Search for the SM Higgs Search for the SM Higgs boson at LHCboson at LHC

Patricia Lobelle Pardo(Universidad de Cantabria)

On behalf of the ATLAS and CMS On behalf of the ATLAS and CMS collaborationscollaborations

XVII International Conference on Deep-Inelastic XVII International Conference on Deep-Inelastic Scattering and Related SubjectsScattering and Related Subjects DIS09, 26th-30th April 2009, Madrid

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid

Higgs current limitsHiggs current limits

• Experimental limits Experimental limits LEPLEP : mH > 114.4 GeV/c2 at 95% CL TevatronTevatron : Excluded the mass range of 160 GeV/c2 to 170 GeV/c2 at 95%CL

Indirect constraintsIndirect constraintsDerived from precise EWK measurements: mH = 90 +36

-27 GeV/c2

(mH < 163 GeV/c2) (mH <191 GeV/c2 if including LEP2 results)

• Theoretical limitsTheoretical limitsFinite and positive Higgs couplings

A light Higgs is favoured by measurementsA light Higgs is favoured by measurements

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 2

SM Higgs production at LHCSM Higgs production at LHCGluon-gluon fusion: Gluon-gluon fusion: dominant mode at LHCdominant mode at LHC Vector Boson FusionVector Boson Fusion

Associated productionAssociated production

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 3

SM Higgs decaysSM Higgs decays At Low massAt Low mass ( m ( mHH< 2m< 2mZ Z ) )

• HH bb bb : BR ~0.85 but huge QCD background

• HH: accessible through VBF

• HH : very important despite the low BR (~0.002 ) due to the excellent /jet separation and resolution

• HHWW*WW*2l22l2 : accesible through gg fusion and VBF, BR~1 at mmHH ~160 GeV/c2

• HHZZ*ZZ*4l 4l : also accesible

For Higher massesFor Higher masses HHWW* WW* 2l2v 2l2v and H HZZ*ZZ*4l4l

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 4

ATLAS and CMSATLAS and CMS

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 5

State of the artState of the art

NEW re

sults w

ith F

ull Sim

ulation!

NEW re

sults w

ith F

ull Sim

ulation!

Plus recent u

pdates on

Plus recent u

pdates on

SM Higgs channels with

SM Higgs channels with

new CMSSW software

new CMSSW software

FrameWork!!

FrameWork!!

• Better detector description and simulation: geometry, material budget, Geant4

• New approaches to match parton showers andmatrix elements (ALPGEN + MLM matching, SHERPA etc.)

• New (N)NLO Monte Carlos for signals & backgrounds(MCFM, MC@NLO)

Sensitivities at NLO!!( NNLO maybe available before data

taking , which will increase x-sections)

• More detailed, better understood reconstruction methods (partially based on test beam results), improved trigger simulation, event reconstruction and analysis tools• Strategies to estimate backgrounds from data and express AlCa streams• Improved statistical treatment (also including treatment of systematic uncertainties)• Tevatron data are extremely valuable for validation, work has started 6P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 6

HH Small branching ratio (0.002 at mH =120 GeV/c2) Final states produced through W, top and b loops Important channel for discovery at low mass (mH<130 GeV/c2) Very clear signature: 2 high Et isolated photonsVery clear signature: 2 high Et isolated photons

• RequirementsRequirements – Good understanding of electromagnetic calorimeter performance and calibration– Excellent /jet separation – Good mass resolution– and Vertex reconstruction

• BackgroundsBackgrounds:– Irreducible: +jets– Reducible: +jets, jets, DY

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 7

HH ATLASATLAS• Vertex determined from extrapolation using calorimeter samplings• Converted photons used to improve vertex determination: 57% Higgs events have ≥1 conversion• Signal divided into categories according to, #jets, #converted photons• Inclusive analysis + search for di-photons with jets• Unbinned maximum-likelihood fit

CMSCMS Background reduction using photon isolation and kinematic Background reduction using photon isolation and kinematic informationinformation MM observed as a peak above background observed as a peak above background Cut-based analysis– Splitting in categories according to and lateral shower shape variable Optimized analysis– Loose sorting and additional kinematical variables– Event-by-event kinematical Likelihood Ratio with background pdf taken from sidebands, signal pdf from MC

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 8

Results for HResults for H ATLASATLAS

• For For mH=120 GeV/c2 and 10 fb-1 :– Inclusive analysis, using event counting σ= 2.6– Combining the 0,1,2 jets analysis +using an unbinned maximum-likelihood fit: Floating (fixed) mass fit : σ=2.8 (3.6)

