1. 1 W boson mass and width measurements at LEP2 Hugo Ruiz, CERN – Aleph On behalf of the LEP Collaborations

2. 2 Outline nIntroduction Measurement of M W (and  w ) by direct reconstruction nSome relevant systematics: –Bose-Einstein correlations –Colour Reconnection nResults and conclusions nProspects

3. 3 Introduction nm W in perspective : –From pp colliders (transverse mass spectra, single Ws, april 2004): m W = 80.452  0.059 GeV (CDF run1+D0 run1+UA2) –Prediction from EW fit: m W = 80.373  0.033 GeV (LEP1,SLD) m W = 80.386  0.023 GeV (+Mtop) nAt LEP2, W’s produced in pairs: 1996-2000: ~40k WW evts  precision measurement nm W measured from direct reconstruction of W decay products

4. 4 The measurement

5. 5 Event selection q q q q q q Semileptonic (qq l ) Hadronic (4q) Leptonic 10%44%46%

6. 6 W reconstruction nHadronic and semileptonic channels: 1.Reconstruct leptons and cluster jets 2.Apply a kinematic fit: –Constraints: E,p conservation In some cases, M 1 =M 2 /  1 =  2 –Effects: alows ‘reconstruction’ of in semileptonic channel resolution ~7GeV  ~3GeV decreases detector systematics nHadronic channel: 3. Pair jets (there is a 3-fold ambiguity): –algorithms provide ~85% of good pairing

7. 7 Reconstructed M W

8. 8 m W and  w extraction nExtraction by fitting distribution of reconstructed M W to either: –MC samples generated with different M W (and  W ) values –Function convoluting BW and detector effects, and then use MC to correct for residual offsets Mass extraction  Assume SM relation between M and  èPerform 1-parameter fit Mass extraction  Assume SM relation between M and  èPerform 1-parameter fit Width extraction  Assume no relation between M and  èPerform 2-parameter fit Width extraction  Assume no relation between M and  èPerform 2-parameter fit nRely on MC simulation: –Lots of data for tuning from LEP1 –Residual discrepancies  systematics

9. 9 Systematic uncertainties for m W Source  M W (MeV) Correlatio ns Detector simulation Mostly from calo energy calibration 15 Channel, Year LEP energy through kinematic fit 17 Channel, Year, Experiment Fragmentation18 Channel, Year, Experiment Interconnection effects 9 (90 MeV in hadr.) Year, Experiment nExpected final statistical error for LEP2 ~ 25 MeV nLargest systematic uncertainties: next sections of the talk

10. 10 Interconnection effects on hadronic events

11. 11 –parton shower (large Q 2, pQCD) nFragmentation (quarks  hadrons): Interconnection effects nHard process: e + e -  4q e+e+ e-e- W+W+ W-W- q _q_q q _q_q nInterconnection effects (not included in standard MC models): –Bose-Einstein correlations: momenta of identical bosons tend to be correlated. d~0.1 fm –Colour reconnection: hadronic interaction between W decays d(W +,W - ) < 1 fm –hadronisation (phenomenological) nEvent simulation:

12. 12 Bose-Einstein Correlations (BEC) Pairs of  0  0,  +  + and  -  - tend to be bunched. nFor calculation of BEC, quantum phases and space-time structure are needed  only phenomenological models available. nEffect on M W : W1W1 W2W2 –Intra-W: not relevant for M inv –Inter-W: Cause discrepancies data-MC in jet overlaps  jet clustering different data-MC  bias.

