Constraints on PDF uncertainties from CDF

1

• Constraints on PDF uncertainties from CDF
• DIS 2006, Tsukuba
• 22.04.2006
• Cigdem Issever
• for the
• CDF Collaboration
• University of Oxford

2
Outline

• Introduction
• Tevatron CDF detector
• EWK Results (95 of the talk)
• Jet results (see talk Inclusive jet production
  at the Tevatron (CDF) of Olga Norniella in
  (HFS-5))
• Conclusion

3
Tevatron
proton-antiproton collisions ?s 1.96 TeV (Run
I ? 1.8 TeV) 36 bunches 396 ns crossing
time Peak luminosity is now 1032 cm-2
s-1 Ultimately 4 9 fb-1 by 2009
4
CDF Recorded Data
Jet results with 1.0 fb-1
1.6 fb-1 delivered 1.2 fb-1 recorded
EWK results
5
CDF RUN II Detector

• Upgraded for RUN II
• New silicon tracking
• New drift chamber
• Increased muon coveraged
• New TOF
• New plug calorimeters

Muon
COT Tracker
Si Detector
PLUG
EM Cal
Had Cal
CDF Data taking effi 80 - 85.
Silicon Detector
6
EWK Physics Input to PDFs

• Motivation
• Test SM (precise measurements)
• Constraints on PDFs
• Search for physics beyond SM
• Important input to LHC
• Outlook
• W forward cross section, 223/pb
• Z ?tt and µµ cross Section, 330/pb
• W charge asymmetry, 170/pb

7
W/Z Gauge Bosons Identification

• At hadronic collider W and Z bosons hadronic
  decays are overwhelmed by QCD background.
• ? identification through leptonic decays

W?e?
Z???
PTgt20GeV
ETgt20GeV
Position of µ consistent with extrapolated track
PTgt20GeV
W signature Isolated Energetic Lepton ET
Z Signature Two Isolated Energetic Leptons
(opposite charge)
8
W cross section in the forward region
Extension into forward region 1.2 lt ? lt 2.8
using calorimeter seeded tracking
Complementary to central
9
W cross section in the forward region
Systematics on A 0.2567
48165
4.8
2
0.07
Axe
10
W cross section in the forward region
223pb-1
s 2.796 /-0.013(stat) 0.095 0.090 (syst)
/- 0.168 (lum.)nb.
NNLO s(pp?W) _at_ 1.96 TeV, Stirling, van Neerven
2.687 - 0.054(Th)
11
Central-to-Forward W vis. cross section ratio

• s(visible)sTOTA where A is the kin. and geo
  accept.
• Strategy assign sys uncertainties but PDF,
  NLO/NNLO effect to svis
• In this way
• Most of the luminosity uncertainty cancels in the
  ratio
• All other uncertainties are uncorrelated
• Accuracy can be used to constrain PDFs

12
Central-to-Forward W vis. cross section ratio

• svis(central) 664.211.7 pb (Etegt25, ETngt25,
  helelt1)
• svis(forward) 71821 pb (Etegt20, ETngt25,
  1.2lthelelt2.8)
• svis(central)/svis(forward) 0.9250.033
• 1 assigned as luminosity syst. (slightly
  overestimate)
• NLO ratios (taking into account correlations
  between central and forward)
• CTEQ 0.92430.037
• MRST01E 0.941370.011
• Most uncertainties will go down with more data ?
  useful to constrain PDFs

13
Z ? µµ cross section (? lt 1) using 337 pb-1
337pb-1
116
66
s261.2 2.7 (stat) 5.8 - 6.9 (sys) 15.1
(lum) pb
NNLO _at_ 1.96 TeV Stirling, van Neerven
s(pp?Z)251.3-0.5(Th)
14
Z ?te th cross section using 349 pb-1

• 316 signal events
• 60 t identification efficiency and 5
  acceptance
• Most systematics are data driven will be reduced
  with more stat.

s265-20(stat)-21(syst)-15(lumi) pb
15
Cross section summary
new
16
W Charge Asymmetry
Asymmetry in W production complicated by unknown
n pz use lepton asymmetry
which convolves W production with V-A decay.
17
W Charge Asymmetry Run II 170pb-1
A as function of ET provides better probe of x
dependence. Statistic allowed two bins. Will be
included into next generation of PDFs.
18
W Charge Asymmetry new method
Lepton asymmetry has turn over at high ? due to
V-A W charge asymmetry does not have this
effect, so we dont purely probe high yW

• Determination of yW with W mass constrain gives
  2 possible solutions.
• Evaluate weight factor F1,2 for each y1,2
  solution.
• Parameterize F1,2 with
• the angular distribution of (1-cosT)2
• with W cross section, s(yW), but this depends on
  asymmetry
• Iterative procedure!!

