1. 1 Probing New Physics with Dijets at CMS Robert M. Harris Fermilab HEP Seminar Johns Hopkins University March 26, 2008

2. Robert Harris, Fermilab2 Outline l Motivation l Introduction è Jets at CMS è New Physics with Jets è Best limits on new physics from Tevatron l CMS Search Plans for Contact Interactions è Jet Rate: Inclusive Jet P T è Jet Angle: Dijet Ratio l CMS Search Plans for Dijet Resonances è Jet Rate: Dijet Mass è Jet Angle: Dijet Ratio l Conclusions

3. Robert Harris, Fermilab3 Credits l Study done at LHC Physics Center (LPC) è A center of CMS physics at Fermilab l CMS approved, publicly available è Initial study completed in April 2006 and published in CMS Physics TDR Vol II à J. Phys. G: Nucl. Part. Phys. 34: 995-1579 à CMS Notes (2006 / 069, 070 and 071) è Update study completed in December 2007 à http://cms-physics.web.cern.ch/cms- physics/public/SBM-07-001-pas-v3.pdf l Demonstration of physics at LPC è Working within the CMS collaboration l Thanks to my many CMS colleagues! è Involved postdocs and grad students: M. Cardaci, S. Esen, K. Gumus, M. Jha, K. Kousouris, D. Mason. P T D R U p d a t e

4. Robert Harris, Fermilab4 CMS ATLAS In 2008 science will start to explore a new energy scale 14 TeV proton-proton collisions will allow us to see deeper into nature than ever before Two large detectors will observe the collisions Large Hadron Collider at CERN Geneva Switzerland

5. Robert Harris, Fermilab5 l ATLAS è 22 x 44 meters, 6000 tons è 2000 physicists, 34 nations l CMS è 15 x 22 meters, 12500 tons è 3000 collaborators, 37 nations

6. Robert Harris, Fermilab6 What will the LHC detectors see? l But we hope to see more than the standard model ! l We expect to see the “standard model”: è The particles and forces already catalogued... è... & perhaps a Higgs particle that remains to be discovered.

7. Robert Harris, Fermilab7 Questions in the Standard Model l Can we unify the forces ?  , Z and W are unified already. è Can we include gluons ? è Can we include gravity ? è Why is gravity so weak ? ? ? l Why three nearly identical generations of quarks and leptons? è Like the periodic table of the elements, does this suggest an underlying physics? è Is it possible that quarks and leptons are made of other particles? l These & other questions suggest new physics beyond the standard model. è We can discover this new physics with simple measurements of jets at the LHC. l The simple picture of the standard model raises many fundamental questions.

8. Robert Harris, Fermilab8 Introduction to Jets

9. Robert Harris, Fermilab9 Jets at LHC in Standard Model Proton q, q, g Proton q, q, g The LHC collides protons containing colored partons: quarks, antiquarks & gluons. q, q, g The dominant hard collision process is simple 2  2 scattering of partons off partons via the strong color force (QCD). Jet Each final state parton becomes a jet of observable particles. The process is called dijet production.

10. Robert Harris, Fermilab10 Experimental Observation of Jets Calorimeter Simulation   ETET 0 1 Jet 1 Jet 2 Dijet Mass = 900 GeV l Dijets are easy to find è Two jets with highest P T in the calorimeter. è A jet is the sum of calorimeter energy in a cone of radius CMS Barrel & Endcap Calorimeters proton  =-1 proton Jet 1 Jet 2  =1  Transverse  =0 

11. Robert Harris, Fermilab11 Jet Response & Corrections Jets in Barrel have uniform response vs  & are sensitive to new physics  Jet response changes smoothly and slowly up to | jet  | = 1.3 Measure relative response vs. jet  in data with dijet balance è Data will tell us what is the region of response we can trust. l Measured jet p T in the calorimeter is less than true jet p T (particles in cone) l Measured jets are corrected so p T is the same as true jet p T è Scales Jet (E,p x,p y,p z ) by à ~1.5 at p T = 70 GeV à ~1.1 at p T = 3 TeV à for jets in barrel region |  |<1.3 |  |<1 Jet Response vs  relative to Barrel CMS Preliminary Jet Response vs p T in Barrel Jets we use (GeV)

