Trigger studies of SM 130 GeV Higgs in the lepton+hadron channel First pass at estimating trigger efficiency Higgs mass reconstruction Catalin Ciobanu,

Trigger studies of SM 130 GeV Higgs

in the lepton+hadron channel

• First pass at estimating trigger efficiency

• Higgs mass reconstruction

Catalin Ciobanu, Dovid Skversky, Kevin Pitts, Tony Liss

Univ. of Illinois

Part I: Trigger Studies

Trigger tables – from Francesca’s talk

0

1

23456

LOW-LUM HIGH-LUM

Notation: 1010000 means the event fired MU20 and J200 triggers

MU20 (20)

2MU6

EM25I (30)

2EM15I (20)

J200 (290)

3J90 (130)

4J65 (90)

J60+xE60 (100+100)

TAU25+xE30 (60+60)

MU10+EM15I

This will hurt efficiency

• Too few ATLFASTB -jets!

103

104

105

104

103

105

ATLFASTB -jets + lepton

Wait! It’s even worse:73% of events with an ATLFASTB -jet have no leptons.Of the 536 evts that do have both a lepton and a -jet, about 300 are below the lepton trigger.

1986/50k= 4% - very poor

Lepton and -jet ET-integral spectra

• Investigated using one additional trigger:

• Lepton + -jet. 3 variables:

– Lepton ET

-ID flag (ATLFASTB! Remember small efficiency) -jet ET

Default Efficiency• Using Erik’s parametrization for the jets

– Not for the jets

– For the triggers use ATLFASTB info, not the ‘truth’

– Given N (N7) choose the most efficient N triggers:

Low luminosity regime:

N= 1: 0000010 0.13598000 (jet+MET)

N= 2: 1000010 0.24079999 (+muon)

N= 3: 1100010 0.32436001 (+electron)

N= 4: 1100011 0.33831999 (+taujet+MET)

N= 5: 1110011 0.34233999 (+1jet)

N= 6: 1111011 0.34255999 (+3jet)

N= 7: 1111111 0.34268001 (+4jet)

High luminosity regime:

N = 1 1000000 0.12954000 (muon)

N = 2 1100000 0.21585999 (+electron)

N = 3 1100010 0.24330001 (+jet+MET)

N = 4 1100011 0.24548000 (+taujet+MET)

N = 5 1110011 0.24738000 (+1jet)

N = 6 1111011 0.24745999 (+3jet)

N = 7 1111111 0.24745999 (+4jet)

Adding +lepton trigger

LOLUM

Max. increase: 0.5%=(34.4-34.3)/34.3

Max. increase: 2.1% = (25.3 - 24.7)/24.7

Plot: 100*[eff (11111111)-eff (11111110)] / eff (11111110)

“Perfect” -jet identification

Max. increase: 6.1% Max. increase: 33.5%= (38.1-28.6)/28.6

Plot: 100*[eff (11111111)-eff (11111110)] / eff (11111110)

Imperfect -jet identification

Plot: 100*[eff (11111111)-eff (11111110)] / eff (11111110)

as -ID efficiency is varied from 10% to 100%.

ET(lep)>5GeV

ET(lep)>5GeV

ET(-jet)>5GeV

perfect ID

ATLFB

Conclusions

• Rough pass at trigger efficiencies for H in the lepton+hadron channel

• Backgrounds?– Top pair production – Bottom pair production– W/Z+jets production– Z

• Some (all?) exist; some may have to be generated• Study improvements to S/B

– Cut on between lepton and -jet

Part II: Higgs Reconstruction

• For this study, to improve statistics, we use the ATLFAST(not-*ATLFASTB*) -jets, i.e. the ‘truth’ -ID:

abs(KFJET[i])=15

Collinear Approximation

Assume the neutrinos (1+2) that go with the lepton move along the same transverse direction as the lepton itself (no need to assume same as well)

Similar for the -jet and the other neutrino 3.

Conserve energy and momentum– Missing energy = transverse comp. of 1+2+ 3

– Solve for the transverse comp. of (1+2) and 3.

– Solve for the longitudinal comp.:

• If complex solutions, then also assume the same (fully collinear approximation). This means giving up the Pz mom. conservation in 21% of the events

e,

1+2

-jet3

Transverse plane

Leptonic and hadronic reconstruction

Needs debugging

Higgs Mass Reconstruction

4.3% events have MH>300

More Higgs plots

Collinear approximation gets better as momentumincreases:

Conclusions II

• We got a mass peak!– There may be room for improvement?

• Fit in the 110-150 range: 131.99.8 GeV

• Clean up the bugs

• New ideas for complex Pz solution cases?