Title: QCD and FSI at LEP
1QCD and FSI at LEP
- aS measurements
- event shapes
- power corrections
- Colour Reconnection
- Bose-Einstein Correlations
- Summary
2aS Motivation
- aS the fundamental, universal QCD parameter
- SM predicts ?s evolution, not absolute value
- Perturbative effects 1/lnQ
- Non-perturbative effects 1/Q
- Test measure different processes, energies
- Intuitive techniques in ee-
- Precision low, O() cf. electroweak O(10?5)
3LEP Luminosity
4Luminosity at LEP2 Energies
- 272 Niobium film and 16 Nb bulk Superconducting
Cavities - Average accelerating field 7.5 MV/m
5ee- ? Hadrons Data
Illustrative numbers, per experiment
- LEP1
- High statistics, ? 0 background or ISR
- LEP 2
- O(10-3) LEP1 statistics, large 4-fermion
contamination, ISR effects - Variable energy, averaged to nominal ?s
6aS from Z0 Parameters
- G(Z?hadrons) sensitive to QCD radiation, gives
- Inclusive
- Negligible hadronisation uncertainty
- Reliable calculations
- Other variables also affected GZ, ,
- Combined fit, all EW data for best measurement
- theory error
0.002
7Sensitivity of Z0 Observables
- most sensitive
- t hadronic width similar measurement, calculation
- Non-perturbative effects
- Estimated OPE 0.0140.005
- Constrained by mass spectra
- Precision 0.003 at mZ
8aS from Event Shapes
- QCD calcs. divergant for colinear or soft gluon
emission - Use robust observables
- e.g. thrust
- Further variables heavy jet mass MH, C-parameter
C, total and wide jet broadenings BT and BW,
diff. 2-jet rate (Durham, y3)
9Measurements
- Select hadronic final states, rejecting
- 4-fermion-like
- Hard ISR
- Observables formed from charged particles,
neutrals, or energy flow objects - 1-T, MH, BW, BT, y3, C-parameter
- 4-fermion background subtracted
- Bin-by-bin corrections
- Acceptance
- Resolution
- ISR contamination
- Fair description of data by MCs
Correction factor
4-fermion background
(MC-data)/data
error
10as Fits
- pQCD predictions corrected for hadronisation
using MC - NLO O(as2) pQCD prediction for event shape
variables, y - Small y behaviour of cumulative cross-section at
LO - Diverges
- At O(aSn), leading term (aSL2)n ? large!
- Saved by resummation, NLLA prediction
Leading, sub-leading logs to all orders in aS
11as2 NLLA Matching
O(aS2 ) calc.
ln R(y)
NLLA calc.
?
- LogR or R matching avoids double counting of
terms - Use modified matching schemes, ensures L0 for
yymax
12Fit Results
Scaled heavy jet mass
Thrust
Total jet B
Wide jet B
L3 data
C-parameter
13Systematic Uncertainties
- Experimental (little correlation, full
covariance) - Event selection, particle reconstruction,
detector corrections vary cuts or models - Background subtraction (4-f criteria, sgg , etc.)
- ISR corrections (LEP2)
- Typically around 1
- Hadronisation (moderate correlation, on-diagonal
covariance) - Model comparisons string (Pythia), cluster
(Herwig), colour dipolestring (Ariadne) - Model parameter variation (Pythia)
- typically around 0.7?1.5
- Theoretical, pQCD (large correlation, on-diagonal
only) - LEP QCD WG devised new prescription
-
14Theoretical Uncertainty
- Uncertainty band obtained (for fixed as),
varying - Renormalisation scale
- Rescaling factor
- L? 1/ln(y.xL)
- Kinematic limit ymax
- Modification degree
- For fixed reference prediction (lnR) find as
variation which covers this band (within the fit
range) - Typically 3.5 ? 5
BT
15Combined Result
?s evolution of as
- LEPQCDWG combination summer 02
- 6 observables
- 8 nominal ?s (L3 ISR)
- 4 experiments
- Results evolved to MZ
16Power Law Approach
- 3 loop level
- (Landau) Pole at ?sL
17Power Law Approach
- Alternative to MC derived hadronisation
corrections - Expect non-pert. effects suppressed 1/Qn
- Parametrise behaviour of as in non-pert. region
D-M-W -
-
- Assumes strong coupling remains finite ?s ? L
- Predictions of form
- means
- distributions
mI 2 GeV a0 universal, same all y
P ? 1/Q Dy differs y shift (T,C,MH2) from
2-j shift/squeeze (BT,BW)
18Power Law Corrections
as(Mz)0.1184 ? 0.0033
(inc. ? 0.0031 (scale))
19Power Law Corrections
- Example of fits to total jet broadening
- Uses data ?s 12-189 GeV
- Note re-analysed JADE data
- 1979-1986, MC resurrected
From P.A. Movilla Fernandez et al.,
Eur.Phys.J.C22(2001)1
20Global (aS,a0) Fit to Means
From P.A. Movilla Fernandez et al.,
Eur.Phys.J.C22(2001)1
?BT?
