1. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 1 MPI@LHC 2010 Rick Field University of Florida Outline of Talk  Discuss the relationship between UE and MB..  Show how well the CDF UE tunes predict the CDF MB data.  PYTHIA 6.4 Tune Z1: New CMS 6.4 MB tune (pT-ordered parton showers and new MPI). Review of Tevatron MB and UE Results Glasgow, Scotland November 2010  Examine versus Nchg for MB and UE at the Tevatron and the LHC.  Review some CDF Run 2 studies that never got published.

2. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 2 QCD Monte-Carlo Models: High Transverse Momentum Jets  Start with the perturbative 2-to-2 (or sometimes 2-to-3) parton-parton scattering and add initial and final- state gluon radiation (in the leading log approximation or modified leading log approximation). “Hard Scattering” Component “Underlying Event”  The “underlying event” consists of the “beam-beam remnants” and from particles arising from soft or semi-soft multiple parton interactions (MPI).  Of course the outgoing colored partons fragment into hadron “jet” and inevitably “underlying event” observables receive contributions from initial and final-state radiation. The “underlying event” is an unavoidable background to most collider observables and having good understand of it leads to more precise collider measurements!

3. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 3 Traditional Approach  Look at charged particle correlations in the azimuthal angle  relative to a leading object (i.e. CaloJet#1, ChgJet#1, PTmax, Z-boson). For CDF PTmin = 0.5 GeV/c  cut = 1. Charged Particle  Correlations P T > PT min |  | <  cut  Define |  | 120 o as “Away”. Leading Calorimeter Jet or Leading Charged Particle Jet or Leading Charged Particle or Z-Boson  All three regions have the same area in  -  space,  ×  = 2  cut ×120 o = 2  cut ×2  /3. Construct densities by dividing by the area in  -  space. “Transverse” region very sensitive to the “underlying event”! CDF Run 1 Analysis

4. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 4 CDF RUN 2 Analysis  Study the charged particles (p T > 0.5 GeV/c, |  | < 1) in the “Transverse 1” and “Transverse 2” and form the charged particle density, dNchg/d  d , and the charged scalar p T sum density, dPTsum/d  d . Charged Particles p T > 0.5 GeV/c |  | < 1 1 charged particle in the “transverse 2” region dN chg /d  d  = 1/(4  = 0.48 Area = 4  /6  The average “transverse” density is the average of “transverse 1” and “transverse 2”. AVE “transverse” (Trans 1 + Trans 2)/2

5. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 5  Distributions  Shows the  dependence of the charged particle density, dN chg /d  d , for charged particles in the range p T > 0.5 GeV/c and |  | < 1 relative to jet#1 (rotated to 270 o ) for “leading jet” events 30 < E T (jet#1) < 70 GeV.  Also shows charged particle density, dN chg /d  d , for charged particles in the range p T > 0.5 GeV/c and |  | < 1 for “min-bias” collisions. Leading Jet Min-Bias 0.25 per unit  -  Log Scale!

6. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 6  Distributions  Look at the “transverse” region as defined by the leading jet (JetClu R = 0.7, |  | 150 o ) with almost equal transverse energies (E T (jet#2)/E T (jet#1) > 0.8).  Shows the  dependence of the charged particle density, dN chg /d  d , for charged particles in the range p T > 0.5 GeV/c and |  | < 1 relative to jet#1 (rotated to 270 o ) for 30 < E T (jet#1) < 70 GeV for “Leading Jet” and “Back-to-Back” events. Refer to this as a “Leading Jet” event Refer to this as a “Back-to-Back” event Subset

7. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 7 Min-Bias “Associated” Charged Particle Density  Use the maximum p T charged particle in the event, PTmax, to define a direction and look at the the “associated” density, dN chg /d  d , in “min-bias” collisions (p T > 0.5 GeV/c, |  | < 1). “Associated” densities do not include PTmax! Highest p T charged particle!  Shows the data on the  dependence of the “associated” charged particle density, dN chg /d  d , for charged particles (p T > 0.5 GeV/c, |  | < 1, not including PTmax) relative to PTmax (rotated to 180 o ) for “min-bias” events. Also shown is the average charged particle density, dN chg /d  d , for “min-bias” events. It is more probable to find a particle accompanying PTmax than it is to find a particle in the central region!

8. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 8 Min-Bias “Associated” Charged Particle Density  Shows the data on the  dependence of the “associated” charged particle density, dN chg /d  d , for charged particles (p T > 0.5 GeV/c, |  | 0.5, 1.0, and 2.0 GeV/c. Transverse Region  Shows “jet structure” in “min-bias” collisions (i.e. the “birth” of the leading two jets!). Ave Min-Bias 0.25 per unit  -  PTmax > 0.5 GeV/c PTmax > 2.0 GeV/c Rapid rise in the particle density in the “transverse” region as PTmax increases!

9. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 9 Min-Bias “Associated” Charged Particle Density  Shows the data on the  dependence of the “associated” charged particle density, dN chg /d  d , for charged particles (p T > 0.5 GeV/c, |  | 0.5 GeV/c and PTmax > 2.0 GeV/c compared with PYTHIA Tune A (after CDFSIM).  PYTHIA Tune A predicts a larger correlation than is seen in the “min-bias” data (i.e. Tune A “min-bias” is a bit too “jetty”). PTmax > 2.0 GeV/c PTmax > 0.5 GeV/c Transverse Region PY Tune A

10. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 10 Min-Bias “Associated” Charged PTsum Density  Shows the data on the  dependence of the “associated” charged PTsum density, dPT sum /d  d , for charged particles (p T > 0.5 GeV/c, |  | 0.5 GeV/c and PTmax > 2.0 GeV/c compared with PYTHIA Tune A (after CDFSIM).  PYTHIA Tune A predicts a larger correlation than is seen in the “min-bias” data (i.e. Tune A “min-bias” is a bit too “jetty”). PTmax > 2.0 GeV/c PTmax > 0.5 GeV/c Transverse Region PY Tune A

11. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 11 Back-to-Back “Associated” Charged Particle Densities  Use the leading jet in “back-to-back” events to define the “transverse” region and look at the maximum p T charged particle in the “transverse” region, PTmaxT. “Associated” densities do not include PTmaxT!  Look at the  dependence of the “associated” charged particle and PTsum densities, dN chg /d  d  and dPT sum /d  d  for charged particles (p T > 0.5 GeV/c, |  | < 1, not including PTmaxT) relative to PTmaxT.  Rotate so that PTmaxT is at the center of the plot (i.e. 180 o ). Maximum p T particle in the “transverse” region!

12. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 12 Back-to-Back “Associated” Charged Particle Density  Look at the  dependence of the “associated” charged particle density, dN chg /d  d  for charged particles (p T > 0.5 GeV/c, |  | 0.5 GeV/c, PTmaxT > 1.0 GeV/c and PTmaxT > 2.0 GeV/c, for “back-to-back” events with 30 < E T (jet#1) < 70 GeV.  Shows “jet structure” in the “transverse” region (i.e. the “birth” of the 3 rd & 4 th jet). Jet#2 Region “Associated” densities do not include PTmaxT! Log Scale! ??

13. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 13 Back-to-Back “Associated” Charged PTsum Density  Look at the  dependence of the “associated” charged particle density, dPT sum /d  d  for charged particles (p T > 0.5 GeV/c, |  | 0.5 GeV/c, PTmaxT > 1.0 GeV/c and PTmaxT > 2.0 GeV/c, for “back-to-back” events with 30 < E T (jet#1) < 70 GeV.  Shows “jet structure” in the “transverse” region (i.e. the “birth” of the 3 rd & 4 th jet). Jet#2 Region “Associated” densities do not include PTmaxT! Log Scale! ??

14. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 14 Back-to-Back “Associated” Charged Particle Densities  Shows the  dependence of the “associated” charged particle density, dN chg /d  d  for charged particles (p T > 0.5 GeV/c, |  | 0.5 GeV/c, PTmaxT > 1.0 GeV/c and PTmaxT > 2.0 GeV/c, for “back-to-back” events with 30 < E T (jet#1) < 70 GeV.  Shows  dependence of the charged particle density, dN chg /d  d  for charged particles (p T > 0.5 GeV/c, |  | < 1) relative to jet#1 (rotated to 270 o ) for “back-to-back events” with 30 < ET(jet#1) < 70 GeV. It is more probable to find a particle accompanying PTmaxT than it is to find a particle in the “transverse” region! “Back-to-Back” “associated” density “Back-to-Back” charge density

15. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 15 Back-to-Back “Associated” Charged Particle Densities  Shows the  dependence of the “associated” charged particle density, dN chg /d  d , p T > 0.5 GeV/c, |  | 0.5 GeV/c, |  | < 1 relative to jet#1 (rotated to 270 o ) for “back-to-back events” with 30 < E T (jet#1) < 70 GeV. Polar Plot 0.5 1.0 1.5 2.0 “Back-to-Back” charge density “Back-to-Back” “associated” density

16. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 16 Back-to-Back “Associated” Charged Particle Densities  Shows the  dependence of the “associated” charged particle density, dN chg /d  d , p T > 0.5 GeV/c, |  | 2.0 GeV/c (not including PTmaxT) relative to PTmaxT (rotated to 180 o ) and the charged particle density, dN chg /d  d , p T > 0.5 GeV/c, |  | < 1, relative to jet#1 (rotated to 270 o ) for “back-to-back events” with 30 < E T (jet#1) < 70 GeV. Polar Plot “Back-to-Back” “associated” density “Back-to-Back” charge density 0.5 1.0 1.5 2.0

17. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 17 Jet Topologies  Shows the  dependence of the “associated” charged particle density, dN chg /d  d , p T > 0.5 GeV/c, |  | 2.0 GeV/c (not including PTmaxT) relative to PTmaxT (rotated to 180 o ) and the charged particle density, dN chg /d  d , p T > 0.5 GeV/c, |  | < 1, relative to jet#1 (rotated to 270 o ) for “back-to-back events” with 30 < E T (jet#1) < 70 GeV. Polar Plot 0.5 1.0 1.5 2.0 QCD Three Jet Topology QCD Four Jet Topology

18. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 18 “Associated” Charge Density PYTHIA Tune A vs HERWIG HERWIG (without multiple parton interactions) too few “associated” particles in the direction of PTmaxT! And HERWIG (without multiple parton interactions) too few particles in the direction opposite of PTmaxT!

19. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 19 “Associated” PTsum Density PYTHIA Tune A vs HERWIG HERWIG (without multiple parton interactions) does not produce enough “associated” PTsum in the direction of PTmaxT! And HERWIG (without multiple parton interactions) does not produce enough PTsum in the direction opposite of PTmaxT! PTmaxT > 0.5 GeV/c

20. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 20 “Associated” Charge Density PYTHIA Tune A vs HERWIG For PTmaxT > 2.0 GeV both PYTHIA and HERWIG produce slightly too many “associated” particles in the direction of PTmaxT! But HERWIG (without multiple parton interactions) produces too few particles in the direction opposite of PTmaxT!

21. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 21 “Associated” PTsum Density PYTHIA Tune A vs HERWIG For PTmaxT > 2.0 GeV both PYTHIA and HERWIG produce slightly too much “associated” PTsum in the direction of PTmaxT! But HERWIG (without multiple parton interactions) produces too few particles in the direction opposite of PTmaxT! PTmaxT > 2 GeV/c

22. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 22 Jet Multiplicity  Shows the data on the number of jets (JetClu, R = 0.7, |  | 3 GeV) for “back-to-back” events with 30 2.0 GeV/c. Max p T in the “transverse” region!  Compares the (uncorrected) data with PYTHIA Tune A after CDFSIM.  Compares the (uncorrected) data with HERWIG (without MPI) after CDFSIM. Data have about equal amounts of 3 and 4 jet topologies! HERWIG (without multiple parton interactions) does not have equal amounts of 3 and 4 jet topologies! Rick Field KITP Collider Workshop 2004

23. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 23  Shows the data on the tower ETsum density, dET sum /d  d , in the “transMAX” and “transMIN” region (E T > 100 MeV, |  | < 1) versus P T (jet#1) for “Leading Jet” and “Back-to-Back” events.  Compares the (corrected) data with PYTHIA Tune A (with MPI) and HERWIG (without MPI) at the particle level (all particles, |  | < 1). “Leading Jet” “Back-to-Back” Rick Field University of Chicago July 11, 2006 Neither PY Tune A or HERWIG fits the ETsum density in the “transferse” region! HERWIG does slightly better than Tune A!

24. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 24 Rick Field University of Chicago July 11, 2006  PYTHIA Tune A fits the charged particle PTsum density for p T > 0.5 GeV/c, but it does not produce enough ETsum for towers with E T > 0.1 GeV.  It is possible that there is a sharp rise in the number of particles in the “underlying event” at low p T (i.e. p T < 0.5 GeV/c).  Perhaps there are two components, a vary “soft” beam-beam remnant component (Gaussian or exponential) and a “hard” multiple interaction component. Possible Scenario?? Two things changed at the LHC! New center-of-mass energies (900 GeV & 7 TeV) AND the charged particles were measured at smaller p T values!

25. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 25 Proton-Proton Collisions  tot =  EL  SD   DD   HC The “hard core” component contains both “hard” and “soft” collisions. “Inelastic Non-Diffractive Component” ND  tot =  EL  IN

26. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 26 The Inelastic Non-Diffractive Cross-Section + + + + … “Semi-hard” parton- parton collision (p T < ≈2 GeV/c) Occasionally one of the parton-parton collisions is hard (p T > ≈2 GeV/c) Majority of “min- bias” events! Multiple-parton interactions (MPI)!

27. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 27 The “Underlying Event” Select inelastic non-diffractive events that contain a hard scattering + + + … “Semi-hard” parton- parton collision (p T < ≈2 GeV/c) Hard parton-parton collisions is hard (p T > ≈2 GeV/c) The “underlying-event” (UE)! Multiple-parton interactions (MPI)! Given that you have one hard scattering it is more probable to have MPI! Hence, the UE has more activity than “min-bias”. 1/(p T ) 4 → 1/(p T 2 +p T0 2 ) 2

28. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 28 Model of  ND + + + + … “Semi-hard” parton- parton collision (p T < ≈2 GeV/c) Allow leading hard scattering to go to zero p T with same cut-off as the MPI! Model of the inelastic non- diffractive cross section! Multiple-parton interactions (MPI)! 1/(p T ) 4 → 1/(p T 2 +p T0 2 ) 2

29. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 29 UE Tunes + + + + … “Underlying Event” “Min-Bias” (ND) Fit the “underlying event” in a hard scattering process. Predict MB (ND)! 1/(p T ) 4 → 1/(p T 2 +p T0 2 ) 2 Allow primary hard-scattering to go to p T = 0 with same cut-off! “Min-Bias” (add single & double diffraction) Predict MB (IN)! All of Rick’s tunes (except X2): A, AW, AWT,DW, DWT, D6, D6T, CW, X1, and Tune Z1, are UE tunes!

30. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 30 MB Tunes + + + + … “Underlying Event” “Min-Bias” (ND) Predict the “underlying event” in a hard scattering process! Fit MB (ND). Most of Peter Skand’s tunes: S320 Perugia 0, S325 Perugia X, S326 Perugia 6 are MB tunes!

31. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 31 MB+UE Tunes + + + + … “Underlying Event” “Min-Bias” (ND) Fit the “underlying event” in a hard scattering process! Fit MB (ND). Most of Hendrik’s “Professor” tunes: ProQ20, P329 are MB+UE! The ATLAS AMBT1 Tune is an MB+UE tune, but because they include in the fit the ATLAS UE data with PTmax > 10 GeV/c (big errors) the LHC UE data does not have much pull (hence mostly an MB tune!). Simultaneous fit to both MB & UE

32. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 32 Charged Particle Multiplicity  Data at 1.96 TeV on the charged particle multiplicity (p T > 0.4 GeV/c, |  | < 1) for “min-bias” collisions at CDF Run 2 (non-diffractive cross-section).  The data are compared with PYTHIA Tune A and Tune A without multiple parton interactions (pyAnoMPI). No MPI! Tune A!

33. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 33 PYTHIA Tune A Min-Bias “Soft” + ”Hard”  Comparison of PYTHIA Tune A with the p T distribution of charged particles for “min-bias” collisions at CDF Run 1 (non-diffractive cross-section). 12% of “Min-Bias” events have P T (hard) > 5 GeV/c! 1% of “Min-Bias” events have P T (hard) > 10 GeV/c!  PYTHIA Tune A predicts that 12% of all “Min-Bias” events are a result of a hard 2-to-2 parton-parton scattering with P T (hard) > 5 GeV/c (1% with P T (hard) > 10 GeV/c)! Lots of “hard” scattering in “Min-Bias” at the Tevatron! Ten decades! p T = 50 GeV/c!

34. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 34 CDF: Charged p T Distribution  Published CDF data on the p T distribution of charged particles in Min-Bias collisions (ND) at 1.96 TeV compared with PYTHIA Tune A. Excess events at large p T ! CDF inconsistent with CMS and UA1! No excess at large p T ! CDF consistent with CMS and UA1! 50 GeV/c! Erratum November 18, 2010

35. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 35 Min-Bias Correlations  Data at 1.96 TeV on the average p T of charged particles versus the number of charged particles (p T > 0.4 GeV/c, |  | < 1) for “min-bias” collisions at CDF Run 2. The data are corrected to the particle level and are compared with PYTHIA Tune A at the particle level (i.e. generator level).

36. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 36 Min-Bias: Average PT versus Nchg =+ +  Beam-beam remnants (i.e. soft hard core) produces low multiplicity and small with independent of the multiplicity.  Hard scattering (with no MPI) produces large multiplicity and large.  Hard scattering (with MPI) produces large multiplicity and medium. The CDF “min-bias” trigger picks up most of the “hard core” component! This observable is sensitive to the MPI tuning!

37. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 37 Transverse: vs Nchg  Shows versus N chg in the “transverse” region (p T > 0.5 GeV/c, |  | < 1) for “Leading Jet” and “Back-to-Back” events with 30 < E T (jet#1) < 70 GeV compared with “min-bias” collisions. “Leading Jet” “Back-to-Back”  Look at the of particles in the “transverse” region (p T > 0.5 GeV/c, |  | vs N chg. Min-Bias Rick Field KITP Collider Workshop 2004

38. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 38 Transverse: vs Nchg  Average p T versus Nchg (p T > 0.5 geV/c, |  | 1 GeV/c) at 7 TeV from ATLAS. Not much change in going from 900 GeV to 7 TeV!

39. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 39 “Transverse 1” Region vs “Transverse 2” Region  Shows the average number of charged particles in the “transverse 2” region versus the number of charged particles in the “transverse 1” region for p T > 0.5 GeV/c and |  | < 1 for “Leading Jet” and “Back-to-Back” events.  Use the leading jet to define two “transverse” regions and look at the correlations between “transverse 1” and “transverse 2”. “Leading Jet” “Back-to-Back”  Shows the average p T of charged particles in the “transverse 2” region versus the number of charged particles in the “transverse 1” region for p T > 0.5 GeV/c and |  | < 1 for “Leading Jet” and “Back-to-Back” events. Rick Field KITP Collider Workshop 2004

40. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 40 “Transverse 1” Region vs “Transverse 2” Region Rick Field KITP Collider Workshop 2004

41. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 41 QCD Monte-Carlo Models: Lepton-Pair Production  Start with the perturbative Drell-Yan muon pair production and add initial-state gluon radiation (in the leading log approximation or modified leading log approximation). “Underlying Event”  The “underlying event” consists of the “beam-beam remnants” and from particles arising from soft or semi-soft multiple parton interactions (MPI).  Of course the outgoing colored partons fragment into hadron “jet” and inevitably “underlying event” observables receive contributions from initial-state radiation. “Hard Scattering” Component

42. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 42 Average PT versus Nchg  Data at 1.96 TeV on the average p T of charged particles versus the number of charged particles (p T > 0.4 GeV/c, |  | < 1) for “min-bias” collisions at CDF Run 2. The data are corrected to the particle leveland are compared with PYTHIA Tune A, Tune DW, and the ATLAS tune at the particle level (i.e. generator level).  Particle level predictions for the average p T of charged particles versus the number of charged particles (p T > 0.5 GeV/c, |  | < 1, excluding the lepton-pair) for for Drell-Yan production (70 < M(pair) < 110 GeV) at CDF Run 2.

43. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 43 Average PT versus Nchg =+  Z-boson production (with low p T (Z) and no MPI) produces low multiplicity and small. +  High p T Z-boson production produces large multiplicity and high.  Z-boson production (with MPI) produces large multiplicity and medium. No MPI!

44. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 44 Average PT(Z) versus Nchg  Data on the average p T of charged particles versus the number of charged particles (p T > 0.5 GeV/c, |  | < 1, excluding the lepton-pair) for for Drell-Yan production (70 < M(pair) < 110 GeV) at CDF Run 2. The data are corrected to the particle level and are compared with various Monte-Carlo tunes at the particle level (i.e. generator level). No MPI!  Predictions for the average P T (Z-Boson) versus the number of charged particles (p T > 0.5 GeV/c, |  | < 1, excluding the lepton-pair) for for Drell-Yan production (70 < M(pair) < 110 GeV) at CDF Run 2.

45. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 45 Average PT versus Nchg  Predictions for the average p T of charged particles versus the number of charged particles (p T > 0.5 GeV/c, |  | < 1, excluding the lepton-pair) for for Drell-Yan production (70 < M(pair) < 110 GeV, P T (pair) < 10 GeV/c) at CDF Run 2.  Data the average p T of charged particles versus the number of charged particles (p T > 0.5 GeV/c, |  | < 1, excluding the lepton-pair) for for Drell-Yan production (70 < M(pair) < 110 GeV, P T (pair) < 10 GeV/c) at CDF Run 2. The data are corrected to the particle level and are compared with various Monte-Carlo tunes at the particle level (i.e. generator level). P T (Z) < 10 GeV/c No MPI! Remarkably similar behavior! Perhaps indicating that MPI playing an important role in both processes.

46. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 46 ATLAS: versus Nchg  Shows the 900 GeV and 7 TeV ATLAS MB corrected data on versus Nchg (p T > 100 MeV/c, |  | < 2.5).

47. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 47 NSD: versus Nchg  Shows the 900 GeV and 7 TeV CMS NSD corrected data on versus Nchg (all p T, |  | < 2.4) compared with Tune Z1. Okay not perfect! But not that bad! CMS Overall average p T

48. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 48 NSD: versus Nchg  Note that for a given multiplicity increase only very slightly with in going from 900 GeV to 7 TeV (2%-4%). However, the average p T increases by ~20% due mainly from the broadening of the multiplicity distribution! increases by 20%

49. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 49 Energy Dependence  Shows the energy dependence of the CMS NSD corrected data on versus Nchg (all p T, |  | compared with Tune X2.  Shows the energy dependence of the CMS NSD corrected data on versus Nchg (all p T, |  | < 2.4) compared with Tune P0, Tune PQ20, and Tune P329. It will be interesting to see if any tune can get this right! Overall average p T

50. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 50 Energy Dependence  Shows the energy dependence of the CMS NSD corrected data on versus Nchg (all p T, |  | compared with Tune Z1. Amazing! First Tune (except PhoJet) to come close here! ATLAS AMBT1 probably does well here also!

51. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 51 NSD Multiplicity Distribution  Generator level charged multiplicity distribution (all pT, |  | < 2) at 900 GeV and 7 TeV. Shows the NSD = HC + DD prediction for Tune Z1. Also shows the CMS NSD data. CMS Tune Z1 Difficult to produce enough events with large multiplicity! Okay not perfect! But not that bad!

52. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 52 MB & UE  Generator level charged multiplicity distribution (all pT, |  | < 2) at 900 GeV and 7 TeV. Shows the NSD = HC + DD prediction for Tune Z1. Also shows the CMS NSD data.  CMS uncorrected data at 900 GeV and 7 TeV on the charged particle multiplicity distribution in the “transverse” region for charged particles (p T > 0.5 GeV/c, |  | 3 GeV/c compared with PYTHIA Tune Z1 at the detector level (i.e. Theory + SIM). CMS Tune Z1 “Min-Bias” “Underlying Event” Difficult to produce enough events with large multiplicity! Difficult to produce enough events with large “transverse” multiplicity at low hard scale!

53. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 53 MB & UE  CMS uncorrected data at 7 TeV on the charged particle multiplicity distribution in the “transverse” region for charged particles (p T > 0.5 GeV/c, |  | 20 GeV/c compared with PYTHIA Tune Z1 at the detector level (i.e. Theory + SIM). Also shows the CMS corrected NSD multiplicity distribution (all pT, |  | < 2) compared with Tune Z1 at the generator. CMS Tune Z1 Amazing what we are asking the Monte-Carlo models to fit!

54. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 54 UE Summary  The “underlying event” at 7 TeV and 900 GeV is almost what we expected! With a little tuning we should be able to describe the data very well (see Tune Z1 later in this talk).  “Min-Bias” is a whole different story! Much more complicated due to very soft particles and diffraction! PARP(90) Color Connections PARP(82) Diffraction  I am surprised that the Tunes did as well as they did at predicting the behavior of the “underlying event” at 900 GeV and 7 TeV! Remember this is “soft” QCD! Warning! All the UE studies look at charged particles with p T > 0.5 GeV/c. We do not know if the models correctly describe the UE at lower p T values! ATLAS has pushed the UE measurements down to p T > 100 MeV!

55. MPI@LHC 2010 Glasgow, November 30, 2010 Rick Field – Florida/CDF/CMSPage 55 Min-Bias Summary  We need a better understanding and modeling of diffraction!  We are a long way from having a Monte-Carlo model that will fit all the features of the LHC min-bias data! There are more soft particles that expected! PARP(90) Color Connections PARP(82) Diffraction  It is difficult for the Monte-Carlo models to produce a soft event (i.e. no large hard scale) with a large multiplicity. There seems to be more “min- bias” high multiplicity soft events at 7 TeV than predicted by the models!  The models do not produce enough strange particles! I have no idea what is going on here! The Monte-Carlo models are constrained by LEP data.