1. 1 Probing New Physics with Jets at the LHC Robert M. Harris Fermilab Los Alamos Seminar November 28, 2007

2. Robert Harris, Fermilab2 Outline l Motivation l Introduction è Jets è Description of Jet Algorithms on Los Alamos Request è New Physics with Jets l CMS Jet Trigger and Jet Backgrounds from QCD l CMS Sensitivity to Signals of New Physics è New Particles: Dijet Resonances è New Forces: Contact Interactions 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 CMS Physics TDR Volume II à J. Phys. G: Nucl. Part. Phys. 34: 995-1579 è CMS Notes (2006 / 069, 070 and 071) l Ph.D thesis of two students è Selda Esen now a postdoc at Brown è Kazim Gumus from Texas Tech è Both mentored at the LPC l Demonstration of physics at LPC è Working within the CMS collaboration l Thanks to many CMS colleagues!

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 collaborators, 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 towers in a cone of radius CMS Barrel & Endcap Calorimeters proton  =-1 proton Jet 1 Jet 2  =1  Transverse  =0 

11. Robert Harris, Fermilab11 Iterative Cone Algorithm l Simplest cone algorithm. Used for this analysis, and for CMS trigger. è Consider towers in calorimeter with E T > seed threshold. Start with highest E T à Include all towers with E T > tower threshold within a cone centered on the seed è Calculate the jet Lorentz vector from the vector sum of tower energies. à Include all towers over threshold in new cone centered on jet momentum vector. à Iterate until towers in jet don’t change. Found a “stable cone” = jet in this algorithm. è Remove all the towers used from the list. Repeat until you run out of seeds. l Cookie cutter cone algorithm è First seed takes all towers in cone E T > Seed Threshold (1.0 GeV at CMS) E T > Tower Threshold (0.5 GeV at CMS) Example Calorimeter Towers Jet 1 Jet 2

12. Robert Harris, Fermilab12 JETCLU Cone Algorithm l First algorithm with splitting and merging, in use at CDF. è Similar to iterative cone, except towers are not removed from list. è Towers are part of all overlapping stable cones found from seeds until the jets are split or merged. à Merged if overlap p T is greater than 75% of lower p T jet. à Otherwise split, with towers going to nearest jet. Jet 1 Jet 2 Here overlap towers are “split” between jet 1 and jet 2. This example gave a larger Jet 2 than the Iterative cone. Overlap

13. Robert Harris, Fermilab13 Midpoint Cone Algorithm l Infrared safer algorithm from the Tevatron and briefly used by CMS. è Also starts with the stable cones found from seeds. è Includes mid-points between stable cones as seeds for more stable cones. è Concludes with usual splitting and merging based on overlap p T. l In this example the midpoint seed has given a single large jet. è Note that jet would have been found regardless of energy between the two seeds. è More infrared safe. à Stable for infrared radiation between two partons. à Up to NLO for dijets. “Infrared” Radiation Jet 1 Midpoint Seeded Cone

14. Robert Harris, Fermilab14 Seedless Infrared Safe Cone Algorithm l Infrared safe algorithm which will be the CMS standard cone algorithm for analysis. l Finds jets from ALL possible stable cones and does NOT use seeds. è Also an iterative cone algorithm with splitting and merging è Similar to midpoint algorithm but infrared safe. à Uses all possible stable cones and includes jets without a high E T seed. l Here it finds same large jet as midpoint plus an extra low energy jet. Jet 1 Jet 2 For high p T dijets and R=0.5 all cone algorithms discussed give very similar results in CMS simulations. There are also K T algorithms, which I am not familiar with, and will not discuss.

15. Robert Harris, Fermilab15 l Many signals are also large  Either large PDFs or  or both. Dijet Rates and Cross Sections l Rate = Cross Section x Luminosity è Luminosity (L) is rate of protons / area supplied by the LHC. è Design L=10 34 cm -2 s -1 ~ 10 fb -1 /month l Cross section from two factors è Parton distributions functions (PDFs) à Probability of finding partons in proton with fractional momentum x à Valence quarks u and d have large PDFs at high x (high dijet mass).  Parton scattering cross section  Jet 1 Jet 2 ^ | jet  | < 1 l QCD dijet cross section is large.   from color force is large ^ ^ Proton PDF(x a ) PDF(x b ) ~10 events/ month at design L ~10 7 events/ month at design L

16. Robert Harris, Fermilab16 Introduction to New Physics with Jets

17. Robert Harris, Fermilab17 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 Searching for more dijets in center of the CMS Barrel than at the edge. è More about this when we discuss sensitivity to contact interactions!   Gold Detector Discovery of Nucleus More of this Than of this Quark Compositeness Signal q q q q q q q q

18. Robert Harris, Fermilab18 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 CDF Run 1 ( Data – Fit ) / Fit time space

19. Robert Harris, Fermilab19 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

20. Robert Harris, Fermilab20 Example Resonance Model: Excited Quarks l If quarks are composite particles then excited states, q*, are expected è Excited quarks are produced when a ground state quark absorbs a gluon.  q* decay to the ground state q by re-emitting a gluon (qg  q*  qg). l Cross section is large because the interaction is from the color force (QCD). è Similar number of events produced as the QCD background ! g q q* q g Initial State Resonance Final State l Excited states are common in nature. Hydrogen Atom ground state Excited State Light from Excited States of Hydrogen

21. Robert Harris, Fermilab21 CMS Jet Trigger & Dijets from QCD

22. Robert Harris, Fermilab22 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 amounts by end of 2008. 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

23. Robert Harris, Fermilab23 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.

24. Robert Harris, Fermilab24 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

25. Robert Harris, Fermilab25 Uncertainties on Dijet Cross Section l Statistical Uncertainties è Simplest measure of our sensitivity to new physics as a fraction of QCD background è < 3% in prescaled region. è As luminosity increases our statistical error shrinks at high mass. l Systematic Uncertainties are large è Dominated by uncertainty in jet energy measurement. è Correlated with dijet mass. à Smooth changes, not bumps.

