ATLAS discovery Potential

1
ATLAS discovery Potential
2
Higgs to 4 leptons
3
H ? 4m with Rome WS Samples
Process Evt. Recon sXBR(fb)(Generator) EffFILT() Events for 30 fb-1
H ? 4m (m130 GeV) 2.4K () (9.4K/4) 0.865 (3.46/4) () (gg .VBF) No filter 21
ZZ ? 4m 8K (32K/4) 39.7(158.8/4)1.3 (for miss. gg-gtZZ) 21 1200
Zbb with Z?2m 33K (65K/2) 24.6 103(49 103/2) (AcerMC PYTHIA) 1.1 738000 -gt after filter 8100
tt ? WbWb with W?m? 12K (48K/4) 177 103(708 103/4) (MC_at_NLO) 0.83 5310000 -gt after filer 44000
() Statistics and cross section for 4m obtained
by (4lepton values)/4 -Cross sections for signal
taken from ATL-COM-PHYS-2004-062
At least a factor 10 in MC statistics w.r.t. data
is needed (after filter) Zbb -gt 100k tt -gt
500k Full 4 lepton samples need another factor 4
4
H -gt ZZ() -gt 4 leptons review

• First step the 4 lepton Pre-Selection
• Select 4 high-Pt leptons with hlt2.5 and 2 with
  Ptgt20 GeV (trigger) and 2 with Ptgt7 GeV and with
  QTOT0
• Build 2 invariant masses (2 combinations) with
  opposite sign leptons
• One pair best match Z mass within resolution
  M12MZ15GeV
• Add a cut on the other pair (dependent on MH)
  M34gt20GeV for MH130 GeV
• After pre-selection for 30 fb-1 and MH130 GeV
  and ZZ-gt4m

Signal (4m only) tt -gt WWbb-gt 4mx Zbb -gt
4mx ZZ-gt 4m
ZZ irreducible Zbb and tt reducible 2
non-isolated leptons coming from heavy flavor
decays
5
H -gt ZZ() -gt 4 leptons reducible background

• Second step reduce the reducible
• Identify leptons coming from heavy flavor decays
• Leptons in b-jet ISOLATION -gt additional energy
  clusters and/or tracks close to them look in a
  cone with aperture DR v(Df2Dh2) 0.2
• Leptons are expected to have large impact
  parameter d0ctB0.5mm
• For 4m rejections 100 for Zbb and 1000 for tt
  achieved with e80

Signal H-gt4m tt -gt WWbb-gt 4lx Zbb -gt 4lx
6
H -gt ZZ() -gt 4 leptons sideband technique

• Third step look for a peak in the invariant mass
  distribution
• Count events in 2s(MH) and compare with BKG
  determined from side-bands
• Significances 5s achieved combining 4m4e2m2e
  masses 130-200 GeV and L30fb-1 With exception
  around 170-180 GeV
• Error coming from Side-bands technique may be
  relevant
• statistics and mass shape !
• Not taken into account up to now to be studied

Signal 130 GeV ZZ-gt 2e2m
4m only L30fb-1 MH130 GeV
Signal 4m only ZZ-gt 4m
7
H -gt ZZ() -gt 4 leptons sidebands technique
Signal 4m only ZZ-gt 4m

• Few comments on significance and side-bands
  technique
• Difficult region 170-190 GeV
• lower BR (WW threshold)
• increase in ZZ background (ZZ on shell
  production)
• May not be straightforward to distinguish the
  signal from ZZ there

4m only L30fb-1 MH130 GeV
8
Signal ZZ -gt 4l

• reduce the irreducible
• Add to M12 and M34 more discriminating variables
• Higgs has spin0 ZZ longitudinally polarized
  while ZZ bkg mainly transverse polarized angular
  correlations induced in final state fZZ and q
• Higgs expected to be produced with larger PT than
  ZZ bkg
• To fully exploit correlations combine
  discriminating variables with NN Technique
• Additional ZZ rejection 2 with efficiency 90
  achieved
• Significance improved 20

f
q
M12
M34
CAVEAT PT for signal and background very
sensitive the higher order QCD correction may be
subject to large uncertainty !
PT4m
9
H -gt ZZ() -gt 4 leptons

• reduce the irreducible
• Mass shape for the background distorted
• Better rejection of the ZZ on-shell background !
• Improvement on expected significance 20

MH130 GeV L30 fb-1 4m only
MH180 GeV L30 fb-1 4m only
10
DC1 performance vs TDR
TDR
DC1
11
H -gt ZZ() -gt 4 m
H-gt4m L30fb-1

• Expected discovery potential
• TDR discovery potential wrong 180-200 GeV
• Deterioration in detector performance with
  respect TDR NOT understood
• Can be recovered thank to additional
  discriminating variables and multivariate
  techniques

12
Single Muon reconstruction performance DC3
Signal Sample (mH130GeV)
13
H -gt ZZ() -gt 4 m Recovering efficiency

• Recovering muon efficiency
• In the final ATLAS layout reduced coverage at h0
  due to services and at h1.1-1.3 due to staging
  of part of the muon stations (EE)
• Typical Pm in this mass range 10-50 GeV
  resolution dominated by the Inner Detector
• Lepton reconstruction efficiency is the most
  important element for this channel
• Significance S/vB vs(M)vLem2
• Idea recover efficiency accepting muon measured
  in the ID but only tagged by m-spectrometer
  or/and TILECAL
• Work in progress

14
H -gt ZZ() -gt 4 leptons Other comments on
sidebands technique

• Comments on side-band technique
• It is assumed that B in the signal region from
  side-band fit with negligible error
• First exercise perform a fit of S and B assuming
  both shapes are known 30 error on background
  normalization !
• Another effect not taken into account in
  significance estimate the Higgs mass is a free
  parameter in the SM we dont know it we need to
  open many mass windows (one every 2 GeV) and the
  background can fluctuate in any of them
  look-where-else effect

H-gt 4 leptons 30fb-1 MH130 GeV
15
Conclusions

• Detector
• Momentum Resolution detector performance worse
  then in TDR for both ID (20) and Muon
  Spectrometer (30-40) but origin not completely
  understood
• Efficiency large differences between two
  packages situation in the EC still very MESSY
• Possibility to recover efficiency with algorithm
  that uses ID Calorimeters/Muon Spectrometer to
  tag muons
• Situation is even worse with electrons
• Physics
• Quite large samples needed to study Muon/electron
  isolation
• Still missing complete studies with Pileup
  Cavern Background
• Use of spin 0 information and large Pt to be
  clarified / completed
• Important to understand lepton isolation from the
  beginning control sample tt-gt l n b jj b