On the Trail of the Standard Model Higgs

On the Trail of the Standard Model Higgs Boson Jason Nielsen Lawrence Berkeley National Lab. Fermilab, February 7, 2006 J. Nielsen, LBNL VERTEX 2004

A Trail Map for the Search • Physical Higgs boson is concrete prediction of electroweak symmetry breaking mechanism in the standard model • Experiments suggest Higgs boson mass likely between 114 and 200 Gev/c 2 • Isolating Higgs boson production at Tevatron requires advanced experimental techniques and theoretical calculations J. Nielsen, LBNL 2

Central Challenges in High-Energy Physics • Unification of the basic forces and the origin of mass for the fundamental particles • Unexpected new physics or extra dimensions not included in Standard Model • Unknown new physics (forces or particles) hinted at by cosmology Particle collisions at the energy frontier enable us to pursue these and other questions about nature J. Nielsen, LBNL 3

Motivation for a Higgs Boson Weak interaction acts at short-range Something must break symmetry and give mass to weak gauge bosons Expect “new” physics below 1 Te. V WL WL WL ? WL M? < 1000 Ge. V A single weakly-coupled Higgs boson: • remnant of EWSB which gives mass to W±, Z bosons • has well-predicted production and decay rates But that also: • leaves problem of quadratically divergent corrections to MH • which leads in turn to supersymmetric theories This talk focuses exclusively on the Standard Model Higgs boson. J. Nielsen, LBNL 4

Results from Recent Electroweak Fits Effects of the Higgs boson are felt via loop interactions Precision measurements are sensitive to the Higgs mass Updated summer 2005 with new Tevatron mt=172. 7± 2. 8 Ge. V m. H < 186 Ge. V/c 2 at 95% CL (or 219 including LEP search) J. Nielsen, LBNL 5

Higgs Searches in e+e- Collisions Combined ALEPH/DELPHI/L 3/OPAL results (April 2003) NN selection ALEPH 3 s excess 68%, 95% (Phys. Lett B 495 (2000) 1. ) m. H > 114. 4 Ge. V at 95% CL J. Nielsen, LBNL 6

Tevatron Run 2 at Fermilab Tevatron achievements in 2005: §Record inst. luminosity 1. 8 E 32 cm-2 s-1 § 24 pb-1 integrated in a single week results shown today use data through Aug 2004 These CDF results use integrated luminosity 320± 20 pb-1 (cf. 110 pb-1 from Run 1) J. Nielsen, LBNL delivered to tape 7

Higgs Production and Decay in pp Collider H bb is decay of choice • but non-res. bb too high WH, ZH are good compromise • lepton cuts QCD background J. Nielsen, LBNL 8

CDF II Detector at the Tevatron J. Nielsen, LBNL 9

Run 2 CDF Silicon Detectors Solid-state detectors: • e-h pairs along particle track • spatial resolution <10 m Three different detector geometries fill the silicon tracking volume • 3 -dimensional space points • 2 -D closest to beams 8 layers, 704 ladders, 722432 channels Total sensor area: 6 m 2 Largest operational silicon detector J. Nielsen, LBNL 10

Modeling CDF Silicon Detector Changes Overall detector status: Holding steady at 85% good Offline noise calibration used to improve clustering 1 0. 8 0. 6 • Realistic simulation uses • Detector configuration • Lists of dead/noisy strips J. Nielsen, LBNL Silicon Hit Occupancy (Layer 3 SVX) 0. 4 0. 2 0 1 2 3 Data Simulation 4 5 6 Phi (radians) 11

Tevatron Cross Section Hierarchy In proton-antiproton collisions at s = 1. 96 Te. V: b-jet pairs from QCD high-energy leptons 1 Particle production rates vary widely: the Higgs is the “needle in the haystack!” J. Nielsen, LBNL 0. 05 12