12. Robert Harris, Fermilab12 Introduction to New Physics with Jets

13. Robert Harris, Fermilab13 Quark Compositeness and Scattering l Three nearly identical generations suggests quark compositeness. è Compositeness is also historically motivated.  Molecules  Atoms  Nucleus  Protons & Neutrons  Quarks  Preons ? l Scattering probes compositeness. l In 1909 Rutherford discovered the nucleus inside the atom via scattering.  Scattered  particles off gold foil.  Too many scattered at wide angles to the incoming  beam è Hit the nucleus inside the atom! l A century later, we can discover quark compositeness in a similar way ! è Rate: more jets at high p T than QCD. è Angle: more dijets in center of the CMS barrel than at the edge. Measured with dijet ratio (more later) è Today we model quark compositeness with contact interactions.   Gold Detector Discovery of Nucleus More of this Than of this Quark Compositeness Signal q q q q q q q q

14. Robert Harris, Fermilab14 Quark Contact Interactions New physics at large scale  è Composite Quarks è New Interactions l Modeled by contact interaction  Intermediate state collapses to a point for dijet mass << . è For example, the standard contact interaction among left- handed quarks introduced by Eichten, Lane and Peskin  Excluded for   < 2.7 TeV (D0) l Observable Consequences è Effects at high p T & dijet mass. à Rate: Higher rate than QCD à Angle: Angular distributions can be very different from QCD. Quark Contact Interaction  M ~  Composite QuarksNew Interactions M ~  Dijet Mass <<  q q q q q q q q L = ± [2  /   ] (q   q) (q   q)

15. Robert Harris, Fermilab15 Dijet Resonances X q, q, g Mass Rate M New particles, X, produced in parton-parton annihilation will decay to 2 partons (dijets). They are observed as dijet resonances: mass bumps. Tevatron has searched but not found any dijet resonances so far: D0 Run 1 time space CDF Run 1 ( Data – Fit ) / Fit

16. Robert Harris, Fermilab16 Why Search for Dijet Resonances? l Experimental Motivation è LHC is a parton-parton resonance factory in a previously unexplored mass region à With the higher energy we have a good chance of finding new physics. è Nature may surprise us with previously unanticipated new particles. à We will search for generic dijet resonances, not just specific models. è One search can encompass ALL narrow dijet resonances. à Resonances narrower than jet resolution produce similar mass bumps in our data. è We can discover dijet resonances if they have a large enough cross section. l Theoretical Motivation è Dijet Resonances found in many models that address fundamental questions.  Why Generations ?  Compositeness  Excited Quarks  Why So Many Forces ?  Grand Unified Theory  W ’ & Z ’  Can we include Gravity ?  Superstrings & GUT  E6 Diquarks  Why is Gravity Weak ?  Extra Dimensions  RS Gravitons  Why Symmetry Broken ?  Technicolor  Color Octet Technirho  More Symmetries ?  Extra Color  Colorons & Axigluons

17. Robert Harris, Fermilab17 Dijet Resonance Details 11 11 22 11 11 1+1+ ½+½+ 0+0+ J P qq0.01SingletZ ‘Heavy Z q1q2q1q2 0.01SingletW ‘Heavy W qq,gg0.01SingletGR S Graviton qq,gg0.01Octet  T8 Octet Technirho qq0.05OctetCColoron qq0.05OctetAAxigluon qg0.02Tripletq*Excited Quark ud0.004TripletDE 6 Diquark Chan  / (2M) ColorXModel Name mainly t - channel QCD Background X q, q, g Dijet Resonance s - channel Signal and background will have very different angular distributions

18. Robert Harris, Fermilab18 CDF Dijet Resonance Search in Run II l Preliminary Results Just Released è Best limits on dijet resonances è Thanks to Ken Hatakeyama ! ModelExcluded (GeV)ModelExcluded (GeV) A or C260 - 1250D290 - 630  T8 260 - 1110W ’280 - 840 q*260 - 870Z ’320 - 740

19. Robert Harris, Fermilab19 CMS Search Plans for Contact Interactions

20. Robert Harris, Fermilab20 Rate: Inclusive Jet p T l Inclusive jet p T is a QCD measurement that is sensitive to new physics.  Counts all jets inside a p T bin and  interval, and divides by bin width and luminosity. l Corrected calorimeter jets (CaloJets) agree with particle level jets (GenJets). è CaloJets before corrections shifted to lower p T than GenJets è Ratio between corrected CaloJets and GenJets is “resolution smearing”: small at high p T. l Simple correction for resolution smearing in real data is to divide rate by this ratio. Resolution Smearing Inclusive Jet Cross Section

21. Robert Harris, Fermilab21 Rate: Jet P T and Contact Interactions l Contact interactions create large rate at high P T and immediate discovery possible è Error dominated by jet energy scale (~10%) in early running (10 pb -1 )   E~ 10% not as big an effect as   = 3 TeV for P T >1 TeV. è PDF “errors” and statistical errors (10 pb -1 ) smaller than E scale error With 10 pb -1 we can see new physics beyond Tevatron exclusion of   < 2.7 TeV. Rate of QCD and Contact Interactions Sensitivity with 10 pb -1 Sys Err. PDF Err.

