A proposal for the W boson mass measurement at CDF

1
A proposal for the W boson mass measurement at CDF

• Yu Zeng
• Duke University

2
Outline

• The Standard Model
• Motivation of W mass measurement
• Historical W mass measurements
• W mass measurement strategy at CDF
• Summary

3
The Standard Model
H?
ZW
4
The Higgs mechanism for boson masses

• The SM electroweak Lagrangian is given by

with covariant derivative

• We can get boson masses via Higgs mechanism

5
The Higgs boson

• In SM, Higgs mechanism gives rise to the Higgs
  boson.
• Mass of the Higgs is not determined by the
  theory.
• Direct searches at LEP sets lower bound -
  MH ? 114.4 GeV 95 CL
• Precision electroweak measurements sets upper
  bound - MH lt 144 GeV 95 CL
• Direct searches are being / will be carried on
  at Tevatron and LHC.

CERN website
6
Motivation

• Constrain the mass of undiscovered Higgs.

7
Theoretical calculation

• Writing out ?r terms (S. Dawson)

8
Relationship Mw, Mtop and MH

• Current Mw, Mt relationship

9
Motivation

• If Higgs is found, precise Mw can be used to
  infer non-SM particles. SUSY as one example.

10
Historical W mass measurements
11
Measurement at CDF (200 pb-1)

• W production and decay via leptons _at_ Tevatron -
  s(p pbargWX) Br(Wgen) 2.7 nb - produced 1
  in 50?106 collisions
• Use data collected from Feb. 2002 Sept. 2003
• - electron channel (L 218 pb-1)
• - muon channel (L 191 pb-1)
• Event selection leads to clean samples -
  mis-identification 0.05

12
Transverse mass fitting results (200 pb-1)
T. Aaltonen et al., CDF Collab.,
hep-ex/0708.3642, submitted for publication in
PRD.
13
Implication for the SM

• Including the CDF W mass measurement (200 pb-1)

Left/Right before/after CDF 200 pb-1 Mw
measurement
14
Implication for Tevatron

• In 2004, the estimated upper limit for Higgs mass
  is 250 GeV, however Tevatron only reach upper
  limit 170 GeV, people think Tevatron has no
  chance to find Higgs.

Now Tevatron is back into the competition.
15
Analysis using 2 fb-1 at CDF

• About 10 times more data to analyze
• - expect smaller statistical and
  systematic uncertainties
• Use Wgen and Wgmn channels to measure W mass -
  branching ratio 11
• Use Zgee and Zgmm channels as control sample -
  branching ratio each 3.3
• Use J/?, ? resonances to calibrate momentum
• Use E/p to calibrate electron energy
• Use information in transverse plane
• - information along beam direction
  incomplete
• Use fast Monte Carlo simulation to extract MW

16
Collider Detector at Fermilab (CDF)
17
The CDF detector
18
The CDF detector (quadrant)

• Select W and Z bosons with central (h lt1)
  leptons

19
W boson production and decay
MW

• Quark-antiquark annihilation dominates (80)
• W transverse motion due to gluon / sea quark
  involved production process (20)

• W event signature
• - a single, isolated, high pT charged
  lepton
• - a large missing energy due to neutrino

20
Electron identification
MW
High pT electron will

• Leave a track in tracking device
• Deposit a significant amount of energy in EM
  calorimeter

Identification can be improved by

• Matching the energy in calorimeter and the
  momentum of the pointing track.
• Cutting on measured energy in areas surrounding
  the electron shower.

21
Muon identification
MW
Muon will

• Leave a track in tracking device
• Leave hits in the muon chambers
• Very little energy deposition in calorimeters

Constraining using beam spot will

• Increase momentum resolution
• Reject cosmic ray events

22
Neutrino inference
MW
Neutrino

• Will not leave direct information
• Can only be inferred in an indirect way using
  momentum conservation.

In transverse plane

• Imbalance of the vector sum of lepton pT and UT
• UT is the transverse momentum carried by hadronic
  recoil

23
The hadronic recoil UT
MW
Three contributions to the recoil

• Jet recoiling off the W
• Underlying energy - multiple interactions -
  remnants of the ppbar collisions
• Bremsstrahlung - photons emitted by lepton which
  are not in the excluded region

24
Measurement strategy
MW

• Use leptonic decay modes (e, m)
• Use quantities in transverse plane - n info
  along beam direction unknown
• Transverse mass

• Extract MW from transverse mass spectrum mT -
  fitting mT (data) with mT (MC)

25
Transverse mass spectrum
MW
(Figure from I. Vollrath PhD thesis)

• W mass information contained in the location of
  Jacobian edge

• Relatively insensitive to PT(W)

• Sensitive to detector response to recoil
  particles.