?1-T?
?MH2/s ?
?C?
?y3?
?BW?
21Global (aS,a0) Fits
- Systematic shift, aSpow vs. aSMC from
distributions - Combined results (JADE, prel., summer 02)
- Universality of a0 at 25 level (1-2 s total
error)
224-jet Rate
- O(aS3) (NLO) predictions for k? Dixon, Signer,
Glover, Nagy, Trocsanyi - ALEPH (xm1)
- as(Mz) 0.1170 ? 0.0001 (stat)
- ? 0.0003 (had.)
- ? 0.0008 (scale)
- 0.1170 ? 0.0013
- xm 0.729 (fit for scale and aS)
- as(Mz) 0.1175 ? 0.0013
Similar to earlier DELPHI results
High sensitivity to aS at LO
23Colour Reconnection Motivation
Winter 2003 Summary
- Single dominant uncertainty
- Final state interactions
- WW?qqqq only
?(W?W?) decay vertices ? 0.1 fm hadronic
scale ? 1 fm ? Large spacetime overlap ?
Colour exchangeW??W? ? DMW bias ? 25-300 MeV
- LEP gives best mW measurement
- Agreement, direct/indirect
24Predicted mW bias
- Comparison of mass bias ADLO
- Analysis of shared MC sample
- Consistent results
- Potentially large problem
25Example Hadronic B Meson Decay
Gustafson, Petterson, Zerwas Phys.Lett.
B209(1988)90
- B ? J/Y X
- Colour suppressed decay
BR.exp 1 BR.fullCR 3-5 BR.noCR 0.3-0.5
c/o P.Abreu
26Charged Particle Multiplicity
- CR may alter nch in qqqq
- W hadronisation modelling
- Compare qqqq/qqlv
- More important for low p
- Studied by all LEP expts.
- ALEPH update, all data
- Conclusion only limited sensitivity to CR
D?nch? 4q ? 2(qqlv) 0.31 0.23 0.10
(within acceptance)
27Particle Flow
- Motivated by simple string picture of CR
- Regions of interest
- Define 4 planes
- Pair jet-jet ? W
- Minimise ??(j2-j3)??(j4-j1)
- Project particles ? planes
- Compare intra-W ? inter-W
- ? string effect
- Define RN ? ? intra-W
- ? inter-W
- (away from jet cores)
28Projection of Particles
- 4 jets, not coplanar
- 6 possible jet-jet regions
- Project onto 4 jet-jet planes intra-W and
inter-W - Consider region away from jet cores
29Event Selection
- Topological
- 4 distinct jets, y34gt0.01
- 2 angles ? 1000
- 1000 ? angles ? 1400
- Large/small not adjacent
- Good jet-jet ? W
- Efficiency 15
- Correct pairing 90
- W mass
- Minimise ??(j2-j3)??(j4-j1)
- Pairing integral to selection
- Efficiency 85 (A), 40 (O)
- Correct pairing 75 (A), 90 (O)
30Particle Density
- Normalised particle density
- Effect of CR shown
- Raw particle density
- Non-planar
- 4-jet structure evident
31Particle Flow
- Ratio intra-W/inter-W particle density Rflow
- no-CR models
- CR models
- Most sensitivity outside jet cores
- Statistically limited
- Combine LEP expts.
extreme case
32Quantitative Measure
- Quantify using ratio of sums, RN
- Different experimental acceptances, normalise to
shared no-CR MC sample before comparison - Very different selections, weight by sensitivity
for each CR model, i
33Combination Systematics
- Each expt. evolves their data to single ?s point
and averages - Systematics considered correlated or uncorrelated
between expts. - WW signal
- Hadronisation model, spread in predictions of
Koralw Jt,Hw,Ar - BEC, ?intra-W no-BE no evidence for
inter-W BEC - (4-jet) Background subtraction
- Z ? qq, vary sqq ?10
- Z ? qqqq, vary sZZ ?15
- Z ? qq hadronisation models
- Energy dependence
- Model dependence of ?s evolution
- Detector effects
34LEP CR Combination SK-I model
- NB Extreme, 100 SK-I
- Vary reconnected fraction in combination
- Preferred Preco 49 in data
- Increases DmW from LEP
r?RN(x)/RN(no-CR)
35Ariadne and Herwig CR models
r?RN(x)/RN(no-CR)
r?RN(x)/RN(no-CR)
36CR from Rapidity Gaps
Planar 3-jet event
- CR reduces particle production in
- central y region
- Increased prob. of rapidity gap
CR suppressed O(1/Nc2)
37CR from Rapidity Gaps
CR and no-CR models describe inclusive data
- Select ee-?qqg
- Anti-tag g by b decays
- 10k g jets, ?Ejet? 23 GeV, P 94
Highest E
Select 4 gluon jets with rapidity gaps
38CR from Rapidity Gaps
39Leading Part of Gluon Jet
No. charged particles
- Isolated, neutral, leading g jet system
- sensitive to CR
- and glueball prodn.