22. Robert Harris, Fermilab22 Dijet Ratio: Simple Angular Measure l Dijet angular distributions are sensitive to new physics. è Contact Interactions & Resonances l Dijet Ratio = N(|  |<0.7) / N(0.7<|  |<1.3)  Number of events in which each leading jet has |  |<0.7, divided by the number in which each leading jet has 0.7<|  |<1.3  Numerator is sensitive to new physics at low cos  *.  Denominator is dominated by QCD at high cos  *. l Simplest measurement of angular distribution  Uses detector variable  è Uses same mass bins as resonance search in rate vs. mass. Jet 1 Jet 2 Numerator Sensitive to New Physics |cos    ~ 0 Denominator Dominated By QCD |cos  *| ~ 0.7, usually Jet 1 Jet 2 (rare) or  = -1.3 - 0.7 0.7 1.3 z z

23. Robert Harris, Fermilab23 Angle: Dijet Ratio from QCD l We have optimized the dijet ratio for a contact interaction search in barrel  Old dijet ratio used by D0 and PTDR was N(|  |<0.5) / N(0.5<|  |<1.0)  New dijet ratio is N(|  |<0.7) / N(0.7<|  |<1.3) l Dijet ratio from QCD agrees for GenJets and Corrected CaloJets è Flat at 0.6 for old ratio, and flat at 0.5 for new ratio up to around 6 TeV. Old Dijet RatioNew Dijet Ratio

24. Robert Harris, Fermilab24 l Optimization dramatically increases sensitivity to contact interactions. è Raising the signal and decreasing the QCD error bars. Old Dijet Ratio (D0 and PTDR  Cuts) 3 5 10  + (TeV) QCD New Dijet Ratio (Optimized in Barrel) 3 5 10  + (TeV) QCD Angle: Dijet Ratio from Contact Interactions

25. Robert Harris, Fermilab25 Dijet Ratio and Systematic Uncertainties l Systematics (red) are small è They cancel in the ratio. l Relative Energy Scale  Energy scale in center vs edge of barrel in . è Estimate +/- 0.5 % is achievable in barrel. è Determined with dijet balance l Parton Distributions è We’ve used CTEQ6.1 uncertainties P T D R

26. Robert Harris, Fermilab26 CMS Search Plans for Dijet Resonances

27. Robert Harris, Fermilab27 Dijet Mass Resolution l First high statistics study of CMS dijet resonance mass resolution. Gaussian core of resolution for |  |<1 and |  |<1.3 is similar. l Resolution for corrected CaloJets è 9% at 0.7 TeV è 4.5% at 5 TeV è Better than in PTDR 2 study. 2 TeV Z’ |η| < 1.3 Corrected CaloJets GenJets Natural Width Resolution

28. Robert Harris, Fermilab28 Rate vs. Dijet Mass and Resonances l Measure rate vs. corrected dijet mass and look for resonances. è Use a smooth parameterized fit or QCD prediction to model background l Strongly produced resonances can be seen è Convincing signal for a 2 TeV excited quark in 100 pb -1 à Tevatron excluded up to 0.87 TeV. QCD Backgound Resonances with 100 pb -1

29. Robert Harris, Fermilab29 Rate: Systematic Uncertainties l Jet Energy è CMS estimates +/- 5 % is achievable by 1 fb -1 è Changes dijet mass cross section between 30% and 70% l Parton Distributions è CTEQ 6.1 uncertainty l Resolution è Bounded by difference between particle level jets and calorimeter level jets. l Systematic uncertainties on the cross section vs. dijet mass are large. è But they are correlated vs. mass. The distribution changes smoothly. P T D R

30. Robert Harris, Fermilab30 Rate: Sensitivity to Resonance Cross Section l Cross Section for Discovery or Exclusion è Shown here for 1 fb -1 è Also for 100 pb -1, 10 fb -1 l Compared to cross section for 8 models l CMS expects to have sufficient sensitivity to  Discover with 5  significance any model above solid black curve è Exclude with 95% CL any model above the dashed black curve. l Can discover resonances produced via color force, or from valence quarks. P T D R