CMSCMS • 5σ discovery between LEP lower limit and 140

GeV/c2 with less than 30 fb−1 of integrated luminosity

• 5σ discovery with event by event estimation of the S/B ratio possible at mH=120 GeV/c2 with 7-8 fb-1

CMS TDR 2006

CSC 2008

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 9

HH→WW*→2l2→WW*→2l2

Signal: 2 high pT isolated leptons, MET and no central jets

• BackgroundsBackgrounds:– WW, tt, W+jets, Z+jets, tW, WZ, ZZ…

Two main discriminants: Two main discriminants: Angle between leptons in the transverse plane : main variable to reject WW (small opening angle for the signal due to

spin correlations) Central Jet Veto for ttbar reduction

No mass peak (undetected neutrinos) Needs a good background understanding

11 fbdtL

Main search channel for range 140<mH<2mZ

-Highest branching ratio for mH >140 GeV/c2: 95% at mH = 160 GeV/c2

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 10

HH→WW*→2l2→WW*→2l2ATLASATLAS ( e final state, H+ 0j, H+2j analysis )• Preselection : •2 opposite-sign isolated and identified leptons• Cuts on mll, MET, Δφll , Z removal •Central Jet veto & b-tag veto• Final selection:2D Fit of transverse mass and Higgs candidate pT in 2 bins of di-lepton azimuthal angle Δφll to extract S/B ratio in signal region

CMS CMS (ee, , e final states, H+0j analysis)• Preselection:•2 opposite-sign isolated and identified leptons•Cuts on mll, MET•Central Jet Veto•Variables used to reduce background : pT of the leptons, mll , Δφll , MET• Final selection:Mass dependent cut based & multivariate analysis(optimization for 1fb-1)• Control regions: fake leptons, background normalization

Analysis strategyAnalysis strategy

11 fbdtL

CMS PAS HIG-08-006CMS PAS HIG-08-006

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 11

Results for HResults for H→WW*→2l2→WW*→2l2

CMS PAS

HIG-08-006

Improvements on Lepton ID , mass dependent cuts, more data driven approaches (compared to TDR) CMS TDR

Updated CMSCMS result:

Sensitivity to a SM Higgs improved using a mass dependent multivariate analysis

1414

ATLASATLAS

– Combined significance above the 5 level for ~ mH >140 GeV/c2

ATLAS results to be scaled by √2 and to 1 fb-1

Cuts have been Cuts have been optimized separately optimized separately for 1fbfor 1fb-1-1 maximizing the maximizing the expected statistical expected statistical significance.significance.

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 12

HHZZ*ZZ*ll• SignalSignal: Very clean signature ( 2 pairs of isolated

leptons 4e, 4μ, 2e2μ) with high branching ratio except at mH ≈ 2 mW

Selection of isolated (tracker + calorimeter) muon and electron pairs with opposite charge Isolation and Impact Parameter Significance main variables to reject Zbb, tt At least one Zll on shell (Mass of the leptons pairs some discrimination against the background) 4-lepton mass reconstruction4-lepton mass reconstruction Peak in the mass spectrum of the signal most discriminating variable against ZZ

• BackgroundsBackgrounds:– Irreducible: ZZ* dominant– Reducible: Zbb, tt, ZW, Z + X

– Excellent mass resolution (1.5 – 2 GeV/c2 for MH= 130 GeV/c2)

– Powerful analysis in a wide mass range Powerful analysis in a wide mass range

CMS PAS

HIG-08-003

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 13

P.Lobelle (U.Cantabria)

Results for HResults for H→ZZ*→4l→ZZ*→4l

CMS update:• Cuts optimised for 1 fb-1 and data driven methods for background optimization control• Significance of about 3σ can be reached for mH ~ 200 GeV/c2 at ∫Ldt = 1 fb-1

• SM Higgs boson can be excluded for mH > 185 GeV/c2 with ∫Ldt = 1 fb-1

• "look-elsewhere" effect: in real search, significance to be de-rated by ~1 sigma

Comparable significance Highly sensitive in the high mass region (200 GeV/c2 < mH < 400 GeV/c2) and in the 150 GeV/c2 region, where the Higgs boson could be discovered with 5 fb−1

3 channel independent +1 combined analysis with several levels of systematic uncertainties treatment

CMS TDR 2006CSC 2008

CMS PAS

HIG-08-003

-1-1

1428/04/09 DIS09 Madrid 14

Important items:• Tau tagging (Likelihood, NN methods)• tt rejection (b-jet ID and veto for lepton-lepton)•Z+jets background, especially at low masses

