Single-Top Cross Section Measurements at ATLAS Patrick Ryan (Michigan State University) Patrick.Ryan@cern.ch. Cross Section and Uncertainties. Introduction to Single-Top. Simulation of Monte Carlo Samples. s-channel Cross Section.
Patrick Ryan (Michigan State University)Patrick.Ryan@cern.ch
Cross Section and Uncertainties
Introduction to Single-Top
Simulation of Monte Carlo Samples
s-channel Cross Section
The measurement of the single-top cross section provides a direct measurement of the CKM Matrix Element |Vtb| and permits verification of Standard Model electroweak coupling. The single-top quark transmits its polarization to its decay products and can provide insight into W-t-b couplings. The single-top quark could also lead to observations of new fields, mediators, and particles which noticeably couple only to heavy fermions. Examples include the Standard Model neutral Higgs, the minimal SUSY charged Higgs, and Flavor Changing Neutral Currents.
The cross section will be calculated with:
Experimental Uncertainties (1fb-1/10fb-1)
- Jet Energy Scale (± 5% / ±1%)
- b-tagging Likelihood (± 5% / ± 3%)
- Luminosity (±5% / ±3%)
- Background cross sections
- ISR / FSR
- PDF and b-quark Fragmentation
Cross Section Uncertainties: NData was generated randomly according to Poisson distribution. NBkg and eSignal were varied for each systematic source by a random value determined by a Gaussian distribution.
Cut-based Analysis: Require 2 jets to reject ttbar and both jets to be b-jets to reject W + Jets and QCD. Cuts on angle btw jets, total jet pT, and Missing ET + pT.
Multivariate Analysis: Require above cuts then discriminate between signal and background using a likelihood function (LF). Input variables to LF chosen according to discrimination power and thresholds set by minimizing uncertainty. There is a set of LFs for each background.
Number of Events
Number of Background Events
Table 1: Monte Carlo samples and their properties
The listed cross sections are theoretical and do not correspond to generator + ME. MCFM was used to derive K-factors in order to scale LO to NLO for W + Jets.
Single-Top Event Pre-Selection
The three single-top processes share a common pre-selection.
Only single-top events with an isolated and high-pT electron or muon in the final state are included in this study. Single-top events with only hadrons in the final state are not considered. The muon and electron channels are exclusive.
- Muons & electrons are reconstructed if:
- ET > 10 GeV and |h| < 2.5
- Isolation ET < 6 GeV in 0.2 cone
- 1 muon or 1 electron with pT > 30 GeV
- Veto events with more than 1 lepton
- Reconstruct jets with
- A cone algorithm with DR = 0.4
- ET > 15 GeV.
- Jet multiplicity between 2 and 4
- At least 2 jets with pT > 30 GeV
- At least 1 b-tagged jet
- Missing ET > 25 GeV
Single-top quarks are produced via the electroweak interaction. At leading order there are 3 production processes; s-channel, t-channel, and Wt-channel. These are shown in Figure 1. Note that each process contains a W-t-b vertex.
Table 5: Results of s-channel multivariate analysis
Figure 4: Likelihood function for ttbar lep + jets
t-channel Cross Section
Main uncertainties are data statistics, b-tagging, ISR/FSR, and bkg cross sections.
Cut-based Analysis: Require b-jet pT > 50 GeV to remove low-pT W + Jets. Require |h| > 2.5 for hardest light jet to remove ttbar (main background) but this cut is not very effective. Results of these cuts are shown in Table 3 for 1fb-1.
Table 6: Uncertainties for s-channel analysis.
Wt-channel Cross Section
Cut-based Analysis: Require one b-jet with pT > 50 GeV. Reject events with more than 1 b-jet (found utilizing a looser weight cut) with pT > 35 GeV to remove ttbar.
Multivariate Analysis: 4 BDTs developed against ttbar (lepton + di-lepton), W + Jets, and t-channel. BDT thresholds set by minimizing total uncertainty. Results are shown below for 1 fb-1 of luminosity.
Table 3: Results of t-channel cut-based analysis.
Multivariate Analysis: Use Boosted Decision Tree (BDT) to remove ttbar instead of cut on Jet |h|. Variables giving a good S/B separation were input into BDT. The BDT output of 0.6 (shown in Figure 3) minimizes total uncertainty and corresponds to S/B = 1.3.
2 2 and 2 3
Figure 3: BDT Output
Triggersselectevents with high pT muons and electrons, which could indicate W decay. Events satisfying any of the following triggers are accepted:
- Muon with pT > 20 GeV
- Isolated Electron with pT > 25 GeV
- Electron with pT > 60 GeV
Trigger efficiencies are shown in Figure 2. Results of pre-selection + trigger are shown in Table 2.
2 2, 2 3, and 2 4
Table 7: Results of Wt-channel cut-based analysis.
Main systematics are ISR/FSR, background cross section, and luminosity.
Background to Single Top
Main systematics are Jet Energy Scale, ISR/FSR, and luminosity.
Figure 1: Single-top production in the s, t, and Wt -channels
Top pair production is the dominant background, with a cross section 3 times higher than that of combined single-top. The single high-pT lepton, 2 b-jets, and missing ET of semi-leptonic top pair decay is most likely to mimic single-top.
W + Jets processes have cross sections many orders of magnitudes higher than the single-top cross sections.
Di-boson events contribute minimally.
QCD will be estimated by data driven methods and is not considered in these studies. Contamination depends on the selections specific to the analyses.
Table 8: Uncertainties for Wt-channel analysis.
For evidence (3s) or discovery (5s):
- t-channel: 5s with 1 fb-1
- s-channel: 3s with 30 fb-1
- Wt-chan: 3s with 1 fb-1, 5s with 10 fb-1
Systematics are the limiting factor for the single-top measurement and have a strong MC dependence in the current analysis.
Table 4: Uncertainties for t-channel analysis.
The single-top cross section is proportional to |fLVtb|2 (where fL is 1 in the SM).
Figure 2: Trigger Efficiencies for single-top events.
Table 2: Results of pre-selection and trigger