31. Robert Harris, Fermilab31 Rate: Discovery Sensitivity for Models l Resonances produced by the valence quarks of each proton è Large cross section from higher probability of quarks in the initial state at high x.  E6 diquarks (ud  D  ud) can be discovered up to 3.7 TeV for 1 fb -1 l Resonances produced by color force è Large cross sections from strong force è With just 1 fb -1 CMS can discover à Excited Quarks up to 3.4 TeV à Axigluons or Colorons up to 3.3 TeV à Color Octet Technirhos up to 2.2 TeV. l Discoveries possible with only 100 pb -1 è Large discovery potential with 10 fb -1 Mass (TeV) E6 Diquark Excited Quark Axigluon or Coloron Color Octet Technirho CMS 100 pb -1 CMS 1 fb -1 CMS 10 fb -1 5  Sensitivity to Dijet Resonances 0 1 2 3 4 5 P T D R

32. Robert Harris, Fermilab32 Rate: Exclusion Sensitivity to Models E6 Diquark Excited Quark Axigluon or Coloron Color Octet Technirho W ’ R S Graviton Z ’ Tevatron Exclusion (Dijets) CMS 100 pb -1 CMS 1 fb -1 CMS 10 fb -1 Mass (TeV) 95% CL Sensitivity to Dijet Resonances 0 1 2 3 4 5 6 l Resonances produced via color interaction or valence quarks. è Wide exclusion possibility connecting up with many exclusions at Tevatron l Resonances produced weakly are harder. è But CMS has some sensitivity to each model with sufficient luminosity. è Z’ is particularly hard. à Weak coupling and requires an anti-quark in the proton at high x. P T D R

33. Robert Harris, Fermilab33 Angle: Dijet Resonances with Dijet Ratio l All resonances have a more isotropic decay angular distribution than QCD  Spin ½ (q*), spin 1 (Z’), and spin 2 (RS Graviton) all flatter than QCD in dN / dcos  *. l Dijet ratio is larger for resonances than for QCD.  Because numerator mainly low cos  *, denominator mainly high cos  * QCD Dijet Ratio vs Mass Dijet Angular Distributions

34. Robert Harris, Fermilab34 Angle: Dijet Resonances with Dijet Ratio l Dijet ratio from signal + QCD compared to statistical errors for QCD alone  Resonances normalized with q* cross section for |  |<1.3 to see effect of spin. l Convincing signal for 2 TeV strong resonance in 100 pb -1 regardless of spin. l Promising technique for discovery, confirmation, and eventually spin measurement. Dijet Ratio for q* Dijet Ratio for Spin ½, 1, 2

35. Robert Harris, Fermilab35 Conclusions l We’ve described CMS search plans for new physics with dijets. l Inclusive jet p T could give a convincing contact interaction signal at startup!  Can discover  + = 3 TeV in 10 pb -1 even if jet energy errors are 10%. l Dijet ratio will probe contact interactions in dijet angular distributions  Can discover  + = 4, 7, 10 TeV in 10, 100, 1000 pb -1 with small systematics. l Dijet mass can be used to discover a dijet resonance up to many TeV. è Axigluon, Coloron, Excited Quark, Color Octet Technirho or E 6 Diquark è Produced via the color force, or from the valence quarks of each proton. l Dijet ratio can discover or confirm a dijet resonance with small systematics. è Gives a convincing signal for a 2 TeV q* with 100 pb -1. è Eventually it will be used to measure the resonance spin. l CMS is preparing to discover new physics at the TeV scale using dijets.

36. Robert Harris, Fermilab36 Backup Slides

37. Robert Harris, Fermilab37 The CMS Detector Hadronic Electro- magnetic Calorimeters Protons

38. Robert Harris, Fermilab38 CMS Barrel & Endcap Calorimeters (r-z view, top half)          HCAL BARREL ECAL BARREL SOLENOID HCAL END CAP ECAL END CAP HCAL END CAP ECAL END CAP Z                     HCAL OUTER 3 m HCAL > 10  I ECAL > 26 0

