Fermilab Tevatron University Toward an Understanding of Hadron-Hadron Collisions Fermilab 2006 Rick Field University of Florida Calorimeter Jet Charged Particle Jet CDF Run 2 From Feynman-Field to the Tevatron Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 1 Toward and Understanding of Hadron-Hadron Collisions 1 hat! From Feynman-Field to the Tevatron st Feynman and Field From 7 GeV/c p0’s to 600 GeV/c Jets. Outgoing Parton PT(hard) Some things we have learned about quark and gluon jets at CDF. Initial-State Radiation Proton Underlying Event Jet algorithms and the “jet” cross section at CDF. Fermilab Tevatron University December 14, 2006 AntiProton Rick Field – Florida/CDF/CMS Outgoing Parton Underlying Event Final-State Radiation Page 2 The Feynman-Field Days 1973-1983 “Feynman-Field Jet Model” FF1: “Quark Elastic Scattering as a Source of High Transverse Momentum Mesons”, R. D. Field and R. P. Feynman, Phys. Rev. D15, 2590-2616 (1977). FFF1: “Correlations Among Particles and Jets Produced with Large Transverse Momenta”, R. P. Feynman, R. D. Field and G. C. Fox, Nucl. Phys. B128, 1-65 (1977). FF2: “A Parameterization of the properties of Quark Jets”, R. D. Field and R. P. Feynman, Nucl. Phys. B136, 1-76 (1978). F1: “Can Existing High Transverse Momentum Hadron Experiments be Interpreted by Contemporary Quantum Chromodynamics Ideas?”, R. D. Field, Phys. Rev. Letters 40, 997-1000 (1978). FFF2: “A Quantum Chromodynamic Approach for the Large Transverse Momentum Production of Particles and Jets”, R. P. Feynman, R. D. Field and G. C. Fox, Phys. Rev. D18, 3320-3343 (1978). FW1: “A QCD Model for e+e- Annihilation”, R. D. Field and S. Wolfram, Nucl. Phys. B213, 65-84 (1983). My 1st graduate student! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 3 Hadron-Hadron Collisions FF1 1977 (preQCD) What happens when two hadrons collide at high energy? Hadron ??? Hadron Feynman quote from FF1 “The model we shall choose is not a popular one, Most of the time the hadrons ooze thatapart we will not duplicate too much of the through each other andsofall (i.e. work of others who are similarly analyzing no hard scattering). The outgoing various models (e.g. constituent interchange particles continue in roughly the same Parton-Parton Scattering Outgoing Parton model, multiperipheral models, etc.). We shall direction as initial proton and assume that the high PT particles arise from “Soft” Collision (no large transverse momentum) antiproton. direct hard collisions between constituent in the incoming particles, which Hadron Occasionally there will bequarks a large Hadron fragment or cascade down into several hadrons.” transverse momentum meson. Question: Where did it come from? We assumed it came from quark-quark elastic scattering, but we did not know how to calculate it! Outgoing Parton high PT meson “Black-Box Model” Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 4 Quark-Quark Black-Box Model No gluons! Quark Distribution Functions determined from deep-inelastic lepton-hadron collisions FF1 1977 (preQCD) Feynman quote from FF1 “Because of the incomplete knowledge of our functions some things can be predicted with more certainty than others. Those experimental results that are not well predicted can be “used up” to determine these functions in greater detail to permit better predictions of further experiments. Our papers will be a bit long because we wish to discuss this interplay in detail.” Quark-Quark Cross-Section Unknown! Deteremined from hadron-hadron collisions. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Quark Fragmentation Functions determined from e+e- annihilations Page 5 Quark-Quark Black-Box Model Predict particle ratios FF1 1977 (preQCD) Predict increase with increasing CM energy W “Beam-Beam Remnants” Predict overall event topology (FFF1 paper 1977) 7 GeV/c p0’s! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 6 Telagram from Feynman July 1976 SAW CRONIN AM NOW CONVINCED WERE RIGHT TRACK QUICK WRITE FEYNMAN Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 7 Letter from Feynman July 1976 Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 8 Letter from Feynman Page 1 Spelling? Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 9 Letter from Feynman Page 3 It is fun! Onward! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 10 Feynman Talk at Coral Gables (December 1976) 1st transparency Last transparency “Feynman-Field Jet Model” Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 11 QCD Approach: Quarks & Gluons Quark & Gluon Fragmentation Functions Q2 dependence predicted from QCD Parton Distribution Functions Q2 dependence predicted from QCD FFF2 1978 Feynman quote from FFF2 “We investigate whether the present experimental behavior of mesons with large transverse momentum in hadron-hadron collisions is consistent with the theory of quantum-chromodynamics (QCD) with asymptotic freedom, at least as the theory is now partially understood.” Quark & Gluon Cross-Sections Calculated from QCD Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 12 High PT Jets CDF (2006) Feynman, Field, & Fox (1978) Predict large “jet” cross-section 30 GeV/c! Feynman quote from FFF 600 GeV/c Jets! “At the time of this writing, there is still no sharp quantitative test of QCD. An important test will come in connection with the phenomena of high PT discussed here.” Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 13 A Parameterization of the Properties of Jets Field-Feynman 1978 Secondary Mesons (after decay) continue Assumed that jets could be analyzed on a “recursive” principle. (bk) (ka) Let f(h)dh be the probability that the rank 1 meson leaves fractional momentum h to the remaining cascade, leaving Rank 2 Rank 1 quark “b” with momentum P1 = h1P0. Assume that the mesons originating from quark “b” are distributed in presisely the same way as the mesons which (cb) (ba) Primary Mesons came from quark a (i.e. same function f(h)), leaving quark “c” with momentum P2 = h2P1 = h2h1P0. cc pair bb pair Calculate F(z) from f(h) and b i! Original quark with flavor “a” and momentum P0 Fermilab Tevatron University December 14, 2006 Add in flavor dependence by letting bu = probabliity of producing u-ubar pair, bd = probability of producing ddbar pair, etc. Let F(z)dz be the probability of finding a meson (independent of rank) with fractional mementum z of the original quark “a” within the jet. Rick Field – Florida/CDF/CMS Page 14 Feynman-Field Jet Model R. P. Feynman ISMD, Kaysersberg, France, June 12, 1977 Feynman quote from FF2 “The predictions of the model are reasonable enough physically that we expect it may be close enough to reality to be useful in designing future experiments and to serve as a reasonable approximation to compare to data. We do not think of the model as a sound physical theory, ....” Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 15 Monte-Carlo Simulation of Hadron-Hadron Collisions FF1-FFF1 (1977) “Black-Box” Model F1-FFF2 (1978) QCD Approach FFFW “FieldJet” (1980) QCD “leading-log order” simulation of hadron-hadron collisions the past today FF2 (1978) Monte-Carlo simulation of “jets” ISAJET HERWIG (“FF” Fragmentation) (“FW” Fragmentation) tomorrow Fermilab Tevatron University December 14, 2006 SHERPA “FF” or “FW” Fragmentation PYTHIA PYTHIA 6.3 Rick Field – Florida/CDF/CMS Page 16 QCD Monte-Carlo Models: High Transverse Momentum Jets Hard Scattering Hard Scattering Initial-State Radiation “Jet” Initial-State Radiation Outgoing Parton PT(hard) Outgoing Parton “Jet” PT(hard) Proton “Hard Scattering” Component AntiProton Underlying Event Final-State Radiation Underlying Event Outgoing Parton Proton “Jet” Final-State Radiation AntiProton Underlying Event Outgoing Parton Underlying Event “Underlying Event” Start with the perturbative 2-to-2 (or sometimes 2-to-3) parton-parton scattering and add initial and finalstate gluon radiation (in the leading log approximation or modified leading log approximation). The “underlying event” consists of the “beam-beam remnants” and from particles arising from soft or semi-soft multiple parton interactions (MPI). The “underlying event” is“jet” an unavoidable Of course the outgoing colored partons fragment into hadron and inevitably “underlying event” background to most collider observables observables receive contributions from initial and final-state radiation. and having good understand of it leads to more precise collider measurements! