Presentazione di PowerPoint

1
Study of the H??? decay channel with the ATLAS
detector
Carminati Leonardo, INFN Milano on behalf of the
ATLAS collaboration

• The ATLAS detector and the electromagnetic
  calorimeter
• The H??? decay channel and inclusive analysis in
  full
• detector simulation.
• Background evaluation and discovery potential
  (LO)
• Ongoing activities NLO signal and background
  studies,
• Hjet and VBF exclusive analysis

2
The ATLAS detector
Length 40m Radius 10m Weight 7000 t El.
Channels 108 Cables 3000 km
Muon Detectors Muon momentum mesurement and
trigger
Inner Detector Tracking and impact parameter
determination
Electromagnetic calorimeters electron/photon
identification, energy and position reconstruction
Hadron Calorimeters jet and ET miss reconstruction
3
The ATLAS em calorimeter

• Lead Liquid Argon sampling calorimeter
• ? 200000 read out channel
• Material in front of the calorimeter Inner
  detector, cryostat, coils ? 2X0 depending
• on eta. A thin presampler to recover energy
• lost in the upstream material

ACCORDION GEOMETRY
Sampling Granularity ?? x ?? Depth
Presampler 0.025 X 0.1
Strips 0.003 X 0.1 5X0
Middle 0.025 X 0.025 16X0
Back 0.05 X 0.025 2 to 10 X0
Back
Middle
Strips
Presampler in front
4
INTRODUCTION the ATLAS em calorimeter

• Eta coverage
• 0 lt ? lt 1.475 barrel
• 1.375 lt ? lt 3.2 endcap
• Cracks in ? direction
• ? ? 0 two half barrel
• 1.4 lt ? lt 1.55 barrel / endcap
• ? ? 2.5 inner - outer wheel
• The accordion geometry ensures full
• azimuthal coverage without cracks

• Calorimeter performance studied in details with
  both simulations and testbeams data
• Energy resolution 10/?E ? 0.7
• Position resolution ? ? 60 mrad/?E -- ? ? 4
  mrad/?E -- ZV ? 16 mm

5
INTRODUCTION the ATLAS em calorimeter

• Electromagnetic calorimeter assembly almost
  finished ongoing electrical and functionnality
  tests, cooling
• Barrel calorimeter finished
• 1 endcap ready
• Calorimeter in the pit in september !!

Insertion of the solenoid in the electromagnetic
calorimeter cryostat
6
SM Higgs production processes

• Standard Model Higgs production processes at the
  Large Hadron Collider
• Dominant production process is the gluon-gluon
  fusion (LO ?20 pb for MH120 GeV)
• Vector Boson Fusion contribution becomes
  important for higher MH (but distinctive
  signature!!) (LO ?4 pb for MH120 GeV)
• Small contribution from WH,ZH and tt(bar)H (LO
  ?2.4 pb for MH120 GeV)

7
SM Higgs decay branching ratios

• H?bb(bar) is dominant decay mode for MHlt2MW only
  events from associated production can be observed
• Nice results have been recently showed for the
  ??- and WW channel from VBF

• H??? is a rare decay mode with BR ? 10-3 (2.186
  10-3 for MH120 GeV)
• The signal should be visible as a small peak
  above the ?? continuum background
• Severe requirements on particle identification
  capabilities of the detector

8
H??? decay channel full detector simulation

• ATLAS full detector simulation for 100 GeV lt mH lt
  140 GeV in most realistic experimental
  conditions
• Expected pileup events added for both low (1033
  cm-2s-1) and high (1034 cm-2s-1) luminosity.
• Electronic noise measured in the testbeams added
  for each em calorimeter cell
• Official results obtained with Pythia 5.7 and
  CTEQ2L. Detector simulations based on GEANT 3
  (Physics TDR)
• New full simulations (2002) with Pythia 6.2 and
  CTEQ5L using updated detector geometry (material
  in front of the calorimeters increased), new
  software framework and 21033 cm-2s-1
• Major effort spent in understanding the new
  detector performance and validating the new
  software framework

9
Photon calibration

• Photons energy and position reconstruction is a
  critical item calibration procedure has been
  carefully optimized
• Sliding window algorithm to find a cluster
• different possible cluster sizes can be
  selected
• 3x3, 3x5 (middle layer granularity)
• Refined energy after corrections
• Corrections for upstream material
• and leakage
• out of cone (shower lateral containment)
• accordion modulation corrections
• Refined position using the following corrections
• S shape for for strips and middle
• Phi offset (middle only)

Example 3x3 cluster size
Shower barycenter
10
Converted photons

• ? 0.30 (depending on ?) probability of photons
  conversion
• Inner Detector will identify and reconstruct
  photon conversions occurring within Rlt80, zlt280
  cm with an 80 efficiency
• Larger cluster size (3x7) for converted (3x5 for
  unconverted)
• Special calibration for converted photons
  depending on the conversion radius

11
Analysis cut and reconstruction

• Events with 1 ? in the electromagnetic
  calorimeter cracks excluded (bad resolution)
• ?lt0.05, 1.4lt?lt1.55 and ?gt2.45
• Transverse momentum cuts (background rejection)
• pT(1)gt40 GeV, pT(2)gt25 GeV
• 0.63 constant term to the photons reconstructed
  energies added (mechanics, calibration, HV
  variationsnot included in the detector
  simulations)
• The direction of both photons is corrected for
  the primary vertex position using a weighted
  least squares fit and a GEANT 3 parametrization
  of the shower depth

