Color Reconnection in W Pair Events

1
Color Reconnection in W Pair Events

• Guillaume Leibenguth
• Université Catholique de Louvain
• Belgium

On behalf of the LEP collaborations
DIS 2003, St. Petersburg
2
Introduction

• W pair production at LEPII
• Color reconnection
• W decay range 0.1 fm
• QCD hadronisation scale 1 fm

e
e
W
?,Z
W
?e
W-
e-
e-
W-
qqqq 45 lvqq 45
WW WW
3
Definition and Goal

• Goal CR main systematic effect on the W mass
  measurement in qqqq channel
• ?Mw (Stat) 35 MeV
• ?Mw(CR) 90 MeV
• (reduce the weight of this channel in the
  combination from 45 to 10 )

4
Color Reconnection

• Interconnections in WW qqqq are
  expected
• in QCD, due to the large space time overlap

Perturbative effects of CR (hard gluon exchange
between quarks from W decays) are suppressed
CR implemented in different hadronisation models
5
Color Reconnection Models

• Pythia (string remodelling), CR occurs for
  overlapping strings
• SK1 flux tube (lateral dimension), reconnection
  is based on the overlap O

K being a free parameter

• SK2 strings are treated as vortex lines,
  reconnection
• if lines crosses
• SK2 like SK2 condition that the string
  length is reduced
• GAL generalised area law (usually called
  Rathsman)

Sjöstrand, Khoze (SK) models based on doubly
resonant W!
6
Color Reconnection Models (2)

• Herwig CR
• CR occurs if cluster size can be reduced
• Ariadne CR2 (Reconnection allowed if)
• String length is reduced
• After cascade evolved down to 2 GeV (i.e. ?w)

7
WW multiplicity

• Measure charged particle multiplicities in the
  fully hadronic (4q) and semi-leptonic (2q)
  channels, compare 4q 2 (2q)
• (note different momentum cuts, i.e. no LEP
  combination)

preliminary
8
Fragmentation Function

• CR is expected to change string potential and
  configuration
• Larger differences between fragmentation models
  are observed

9
Particle Flow Definition

• In order to compare the particle production
  from the Ws, one has to take into account that
  the decay products are in different planes.
• (The angle?? is defined with respect to the
  jet direction.)

10
Particle Flow Projection

Rescale angle ?resc ?j /?j1 Define Ratio R
(AB)/(CD)


W1, W2 show the intra-jet region, CR changes
inter-jet activities (i.e. region C and D)
11
Particle Flow Method

• Integrate the most sensitive region,
• use the ratio

Rn
12
Results of Particle Flow Analysis
For LEP combination, normalize to r
Rdata/RnoCR
Distance between CR and no-CR sensitivity
Fit no-CR cannot be ruled out
SK1 100 is excluded
13
Results of Particle Flow Analysis
(Main uncertainty hadronization
modelling(0.0081), background (0.0031))
Rdata smaller than predicted by usual Ariadne
or Herwig!
14
Reminder W Mass Measurement

• Standard analysis direct reconstruction of
  all decay products. It assumes that both W bosons
  decay independently.

W mass best statistical variable. Use it as a
discriminating variable for Color Reconnection.
15
Reduce CR bias

• Remove low energy particles Pcut
• or hybrid cone algorithm

16
and measure Mw

• CR from particle flow
• combined

• Mw alternative estimator

17
Problems Remaining

• ?( Mw(qqqq) , Mw (lvqq)) 22 ? 43 MeV (no FSI
  syst)
• Ariadne 2 does not behave as expected

A problem with the gluon energy parameter?
Use of the Z0 peak data.
18
At the Z0 Peak

• Define rapidity
• Rapidity gap event
• Two populated regions are separated by an
  empty region
• ? look for CR effect

Where is the 3-momentum component wrt
thrust or jet axis
19
Rapidity Gap Analysis

• Select 3 jets events, with defined gluon jet

7 GeV
Ejet lt 35 GeV
Cut on the smallest particle rapidity (a) or on
the largest rapidity difference (b)
20
Results

• Distribution of the charged particle
  multiplicity and total electric charge of the
  leading part of the jet

Ariadne 2-CR, Rathsman-CR Large excess at
Qleading 0 Re-tuning leads to large
chi2 Rathsman and Ariadne 2 are disfavored.
21
Conclusions

• ?(Mw(qqqq),Mw(lvqq)) 22 ?? 43 MeV
• No color reconnection effects
• are observed at the level of 100 MeV.
• Ariadne 2
• LEP collaborations are working on this issue.
• Final publications? Maybe this summer