Gg→H for different mcs: uncertainties due to jet veto G. Davatz, eth zurich

gg→H for different MCs: uncertainties due to jet veto

• G. Davatz, ETH Zurich

• gg→H→WW→ll

• Higgs discovery channel between 2MW and 2MZ
• Dominant background: nonresonant WW, ttbar and Wtb

• jet veto crucial to reduce top-background

• → get uncertainty of jet veto for different Monte Carlos

MCs compared : PYTHIA 6.319, HERWIG 6.505 + ME correction*, MCatNLO 2.31 and CASCADE 2.009

• NO underlying events

• M(Higgs) = 165 GeV, M(top) =175 GeV

• CASCADE 2.009 (CCFM hadron level MC) with PYTHIA final state parton shower

• HERWIG: gg→ H : no hard ME Corrections,

• here:preliminary version with ME corrections used (exact )

• PYTHIA, MCatNLO : with ME Corrections (PYTHIA: m(top)→∞, MCatNLO exact)

for this study:

• for this study:

• Cone algorithm

• pT jet>20 GeV, | jet<4,5, R=0.5,

• pT seed>1 GeV

• pT Higgs balanced by one or more jets

• → similar but not identical pt spectrum

• Apply jet veto of 30 GeV

• → get the efficiency

• gg→H→WW→llselection

• (GD et al jhep05(2004)009 ) shows:

• small pT Higgs region most important

In low pT region, HERWIG, MCatNLO and PYTHIA are now very similar

• In low pT region, HERWIG, MCatNLO and PYTHIA are now very similar

• The total efficiencies for HERWIG, MCatNLO, PYTHIA and CASCADE vary around 10%

• In the region of interest for the gg→H→WW→llsignal selection (up to pTH 80 GeV),

• the difference for HERWIG, MCatNLO and PYTHIA are smaller than 2% !

• If we smear the ET of the jet to get realistic CMS efficiency for jet veto with

• jet resolution: ET / ET = 118% / sqrt(ET) + 7%, the difference in the

• efficiencies between smeared and not smeared case is smaller than 1%

Including higher order corrections (by reweighting) leads to about same efficiency uncertainty as without reweighting

• Including higher order corrections (by reweighting) leads to about same efficiency uncertainty as without reweighting

• Including UE, the difference in the efficiency between PYTHIA with and without UE

• is smaller than 1% (tested CDF tune A and ATLAS tune)

• The uncertainty of the efficiency for different scales in MCatNLO

• is lower than 1%

• Results with CASCADE have to be treated carefully

• Ongoing work

• Staying with ATLAS tune, change pT cut for UE within 3of fit error as

• proposed by Paolo Bartalini

• …

backup

Efficiency after smearing (pythia, mcatnlo, herwig without ME correction)

Including Underlying events in Pythia

Including HO corrections

Analyse of Cascade efficiency shape

Efficiency for the jet veto with Herwig without ME correction

Herwig with ME correction

if hard ME corrections included *→ more jets with high pt →total  the same → less jets with low pt → overall efficiency ≈ 0.55 (10% smaller than for HERWIG 6.505)

Different pt Higgs regions with Herwig without ME correction