Standard Model Higgs Boson Searches Given gg H bb is difficult, what channels are promising? For Higgs mass < 140 Ge. V • WH l bb and ZH bb • Trigger on leptons or missing energy from • Look for resonant bb decay in mass distribution For Higgs mass > 140 Ge. V • gg H WW and WH WWW • Trigger on leptonic decay of W bosons • Distinctive angular spectrum between leptons J. Nielsen, LBNL 13

Promising Channels for Higgs at Tevatron Studies before Run 2 highlighted low-mass SM Higgs search • Follow up on possible LEP excess at 115 Ge. V • Extend reach toward 130 Ge. V by combining CDF and D 0 Update indicated workshop results are feasible Benchmark with current physics analyses using 350 pb-1 J. Nielsen, LBNL 14

WH l bb Channel Overview Large H bb branching fraction Trigger on high-ET lepton: • factor of 5 less than gg H Key is understanding composition of lepton + 2 jets sample • Challenge of physics objects: leptons, jets, missing ET • Constrain W+jets production, including W + heavy flavor • Understand fakes from detector and algorithms • Overlap with top quark sample, especially single top To be published in PRL J. Nielsen, LBNL 15

b Tagging Tool: Secondary Vertexing B hadrons in WH signal events are long-lived and massive Vertex of displaced tracks Form vertex of displaced tracks • Sec vtx err: 190 microns Cut on decay length significance Event is tagged if any jet is tagged Also check efficiency in data events vs. efficiency in simulation J. Nielsen, LBNL 16

Jet Energy Measurement and Calibration Important handle in Higgs search: bb pair mass resonance Current resolution is 17 Ge. V • defines mass window • direct factor in acceptance • luminosity equivalence Working to improve resolution • track + calorimetry (H 1) • lookup tables (hyperball) • neural network function J. Nielsen, LBNL 17

Validation with Top Quark Physics “Yesterday’s discovery is today’s tool” Pair production at Tevatron Develop and validate techniques and calculations in W+jets But also: • only quark at electroweak symmetry breaking scale • constrains Higgs mass; fertile ground for new physics J. Nielsen, LBNL 18

Top Quark Pair Production Cross Section Search for top production in Wb. Wb lvbjjb sample: 1. test b-tagging algorithms and efficiency 2. test W reconstruction requirements (lepton, neutrino) 3. test calculation of other physics with same final signature J. Nielsen, LBNL PRD 71, 052003 (2005) PRL in preparation with more data 19

WH l bb Event Selection at CDF W boson • Exactly one isolated e/ with ET>20 Ge. V • Total missing ET>20 Ge. V Jets • Exactly two jets with ET>15 Ge. V and | |<2. 0 • At least one jet must be tagged with secondary vertex • No extra jets with 8 < ET < 15 Ge. V or with | |>2. 0 Rejection of top pair events with dilepton signature: • Veto events with a second isolated track p. T> 20 Ge. V J. Nielsen, LBNL 20

WH l bb Signal Acceptance Calculated using PYTHIA Monte Carlo and full simulation Includes effects from: • all selection criteria • simulation/data difference • trigger efficiencies lepton ID: 30% kinematic: 40% b-tagging: 45% J. Nielsen, LBNL Significant uncertainties: • b-tagging efficiency (8%) • ISR, FSR modeling (5%) • jet energies (3%) 21

Backgrounds to WH Production Presented in rough order of increasing contribution to the vertex-tagged lepton + jets sample: non-W QCD false isolated leptons or false missing energy top quark production physics background mistags in W events false b-tags w/ lepton + MET W + heavy flavor physics background Each background estimate is a miniature analysis unto itself. Techniques can be spun off to measure other physics processes. J. Nielsen, LBNL 22

False tags in W+jets events Fraction of light flavor jets are “mistagged” with displaced vertex • Light flavor Lxy distribution is roughly symmetric about 0 • Use negative Lxy distribution to predict positive side “mistags” Correction factor (27 ± 13)% accounts for material interactions, long-lived light flavor particles (KS, L) J. Nielsen, LBNL 23