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 17 Monte-Carlo Simulation of Quark and Gluon Jets ISAJET: Evolve the parton-shower from Q2 (virtual photon invariant mass) to Qmin ~ 5 GeV. Use a complicated fragmentation model to evolve from Qmin to outgoing hadrons. Q2 HERWIG: Evolve the parton-shower from Q2 (virtual photon invariant mass) to Qmin ~ 1 GeV. Form color singlet clusters which “decay” into hadrons according to 2-particle phase space. MLLA: Evolve the parton-shower from Q2 (virtual photon invariant mass) to Qmin ~ 230 MeV. Assume that the charged particles behave the same as the partons with Nchg/Nparton = 0.56! hadrons CDF Distribution of Particles in Jets MLLA Curve! Field-Feynman 5 GeV 1 GeV 200 MeV Fermilab Tevatron University December 14, 2006 = ln(Ejet/pparticle) Rick Field – Florida/CDF/CMS Page 18 Distribution of Particles in Quark and Gluon Jets Both PYTHIA and HERWIG predict more charged particles than the data for quark jets! CDF Run 1 Analysis x = 0.37 0.14 0.05 0.02 0.007 pchg = 2 GeV/c Momentum distribution of charged particles in gluon jets. HERWIG 5.6 predictions are in a good agreement with CDF data. PYTHIA 6.115 produces slightly more particles in the region around the peak of distribution. Momentum distribution of charged particles in quark jets. Both HERWIG and PYTHIA produce more particles in the central region of distribution. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 21 Evolution of Charged Jets “Underlying Event” Charged Particle Correlations PT > 0.5 GeV/c |h| < 1 Charged Jet #1 Direction “Transverse” region very sensitive to the “underlying event”! Look at the charged particle density in the “transverse” region! 2p “Toward-Side” Jet “Toward” CDF Run 1 Analysis Away Region Charged Jet #1 Direction Transverse Region “Toward” “Transverse” Leading Jet “Transverse” Toward Region “Transverse” “Transverse” Transverse Region “Away” “Away” Away Region “Away-Side” Jet 0 -1 h +1 Look at charged particle correlations in the azimuthal angle relative to the leading charged particle jet. Define || < 60o as “Toward”, 60o < || < 120o as “Transverse”, and || > 120o as “Away”. All three regions have the same size in h- space, hx = 2x120o = 4p/3. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 22 “Transverse” Charged Particle Density “Transverse” region as defined by the leading “charged particle jet” “Toward” “Transverse” “Transverse” “Away” 1.25 1.25 1.25 "Transverse"Charged ChargedDensity Density Charged Density "Transverse" "Transverse" Charged Particle Jet #1 Direction "Transverse" "Transverse" Charged Charged Particle Particle Density: dN/dhd Particle Density: Density: dN/dhd dN/dhd CDF Run 1 Min-Bias CDFRun Run11Published Min-Bias CDF CDF Run 1 JET20 CDF Run 1 Published CDFRun Run21Preliminary JET20 CDF CDF Run 2 Preliminary PYTHIA Tune CDF RunA2 CDFPreliminary Run 1 Data CDF CDF Preliminary CDFdata Preliminary uncorrected 1.00 1.00 1.00 data uncorrected data uncorrected data uncorrected theory corrected 0.75 0.75 0.75 0.50 0.50 0.25 0.25 1.8 TeV |h|<1.0 PT>0.5 GeV |h|<1.0 |h|<1.0 PT>0.5 PT>0.5 GeV/c GeV/c |h|<1.0 PT>0.5 GeV 0.00 0.00 00 10 5 20 10 30 30 40 4015 50 50 20 60 60 70 702580 80 30 40 130 50 90 90 100 10035110 110 120 120 13045140 140 150 150 PT(charged PT(charged jet#1) jet#1) (GeV/c) (GeV/c) (GeV/c) Excellent agreement between Run 1 and 2! Shows the data on the average “transverse” charge particle density (|h|<1, pT>0.5 GeV) as a function of the transverse momentum of the leading charged particle jet from Run 1. Compares the Run 2 data (Min-Bias, JET20, JET50, JET70, JET100) with Run 1. The errors on the (uncorrected) Run 2 data include both statistical and Tune correlated PYTHIA A was tuned to fit the “underlying event” in Run I! systematic uncertainties. Shows the prediction of PYTHIA Tune A at 1.96 TeV after detector simulation (i.e. after CDFSIM). Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 23 Charged Multiplicity in Charged Particle Jets PYTHIA predict more charged particles than the data for charged jets! Nchg (jet#1) versus PT(charged jet#1) <Nchg> (Jet#1, R=0.7) 12 CDF Run 1 Analysis 1.8 TeV |eta|<1.0 PT>0.5 GeV 10 8 6 Includes charged particles from the “underlying event”! 4 CDF 2 data uncorrected theory corrected 0 0 5 10 15 20 25 30 35 40 45 50 PT(charged jet#1) (GeV) Herwig Isajet Pythia 6.115 CDF Min-Bias CDF JET20 Plot shows the average number of charged particles (pT > 0.5 GeV, |h| < 1) within the leading charged particle jet (R = 0.7) as a function of the PT of the leading charged jet. The solid (open) points are Min-Bias (JET20) data. The errors on the (uncorrected) data include both statistical and correlated systematic uncertainties. The QCD “hard scattering” theory curves (Herwig 5.9, Isajet 7.32, Pythia 6.115) are corrected for the track finding efficiency. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 24 Run 1 Fragmentation Function Density F(z)=dNchg/dz 100.0 Charged Momentum Distribution Jet#1 PT(jet#1) > 30 GeV <NchgJet#1> = 8.0 CDF PT(jet#1) > 5 GeV <NchgJet#1> = 4.0 data uncorrected 1.8 TeV |eta|<1.0 PT>0.5 GeV 10.0 CDF Run 1 Analysis Includes charged particles from the “underlying event”! 1.0 PT(jet#1) > 2 GeV <NchgJet#1> = 2.5 0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 z = p(charged)/P(charged jet#1) CDF Run 1 data on the momentum distribution of charged particles (pT > 0.5 GeV and |h| < 1) within chgjet#1 (leading charged jet). The points are the charged number density, F(z) = dNchg/dz, where z = pchg/P(chgjet#1) is the ratio of the charged particle momentum to the charged momentum of chgjet#1. The integral of F(z) is the average number of particles within chgjet#1. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 27 Run 1 Fragmentation Function Charged Momentum Distribution Jet#1 Density F(z)=dNchg/dz 100.0 PT(jet#1) > 5 GeV PYTHIA does not agree at high z! CDF data uncorrected theory corrected 10.0 1.8 TeV |eta|<1.0 PT>0.5 GeV CDF Run 1 Analysis 1.0 Herwig Isajet Pythia 6.115 CDF Min-Bias 0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 z = p(charged)/P(charged jet#1) CDF Run 1 data from on the momentum distribution of charged particles (pT > 0.5 GeV and |h| < 1) within chgjet#1 (leading charged jet) for PT(chgjet#1) > 5 GeV compared with the QCD “hard scattering” Monte-Carlo predictions of HERWIG, ISAJET, and PYTHIA. The points are the charged number density, F(z) = dNchg/dz, where z = pchg/P(chgjet#1) is the ratio of the charged particle momentum to the charged momentum of chgjet#1. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 28 Run 1 Fragmentation Function Charged Momentum Distribution Jet#1 Density F(z)=dNchg/dz 100.0 PYTHIA does not agree at high z! CDF PT(jet#1) > 30 GeV data uncorrected theory corrected 10.0 1.8 TeV |eta|<1.0 PT>0.5 GeV CDF Run 1 Analysis 1.0 Herwig Isajet Pythia 6.115 CDF JET20 0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 z = p(charged)/P(charged jet#1) Data from Fig. 3.8 on the momentum distribution of charged particles (pT > 0.5 GeV and |h| < 1) within chgjet#1 (leading charged jet) for PT(chgjet#1) > 30 GeV compared with the QCD “hard scattering” Monte-Carlo predictions of HERWIG, ISAJET, and PYTHIA. The points are the charged number density, F(z) =dNchg/dz, where z = pchg/P(chgjet#1) is the ratio of the charged particle momentum to the charged momentum of chgjet#1. