1. M. Poli Lener XI Spring School - "Bruno Touschek" 1 Luminosity measurements with dimuon and single muon reconstruction of Z 0 and W decays OUTLINE:  LHCb apparatus & trigger;  Theoretical uncertainty of Z 0 and W production cross section;  Pythia settings and MC samples;  Performance of the dimuon luminometer (Z 0   );  Performance of the single muon luminometer (W   & Z 0   );  Conclusion M. Poli Lener Details of this work are published in CERN-THESIS-2006-013

2. M. Poli Lener 2 LHCb spectrometer 1 MHz Calorimeters Muon system Pile-up system Level-0: p T of p T of , e, h,  Rough p T ~ 20% 40 MHz 2kHz output HLT: Final state reconstruction Full detector information 40 kHz Level-1: Impact parameter Vertex Locator Trigger Tracker Level 0 objects 380 mrad 15 mrad

3. M. Poli Lener 3 Luminosity measurements at LHCb Relative Luminosity:  Correct for systematic effects Reconstruction and trigger efficiencies  Control the stability of the hardware  Stability of colliding beam conditions Motivations: Absolute luminosity :  Measure (and publish) cross section: - bb inclusive production - prompt charm - weak boson production - constrain Parton Distribution Functions from EW processes  Measure absolute BR of B s Two approaches have been investigated to perform luminosity measurements at LHCb by measuring: 1. vertices of beam-gas interaction through the VELO detector (*) 2. event rates of physical channels with a well known and sizeable cross section (*) L. Ferro-Luzzi, CERN-PH-EP/2005-023

4. M. Poli Lener 4 Theoretical uncertainty Two physical channels are investigated to perform an “on-line” luminometer at LHCb due to theoretical accuracy (~ 4%) and sizeable cross sections at  s = 14 TeV  W B.R.(W   )  10 x  z B.R.(Z   +  -) W.L.van Neerven et al., Nucl. Phys. B382 (2000) 11  x B.R. (nb) W.J. Stirling et al., Eur. Phys. J. C18 (2000) 117 MRST 99, 00 PDF sets NNLO QCD  5%

5. M. Poli Lener 5  The diagrams for the boson V (Z 0 and W) production are: Pythia settings  PDF CTEQ4L is used  The initial state radiation are switched off  Only Z 0 neutral current  interference in matrix elements of Z 0 /  * and  * are disabled  A polar angle  400 mrad is required to the leptons decaying from bosons annihilation V QCD radiation V V QED radiation V Compton scattering (LO)(NLO)

6. M. Poli Lener 6 Z 0   +  - decay process (dimuon luminometer) 25 kevents of Z 0   Single muon coming from W and Z 0 decay (single-muon luminometer) 50 kevents of W ±   ±  5 kevents of Z 0   (with a  not reconstructed) Monte Carlo Samples S z   = L int x  2  tot x (  Z x B.R.) 2  where:  2  tot = (  gen x  rec x  sel x  trig ) 2  (  Z x B.R.) 2   2 nb S 1  = L int x  1  tot x (  x BR) 1  where:  1  tot = (  gen x  rec x  sel x  trig ) 1  (  x BR) 1  tot = (  Z x BR  ) +(  W x BR  )  22 nb The annual signal yield will be, assuming L int =2 fb -1 (1 y =10 7 s & = 2x 10 32 cm -2 s -1 ) : The performances of these two physical processes can be compared

7. M. Poli Lener 7 Performance of the dimuon luminometer

8. M. Poli Lener 8 Acceptance: 4  vs 400 mrad Z0+-Z0+-Z0+-Z0+- Number of events Acceptance (%) Generated in 4  3780 Found in 400 mrad 1087 28.8 ± 0.7 In order to evaluate the generation efficiency (  gen ) in [0, 400] mrad a small pre-production of  4 kevents of Z 0   +  - have been generated in 4    1 vs   2 For the next, I will assume the W   and Z   decays have the same geometrical acceptance efficiency Future work

9. M. Poli Lener 9 Dimuon selection algorithm The signal is represented by a couple of muons (  ln L  > -8) with:  opposite charge  low significance (IP/  IP ) < 5  high p T > 10 GeV/c The strategy of the selection algorithm is a compromise between signalbackground a high efficiency on the signal and a large rejection of the background sources (*) N. Kidonanakis et al., hep-ph/0410367 All the production cross section have been evaluated at NNLO (*) These cuts together with the large di-muon invariant mass are able to totally rejects ~ 15x10 6 of minimum bias and ~ 8x10 6 of b inclusive events. The first two physical channels (Z 0   +  - & tt  W + W - ) are not yet generated: Z 0   +  - decay could be rejected requiring an IP cut due to c  (tau) ~ 100  m, while the tt  W + W - contribution to the signal is at most ~4‰ considering their cross section x B.R.  8 pb against 2 nb of the signal Background processes 60

10. M. Poli Lener 10 Dimuon luminometer efficiencies & results The total signal efficiency  2  tot = (  gen x  rec x  sel x  trig ) 2  can be computed 380 mrad 16 mrad Dimuon invariant Mass GeV/c 2 Asymmetric distribution due to radiation in the final state