26
Measurement strategy
MW
Data
Binned Likelihood Fit
W boson mass
NLO event generator Detector response
simulation Hadronic recoil modeling
pi(m) is the predicted probability in bin i, ni
is of data entries in bin i.

Backgrounds
W mass templates, bule for 80 GeV, red for 81 GeV
27
Projected W mass precision
MW

• Statistical uncertainties are expected to
  decrease as N-1/2
• Systematic uncertainties from measurements that
  are obtained from control data samples (expected
  to decrease as N-1/2)
• Systematic uncertainties from theoretical
  calculations (unchanged)
• Assuming a constant theoretical uncertainty of 20
  MeV (blue line).

D. Waters, Wmass Workshop 2007
28
Summary

• Precise W mass measurement, in conjunction with
  the top quark mass measurement, can constrain the
  Higgs boson mass.
• Use transverse information of Wgen and Wgmn
  channels for MW measurement at CDF.
• MW is extracted by fitting mT(data) vs. mT(MC).
• With 10 times more data, we expect to reach the
  25 MeV uncertainty goal in MW.

29
Backup slides
30
Choices of SM parameters
MW
Physical Quantity No.
Fermion masses (6 quark 3 lepton) 9
Higgs Boson 1
Quark weak mixing parameter 4
Strong CP violation parameter 1
Strong interaction coupling constant 1
Fundamental EWK parameters 3
Neutrino masses 3
Neutrino mixing parameter 4
Can be chosen from
31
Choices of electroweak parameters
MW
Choice 2.
Choice 1.
Follow the pattern that parameters are masses and
coupling constants.
Choose parameters measured most precisely.
32
Why two coupling constants
MW
Thus, only two coupling constants
1) ae2/(4phc)1/137 2) aS for
strong coupling
33
Boosts along beam axis

• Define rapidity
• Boosts along the beam axis z (so cos?1) with
  vbb will change y by a constant yb
• Boost of velocity bb along z axis
• - pz ? g(pz bb E)
• - E ? g(E bb pz)
• - Transform rapidity
• Pseudo-rapidity neglecting mass (b1)

34
Particle identification
MW

• Particle detectors measure long-lived particles
  produced from high energy collisions electrons,
  muons, photons and stable hadrons (protons,
  kaons, pions)
• Quarks and gluons do not appear as free
  particles, they hadronize into a jet.

35
Some facts about CDF detector
MW

• Central Out Tracker - Hit position resolution
  140 mm - Momentum resolution s(pT)/pT
  COT alone 0.15 pT-1 COT beam
  constrained 0.15 pT-1
• Central Calorimeter - CEM energy resolution
  s(E)/E 13.5/sqrt(Esinq) - CHA energy
  resolution s(E)/E 0.5/sqrt(E)

36
Main backgrounds
MW

• For Wgmn - largest background comes from
  Zgmm - Wgtngmnn events - cosmic rays -
  kaon decays in flight - events where one jet
  contains one non-isolated m
• For Wgen - Zgee- - Wgtngem - events
  where one jet contains one non-isolated e

37
Event Selection for W Z
MW

• Select clean W Z samples to get maximum ratio
  of S/N - trigger info pt(e, m) gt18 GeV -
  central lepton selection hlt1 - final
  analysis pt(e, m) gt30 GeV - W boson further
  requires uTlt15 GeV, Et gt30GeV - Z
  boson two oppositely charged leptons with
  opposite

38
Helicity and Handedness
MW

• Helicity - spin projection (l) of a particle
  along its direction of motion - e.g., l1/2 for
  e- l-1/2 for e - ? A? A-?-
• Handedness - a particle state projected out by
  (1?g5)/2 - ?R (1g5)/2 ? ?L (1-g5)/2 ? - ?
  ?R ?L

39
V-A nature of W boson decay
MW

• The fact that W couples only to left-handed
  quarks/leptons or right-handed anti-quarks/anti-le
  ptons are confirmed by experimentally observed W
  decay asymmetry.

• cos? -1 is favored, which means e goes
  predominantly in the direction opposite to the
  original proton.

40
Jacobian edge

• First work with ds/dpTe with pT(W)0 case
• - assume pT(W)0 - Since
  mT24(pTe)2, transfer to ds/dmT
• Spreading around MW is due to ?W, pT(W) no equal
  to zero.

41
dmT/mT
42
Diagram