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 29 The “Transverse” Regions as defined by the Leading Jet Jet #1 Direction “Transverse” region is very sensitive to the “underlying event”! Charged Particle Correlations pT > 0.5 GeV/c |h| < 1 2p “Toward-Side” Jet CDF Run 2 Analysis Jet #1 Direction “Toward” “Transverse” “Transverse” “Away” Away Region Transverse Region 1 “Toward” “Trans 1” Look at the charged particle density in the “transverse” region! Leading Jet “Trans 2” Toward Region Transverse Region 2 “Away” Away Region “Away-Side” Jet 0 -1 h +1 Look at charged particle correlations in the azimuthal angle relative to the leading calorimeter jet (JetClu R = 0.7, |h| < 2). o o o o o Define || < 60 as “Toward”, 60 < - < 120 and 60 < < 120 as “Transverse 1” and o “Transverse 2”, and || > 120 as “Away”. Each of the two “transverse” regions have o area h = 2x60 = 4p/6. The overall “transverse” region is the sum of the two o transverse regions (h = 2x120 = 4p/3). Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 30 Charged Particle Density Dependence Refer to this as a “Leading Jet” event Jet #1 Direction Charged Particle Density: Density: dN/dhd dN/dhd Charged Particle 10.0 10.0 Subset “Transverse” “Transverse” “Away” Refer to this as a “Back-to-Back” event Jet #1 Direction “Toward” “Transverse” “Transverse” Charged Particle Particle Density Density Charged “Toward” CDF CDF Preliminary Preliminary 30 << ET(jet#1) ET(jet#1) << 70 70 GeV GeV 30 Back-to-Back data data uncorrected uncorrected Leading Jet Min-Bias "Transverse" "Transverse" Region Region 1.0 1.0 Jet#1 Jet#1 Charged Charged Particles Particles (|h|<1.0, (|h|<1.0, PT>0.5 PT>0.5 GeV/c) GeV/c) 0.1 0.1 00 30 30 60 60 90 120 “Away” 150 180 210 210 240 240 270 270 300 300 330 330 360 360 (degrees) Jet #2 Direction Look at the “transverse” region as defined by the leading jet (JetClu R = 0.7, |h| < 2) or by the leading two jets (JetClu R = 0.7, |h| < 2). “Back-to-Back” events are selected to have at least two jets with Jet#1 and Jet#2 nearly “back-to-back” (12 > 150o) with almost equal transverse energies (ET(jet#2)/ET(jet#1) > 0.8) and with ET(jet#3) < 15 GeV. Shows the dependence of the charged particle density, dNchg/dhd, for charged particles in the range pT > 0.5 GeV/c and |h| < 1 relative to jet#1 (rotated to 270o) for 30 < ET(jet#1) < 70 GeV for “Leading Jet” and “Back-to-Back” events. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 31 “Transverse” Charge Density PYTHIA Tune A vs HERWIG “Leading Jet” Jet #1 Direction "AVE Transverse" Charge Density: dN/dhd 1.0 “Transverse” “Transverse” “Away” “Back-to-Back” Jet #1 Direction “Toward” “Transverse” “Transverse” "Transverse" Charge Density “Toward” CDF Preliminary data uncorrected theory + CDFSIM 0.8 Leading Jet PY Tune A 0.6 0.4 HW Back-to-Back 0.2 1.96 TeV Charged Particles (|h|<1.0, PT>0.5 GeV/c) 0.0 0 50 “Away” 100 150 200 250 ET(jet#1) (GeV) Jet #2 Direction Now look in detail at “back-to-back” events in the region 30 < ET(jet#1) < 70 GeV! Shows the average charged particle density, dNchg/dhd, in the “transverse” region (pT > 0.5 GeV/c, |h| < 1) versus ET(jet#1) for “Leading Jet” and “Back-to-Back” events. Compares the (uncorrected) data with PYTHIA Tune A and HERWIG after CDFSIM. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 32 Charged Particle Density PYTHIA Tune A vs HERWIG HERWIG (without multiple parton interactions) produces too few charged particles in the “transverse” region for 30 < ET(jet#1) < 70 GeV! Charged Particle Density: dN/dhd Charged Particle Density: dN/dhd 10.0 1.0 CDF Preliminary "Transverse" Region data uncorrected theory + CDFSIM 0.1 0 30 60 90 120 150 180 210 (degrees) Jet#1 HERWIG 1.0 CDF Preliminary "Transverse" Region data uncorrected theory + CDFSIM PYTHIA produces too 0.1 many 0 30 60 240 270 300 330 360 particle in the “away-side” jet! 90 120 150 180 210 Jet#1 240 270 300 330 360 330 360 (degrees) Data - Theory: Charged Particle Density dN/dhd Data - Theory: Charged Particle Density dN/dhd 1.0 1.0 CDF Preliminary data uncorrected theory + CDFSIM Back-to-Back 30 < ET(jet#1) < 70 GeV CDF Preliminary PYTHIA Tune A data uncorrected theory + CDFSIM 0.5 Data - Theory 0.5 Data - Theory 30 < ET(jet#1) < 70 GeV Charged Particles (|h|<1.0, PT>0.5 GeV/c) Back-to-Back Charged Particle Density Charged Particle Density PY Tune A 10.0 30 < ET(jet#1) < 70 GeV Charged Particles (|h|<1.0, PT>0.5 GeV/c) Back-to-Back 0.0 "Transverse" Region -0.5 Back-to-Back 30 < ET(jet#1) < 70 GeV 0.0 "Transverse" Region -0.5 Jet#1 Charged Particles (|h|<1.0, PT>0.5 GeV/c) HERWIG Jet#1 Charged Particles (|h|<1.0, PT>0.5 GeV/c) -1.0 -1.0 0 30 60 90 120 150 180 210 240 270 300 330 360 0 30 60 120 150 180 210 240 270 300 (degrees) (degrees) Fermilab Tevatron University December 14, 2006 90 Rick Field – Florida/CDF/CMS Page 33 Jet Algorithms Clustering algorithms are used to combine calorimeter towers or charged particles into “jets” in order to study the event topology and to compare with the QCD Monte-Carlo Models. We do not detect partons! The outgoing partons fragment into hadrons before they travel a distance of about the size of the proton. At long distances the partons manifest themselves as “jets”. The “underlying event” can also form “jets”. Most “jets” are a mixture of particles arising from the “hard” outgoing partons and the “underlying event”. Since we measure hadrons every observable is infrared and collinear safe. There are no divergences at the hadron level! Every “jet” algorithms correspond to a different observable and different algorithms give different results. Studying the difference between the algorithms teaches us about the event structure. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 36 Jet Corrections & Extrapolations Calorimeter Level Jets → Hadron Level Jets: Hadron ← Parton We measure “jets” at the “hadron level” in the calorimeter. We certainly want to correct the “jets” for the detector resolution and efficiency. Also, we must correct the “jets” for “pile-up”. Must correct what we measure back to the true “hadron level” (i.e. particle level) observable! Particle Level Jets (with the “underlying event” removed): Useless without a model of hadronization! Outgoing Parton I do believe wemodel shoulddependent extrapolate Do we want to not make further corrections? the data to the parton level! We should Do we want to try and subtract the “underlying event” from the publish what we measure (i.e. hadron level observed “particle level” jets. with the “underlying event”)! This cannot really be done, but if you trust the Monte-Carlo with event” theory you we should modeling ofTo thecompare “underlying can do it by using the “extrapolate” the parton level to the Monte-Carlo models (use PYTHIA Tune A). (i.e. add hadronization and This is nohadron longerlevel an observable, it is a model dependent the “underlying event” to the parton level)! extrapolation! HERWIG, MC@NLO Hadron LevelPYTHIA, Jets → Parton Level Jets: PT(hard) Initial-State Radiation Proton AntiProton Underlying Event Outgoing Parton Underlying Event Final-State Radiation Fermilab Tevatron University December 14, 2006 Do we want to use the data to try and extrapolate back to the parton level? What parton level, PYTHIA (Leading Log) or fixed order NLO? Next-to-leading order This also cannot really be done, but again if you trust the Monteparton level calculation Carlo models you can try and do it by using the Monte-Carlo 0, 1, 2, or 3 partons! models (use PYTHIA Tune A) including ISR and FSR. Cannot extrapolate the data to fixed order NLO! Rick Field – Florida/CDF/CMS Page 37 Good and Bad Algorithms Calorimeter Jet Particle Jet In order to correct what we see in the calorimeter back to the hadron level we must use an algorithm that can be defined at both the calorimeter and particle level. If you insist on extrapolating the data to the parton level then it is better to use an algorithm that is well defined at the parton level (i.e. infrared and collinear safe at the parton level). If you hadronize the parton level and add the “underlying event” (i.e. PYTHIA, HERWIG, MC@NLO) then you do not care if the algorithm is infrared and collinear safe at the parton level. You can predict any hadron level observable! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Infrared Safety (Parton Level) Soft parton emission changes jet multiplicity Collinear Safety (Parton Level) below threshold (no jets) above threshold (1 jet) Page 38 Four Jet Algorithms Towers not included in a jet (i.e. “dark towers”)! Bad JetClu is bad because the algorithm cannot be defined at the particle level. The MidPoint and Modified MidPoint (i.e. Search Cone) algorithms are not infrared and collinear safe at the parton level. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 39 KT Algorithm kT Algorithm: Begin For each precluster, calculate di pT2,i For each pair of preculsters, calculate ( y y j ) 2 (i j ) 2 dij min( pT2 ,i , pT2 , j ) i D2 Find the minimum of all di and dij. Merge i and j yes Minumum is dij? Cluster together calorimeter towers by their kT proximity. Infrared and collinear safe at all orders of pQCD. No splitting and merging. No ad hoc Rsep parameter necessary to compare with parton level. Every parton, particle, or tower is assigned to a “jet”. No biases from seed towers. Favored algorithm in e+e- annihilations! no Will the KT algorithm be effective in the collider environment where there is an “underlying event”? Move i to list of jets yes Any Preclusters left? Raw Jet ET = 533 GeV KT Algorithm Raw Jet ET = 618 GeV no End Outgoing Parton PT(hard) Initial-State Radiation Proton AntiProton Underlying Event Underlying Event CDF Run 2 Outgoing Parton Final-State Radiation Fermilab Tevatron University December 14, 2006 Only towers with ET > 0.5 GeV are shown Rick Field – Florida/CDF/CMS Page 40 KT Inclusive Jet Cross Section KT Algorithm (D = 0.7) Data corrected to the hadron level L = 385 pb-1 0.1 < |yjet| < 0.7 Compared with NLO QCD (JetRad) corrected to the hadron level. Sensitive to UE + hadronization effects for PT < 300 GeV/c! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 41 Search Cone Inclusive Jet Cross Section Modified MidPoint Cone Algorithm (R = 0.7, fmerge = 0.75) Data corrected to the hadron level and the parton level L = 1.04 fb-1 0.1 < |yjet| < 0.7 Compared with NLO QCD (JetRad, Rsep = 1.3) Sensitive to UE + hadronization effects for PT < 200 GeV/c! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 42 Hadronization and “Underlying Event” Corrections Compare the hadronization and “underlying event” corrections for the KT algorithm (D = 0.7) and the MidPoint algorithm (R = 0.7)! We see that the KT algorithm (D = 0.7) is slightly more sensitive to the underlying event than the cone algorithm (R = 0.7), but with a good model of the “underlying event” both cross sections can be measured at the Tevatrun! Note that DØ does not make any corrections for hadronization or the “underlying event”!? MidPoint Cone Algorithm (R = 0.7) The KT algorithm is slightly more sensitive to the “underlying event”! Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Page 43 Summary and Conclusions Neither HERWIG or PYTHIA describe Charged Particle kT Distribution in Jets perfectly the distribution charged particles Was this Particle measured in Run 1? in quark and gluon jets at the Tevatron! Jet To learn about the fragmentation function at large z we should Comparison Only compare the inclusive “jet” cross-sectionShape to the inclusive charged particle cross section! We have events with 600 GeV “jets” so we must have events In 1 fb-1 we have thousands of with 300 GeV/c charged particles! charged tracks with p > 100 GeV/c! T I wish I could show you the following: Charged Particle PT Distribution 100,000,000 CDF Run 2 Pre-Preliminary Number in 1 GeV/c Bin A lot of work has been done in comparing to analytic MLLA calculations (Korytov and students), but more work needs to be done in improving the fragmentation models in HERWIG and PYTHIA! CDF measured fragmentation functions at different Q2 compared with PYTHIA and HERWIG. The kT distribution of charged particles within “jets” compared with PYTHIA and HERWIG. The ratio of the inclusive charged particle cross-section to the inclusive “jet” cross-section compared with PYTHIA and HERWIG. Fermilab Tevatron University December 14, 2006 Rick Field – Florida/CDF/CMS Jet 100 Trigger 10,000,000 1,000,000 100,000 1.96 TeV 10,000 Charged