Physics Background: W + heavy flavor Real W + real tagged heavy flavor = physics background! Since Run 1, little confidence in absolute normalization from W+HF jets Monte Carlo calculations • Use W+jets data for normalization • Calculate fraction of W + heavy flavor using Monte Carlo Viable NLO W + heavy flavor MC is not yet available Use ALPGEN Monte Carlo: • Matrix element LO calculation • Includes W+HF/W+ n jets processes • Feed to HERWIG parton shower J. Nielsen, LBNL 24

Monte Carlo Matching Prescriptions Matrix element describes hard, separated partons Parton shower models emission of soft gluons Problem: PS radiation in W + n jets process results in events which are also produced by ME in W + (n+1) process Solution: Mangano’s prescription requires each ME parton match a final state particle-level “jet” and no extra jets; otherwise reject event 1 2 3 4 J. Nielsen, LBNL 3 1 a b 2 25

Estimate of W+HF Contribution • Calibrate calculations using jet processes without W boson • Gluon splitting in 3 -jet sample similar to W+HF diagrams Calibrated correction factor 1. 5± 0. 4 consistent with phenomenological K-factor (PRD 71, 052003 (2005)) Additional factor 1. 2 from fit to W+1 -jet data (normalization) Final calculation of true W+HF backgrounds is then: J. Nielsen, LBNL 26

Summary of Background Contributions Use W+1 -jet bin to constrain W+HF bkgd Top pair cross section Measured from the W+3, 4 -jets events Number of W+2 -jet Events consistent with Background predictions J. Nielsen, LBNL 27

bb Dijet Invariant Mass Distribution J. Nielsen, LBNL 28

CDF WH Production Limits J. Nielsen, LBNL 29

Cross-Check with Double-Tagged Events Fluctuation at 100 Ge. V not present in this data sample Slightly worse expected limits due to reduced acceptance J. Nielsen, LBNL 30

Search for ZH bb Production Powerful search in low-mass Higgs regions; ID missing energy b-jet Missing ET Check background estimates in control regions: leptons, MET-jet y x b-jet Selection for bb: • at least 2 jets (>25 Ge. V) • missing ET > 70 Ge. V • at least 1 b-tagged jet Major background: 180 o jet J. Nielsen, LBNL jet Fake Missing ET 31

CDF ZH bb Candidate Event Dijet mass mjj= 82 Ge. V Missing ET=145 Ge. V Two b-tagged jets: • Jet 1 ET= 100. 3 Ge. V • Jet 2 ET= 54. 7 Ge. V J. Nielsen, LBNL Could be example of ZZ 32

ZH bb Search Results Includes sliding mass window dependent on Higgs mass Slightly more sensitive than WH at low mass J. Nielsen, LBNL 33

Search for gg H WW Most useful for m. H > 135 Ge. V Select events with • leptons 20, 10 Ge. V • missing ET > 50 (CDF), 20 (DØ) • low-energy jets allowed • Interesting angular correlation due to scalar Higgs • Different from SM W+W- bkgd decay angular correlation! • Use angle between dileptons as final discriminant J. Nielsen, LBNL W+ W- e+ e- 34

gg H WW Search Results Sliding cuts for different m. H • Example result for m. H=160 • W+W- still dominant background • Fit for signal cross section Exclusion similar to WH, ZH • extends to higher masses • still limited by statistics • work to extend acceptance J. Nielsen, LBNL 35

Summary of Tevatron SM Higgs Searches J. Nielsen, LBNL 36

Summary of Tevatron SM Higgs Searches J. Nielsen, LBNL 37

Improvements Underway Identified potential to approach SM cross section exclusions Improvement Tools common to CDF analyses Analysis-specific improvements WH l bb ZH bb Mass resolution 1. 7 Continuous b-tag (NN) 1. 5 Forward b-tag 1. 1 Forward leptons 1. 3 1. 0 Track-only leptons 1. 4 1. 0 1. 75 WH signal in ZH 1. 0 2. 7 CDF+DØ combination 2. 0 NN Selection And don’t forget factor of 10 more data! Challenge to develop and apply improvements to Higgs searches J. Nielsen, LBNL 38

