknowledgments
I would like to thank the CDF and DØ collaborations for presenting these results, in particu-lar the top conveners Kirsten Tollefson, Robin Erbacher, Elizabetha Shabalina, Ulrich Heintz,Michele Weber for their help in preparing the talk, rehearsals and very useful comments, KevinLannon for a last minute plot and the organizers for another great Moriond!
References
1. CDF Collaboration, F. Abe et al. , Phys. Rev. Lett. , 2626 (1995); DØ Collaboration,S. Abachi et al. , Phys. Rev. Lett. , 2632 (1995).2. M. Cacciari et al., JHEP , 68 (2004).3. N. Kidonakis and R. Vogt., Phys. Rev. D , 114014 (2003).4. DØ Collaboration, DØ Note 5371 CONF - 03/07 (2007).5. CDF Collaboration, CDF-Note available soonavailable soon