Measurement of the Top Mass from the Leptons PT in the Dilepton Channel at CDF

1
Measurement of the Top Mass from the Leptons PT
in the Dilepton Channel at CDF
XXV Workshop on Recent Developments in High
Energy Physics Cosmology National Technical
University, Athens, Greece

• Victoria Giakoumopoulou
• 30 March 2007

2
Outline

• Why measure the top quark
• Motivation
• Present Status
• Methodology
• Dilepton Event Selection
• PT spectra
• Top Mass with the Maximum Likelihood Method
• Systematic errors
• Conclusions prospects

3
Why study the top quark

• Currently the heaviest particle of the SM
• Accurate measurement of the top quark mass
• ?better localization of SM
  Higgs mass

??W a Mt2
??W a lnMH
4
Fermilab
Main Injector
Tevatron
5
CDF detector
6
Top Pair Production Decay
W decays define channel Dilepton 11.12
Leptonjets 44.44 All-hadronic 44.44
7
Measurements of the Top Mass
CDF D0 March 07 (170.91.1(stat)
1.5(syst))GeV/c2
CDF (170.51.3(stat) 1.8(syst))GeV/c2
Moderate (CDF RUN II) or very high (LHC)
statistics of top production are expected
soon The statistical error will decrease The
systematic error will dominate
8
Sensitivity to the Top Mass

• The main contribution in the systematic
  uncertainty is associated with jets
• Proposal (University of Athens CDF/ATLAS Group)
  ? Seek a variable that
• does not depend on the jet energy scale
• does not depend on b-tagging
• it is the same for both the Dilepton and the
  Leptonjets channels
• Such a variable could be the leptons (µ/e) PT/ET
• This can be calibrated against Z?ll to within 50
  MeV

9
Dilepton Signal

• Expect to observe
• two high PT leptons of opposite sign
• large missing ET from the two vs
• two or more jets

Signal
Background

• DrellYan Z?µµ, ??ee
• Fakes
• Diboson WW/WZ/ZZ
• Z?tt

• WW ? 16.13
• WZ ? 4.1
• ZZ ? 1
• Z?tt ? 11.48
• DY ? 28.04
• Fakes ? 38.26

We expect _at_ 1200 pb-1
23.11 background events 55.95 signal
events
S/B2.42
CDF note 8741
10
Methodology Outline

• Estimation of the Top Mass using the Lepton PT
  information
• Create lepton PT templates for the signal for
  several top masses
• PT is linearly dependent to the Top Mass
• Create lepton PT templates for backgrounds
• Fit the PT templates with Gamma x Fermi function
• Implement Maximum Likelihood method to estimate
  the Top Mass

11
PT sensitivity to Top Mass
PT sensitive to the top quark mass Linear
dependency ltPTgt ?MT? SLOPE ?13.52
12
Shape of Leptons PT

• Parameterization of PT
• Fit 30 PT histograms for MT155-192 GeV/c2
• Find the optimum mass dependent parameters
  pa1MTa2 and qa3MTa4
• Fit PT of combined background
• Find the optimum p and q parameters

13
Likelihood

• N number of data events
• nbexp expected number of background events
• snb expected background uncertainty
• ns number of estimated signal events
• nb number of estimated background events

14
Example of mass estimationMC_at_1200pb-1
Input Mass 175 GeV
79 signal bg events
15
Sanity tests - only Signal
MC_at_1200pb-1 200PE
ltPullgtp1Minp0
ltPull_RMSgtp1Minp0
ltMFgtp1Minp0
p0-0.190.25 p10.0010.001
p0-0.960.22 p10.0000.001
p0-1.94.5 p11.020.03
16
Fitted Mass Signal Background
MC_at_1200pb-1 200PE
p00.0170.255 p1-0.00040.001
p00.955.7 p10.9980.033
p00.910.23 p10.00030.001
17
Sources of systematics

• Gluon radiation in the initial state (ISR)
• Estimated by using the official CDF Pythia
  samples with more and
• less ISR
• Choice of Parton Distribution Functions (PDF)
• Estimated using the sets CTEQ5L and MRST75
• Choice of Monte Carlo generator
• Estimated by comparing the results of Pythia
  and Herwig with the
• same input top mass
• Uncertainty associated with the measurement of
  leptons PT
• Uncertainty in the expected background shape

18
Measurement of lepton PT

• Possible pT scale uncertainty and non-linearity
• Tune using Z ? dilepton data (CDF note 8632)
• Bremsstrahlung from the detected electrons
• Found to be negligible

Preliminary estimate of the PT scale uncertainty
on the Mtop is lt 1 GeV
Adapted from CDF Joint Physics Group
19
Systematic errors
20
Projected statistical errors - Tevatron
21
LeptonJets (LJ) Blessed analysis by CDF _at_
340pb-1
Signal

• Expect to observe
• high PT lepton
• large MET
• 4 jets

t?Wb? lvb t?Wb?qqb
Background

• Wjets qq ? W ? l? (80)
• Jets faking leptons (20)

Composition _at_ 340pb-1

• Signal 70 expected events
• Background 125 expected events
• (100 W 15 fake e 10 fake µ)
• Data 197 events

22
LJ Blessed result _at_ 340pb-1
Mtop(220 47.4(stat) 9.3 (syst) ) GeV/c2
23
Summary - Prospects

• New method to measure the Mtop from the lepton(s)
  PT has been proposed and developed by the UoA
  group
• Method blessed (approved) by CDF
• Present systematic error 4 GeV
• Projected combined Ljets Dilepton uncertainties
  on MT for ?Ldt10fb-1 (Tevatron end of RUN II)
• 5 GeV statistical
• 1-2 GeV systematic
• Bright prospect for Mtop at LHC

24

• CDF Notes
• CDF/DOC/TOP/PUBLIC/8741
• CDF/PHYS/TOP/PUBLIC/8632
• CDF/PHYS/TOP/PUBLIC/8617
• CDF/PHYS/TOP/PUBLIC/8616
• CDF/PUB/TOP/PUBLIC/7888
• ?JINR/E1-2005-104 Particles and Nuclei, Letters
• CDF/PHYS/TOP/CDFR/7685
• CDF/PHYS/TOP/CDFR/7684

• UoA CDF group
• Victoria Giakoumopoulou
• Athanasios Staveris-Polykalas
• Constantinos Vellidis
• Arkadios Manousakis-Katsikakis
• Nikolaos Giokaris

• Work partly supported by
• EPEAEK II program in the framework of the
  projects
• Heraklitos
• Pythagoras
• GSRT
• ELKE - UoA

25
BACKUP
26
Consistency of Fit

• ltPTgt vs Mtop from ltPTgt ?MT?, where
  ?0.1327 and ?32.56

• ltPTgt from G x Fermi model

Perfect agreement !! Model works fine
27
Luminosity
28
Combined top masses _at_ 1.2fb-1
29
PE example
30
Statistical Error _at_Tevatron/LHC (LJ)