colour reconnection in w w
DESCRIPTION
Colour Reconnection in W W . “Particle Flow” CR analysis used by ADLO Basis for combination by LEP WG L published -1.5yr, D ~end 2004, A perhaps not? O -> Ed. Board (tomorrow) Outline Particle flow method Particle flow distributions in data Quantitative measures - PowerPoint PPT PresentationTRANSCRIPT
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 1
Colour Reconnection in WColour Reconnection in WWWColour Reconnection in WColour Reconnection in WWW
“Particle Flow” CR analysis used by ADLO
Basis for combination by LEP WG L published -1.5yr, D ~end 2004, A
perhaps not? O -> Ed. Board (tomorrow)
Outline Particle flow method Particle flow distributions in data Quantitative measures Inclusive multiplicity (Miriam Watson) Systematics Summary
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 2
Colour Reconnection in WColour Reconnection in WWWColour Reconnection in WColour Reconnection in WWW
WWqqqq, default selection
W-j-j association,WWJPLH
4 planes defined by jet axes
from 4-c fit
Most energetic jet jet 1; jet 2 from same W
Choose jets 3, 4 to minimise angles: (j2-j3)+ (j4-j3)
Particles projected onto 4 planes
Interested in particles “between” jets
True for >1 plane? Assign to closest in angle
Highest energy jet
j1 j2
j4 j3
W2
W1
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 4
All data 189-206 GeVAll data 189-206 GeVAll data 189-206 GeVAll data 189-206 GeV
Normalised particle density no-CR models (upper) Sk CR models (lower)
(to be added, with reduced data sample, 189/200/206)
Correlations between bins
Compare intra-W with inter-W
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 5
Ratio intra-W/inter-W particle density: Rflow
no-CR models (upper)
CR models (lower)
Most sensitivity outside jet cores
Particle FlowParticle FlowParticle FlowParticle Flow
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 6
Quantitative measureQuantitative measureQuantitative measureQuantitative measure
Quantify using ratio of sums, RN Inverted since PN448 for ADLO consistency
bin-bin correlations important, integrate event-by-event Range of integral must avoid jet-cores
D assign error as variance in MC subsets, AL use empirical correlation matrix from MC, O calculate error.
8.0
2.0 RR
8.0
2.0 RR
regionsW -interdχdχd1
regionsW -intradχdχd1
nN
nN
R
event
eventN
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 10
Inclusive MultiplicityInclusive MultiplicityInclusive MultiplicityInclusive Multiplicity
First OPAL analyses used inclusive measurements (Nch, xp, etc.)
Main measure used by theorists in all CR papers
From all data, 189-206GeV
n4q: 38.76 +- 0.13 +- 0.27
nqqlv: 19.39 +- 0.11 +- 0.09
=-0.04 +- 0.25 +- 0.02
nqq: 19.39 +- 0.06 +- 0.11
206 GeV206 GeV
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 11
Inclusive MultiplicityInclusive MultiplicityInclusive MultiplicityInclusive Multiplicity
Taus not used due to poorly defined tails in distribution
206 GeV206 GeV
qq onlyqq only
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 12
EEcmcm evolution of evolution of multiplicitiesmultiplicities
EEcmcm evolution of evolution of multiplicitiesmultiplicities
Measure <Nch> for 4q, qqlv, difference
Points are fully corrected data (stat. errs. shown)
Curves hadron level predictions
Data show No significant Ecm
dep. No diff. 4q/qqlv
Average to giveW->qq multiplicity
4q4q qqlvqqlv
4q-2*qqlv
4q-2*qqlv
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 13
Summary of systematicsSummary of systematicsSummary of systematicsSummary of systematics Harmonise systematics in the two analyses as appropriate
Average of all s made, assuming Flat energy dependence Energy dependence correctly described by Koralw and Jetset
This case, average to 199.52 GeV and compare with full set CR models
WW Hadronisation Model differences between {Jt,Hw,Ar, Jt/}
RN – model differences Multiplicity, unfold JT as if background free data, all models
BEC, {intra-W — no-BE}
Background subtraction Z qq, vary production cross-section 5% qqqq, 20%
qqlv Z qq Hadronisation model, max. effect from default to
kk2f+{Py,Hw,Ar} and Py+Py ZZ, vary production cross-section 11% (ZZ PR) 4-f modelling: use KandY in place of Koralw/grc4f for
non-WW-like 4f (+correct WW by -2.5%)
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 14
Summary of systematicsSummary of systematicsSummary of systematicsSummary of systematics
Detector effects Variation of track quality cuts in data/MC
Unfolding method Compare direct multiplicity method with (main) xp
measurement
Energy dependence Difference between Jetset parametrisation and s indep.
Cross-check or RN with qqlv events
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 15
qqlv events for particle qqlv events for particle flowflow
qqlv events for particle qqlv events for particle flowflow
Normalised particle density, using qqlv events no-CR models
Data at s~200 GeV
Compare intra-W with inter-W regions
inter interintra
intra
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 16
Summary of particle flow Summary of particle flow systematicssystematics
Summary of particle flow Summary of particle flow systematicssystematics
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 17
Summary of multiplicity Summary of multiplicity systematicssystematics
Summary of multiplicity Summary of multiplicity systematicssystematics
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 18
Summary of data, CR sensitivitySummary of data, CR sensitivitySummary of data, CR sensitivitySummary of data, CR sensitivity
Predicted stat. sensitivity
Assumes CR models as at 199.5 GeV
Extreme scenario of SKI excluded
Most models completely compatible with data Herwig the least favoured of non-CR models
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 20
Compare average RN with SKI
Best agreement for 34% events reconnected
Scan of reconnection probability in SKIScan of reconnection probability in SKIScan of reconnection probability in SKIScan of reconnection probability in SKI
Better to scan in % reconnected
than model parameter kI:
reconnected fraction asymptotic with kI
% CR
OPAL Thursday Meeting, 21-Oct-2004
Nige Watson 21
SummarySummarySummarySummary Updated analysis using 189—208 GeV data
Data (just) exclude extreme case SKI Data and models with/without CR compatible Now use 2 phase Ar1 model which we implemented for qqlv
Data most consistent with SKI when 34% events reconnected Limits weaker than previously due to detector corrections May be changed during (by) editorial process
We find enhanced sensitivity (~x2) with Ar2/Ar1-2-phase model
Inclusive multiplicity analysis included (Miriam)
Editorial board waiting…