Neural Network Event Selection Studies Two advantages over selection with simple rectangular cuts: 1. Use new event variables with weak or correlated effect 2. Exploit correlations between event variables Run 1 CDF study of expected WH l bb limits: s=12. 6 pb L=90 pb-1 Standard selection s=9. 3 pb L=150 pb-1 s=13. 0 pb L=95 pb-1 Use NN selection s=10. 0 pb L=95 pb-1 Using fit to NN output distro gives additional improvement DØ collaboration estimated 75% improvement in ZH channel J. Nielsen, LBNL 39

Large Hadron Collider at CERN Next generation collider: startup scheduled for 2007 Italy p 14 Te. V Luminosity target: 1034 cm-2 s-1 p Increased production of heavy particles like Higgs, top quark CMS J. Nielsen, LBNL ATLAS More particles at higher energy requires new detector design and technology 40

Prospects for SM Higgs at LHC Should discover SM Higgs regardless of mass value! Low-mass Higgs channels: • count on H (sm =1. 5 Ge. V/c 2 • add WBF H WW, H tt • tt. H needs ttjj background study High-mass Higgs channels: • golden mode 4 e/ opens >2 m. Z Experimental techniques: Learn from Tevatron results! J. Nielsen, LBNL 41

How the Tevatron Informs LHC Higgs Real-world Tevatron experience and detailed LHC MC studies • utilize the best of both worlds! Some examples: Selection of weak boson fusion depends on central jet veto • check efficiency using top events? Tools for matrix element-parton shower matching • NLO vs LO comparison Techniques for measuring b-tag eff, t ID in hadron collider environment J. Nielsen, LBNL 42

Summary • Expect Standard Model Higgs boson mass to be below 200 Ge. V/c 2 • Increased Tevatron luminosity and upgraded detectors offer possibility to probe Higgs boson masses above 114 Ge. V LEP exclusion • First analyses highlight challenges and provide benchmark for sensitivity estimates • Identified to-do list for improving search strategies • Chance to explore interesting low-mass region before LHC turn-on J. Nielsen, LBNL 43

J. Nielsen, LBNL 44

Calibrating the b-tagging Efficiency Select soft (p. T>8 Ge. V) electron events to enrich HF, then require tag to reduce charm contributions Measure ratio of single- and double-tagged events (in both data and simulation) “Scale factor” measured on this sample is applied to other samples, even if the efficiencies differ J. Nielsen, LBNL 45

Jet Events without W Bosons Fake isolated (I<0. 1) lepton along with fake missing ET • for QCD background these fakes are uncorrelated Signal region D predicted by Model event kinematics from sideband J. Nielsen, LBNL 46

Even Z bb is Hard! J. Nielsen, LBNL 47

Calculating Heavy Flavor Fractions - Wc Processes we are interested in: Wbb, Wcc, Use ALPGEN + HERWIG samples stitched together after MLM matching prescription, ex. Wc + 0, 1, 2 p Not sensitive to matching criteria Back to reality: Does this MC calculation really reflect the data? J. Nielsen, LBNL 48

One Provocative Candidate Event HZ bbbb selection ECM=206. 7 Ge. V 3 NN b-tagged jets Reconstructed m. H = 110 ± 3 Ge. V/c 2 J. Nielsen, LBNL 49

Theoretical Constraints on Higgs Mass Hambye & Riesselmann Higgs Self-Coupling Diverges with Increasing Energy Allowed “Mexican Hat” Brim Turns Over Finding a Higgs boson with m. H~120 Ge. V is evidence for Te. V-scale New Physics J. Nielsen, LBNL 50