- Unable to tune away effects cannot describe
inclusive Z0 and rapidity gap data - (Current implementation of) Ariadne/Rathsman CR
models strongly disfavoured
Net charge
40Bose-Einstein Correlations
- Enhanced identical boson pairs (?? or ??),
small Q2 -(p1p2)2 - Firmly established phenomena, LEP1 and intra-W
- Studied using 2-particle correlation function
- R(p1, p2) ?2(p1, p2)/?0(p1, p2) Many
experimental factors! - Essential question do BEC exist between W and
W ?? - Problem 1
- Reference ?0 should be identical to ?2,but
without BEC - Unlike-sign data, / ratio of same in
MC (resonances) - Like-sign MC without BEC (MC modelling)
- Event mixing
- Problem 2
- Non-pQCD amplitudes unknown, resort to models
- Phenomenological parametrisation R(Q) ? 1
?exp(-r2Q2)
Source radius
BE strength
41B-E Analysis Method
- Idea Chekanov, De Wolf, Kittel,
Eur.Phys.J.C6(99)403Â - If W and W decays uncorrelated, 2-particle
density - ?2WW(p1,p2) ?2W(p1,p2) ?2W(p1,p2)
2?1W(p1) ?1W(p2) - Subtract background
- Form ratio Dlhs/rhs
- D/ D(data) / D(MC, intra-W BEC) remove
potential residual bias - D ? D/ ? 1 ? non-independent W decays
- inter-W BEC (or similar effect)
?MIXWW(p1,p2) mix 2 x qql? ? BEC ? 0
Assume ?2W?2W-?2W Estimate from qql?
42(Final) L3 results
Phys.Lett.B547(2002)139
- MC tuned to Z-gtqq (udsc)
- LS/ULS behaviour OK
- D (ratio data/MC) ? D
- Enhancement predicted for inter-W BE model
- No effect seen in data
- Quantify effect using BE32
Like-sign
Unlike-sign
43Combination ?s189-207 GeV
- Fit to
- L0.008 ? 0.018 ? 0.012
- k0.4 ? 0.4 ? 0.3 fm
- Inter-W BE gives
- L0.098 ? 0.008 (stat.)
- This model diagrees with data 3.8s
- BE models a problem...
- Delphi/L3 not inconsistent
- Combined LEP results very soon now (D, L so far).
44Summary
- as
- New LEP average, LEP1LEP2 event shapes
- Improved prescription for theoretical
uncertainties - V. precise as from 4-jet rate, mean valuespower
corrections - Colour Reconnection
- First combination, Summer 2002, all data
- Extreme case SKI 100 excluded (favour Preco
49) - Data/models compatible, with/without CR.
- Impact of qqqq channel on LEP mW 9!
- New Z0 rapidity gap study, further constrains
models - Bose-Einstein
- L3 final results, no support for inter-W BEC
- Delphi/L3 not inconsistent, wait for A/O!
-
45Theoretical Uncertainty
- Thrust has large, well-behaved fit region
- Uncertainty band obtained (for fixed as) via
variations - Renormalization scale
- Rescaling factor l1/ln(xL y)
- kinematic constraint
- Modification degree
- For fixed reference prediction (LogR) find as
variation which covers this band (within the fit
range) - Typically 3.5 ? 5
46LEP results
47W Mass
- Dominant systematics
- Hadronisation 18MeV
- Compare models
- Use LepI data MLBZ D
- LEP Energy 17MeV ?MWMW?Ebeam/Ebeam
- Resonant depolarisation, NMR probes/flux loop
- LEP spectrometer
- QCD Final State Interactions
- Colour Reconnection
- Bose-Einstein Correlations
qqqq channel weight 10!
48MW, qqqq vs. qqlnl
DMW(qqqq-qqln) 2243 MeV
(in fit neglecting BE/CR effects)
49WW ? lnl lnl
- Typical (perfect) event
-
- 2 charged leptons
- ? 2 neutrinos
- Branching fraction 11
- Typical efficiency 80
- purity 90
- ? Underconstrained, limited impact on mW
measurements
50WW ? qqlnl
- Typical (perfect) event
-
- 2 well-separated hadronic jets
- 1 charged lepton, 1 hard n
- Branching fraction 44
- Typical efficiency 85
- purity 90
- ? Reconstruct n from (E, p) constraints
51WW ? qqqq
- Typical (perfect) event
-
- 4 well-separated hadronic jets
- Branching fraction 45
- Typical efficiency 87
- purity 80
- ? Jet-jet ? W ambiguity
- ? Colour Reconnection, BEC (FSI)