Muon Fake Rates

1
Muon Fake Rates

• Concept determine rate of
• fake muons due to pion/kaon punch-through (PT)
• muons out of pion/kaon decay-in-flight (DIF)
• METHOD 1 Use source of identified high-pt
  pions/kaons
• reconstruct Ks pp
• reconstruct D0 Kp
• METHOD 2 Use generic tracks (min bias)
• METHOD 3 Use simulated pions and kaons (fake
  event)

2
Ks Selection

• Use B_CHARM dataset
• L1 (2 XFT tracks, opp. charge, Df lt 1350 , SPt
  gt 5.5) or (7 tracks w/ Pt gt 2)
• L2 2 SVT tracks, c2 lt 25, Pt gt 2, 100 mm lt d0
  lt 1 mm
• L3 2 SVX-COT hybrid tracks, Pt gt 2, h lt 1.2,
  120 mm lt d0 lt 1 mm,
• opp. charge, DZ0 lt 5 cm, 20 lt Df lt
  900, Lxy gt 200 mm, SPt gt 5.5
• Offline selection
• Track cuts Nax gt 24, Nst gt 24, Z0 lt 60 cm,
  h lt 1.2, Pt gt 2
• Track pairs opp. charge, DZ lt 5 cm, 20 lt
  Df lt 900, 1 leg Pt gt 5
• Secondary vertex fit CTVMFT algorithm, x2 lt 8,
  Lxy gt 1 cm,
• d0(Ks) lt
  0.3 mm (SVX) / d0(Ks) lt 0.3 cm (COT)

3
Ks Peak BCHARM sample
Peak 0.46 lt Mpp lt 0.53 SB 0.425 lt Mpp lt 0.46
0.53 lt Mpp lt 0.565
Mpp GeV

4
Pions Kinematics
d0 (COT)
d0 (SVX)
peak
sideband
eta
d0 (SVX) both legs have silicon hits d0 (COT)
only one leg or none have silicon hits

Pt GeV
5
Muon Cuts

• Use same cuts as in s(W mn) analysis
  (summer)
• Tracking Nax gt 24, Nst gt 24
• Calorimeter, isolation E(Rlt0.4)corrected for
  track lt 2
• (will apply this later, study fake rate as
  function of iso)
• Calorimeter, MIP Eem lt max(2, 20.0115 (Pt
  100))
• Ehad lt max(6,
  60.028 (Pt 100))
• Matching DxCMU lt 3 cm, DxCMP lt 6 cm, (no
  cut for CMX)

6
Fake Rate Ks

• Ks in peak 74,843
• (Ptgt2) 149,686
• is m 598 0.40 - 0.02
  Muon fake rate from
• is CMU m 405 0.27 - 0.01
  PT and DIF
• is CMP m 183 0.12 - 0.01
  
• is CMUP m 133 0.09 - 0.01
  peak SB
• is CMX m 146 0.10 - 0.01
  all m 0.38 - 0.02
• 
  CMU m 0.26 -
  0.01
• Events in SB 2,925
  CMP m 0.11 - 0.01
• p (Ptgt2) 5,850
  CMUP m 0.08 - 0.01
• p is m 52 0.89 -
  0.12 CMX m 0.09 - 0.01
• p is CMU m 26 0.44 - 0.09
• p is CMP m 19 0.32 - 0.07
• p is CMUP m 13 0.22 - 0.06
• p is CMX m 21 0.36 - 0.08

7
Fake Rate differential distributions
Muon phi degrees

8
Fake Rate different datasets
ELECTRON
BTST2
BCHARM
PHOTON
DIPHOTON
ALL

9
Fake Rate data and MC
DATA
MC
10
Other samples
electron sample
photon sample

Mpp GeV
Mpp GeV
11
Physics Bias ?
Study physics bias in different data samples
fakerate dependent on business of
the event In muon enriched samples the pion
track gets accidentally matched to the stub from
a real muon nearby Check 1 loosen / tighten
Dx cut on CMU, CMP fake rate / Check 2
difference between isolated/non-isolated
tracks/muons ?

12
Muon Matching
Tendency also seen in electron datasetbut within
statistical error
Dxlt15 cm Dx lt 3 / 6 cm Dx lt 1 cm
ELECTRONS
BCHARM
BCHARM dataset (two-track trigger) is muon
enriched (heavy flavor)

13
Track Isolation
pions from Ks pions from D0 kaons from D0

BCHARM
PHOTONS
ELECTRONS
DIPHOTONS
14
D0 Selection

• Based on same BCHARM dataset as Ks selection
• Selection close to the one in the D0 branching
  ratio analysis
• Use only SVT matched offline tracks
• foffline fSVT lt 0.015 , curvoffline
  curvSVT lt 0.00015 , c2SVT lt 25 , NSVXrf gt 3
• Extra cleanup
• d01 d02 lt 0 (for K-p pair)
• not used reconstruct D peak (D-gtD0 ps) , cut
  on q(ps) ! q(K) ,
• cut on Dm m(Kpps)-m(Kp)

15
D0 Peak
Peak 1.83 1.90 GeV Sidebands 1.765 1.80
1.93 1.965 GeV
MKp GeV
16
Fake Rate D0

17
Fake Rate diff. distr.

• from Ks
• from D0
• K from D0

18
Fake Rate charge dependence

• from Ks
• from D0
• K from D0

19
Comparison to Run1 Numbers

• Method in Run1
• use inclusive photon sample, look for isolated
  high-pt tracks therein
• compute probability for PT per track from
  hadronic interaction lengths
• compute probability for DIF per track from ctp /
  ctK
• sum over all high-pt tracks in sample
• see Thesis by J.Berryhill, PRD of note 5569 no
  dedicated note found

• Results
• In Run1
• 1.87 - 0.99 PT background events
• 1.08 - 1.20 DIF background events
• 2.95 - 2.19 background events DIFPT
• Using (0.49 - 0.04) fake rate for Run2 (1/3 K
  2/3 p)
• Combine with 398 high-Pt tracks
• 1.95 - 0.16 - 0.88 background events
  estimated with this method (DIFPT)

20
METHOD 3 Single Pions and Kaons - 1
Simulate single pions and kaons with fake-event
generator 10 particles per event flat pt
spectrum apply track and muon cuts as in METHOD
1 Pions
Kaons 0.24 - 0.02
(all tracks) 0.58 - 0.03
0.31 - 0.06 (ptlt5 GeV)
0.68 - 0.09 55.5 are DIF
67.9 are DIF

21
METHOD 3 Single Pions and Kaons - 2