VBF HVBF H→→

CMS PAS HIG-08-008

CSC 2008

3 final states : •Lepton-lepton•lepton-hadron (CMS only this) •hadron-hadron

• Invariant mass of the pair reconstructed via the collinear approximation (breaks down when the 2 taus are back to back)

Important channel for range 115-145 GeV/c2

Signal: 2 leptons or t-jets in central region, MET and 2 forward jets

• BackgroundsBackgrounds:– Zjj, tt, Z/*->

ATLASATLAS lepton-lepton + lepton-hadron final states 5achieved in the range 115-120 GeV/c2 with 30 fb-1

15P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 15

SM channel combinationSM channel combination

• With With 10 fb10 fb-1-1 ((normally normally considered as one LHC year at low luminosity), considered as one LHC year at low luminosity), 5σ discovery for m5σ discovery for mHH in [~120, ~500] GeV/c in [~120, ~500] GeV/c22

• With With 5 5 fbfb-1-1 5σ discovery for m 5σ discovery for mHH in [~130, ~450] GeV/c in [~130, ~450] GeV/c22

Combining results from both experiments, around half of this luminosity

5fb5fb-1-1

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 16

SM Higgs searches at SM Higgs searches at √s<14TeV√s<14TeV

LHC will start working with center of mass LHC will start working with center of mass energy lower than 14 TeV around 10 TeVenergy lower than 14 TeV around 10 TeV

► Main Effect: cross section changes

Different energy of LHC has two effectsDifferent energy of LHC has two effects:

● Cross section for signals (and background)goes down

● Signal (Higgs production) goes down slightly faster:Higgs is mainly produced from gg and backgrounds from qq

Efficiency and AcceptanceEfficiency and Acceptance:● Higgs becomes relatively “heavier”, i.e. decay products become relatively more central for smaller LHC energies

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 17

CMS CMS projection for LHC@10TeVprojection for LHC@10TeV• LHC 2009 - 2010 luminosity LHC 2009 - 2010 luminosity

performance – estimate: performance – estimate: ~200 pb-1 of “good data”

• StrategyStrategy:– First understand detectors,– do SM measurements,– then search for the Higgs…

• Signal and bkgd yields re-scaled 14→10 Signal and bkgd yields re-scaled 14→10 TeVTeV: loss of a factor of 1.5 in sensitivity, or a factor of ~2 in luminosity

• With roughly ~200 pbWith roughly ~200 pb-1-1:: Reach sensitivity for a SM Higgs with

mH~160-170 GeV/c2 (comparable to the current Tevatron sensitivity) ( but region region just excludedjust excluded by Tevatron) by Tevatron)

CMS CMS

AN-2009/020 AN-2009/020

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 18

ConclusionsConclusions If the Higgs boson is there, ATLAS and CMS are ready ATLAS and CMS are ready to find it… …unless it is discovered or excluded first at the Tevatron!!!

For a SM Higgs, with a combination of ATLAS & CMS with a combination of ATLAS & CMS @14TeV, between ~ 1 and 5 fb-1 are needed depending on mass value. Benchmark luminosities:

~~0.1 fb0.1 fb-1-1 exclusion limits will start carving into SM Higgs cross section ~0.5 fb0.5 fb-1-1- 1 fb- 1 fb-1-1 discoveries start to become possible in the region near the one

excluded by Tevatron ( MH~160-170 GeV/c2) ~5 fb~5 fb-1-1-10 fb-10 fb-1-1 SM Higgs could be discovered (or excluded) in full mass range

(~110-500 GeV/c2)

LHC efforts focused to get as much luminosity as possible as much luminosity as possible to get sensitivity to a SM Higgs boson.

ATLAS and CMS are doing their best to commission and understand the detectors to be ATLAS and CMS are doing their best to commission and understand the detectors to be ready when collisions startready when collisions start

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 19

Post-discovery questions that would need to be answered…Post-discovery questions that would need to be answered…– Is it a Standard Model Higgs ?– What are the couplings to the rest of particles ?– Is there only one Higgs? MSSM, other models?– What is its precise mass, width, quantum numbers ?– Does it generate electroweak symmetry breaking and give mass to fermions?– Higgs discovery also raises the “hierarchy” problem…

LHC will answer these fundamental questions in the next yearsLHC will answer these fundamental questions in the next years

Ready for Higgs discovery, but if it is not found, detectors are able to search for objects with similar signatures…

P.Lobelle (U.Cantabria) 28/04/09 DIS09 Madrid 20