39. Robert Harris, Fermilab39 Trigger and Luminosity l Collision rate at LHC is expected to be 40 MHz è 40 million events every second ! è CMS cannot read out and save that many. l Trigger chooses which events to save l Two levels of trigger are used to reduce rate in steps è Level 1 (L1) reduces rate by a factor of 400. è High Level Trigger (HLT) reduces rate by a factor of 700. l Trigger tables are intended for specific luminosities è We’ve specificied a jet trigger table for three luminosities è L = 10 32 cm -2 s -1. Integrated luminosity ~ 100 pb -1. à LHC schedule projects this after ~1 months running. è L = 10 33 cm -2 s -1. Integrated luminosity ~ 1 fb -1. à LHC schedule projects these after ~ 1 year of running. è L = 10 34 cm -2 s -1. Integrated luminosity ~ 10 fb -1. à One months running at design luminosity. 4 x 10 7 Hz 1 x 10 5 Hz 1.5 x 10 2 Hz Event Selection CMS Detector L1 Trigger HLT Trigger Saved for Analysis

40. Robert Harris, Fermilab40 Path L1HLTANA E T (GeV) Pre- scale Rate (Hz) E T (GeV) Rate (Hz) Dijet Mass (GeV) Low252000146602.8 Med6040971202.4330 High1401442502.8670 Super4501146002.81800 L = 10 32 100 pb -1 Ultra2701194002.61130 L = 10 33 1 fb -1 Add New Threshold (Ultra). Increase Prescales by 10. Mass values are efficient for each trigger, measured with prior trigger L = 10 34 10 fb -1 Add New Threshold (Super). Increase Prescales by 10. Jet Trigger Table and Dijet Mass Analysis l CMS jet trigger saves all high E T jets & pre-scales the lower E T jets. è Prescale means to save 1 event out of every N events. As luminosity increases new trigger paths are added Each with new unprescaled threshold.

41. Robert Harris, Fermilab41 Trigger Rates & Dijet Cross Section (QCD + CMS Simulation) l Include data from each trigger where it is efficient in dijet mass. è Stop analyzing data from trigger where next trigger is efficient l Prescaled triggers give low mass spectrum at a convenient rate. è Measure mass down to 300 GeV è Overlap with Tevatron measurements. l Trigger without any prescaling saves all the high mass dijets l Expect the highest mass dijet event to be è ~ 7.5 TeV for 10 fb -1 è ~ 5 TeV for 100 pb -1 è LHC will open a new mass reach early! l Put the triggers together to form a cross section. |jet  |<1 Prescaled Trigger Samples P T D R

42. Robert Harris, Fermilab42 |  |<1.3 |  |<1 Jet response vs  relative to |  |<1.3 CMS Preliminary Jet  Region l Barrel jets have uniform response & sensitive to new physics  Jet response changes smoothly and slowly up to | jet  | = 1.3  CaloTowers with |  |<1.3 are in barrel with uniform construction.  CaloTowers with 1.3<|  |<1.5 are in barrel / endcap transition region  Some of our analyses use | jet  |<1.3, others still use | jet  |<1  All are migrating to | jet  |<1.3 which is optimal for dijet resonances Measure relative response vs. jet  in data with dijet balance è Data will tell us what is the region of response we can trust. Barrel Jet (|  |<1.3) Probe Jet (any  ) Dijet Balance  = 1.3 HB HE Hcal towers and  cuts Transition Region  = 1

43. Robert Harris, Fermilab43 Dijet Event Cleanup l Dijet events do not usually contain large missing E T  A cut at MET /  E T 99% efficient for P T > 100 GeV è Won’t change the QCD background to new physics. l Most unphysical background contain large missing E T è Catastrophic detector noise, cosmic ray air showers, beam-halo backgrounds  A simple cut at MET /  E T < 0.3 should remove most of these at high jet P T. è This cut is our first defense, simpler and safer than cutting on jet characteristics. 99% Efficiency Cut & Chosen Cut MET /  E T for QCD Dijets and Cut

44. Robert Harris, Fermilab44 Dijet Resonances: Optimization of  cut QCD cross section rises dramatically with |  | cut due to t-channel pole.  Z’ signal only gradually increases with |  | cut  optimal value at low |  |. Optimal cut is at |  | < 1.3 for a 2 TeV dijet resonance. è Optimization uses Pythia Z’ angular distribution for the resonance.  cut and cross section  cut and sensitivity

45. Robert Harris, Fermilab45 CDF Run II Resonance Search l Dijet resonance shapes are similar l Separate limits for W’ and Z’ models l Systematic uncertainties reduced from run 1. è Much lower jet energy scale error