26. Robert Harris, Fermilab26 CMS Sensitivity to Dijet Resonances

27. Robert Harris, Fermilab27 Resonances and Background (CMS Simulation) l QCD cross section falls smoothly as a function of dijet mass. l Resonances produce mass bumps we can see if xsec is big enough.

28. Robert Harris, Fermilab28 Resonances and QCD Statistical Errors l Many resonances give obvious signals above the QCD error bars è Resonances produced via color force à q* (shown) à Axigluon à Coloron  Color Octet  T è Resonances produced from valence quarks of each proton à E6 Diquark (shown)  u d  D  u d l Others may be at the edge of our sensitivity.

29. Robert Harris, Fermilab29 Statistical Sensitivity to Dijet Resonances l Sensitivity estimates è Statistical likelihoods done for both discovery and exclusion 5  Discovery  We see a resonance with 5  significance à 1 chance in 2 million of effect being due to QCD. l 95% CL Exclusion è We don’t see anything but QCD è Exclude resonances at 95% confidence level. Plots show resonances at 5  and 95% CL è Compared to statistical error bars from QCD. 5 TeV 2 TeV 0.7 TeV 2 TeV0.7 TeV

30. Robert Harris, Fermilab30 Sensitivity to Resonance Cross Section l Cross Section for Discovery or Exclusion è Including systematics. è Shown here for 1 fb -1 è Also done 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.

31. Robert Harris, Fermilab31 Sensitivity to Dijet Resonance Models l Discover models up to 5 TeV in 10 fb -1 è 1 month at design luminosity = 10 fb -1 è Discovery up to 2.5 TeV with 100 pb -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 l Wide exclusion sensitivity è Extending Tevatron exclusions (<1 TeV) E6 Diquark Excited Quark Axigluon or Coloron Color Octet Technirho W ’ R S Graviton Z ’ Published Exclusion (Dijets) CMS 100 pb -1 CMS 1 fb -1 Mass (TeV) 95% CL Sensitivity to Dijet Resonances 0 1 2 3 4 5 6 CMS 10 fb -1

32. Robert Harris, Fermilab32 CMS Sensitivity to Quark Contact Interactions

33. Robert Harris, Fermilab33 Contact Interactions & Compositeness New physics at a large scale  è For example composite quarks.  Intermediate state looks like a point for dijet mass << . è Giving a contact interaction. M ~  Composite Quarks New Interactions M ~  q q q q Quark Contact Interaction  q q q q l Increases rate at high dijet mass. è But the signal in rate alone is hard to find due to uncertainties in jet energy & parton distributions. Dijet Mass <<  l We will use a ratio of two rates. è Look at angles like Rutherford did !

34. Robert Harris, Fermilab34 Dijet Ratio and New Physics l Dijet Ratio = N(|  |<0.5) / N(0.5<|  |<1)  Number of events in which each leading jet has |  |<0.5, divided by the number in which each leading jet has 0.5<|  |<1.0 l Simplest measurement of angle dist. è Most sensitive part for new physics è It was first introduced by D0 Jet 1 Jet 2 Numerator Sensitive to New Physics Signal Denominator Dominated by QCD Background Jet 1 Jet 2 (rare) or  = -1 - 0.5 0.5 1 z z

35. Robert Harris, Fermilab35 Dijet Ratio and Uncertainties (Smoothed CMS Simulation) l QCD Background Simulation is flat l Signal rises with mass è Clear statistical sensitivity to contact signal We find 5  discovery and 95% CL exclusion sensitivities for . è Including both the statistical and systematic uncertainties. l Small systematic uncertainties. è Cancel in the ratio

36. Robert Harris, Fermilab36 Left-Handed Quark Contact Interaction (Eichten, Lane & Peskin)  + for 100 pb -1 (TeV)  + for 1 fb -1 (TeV)  + for 10 fb -1 (TeV) 95% CL Exclusion6.210.414.8 5σ Discovery4.77.812.0 CMS Sensitivity to Contact Interactions Published Limit (D0):  + > 2.7 TeV at 95% CL è CMS could quadruple this limit in 2008.  can be translated roughly into the radius of a composite quark.  h =  x  p ~ (2r) (  / c)  r = 10 -17 cm-TeV /   For  ~ 10 TeV, r ~ 10 -18 cm Composite Quark Preon r

37. Robert Harris, Fermilab37 Conclusions l The LHC will probe exciting new physics with simple measurements of jets. l We have presented a jet trigger table and dijet analysis developed at LPC. l CMS can discover a strongly produced dijet resonance up to several TeV. è An excited quark, an E 6 diquark, or even an unanticipated new particle! CMS can discover a quark contact interaction up to  + =12 TeV with 10 fb -1. è Corresponds to a quark radius of order 10 -18 cm if quarks are composite. l The LHC Physics Center at Fermilab is active in jet physics analysis. è Mentoring the postdocs and graduate students who will analyze first CMS data. è Well integrated in the CMS physics organization. l We are looking forward to exciting discoveries at the TeV energy scale!

38. Robert Harris, Fermilab38 Dijet Resonance Cross Sections l Resonances produced via color force, or from valence quarks in each proton, have the highest cross sections.