Estimating the isolated lepton rate in multi-jet events Manuel & Alexander Ivo, Stan, Martijn, Auke, Els et al.
Niphad Meeting idea: How many QCD-events pass the single isolated lepton selection? use ttbar-events as estimate understand topology to translate to QCD depending on: # jets, Pt, eta, b-content ? two groups: non-prompt and fake extra electron per jet: QCD-background in commisioning analysis > 100* 10 -5
Niphad Meeting Searching for isolated leptons in ttbar Muon: etcone20 < 6 GeV Pt > 20 GeV Staco onlyHighPt = true only combined medium Electron: etcone20 < 6 GeV Pt > 20 GeV no TRT, no trackmatch exclude 1.37 < eta < 1.52 Look for electrons in ttbar(muon) and muons in ttbar(electron) 5200 home-grown topview ntuple
Niphad Meeting Full simulation versus fast simulation Medium electron # Good lepton # Good Jet # extra leptons per jet * FullSim840601, ± 5 FastSim268607, ± 3 Muons# Good lepton # Good Jet # extra leptons per jet * FullSim634631, ± 4 FastSim23599, ± 0.8 There is clearly a difference! Where does it come from?
Niphad Meeting Breakdown: Muons in Electron-Events non-prompt: (points to MC muon) - ancestor is a b-quark 96% - ancestor is a light-quark 3 % - ancestor is a gluon 1 % fake: (does not point to a MC muon) note: fake rate in Staco is 7x smaller than MuID 3% 97% Definitions: - ‘Points to’ = dR < ‘Ancestor’ = Parent in MC truth list
Niphad Meeting Breakdown: Electrons in Muon-Events non-prompt: (points to MC electron) fake: (points to cluster and track creating particle) muon fake: (points back to muon) disregarded! - ancestor is a b-jet 94 % - ancestor is a light jet 6 % - ancestor is a b-jet 12 % - ancestor is a light jet 88 % 61 % 39 %
Niphad Meeting Dependencies: p T electrons: investigate originating jet muons: investigate originating parton p T of b-partons# b with muon (p T ) / # all b (p T ) decrease after 60 GeV due to absolute isolation cut high p T muons get cut away harder
Niphad Meeting Dependencies: η[1] η of extra muonsη of b-partons Note the peaks at |η| = 1.5. Not a physics origin as can be seen in the b-parton spectrum. Etcone is underestimated here.
Niphad Meeting Dependencies: η[2] etcone of non-promptsetcone of prompts in the “crack” region etcone is underestimated no consequence for prompts, “peaks” for non-prompts
Niphad Meeting Dependencies: Jet Multiplicity jet multiplicity in ttbar electron events muons mainly from b- parton: always 2 in ttbar should not see dependency on jet multiplicity. jet multiplicity# non-prompt / event 3(467 ± 39) * (454 ± 32) * (352 ± 32) * (247 ± 38) *10 -5 decrease by almost factor 2 isolation
Niphad Meeting Consistency check : fully hadronic ttbar P T spectrum, η spectrum, jet multiplicity similar. Rates are comparable. Note: electron per jet muon per parton originsemi-leptonicfully hadronic b jet169 ± 8169 ± 10 light jet67 ± 469 ± 4 overall107 ± 499 ± 4 extra electrons per jet *10 -5
Niphad Meeting Prediction of extra lepton rate in di-jet sample [1] different topology, different quark-content no b-quarks from the hard scatter fake rate per jet: 3.5 *10 -5 non-prompt per light jet: 6.3 * ,351,300 jets predicted: 328 ± 29 observed: 330 low statistics for muons since only light jets:
Niphad Meeting Prediction of extra lepton rate in di-jet sample [2] multiply the light jet spectrum with the probability to produce a high-P T electron predicted 1669 … fullSim result 1520 ! x = Predicted # elec Integral = 1669 Probability for jet to produce P T > 20 GeV elec P T spectrum jets
Niphad Meeting Conclusion and outlook we understand the origin of extra leptons in multi-jet events investigated dependencies on p T, η and jet-multiplicity able to predict number of extra leptons in di-jet (assuming η distribution of jets to be similar) plan to make a prediction based on p T and η use the QCD fastsim sample (in production) to predict the number of isolated high- p T leptons present results of ATLFast II in meeting, hope to contribute