12
Direction reconstruction

• Low luminosity
• Use calorimeter ? measures from strips
• and middle Zv measure from Inner
• Detector (?z40 ?m)
• High luminosity
• No use of ID. Photons direction obtained
• with calorimeter information only crucial
• role for fine ? - segmented strips layer
• Conversion vertex also used in the direction
  reconstruction
• For a conversion occurring in the Rlt40 cm and
  ?Z?lt220 cm (early conversions) the vertex is
  reconstructed with the Inner Detector
• Primary vertex is determined and the photons 4 -
  momenta recomputed

13
Mass peak reconstruction

• ?? invariant mass has been reconstructed for each
  MH
• Invariant mass resolutions have been evaluated
• The acceptances have been computed taking into
  account geometrical acceptance, photon
  identification efficiency (80), mass bin (?1.4
  sigma)

Mass resolution Acceptance efficiency Nr. of events
100 GeV 1.31 0.23 1045
120 GeV 1.43 0.26 1283
130 GeV 1.55 0.28 1186
140 GeV 1.66 0.28 950

• (All numbers in the table refer to 100 fb-1 of
• integrated luminosity collected at 1034 cm-2
  s-1)

14
Background

• Irreducible background consists of genuine
  photons pairs continuum. For the standard
  analysis (a la TDR) Born,Box diagrams (BornBox
  ?0.7 pb/GeV for MH120 GeV ) and an estimation of
  the quark bremsstrahlung processes have been
  considered (? 50 of the total Bornbox)
• The reducible background comes from jet-jet and
  gamma-jet
• events in which one or both jets are
  misidentified as photons
• (Reducible / irreducible cross section
  ?2x106(jj) 8x102(?j))
• Excellent jet rejection factor (?103) for 80 ?
  efficiency

100 GeV 120 GeV 130 GeV 140 GeV
Expected background in the mass bin (100 fb-1) 41400 35000 29000 20600
15
Photons/jet separation

• ?/jet separation criteria mainly based on
  calorimeters information.
• Leakage of electromagnetic showers in the
  hadronic calorimeter
• Second sampling of the electromagnetic
  calorimeter information different transverse
  development of electromagnetic and hadronic
  showers.
• Shower shapes in ? and ? E37/E77, E33/E37
• Shower width in ? direction
• First sampling of the electromagnetic calorimeter
  information only jets with a little hadronic
  activity survive. ?/?0 separation
• exploiting the fine segmentation of the
  strips
• look for substructures in strips
• Shower width in ?
• fraction of en. in the core ( (E(?3)-E(?1))/E(?1)
  )

16
Photons/jet separation
For an 80 photons identification efficiency
(flat in ? and pT) a jet rejection factor of ?
3000 can be achieved for ET ? 50 GeV (Higgs
photons ET mean)
MH120 GeV Higgs on top of the irreducible
background (TDR)

• In the TDR analysis the reducible
• background after identification cuts has
• been estimated as fraction of the total
• irreducible background ? 20 for ?/j
• and ? 15 for jj

17
RESULTS discovery potential

• 100 fb-1 collected in high luminosity conditions
  (1034 cm-2 s-1)
• 30 fb-1 collected in low luminosity conditions
  ((2)1033 cm-2s-1)
• Last official results reported in physics TDR
  (PYTHIA 5.7 with CTEQ2L)
• Latest simulations (PYTHIA 6.2 and CTEQ5L) and
  analysis confirms the published results with a
  slight degradation (lt 10)

18
NLO calculation
Recent theoretical computations on the background
allow a complete study of the channel at NLO

• Signal cross section
• Gluon-gluon fusion K factor ? 2.0 , vector boson
  fusion ? 1.1, associated production ? 1.2
• Irreducible background cross section
• Ongoing activities in background understanding
  using NLO tools (DIPHOX and ResBos)

K factor ?2 for signal and ?1.7 for background
improvement of the significance is expected
although with large uncertainties. Real
background will be measured directly on the data!!
19
Ongoing activities

• Vector Boson Fusion(no full simulation yet)
• Two high PT jets (gt 20 GeV) with large ??
  separation (??ijgt3.5)
• Central jet veto no additional jets with PTgt20
  GeV observed in the ?lt3.2 region
• PT of both photons gt 25 GeV and ? between the two
  tagged jets
• Current indications S/?B ? 2.5 for 30 fb-1
• Hjet analysis (no full simulation yet)
• Jet with ETgt40 GeV and cut on M(?-?-jet)gt400 GeV
• Current indications S/?B ? 4.0 for 30 fb-1
• Combined analysis not finalized yet ? 20
  significance improvement is expected combining
  various channels

20
Conclusion and perspectives

• Full simulation signal analysis has been
  performed from MH100 to 140 GeV in the most
  realistic conditions. Detector performance
  studied in details.
• In a conservative LO approach the discovery
  potential above 5 sigma for 100 fb-1 of
  integrated luminosity in 100 GeV lt MH 140 GeV
  published in physics TDR is confirmed.
• Ongoing activities on NLO signal and background
  considerable increase of the discovery potential
  is expected
• Promising studies on VBF and Hjet in fast
  simulation will be studied in full detector
  simulation. The discovery potential should be
  improved properly combining various analysis.