13. 13 BEC in W + W - events nBEC effects experimentally established in Z jets at LEP1 nInter-W BEC? Analyses performed in 4 LEP experiments to search/limit them –Observable: distance in p-space between pairs of charged pions: Q 2 ij =-(p i -p j ) 2  Inter-W BEC correlations disfavoured  Limit on systematic:  M W ~ 15 MeV L3 0 1 Q(GeV) LEPWW/FSI/2002-02 fraction of model seen

14. 14 Colour Reconnection nSeveral phenomenological models used to study the effect on m W, amongst them: –At parton shower: Ariadne2: formation of some inter-W dipoles,  M W ~ 70 MeV –At hadronisation: Herwig-CR: hadrons created from inter-W parton pairs,  M W ~ 40 MeV Rathsman: reduce string tension by reconnecting,  M W ~ 40 MeV Jetset SK1: allow formation of inter-W strings,  M W up to 400 MeV, depending on a free parameter nDedicated analyses try to observe / limit CR effects from data on 4 LEP experiments

15. 15 The particle flow analysis nCR models predict a modified particle flow in W + W - events: CR:No CR: W-W- W+W+ W-W- W+W+ nObservable: ratio of particle flow between the inter and intra-W regions: (A + B) / (C + D) A B C D Data - SK1 (extreme parameter) - Jetset

16. 16 Results from particle flow  Try to make analyses more robust to CR effects CR Prob nFor SK1: –Extreme values discarded –Preferred value of the parameter corresponds to  M W ~ 100 MeV!! ‘Asymmetry’ from experiments combined in a  2. nCannot discard models like: –Ariadne2 –Herwig-CR –Rathsman

17. 17 Towards a less CR sensitive analysis

18. 18 Principle nInterconnections mainly occur between low momentum particles in the inter-W region. nIdea: modify clustering algorithm to dismiss information from those particles. This implies: –“purer” information –loss of statistical precision nMany variations of jet algorithms tried, mainly: Cones: perform angular cut around jet direction P-cuts: remove low momentum particles

19. 19 Reduction of  M W nExample: SK1  M W (MeV) ModelStandard Cone R=0.5 rad SK1, k I ~210040 Herwig3510 AR 25020 Rathsman4015 nGood reduction factors are obtained for all available models nExample: Cone (R=0.5 rad), with a statistical loss of ~ 25%:  parameter

20. 20 A by-product: measure CR? nThe difference between M W measured with ‘robust’ and standard analyses is sensitive to CR effects: DELPHI, Cone algorithm R=0.5 nDELPHI preliminary: –Exclude extreme scenarios. –Minimum at ~1.3, P~0.5 Cone radius (rad) ALEPH SK1 k=2.13 M W (GeV) MWMW

21. 21 Results

22. 22 m W from direct reconstruction Non-4q 4q  m W = 22 ± 43 MeV Results in CERN-EP/2003-091, LEPEWWG/2003-02 still with standard jet algorithms

23. 23 m W at LEP2 m W = 80.412 ± 0.042 GeV 0.029 stat, 0.031 syst nLEP2 combination:nWorld average: m W = 80.425 ± 0.034 GeV

24. 24  w at LEP2  W = 2.150 ± 0.091GeV 0.068 stat, 0.060 syst  W = 2.150 ± 0.091GeV 0.068 stat, 0.060 syst nCombination: Detector: Detector: 29 MeV Frag: Frag: 30 MeV FSI: FSI: 37 MeV Detector: Detector: 29 MeV Frag: Frag: 30 MeV FSI: FSI: 37 MeV nWorld average:  W = 2.133 ± 0.069 GeV

25. 25 Conclusions nThe combination of the results of the LEP experiments gives: m W = 80.412 ± 0.042 GeV  W = 2.150 ± 0.091 GeV m W = 80.412 ± 0.042 GeV  W = 2.150 ± 0.091 GeV nConsistency of m W within SM:

26. 26 Prospects nPublication status: –BEC studies already published for L3 Phys.Lett.B 547 (2002) and ready for OPAL (CERN-PH-EP/2004-008, to Eur. Phys. Journal C). –CR studies ready for publication for L3 (CERN-EP/2003-12, to Phys. Lett. B). –All the rest will be ready by this autumn/end of this year nExpected developments related with color reconnection: –Final estimation of effect –Analysis with improved robustness. If all experiments use them: Total error in hadronic channel: ~90  ~60 MeV. Total error from decrease by ~3 MeV Weight of hadronic channel in combination: 0.1%  0.3%. + Learn something about colour reconnection