??af??e?a 1

1
SM Higgs? 4 leptons and MSSM Higgs?µµ- in ATLAS
Dimitris Fassouliotis C. Kourkoumelis K.
Nikolopoulos Univ. of Athens
XXVI Work1shop on recent Developments in High Energy Physics and Cosmology
Ancient Olympia, 16-19 April 2008
2

• Improvements from previous ATLAS studies
• (N)NLO calculations for the cross sections
• Fully simulated data samples
• Realistic detector description (as installed)
• Trigger study
• Improved event selection
• Background extraction from data
• Careful study of systematic uncertainties
• Improved statistical techniques for the
  estimation of the significance

3
The ATLAS Detector
Hadron Calorimetry Fe/Sci Cu/LAr s/E60/?E3
E/M Calorimetry Pb/LAr s/E10/?E
Inner Detector2 T solenoid Pixels, Si Strips
and Transition Radiation Tracker
Muon Spectrometer Air-core toroid, Precision
drift chambers and trigger chambers.PT
resolution 10 at 1 TeV
4
Current SM Higgs Limits LEP direct
search mHgt114.4 GeV _at_95 CL Tevatron Direct
search LEP, SLD, Tevatron e/w fit mHlt182 GeV
5
Why H ? ZZ ? ll-ll-

• Significant discovery potential over a wide mass
  range
• Clean channel involving only electrons and muons
  ? first objects to understand at LHC start-up
• Full event reconstruction ? Mass peak
• Background estimation from sidebands
• Benchmark channel for detector performance

6
Backgrounds
Main Backgrounds Irreducible ZZ/?
?4leptons Reducible Zbb?4leptons tt
?WbWb ? 4leptons (McAtNLO) Secondary backgrounds,
like WZ?3leptonsfake lepton, are not an issue.
Zbb?4leptons
ZZ/??4leptons
15 of gg?Zbb
AcerMC / rescale to NLO MCFM ? K1.42Remaining
uncertainty 16
qq?ZZ/?Pythia / Rescale to NLO
MCFMltKgt1.34Rescale for gg box 30 of qq
The ZZ/? background dominates after selection
7
Event preselection
2µ, 2e and µe triggers select almost 100 the
signal events surviving selection
low statistics four lepton final state ?
excellent lepton identification needed
ATLAS Preliminary
Muons
Electrons
ATLAS Preliminary
2 leptons PT gt 20 GeV, ?lt2.5 2 leptons PT gt
7 GeV, ?lt2.5 On shell di-lepton candidate
consistent with Z mass Off shell di-lepton
candidate above lower threshold.
Reduce non-resonant tt background
Reduce low mass combinations
8
Background Rejection Lepton Isolation
Calorimeter Isolation
Both Zbb and tt have leptons close to b-jets.?
Activity around the lepton in Calorimeter and
Tracker
ATLAS Preliminary
Isolation ET / PT in ?R around lepton ?R ?
compromise between physics and pile-up
Tracker Isolation
?R(??2?f2)1/2
1
9
Background Rejection H?4µ Lepton Isolation
Calorimeter Isolation
Track Isolation
Normalized Isolation criteria are more powerful
Despite the obvious correlation, a Fisher
discriminant can provide improvement wrt track
isolation for signal efficiencies lt95
10
Background Rejection Impact Parameter / Vertex
fit
Displaced vertices are present in both reducible
backgrounds Impact Parameter Significance
d0/sd0 Transverse impact parameter resolution
O(10 30 µm) Transverse primary vertex spread 15
µm, should be taken into account
Isolation Impact Parameter CriteriaO(102)
Rejection for Zbb O(103) Rejection for tt for
signal efficiency O(90)
11
Mass Reconstruction - Constraint fit on Z mass
Z constraint fit? 10 Improvement in the mass
reconstruction resolution
Mass Width
12
Mass Distributions
All topologies included After event selection
signal can be clearly observed above background
Rencontres de Physique La Thuile Feb 2008 K.
Nikolopoulos
13
MSSM Higgs?µµ-

• Associated Production dominates for high tanb
  values 10
• H/A mass degenerate for mA gt130 GeV/c2 (h/A for
  mA lt130 GeV/c2)
• Observed signal is the sum of all
  degenerate states

14
Why MSSM Higgs?µµ-

• Very efficient muon identification /
  reconstuction
• Complete final state
• Excellent Higgs mass resolution
• Not visible in SM, enhanced in MSSM

15
MSSM Higgs?µµ- Backgrounds
Backgrounds several hundred times higher than
Signal Zjets dominates at low masses Reducible
ttbar important at higher masses
16
MSSM Higgs?µµ- Event Selection
Preselection µµ-, isolated, pTgt20 GeV/c,
?lt2.7 ETmisslt40 GeV
Analysis 1 0 b jets
Analysis 2 1 b jets
17
MSSM Higgs?µµ- Mass distribution
0 b jets analysis (inclusive analysis similar)
1 b jets analysis S/B improved by 10
18
MSSM Higgs?µµ- Background extraction from data
Using ee-, µ e signatures
Function to fit the background Breit-Wigner
exponential
Using the side bands
19
MSSM Higgs?µµ- Systematic uncertainties
Experimental Conservative ATLAS performance
estimates for L 1 fb-1
Theoretical
20
MSSM Higgs?µµ- Significance Evaluation
21
MSSM Higgs?µµ- Significance Evaluation
Profile Likelihood formulation
22
MSSM Higgs?µµ- Discovery Potential
23

• Conclusions
• ATLAS detector is well suited for Higgs SM or
  MSSM discovery, starting this year.
• Discoveries possible with a few fb-1
• Still, great effort must be put on understanding
  the detector with the first data