RHIC-PV, April 27, 2007 M. Rijssenbeek 1
The Measurement of W’sat the CERN and FNAL hadron
colliders
• W’s at RHIC !• W’s at CERN – UA2• W’s at FNAL - CDF
RHIC-PV, April 27, 2007 M. Rijssenbeek 2
W’s at RHIC !• Measurement of W’s in polarized-proton collisions at
RHIC:– measure the u and d-quark and anti-quark
contributions to the proton spin as function momentum fraction x.
– use the W charge and the V–A structure of W production & decay to eν and μν to select quark flavor and quark helicity.
( )d x ( )u x
For W–: u↔d
RHIC-PV, April 27, 2007 M. Rijssenbeek 3
Non-Hermetic Detectors• RHIC detectors are not fully hermetic… • electron and muon acceptance of the RHIC
detectors varies strongly over the rapidity range…• Thus:
– missing transverse energy cannot be used to clean up the W signal
– Trigger has to rely exclusively on high pT electrons and muons
– The backgrounds to the W from Z→e/μ and QCD/fakes may be significant
– enough Z-statistics for measurement/tuning of corrections?
• Detailed simulations will be crucial to determine acceptance corrections, efficiencies, and backgrounds– limited Z acceptance makes tuning of the simulations with
Z→ee/μμ more difficult
RHIC-PV, April 27, 2007 M. Rijssenbeek 4
a “Non-Hermetic” Detector: UA2 – vs.1• Central tracking + Preshower + EM
Calorimetry; |η|<1
• Forward spectrometer + PS + EM Calorimetry; 1<|η|<3
a non-hermetic detector…
UA2 collaboration: M Banner et al., Phys. Lett. 122B (1983) 476.
UA2 collaboration: P Bagnaia et al., Phys. Lett. 129B (1983) 130.
RHIC-PV, April 27, 2007 M. Rijssenbeek 5
W and Z in UA2• strong quality selections on
electron candidates necessary:– isolation, shower shape,
preshower signal, track-PS-cluster match
– even so: cut on missing pT for final sample…
all good EM cluster pairs
+ Track & PS match
Z
W
for eTE p 0.8e
TT E
p e
TT E
eTM
eTM
W
vs and eTE
RHIC-PV, April 27, 2007 M. Rijssenbeek 6
W and Z in UA2 – vs.2• 1987: UA2 Upgrade program with hermetic calorimetry
Tp
QCD
UA2 collaboration: J.Alitti et al., Z.Phys.C47 (1990) 11.
RHIC-PV, April 27, 2007 M. Rijssenbeek 7
W‘s at Fermilab• the measurement of the W mass with a 0.05%
accuracy (50 MeV or 100×Me) requires the ultimate understanding of the detector!
simulations and cross checks & tuning with the data itself…
RHIC-PV, April 27, 2007 M. Rijssenbeek 8
Simulations• the state of simulations of W and Z production
(and decay) has much advanced over the past decades– forced by very high statistics W/Z samples for mass
determination from LEP and Tevatron– much QCD calculational progress– improved detector simulations: showering – availability of raw computing power allows more detail
and increased sophistication
• State-of-the-Art: – RESBOS: NLO W and Z production – CTEQ6M: NLO pdf’s with uncertainties
Note: pTe/μ is quite sensitive to boson recoil
MT less so, but is sensitive to yW/Z
• CRUCIAL for all precision W/Z measurements
RHIC-PV, April 27, 2007 M. Rijssenbeek 9
FNAL Example: CDF W Mass(from seminar by Dr. David Waters, UC
London)• CDF is a modern hermetic detector:
hermetic detection of e, γ, jets; less hermetic for μ
– recent MW measurement (e+μ) is single best in the world: MW
CDF = 80413 ± 34 (stat) ± 34 (syst) MeV
cfr: WA 2006: 80392±29 MeV
65 MT 90
2 (1 cos( ))e eT T TM p p
CDF Note 8665, Jan 17, 2007; CDF http://www-cdf.fnal.gov/
RHIC-PV, April 27, 2007 M. Rijssenbeek 10
W/Z Production in pp & Decay Modeling
+ Corrections:Higher orders (EW,QCD)Non-perturbative
Leading Order picture:
dpp W / Z ll
[ f iq (x p ) f j
q (x p )i, ju,d ,s,(c,b ) f i
q (x p ) f jq (x p )]d
qq W / Z ll dx pdx p
rapidity distribution
dqq W / Z ll
(s , l,l ) couplings
1
(s - MW/Z
2 )2 (W/Z
s /MW/Z )2
angular & mass distributions :
p
p
W /Z
l
l
pT distribution
RHIC-PV, April 27, 2007 M. Rijssenbeek 11
W Production Modeling: pT
• Use the best theoretical model on the market :– RESBOS NLO QCD + resummation + non-perturabtive.
• Constrain the parameters g1, g2, g3 and lineshape with the Z data:
W : 7 MeV
MW : 3 MeV
1 3 RE2 SBOS~ (1 ) ( , , )ZTZ
T
dp fB gg g
dp
g20.640.05
b 0.00140.0010 GeV-1
(Landry et al., 2003)
RHIC-PV, April 27, 2007 M. Rijssenbeek 12
W→μν Backgrounds• mostly Z→μμ: easy to lose a muon (at CDF !)• But can estimate this background very reliably.
RHIC-PV, April 27, 2007 M. Rijssenbeek 13
Decay-In-Flight Background in W→μν
• difficult background: very flat in transverse mass
• Use track quality: 2 and track impact parameter
fin
al
cu
t v
alu
e
/NDF
xxx
x
xx
x
x
K,fake
high-pT track
W : 27 MeV
MW : 5 MeV
Vary normalization &
shape:
Z provides the template for real
muons
High impact parameter cuts provide the DIF template
RHIC-PV, April 27, 2007 M. Rijssenbeek 14
W→eν Backgrounds
Z’s ~ negligible
QCD background dominates
RHIC-PV, April 27, 2007 M. Rijssenbeek 15
QCD Background in W→eν • Multijet events: large σ Rfake(jete)
Rfake(jetET)
fin
al
cu
t v
alu
e
QCD template from a
background-rich “anti-electron“
sample
W : 32 MeV
MW : 7 MeV
Vary normalization &
shape:
RHIC-PV, April 27, 2007 M. Rijssenbeek 16
CDF W Mass 2007:• sample size analyzed: 200/pb
• event selection: pTe,μ>30 GeV, pT
ν>30 GeV
• Uncertainties in MW for fit to pTe/μ:
RHIC-PV, April 27, 2007 M. Rijssenbeek 17
W/Z Physics result: Van Neerven Plot:
• MW/MZ with Mtop constrain the SM Higgs range; 200/pb
• Current data sets: 2/fb (CDF, DØ)
• FNAL Expectation: δMW≈ 30 MeV/expt
RHIC-PV, April 27, 2007 M. Rijssenbeek 18
My Conclusions• A precise measurement of Δu and Δd will bring new
understanding of the spin structure of the baryons…
• However: in order to obtain the required precision, the measurement will need sophisticated simulations to understand and model the detector acceptance, efficiencies, and backgrounds – the physics and detector models must be tuned and
checked with measurements of the Z (<10% of W statistics)
– Dominant backgrounds must be measured with the data itself
these simulations must be done beforehand to prove the measurement capability with RHIC’s “non-hermetic” detectors…