Hgg analysis at ATLAS

1
H?gg analysis at ATLAS
M. Escalier, B. Laforge
LPNHE Paris CNRS/IN2P3-Universités Paris 67

• Status of TDR analysis
• Signal background generation
• Kinematic variables
• Likelihood analysis
• Conclusion and outlooks

Special thanks to Arthur for let me use his
laptop for some update
Thanks L. Fayard, G. Unal (discussion, feedback)
2
Status of TDR

• Leading Order CTEQ 2L
• no K factor
• Pythia 5.700
• ISR, FSR
• Atlfast 2.0

from full reconstruction
3
Signal/?B
S/vB6.5, signal background computed at LO
4
Signal H?gg
VBF

• direct production

gg fusion

• associated production

WH,ZH,ttH
5
Signal
pdfCTEQ6L1 (comment of M. Spira) at Les Houches
Using correction of HDecay for decay H?gg
6
Background

• Irreductible background

NLO calculation using Diphox 1.2, pdfCTEQ6M

• reductible background (LO Pythia), pdfCTEQ6L1

g jet, 1 jet misidentified
jet jet, 2 jets misidentified
7
Higgs mass resolution
With Pythia/ATLFAST
s1.34 GeV
Next slides use of s of full simu TDR for window
1.4 s
Mgg (GeV)
8
Diphox definitions
Thanks J. Ph. Guillet for help with Diphox
9
Analysis Requirements

• CTEQ 6L1 pdf (LO fit), CTEQ 6M pdf for diphox
  (NLO)
• Pythia 6.210, Atlfast 2.53, Diphox 1.2
• Cuts
• photon candidates pT1gt40 GeV, pT2gt25 GeV
• hlt2.4
• transition crack of h1.45, Dh0.15 considered
• Photon efficiency 80 (Cf. M. Wielers)
• Jet rejection factor 2900 (Cf. M. Wielers)
• photon Isolation cone R0.4 ETlt15 GeV

In discussion (F. Derue 1200-1600) cf. Athens
4th Atlas Worshop
10
Branching ratio depends on b mass and on QCD
corrections
8.5
56
Branching ratio Pythia running aQED for pythia
(I mean by default) HDecay QCD corrections more
relevant No corrections aQED (better because
real photons q20)

• gg fusion compatible with papers
• R.Harlander and W.Kilgore   Phys Rev Lett
  88,201801,2002
• C.Anastasiou and K.Melnikov Nucl Phys
  B646,220,2002
• K factor 1.7-2.5 depending on higgs mass
• Ravindran, Smith, Van Neerven
• KNLO1.76
• VBFcompatible with talk of C. Oleari

Total production cross-section What is the NLO
Prediction ? HiGlu package (M. Spira et. al) gg
fusion K1.8 (using CTEQ6) VBF K1.04 (using
CTEQ6)
11
Problem of resummation C. Balazs has done
resummation, use/compare his soft. later
Ptgg

• Ptgglt30 GeV
• Ptgggt30 GeV ?likelihood analysis

12
Ptgg distribution
fsignal
fbackground
13
Cos q
fsignal
fbackground
14
Likelihood ratio R
Search for optimal cut
15
Results
8.8
6
8.25
16
Conclusion

• Background
• Irreductible background
• Use of NLO irreductible background
• Need resummation
• (but box is LO in Diphox)
• ?use Dixon calculations (L. Fayard, G. Unal)
• Does NNLO computation is possible ?
• reductible background
• Know better the rejection factor (g jet)
• Then look for NLO computation of g jet, jet jet

17

• Signal
• gg fusion and VBF comp. at NLO (but Kf(Pt)?need
  MC at NLO)
• gg fusion Ravindran, Smith, Van Neerven
• KNLO is not smal (1.76), KNNLO1.16 (use of KNNLO
  in the future)
• NNLO of VBF ?
• likelihood analysisgain not so much important
• use neural network ?
• Understand background of gg jet jet with 2
  tagged jets
• ?jet veto in central region