11. M. Poli Lener 11 Performance of the single muon luminometer

12. M. Poli Lener 12 Single muon selection algorithm The signal is given by single muon events coming from W or Z 0 (with a  not reconstructed) The first three physical channels (W   , Z 0   +  -, tt  W + W - ) are not yet generated: W    and Z 0   +  - decays could be rejected requiring an IP cut due to c  (tau) ~ 100  m, while the tt  W + W - contribution to the signal is at most ~ 4‰ considering their cross section x B.R.  ~70 pb of the BG against ~ 22 nb of the signal The minimum bias events are not taking into account because ~ 99% events are rejected with the previous “smooth” selection cuts All the production cross section have been evaluated at NNLO (*) Background processes (*) N. Kidonanakis et al., hep-ph/0410367 320 3520

13. M. Poli Lener 13 Signal Background p T spectra Single muon selection algorithm Selection  vertex reconstructed with IP/sigma < 3  p T cut IP/  IP Signal Background To achieve a systematic uncertainty below 4%, a S/B ratio > 25 is needed 22 nb 500  b  conservative p T > 30 GeV/c The single muon selection algorithm is applied on background (~ 8x10 6 bb inclusive) and signal events

14. M. Poli Lener 14 Single muon luminometer efficiencies & results The total signal efficiency  1  tot = (  gen x  rec x  sel x  trig ) 1  can be calculated

15. M. Poli Lener 15 Comparison of luminosity measurements S z   = L int x  2  tot x (  Z x B.R.) 2  with:  2  tot = 14.3% (  Z x B.R.) 2   1.86 nb S 1  = L int x  1  tot x (  x BR) 1  with:  1  tot = 6.1 % (  x BR) 1  tot = 22.13 nb The performances of these two samples can be compared The final annual yield ( L int = 2 fb -1 ) is 5.3x10 5 selected & triggered events  bandwidth of 53 mHz  Z 0   +  - event every ~ 20 s 2.7x10 6 selected & triggered events  bandwidth of 270 mHz  Z 0 or W muon decays every ~ 4 s To perform an “on-line” luminosity measurement with an uncertainly < 4%, 700 events must be collected during data taking ~ 3 1/2 hours~ 45 minutes

16. M. Poli Lener XI Spring School 16 Spares

17. M. Poli Lener XI Spring School 17 L1 Trigger algorithms A new L1 specific algorithm, based on a IP  10 GeV, is introduced in the L1Decision package (v4r5)

18. M. Poli Lener XI Spring School 18 The addition of the new L1 specific algorithm, called low IP muon - reaches a L1 efficiency on the Z 0   signal up to ~ 85% comparable to that obtained with other dimuon processes such as the B 0 s → J/  (µµ)  - requires a limited bandwidth in order to not upset the L1 streaming. The bandwidth can be computed looking at the muons coming from the bb inclusive events which pass L0&L1 trigger (without any selection cuts)  a negligible value of ~ 50 Hz is obtained L1 Trigger algorithms & results

19. M. Poli Lener XI Spring School 19 Used by the dimuon luminometer HLT Trigger data flow

20. M. Poli Lener XI Spring School 20 Parton Distribution Function New PDF sets have been recently updated considering the more recent data from H1 and ZEUS at HERA and CDF and D0 at Tevatron:  Alekhin (*)  CTEQ6 (**)  MRST2004 (***)  ZEUS2005 All these PDFs estimate an uncertainty on the Z and W boson production cross sections of  2÷3 % (*) S.I Alekhin, hep-ph/0508248 (**) J.Pumplin et al., A.D. Martin et al, hep-ph/0201195 (***) A.D. Martin et al, hep-ph/0507015

21. M. Poli Lener XI Spring School 21 ========================================================== I I I I I Subprocess I Number of points I Sigma I I I I I I------------------------------------------------I---------------------------------I (mb) I I I I I I N:o Type I Generated Tried I I I I I I ========================================================== I I I I I 0 All included subprocesses I 7047 120254 I 3.010E-05 I I 1 f + fbar -> Z0 I 2088 10232 I 8.917E-06 I I 15 f + fbar -> g + Z0 I 2596 72079 I 1.127E-05 I I 19 f+ fbar -> gamma + Z0 I 29 743 I 1.260E-07 I I 30 f + g -> f + Z0 I 2334 37200 I 9.787E-06 I I I I I ========================================================== Pythia results

22. M. Poli Lener XI Spring School 22 = 0.1 = 2.5*10 -4 From S. de Capua PhThesis http://sdecapua.home.cern.ch/sdecapua/ Parton momentum distributions

23. M. Poli Lener XI Spring School 23 UP & UPbar distributions vs PDF sets (Q 2 =10 4 GeV 2 )

24. M. Poli Lener XI Spring School 24 DOWN & DOWNbar distributions vs PDF sets (Q 2 =10 4 GeV 2 )

25. M. Poli Lener XI Spring School 25 STRANGE & CHARM distributions vs PDF sets (Q 2 =10 4 GeV 2 )

26. M. Poli Lener XI Spring School 26 BOTTOM & GLUON distributions vs PDF sets (Q 2 =10 4 GeV 2 )

27. M. Poli Lener XI Spring School 27 Single muon selection algorithm The single muon selection algorithm is applied on background (~ 8x10 6 bb inclusive) and signal events Pre-selection  particles identified as muons   ln L  > -2 (standard  ln L  > -8 )  well reconstructed tracks   2  -track < 2.5  ln L  hypothesis Signal Background  2  track Signal Background