19
W charge asymmetry new method
Iterative procedure

• Smaller statistical errors
• Greater sensitivity
• No additional systematics due to new method

20
Midpoint jet cross section
Good agreement with NLO
More details see talk of Olga Norniella in
(HFS-5) Jets 1
21
Results with KT Data/NLO 1fb-1
IR and CL safe No splitting or merging
Measurements in the forward region will allow to
reduce the PDFs uncertainties
22
Conclusions

• New cross section measurements from CDF
• W ? e? in forward region (1.2 lt ? lt 2.8) using
  223 pb-1
• Central-to-forward W vis cross section ratio
• Z ? µµ using 337 pb-1
• Z ? te th using 349pb-1
• Inclusive Jets with Mitpoint using 1.04 fb-1
• Inclusive Jet s with Kt algorithm using 0.96 fb-1
• Excellent base for next set of analyses
• ds/dy for W ? e?
• ds/dpt for Z ? µµ
• Tau widely used in SM measurements and SUSY,
  Higgs
• New generation of WZ measurements
• (R, W Charge Asymmetry, )
• on the way !!

23
Backup Slides
24
W forward cross section
25
Z ? µµ cross section (? lt 1)
26
Z ?te th cross section

• taus difficult to reconstruct at hadron
  colliders
• Z?tt exploits event topology to suppress
  backgrounds (QCDWjet)

• CDF strategy for hadronic tau reconstruction
• Charge tracks ? define signal and isolation cone
  (shrinking cone vs. E)
• isolation require no tracks in isolation cone
• Hadronic calorimeter cluster (to suppress e
  background)
• p0 required in isolation cone (identified by
  shower maximum detector)

30o

• Z?tt event selection
• t?e electron isolated track (ETgt10 GeV)
• t?h PT(seed) gt 6 GeV PT(signal)gt15 GeV
• remove backgrounds by event topology cuts

27
Z?tt cross section
28
Z?tt cross section -- Systematics
29
W Asymmetry new method
Leading order W production
from Bo Young Han
30

• I. The angular distribution of ( )2
  from W production

• in Collin-Soper frame
• The W production Probability from angular
  distribution

ratio of two angular distributions at each
rapidity
from Bo Young Han
31

• II. Weight must also depend on W- cross-section.
• But cross-sections depend on W asymmetry!
• This method must be iterated.
• III. Iteration procedure

measuring asymmetry
reconstruction
Input data
if no,
min( )
the closest asymmetry to data
F1
Fn
assumed sample
new assumed sample
Yes
No
from Bo Young Han
32
Sensitivity Study

• , 400pb-1 MC data generated by Pythia
• Selecting W events
• high PT electron ET gt 25 GeV
• Missing ET gt 25 GeV
• Used CTEQ6M errors PDF 40 sets for PDF
  uncertainty
• Comparison of statistical uncertainty between
  lepton and W boson asymmetry
• Our method has statistical sensitivity to probe
  PDFs

from Bo Young Han
33
Systematic Uncertainty

• Weight Factors depend on Q(yW, PtW) and s(yW)
• Ratio of two angular distributions, Q(yW, PtW)
• PDF dependence
• W cross section, s(yW)
• PDF dependence

from Bo Young Han
34
Systematic Uncertainty (cont.)

• The uncertainties from the energy measurement
• Energy scale
• Energy resolution (not yet)
• Electron ET scale
• 0.1(1s) ? lt 1.1
• 0.15(1s) ? gt 1.1
• Missing ET scale
• W boson Recoil energy tuning

from Bo Young Han