Top Dilepton Cross Section in 2.0 fb-1 using  the DIL Selection

 

Authors :

Chang-Seong Moon
Seoul National University

Naoki Kimura
University of Tsukuba

Alexei Varganov, David Mietlicki, Monica Tecchio
University of Michigan

Victoria Giakoumopoulou
University of Athens

Andrew Beretvas
FNAL

 

Abstract :

We report on a measurement of the ttbar cross section using 2.0 fb-1 of data collected by CDF II. We use ttbar dilepton events reconstructed with the Dilepton selection assuming Mtop = 175 GeV/C2. Signal selection and background estimates are based on the methods developed previously for 1.2 fb-1 measurement. We measure

6.78 ± 0.96stat  ± 0.42syst ± 0.42lumi pb.

where the first uncertainty is statistical, the second is the convolution of acceptance and background systematics and the third is due to the uncertainty in  the luminosity measurement.


The analysis and results were blessed at the 3-APR-2008 top meeting and have been documented in CDF Public Conference Note.

 

Analysis :

The DIL selection for the 2.0 fb-1 is based on the GEN6 DIL selection of 1.2 fb-1 analysis.

  1. 2 leptons requirement : First lepton should be isolated electron of ET, or muon with PT, greater than 20 GeV are selected. a second electron of ET, of muon of PT, greater than 20 GeV is also required using looser identification cuts and no requirement on isolation. Events with more than two leptons in the final state are rejected.
  2. At least 2 jets : ttbar candidate events is obtained by requierd at least 2 jets with ET >15 GeV.
  3. MET cut : Missing transverse energy is required missing ET > 25 GeV.
  4. Lcut : If any letpton or jet is closer than 20o from the missing ET direction, Missing ET should be ET > 50 GeV.
  5. Z-veto : This cut removes possible ee/mm contamination from Z events in the 76-106 GeV invariant mass window. In previous top DIL cross-section analysis, we were using a cut on Jet Significance, which is now replaced with a cut on Mert Significance. Optimization studies show that we can sligltly improve the S/B by rejecting 25% of the Z background while maintaning over 95% of the signal.
  6. Summed transverse energy HT > 200 GeV and the two leptons to be of opposite charge.
  7. Only Stubless CMIO muons are allowed. Stubbed CMIO muons pointing to Miniskirt/Keystone CMX regions are now considered as CMX (trigger) muons. The CMX fiduciality is now solely defined according to the Muon Fiducial tool, i.e. no extra cuts are applied to the muon (eta,phi)  direction as a way to separate Arches CMX muons from Miniskirt/Keystone CMX muons.
  8. Jet are "truly" corrected up to L5.
  9. Larry curvature corrections and plug energy corrections are now applied by TopEventModule while creating the TopNtuples. This eliminate the need to apply these correction at ntuple level before calculating quantities like MET and HT which are used by the top DIL selection

 

Results :

The following table contains a summary of the background estimates for each jet bin before the HT and Opposite Sign requirements are applied and in the 2 jet bin after applying only the HT cut. The last column contains the candidate events with all cuts applied. This table also shows the total background expectation, the ttbar expectation for the cross section which is 6.7 pb , their sum (labeled as "Total SM expectation") and the number of candidate events in 2.0 fb-1 of data.

 

The following table shows the total number of background, SM expectation and 2.0 fb-1 data candidate events, divided by lepton flavor contribution.

 

The following plot eps presents the background and signal composition of the Inclusive 2.0 fb-1 DIL events vs jet multiplicity after the Z-veto, MET and L-cut have been applied.

Selected kinematic variables for background and top signal predictions, overlaid to data for the lepton transverse energy spectrum, the dilepton invariant mass, MET and HT,   distributions in 2 fb-1 top DIL candidate events. The corresponding ps files are found here: Lepton Pt, Dilepton invariant Mass, Missing ET, HT

 

 

 

 

Correlation between pre-tag and b-tag analysis :

We investigated signal candidate correlation between pre-tag(145 events) and b-tag(90 events) top dilepton analysis. Of 90 events of btag signal candidate for 2.0fb-1 data, 41 events are found in the pre-tag analysis,  18 events are removed by difference of cut(Z-veto, HT,  Lcut) , 31 events are removed by difference of other selection.

 

Further detail :

More information can be found at the documentation/Q&A page for this analysis, or by contacting any one of the good people listed at the top of this page.