Particles (|h|<1.0) 1,000 0 10 20 30 40 50 60 70 80 90 Charged Particle PT (GeV/c) Sergo’s latest “blessing” from the CDF-QCD group! Page 44 100 Inclusive & Exclusive 3-Jet Study CDF analysis using 1fb-1. At least 1 Jet (“trigger” jet) (PT > PTtrig, |h| < 1.0) Exactly 3 jets (Exclusive) (PT > PTmin, |h| < 2.5) Charged Particle Jets RDF Algorithm R = 0.4 More than 2 jets (Inclusive) (PT > PTmin, |h| < 2.5) Jet1 MidPoint R = 0.4 and JetClu R = 0.4 Jet4 Proton Anti-proton Jet3 Fermilab Tevatron University December 14, 2006 CDF Run 2 Jet2 Rick Field – Florida/CDF/CMS Order Jets by PT Jet1 highest PT, etc. Page 45 MIT Search Scheme 12 Data Exclusive 3 Jet Final State Challenge (MITCDF JetClu R=0.4) At least 1 Jet (“trigger” jet) (PT > 40 GeV/c, |h| < 1.0) Normalized to 1 PYTHIA Tune A Exactly 3 jets (PT > 20 GeV/c, |h| < 2.5) R(j2,j3) Order Jets by PT Jet1 highest PT, etc. Bruce Knuteson Khaldoun Makhoul Georgios Choudalakis Fermilab Tevatron University December 14, 2006 Markus Klute Conor Henderson Rick Field – Florida/CDF/CMS Ray Culbertson Gene Flanagan Page 46 CDF Data (MIT JetClu R=0.4) Exc3J R(j2,j3) Normalized The data have more 3 jet events with small R(j2,j3)!? Let Ntrig40 equal the number of events Exclusive 3-Jet Production: R(j2,j3) with at least one jet with PT > 40 GeV/c and |h| < 1.0 (this is the “offline” trigger). 0.16 Let N3Jexc20 equal the number of events 0.12 with exactly three jets with PT > 20 GeV/c and |h| < 2.5 which also have at least one jet with PT > 40 GeV/c and |h| < 1.0. Let N3JexcFr = N3Jexc20/Ntrig40. The is the fraction of the “offline” trigger events that are exclusive 3-jet events. Initial-State Radiation AntiProton Underlying Event Underlying Event Outgoing Parton “Jet 3” Initial-State Radiation “Jet 2” Data R(j2,j3) pyAW pyDW pyBW data uncorrected generator level theory data corrected using “jet corrections”! 0.08 Normalized to N3JexcFr 0.04 0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 R(j2,j3) 1.96 TeV compared with PYTHIA Tune AW (PARP(67)=4), Tune DW (PARP(67)=2.5), Tune BW (PARP(67)=1). PARP(67) affects the initial-state radiation which contributes primarily to the region R(j2,j3) > 1.0. R > 1.0 Fermilab Tevatron University December 14, 2006 5.0 The CDF data (MIT JetClu R=0.4) on dN/dR(j2,j3) at Outgoing Parton “Jet 1” Proton dN/dR(j2,j3) CDF Run 2 Preliminary Rick Field – Florida/CDF/CMS Page 47 Exc3J R(j2,j3) Normalized Let Ntrig40 equal the number of events Exclusive Exclusive 3-Jet 3-Jet Production: Production: R(j2,j3) R(j2,j3) 0.16 0.80 0.16 with at least one jet with PT > 40 GeV/c and |h| < 1.0 (this is the “offline” trigger). with exactly three jets with PT > 20 GeV/c 0.08 0.40 0.08 and |h| < 2.5 which also have at least one UF-MIT 0.04 jet with PT > 40 GeV/c and |h| < 1.0. (and Steve 0.20 0.04Mrenna) are working to understand Normalized to 1 Let N3JexcFr = N3Jexc20/Ntrig40. theThe CDFisinclusive and exclusive 0.00 0.00 0.0 0.5 1.0 1.5 2.0 2.5 the fraction of the “offline” trigger events 3-jet0.0 0.5 1.0 1.5 2.0 2.5 data! R(j2,j3) R(j2,j3) that are exclusive 3-jet events. Final-State Radiation AntiProton Underlying Event Final-State Radiation “Jet “Jet 2” 3” R < 1.0 Fermilab Tevatron University December 14, 2006 3.0 3.0 3.5 3.5 4.0 4.0 4.5 4.5 5.0 5.0 1.96 TeV compared with PYTHIA Tune DW (PARP(67)=2.5) and HERWIG (without MPI). Proton Outgoing Parton Normalized to N3JexcFr The CDF data (MIT JetClu R=0.4) on dN/dR(j2,j3) at Outgoing Parton “Jet 1” Underlying Event CDF Data Data Data R(j2,j3) R(j2,j3) hw05 pyDW pyDW pyDW pyDWnoFSR hw05 pyBW data uncorrected datauncorrected uncorrected data generator level theory generatorlevel leveltheory theory generator dN/dR(j2,j3) dN/dR(j2,j3) dN/dR(j2,j3) Let N3Jexc20 equal the number of events 0.12 0.60 0.12 CDF Run Run 22 Preliminary Preliminary CDF Final-State radiation contributes to the region R(j2,j3) < 1.0. If you ignore the normalization and normalize all the distributions to one then the data prefer Tune BW, but I believe this is misleading. Rick Field – Florida/CDF/CMS Page 48