b-tagging , leptons and missing energy in atlas after first data
DESCRIPTION
B-tagging , leptons and missing energy in ATLAS after first data. Ivo van Vulpen (Nikhef). on behalf of the ATLAS collaboration. Studying top quarks pairs – the building blocks. Jets are discussed in a separate talk. 2/40. ATLAS data-sets. 2009: √s = 900 GeV. 2010: √s = 7 TeV. - PowerPoint PPT PresentationTRANSCRIPT
B-tagging, leptons and missing energy in ATLAS after first dataB-tagging, leptons and missing energy in ATLAS after first data
Ivo van Vulpen (Nikhef)
on behalf of the ATLAS collaboration
Studying top quarks pairs – the building blocks
Jets are discussed in a separate talk
2/40
ATLAS data-sets
2009: √s = 900 GeV
Lumi = 12 μb-1 (stable beams) 538,000 collision candidates
Main focus of this talk is onthis data-set
2010: √s = 7 TeV
3/40
Public ATLAS results:
https://twiki.cern.ch/twiki/bin/view/Atlas/AtlasResults
Integrated luminosityIntegrated luminosity
You are here
UpcomingATLAS results
B-tagging
B-tagging in top analyses
- Reduce W+jets & QCD bckg.- Reduce jet-combinatorics
Complications: - JESb-jet difficult (vital for Mtop)
- multi-jet environment
B-tagging in top analysesB-tagging in top analyses
6/40
Inner detector
Pixel detector
Transition Radiation Tracker
SemiConductor Tracker
ATLAS inner detector
3 systems in 2 T solenoid|η|< 2.5
7/40
barrel: 4 layers (x2)
barrel: 3 layers
barrel: ≥ 30 straws
Tracking performance
SCT - number of hitsSCT - number of hits
Events with: - Minimum bias trigger, primary vertex ≥ 3 tracks - Tracks with:PT > 0.5 GeV, Npixel ≥ 1, NSCT ≥ 6, |d0|<1.5 mm, |z0 sinθ|<1.5 mm
- Simulation: reweighted beamspot and corrections for disabled modules in data
Minimum bias events at √s=900 GeV
Track reconstruction efficiencyTrack reconstruction efficiency
Monte Carlo
8/40
Excellent agreement:simulation reproduces track properties
Alignment tracking detectors
residual
residual
Pixel –x residualsPixel –x residuals
9/40
Radius inATLAS [mm]
122.588.550.5
371299
443
514
MC: 38 μmMC: 38 μm
SCT –x residualsSCT –x residuals
Data: 43 μmData: 43 μm
MC: 23 μmMC: 23 μm
Data: 28 μmData: 28 μm
Tracks with: PT > 2 GeV, Nsilicon ≥ 6, |d0|<10 mm
barrel
barrel
Primary vertices Pile-up event
Vertex resolution ~75 μm
Luminous region: σX=45 μm, σY=70 μm
Vertexing
10/40
~2.5 cm
Secondary vertices
Ks invariant massKs invariant mass
11/40
Λ invariant massΛ invariant mass
Impact parametersignificance
d0
primary vertex
secondary vertex
jet
12/40
Flight length significance: L/σLFlight length significance: L/σL
B-tagging algorithms:
- jet tracks (combined) incompatibility with originating from primary vertex - soft lepton tagging - secondary vertices & their properties
d0 significance: d0/σd0d0 significance: d0/σd0
‘Typical’ b-jet candidate
primary vertex
secondary vertex
√s=7 TeV
B-jet details: - PT = 19 GeV (EM scale)
- 4 b-tag quality tracks Prob(PV compatibility) 9∙10-6
- Secondary vertex: o 50 sigma in 3d away from PV o Mass sec. vertex = 3.9 GeV
13/40
electrons
Electrons in top analyses
isolated & high-PT
- dominant trigger stream- reduce QCD multi-jet bckg. - handle on leptonic W-boson
Complications: - extrapolation from Ze+e- to top multi-jet environment - fake and non-prompt electrons o e/jet ~ 10-5 at PT=40 GeV
o identify b/c-decays: isolation
Electrons in top analysesElectrons in top analyses
15/40
Electromagnetic calorimeter& particle identification
Transition Radiation Tracker
Electromagnetic calorimeter
e/π separation
Liquid Argon / Lead|η|< 3.220-30 X0
16/40
L1 rates: normalised to L=1027 cm-2s-1
L1 efficiency w.r.t. offline (~3 GeV)L1 efficiency w.r.t. offline (~3 GeV)
EM calorimeter trigger- level 1 -
17/40
Signature Rate (1x1031)
L1: e10 5 kHz
EF: e10 21 Hz
Nominal settings / rates
ATLAS detector paper
EM & e/γ trigger- level 2 and event filter -
Shower shapes at L2Shower shapes at L2L2 rate versus ET L2 rate versus ET
Note: - Already at Level 2 fairly good agreement on e/γ specifics - Highest level (Event Filter) trigger offline - more refined
900 GeV
7 TeV
18/40
879 electron candidates: electrons from photon conversions and fake electrons (hadrons)
Electrons in 900 GeV data
EtaEtaTransverse energyTransverse energy
γ/π0 – separation bulk energy leakage-estimate 19/40
Photon conversions
Pixel layers
SCT layer 1beampipe
Conversion radiusConversion radius
Eta of candidateEta of candidate
20/40
- Clean electron sample- Good material description in MC
Electron classifications and data
Loose
Medium
Tight
ECAL: shower shapes Track: loose track
Loose 94.3 % ~ 1Medium: 90.0 % ~ 7Tight: 71.5 % ~ 140
Efficiency(Z e+e-)
Expectations, √s=10 TeV (MC)
Jet rejection (x 103)ECAL: strip layer info
Track: tighter quality + matching
Track: B-layer hit + tight matchTRT: high-threshold hits
21/40
Calorimeter information
Fraction of energy in strip layer
electrons
hadrons
Fraction of energy in strip layer
Loose
Medium
Tight
ECAL: shower shapes Track: loose track
ECAL: strip layer infoTrack: tighter quality + matching
Track: B-layer hit + tight matchTRT: high-threshold hits
22/40
Strip layer infoStrip layer info
Strip layer infoStrip layer info
Tight electrons
Loose
Medium
Tight
ECAL: shower shapes Track: loose track
ECAL: strip layer infoTrack: tighter quality + matching
Track: B-layer hit + tight matchTRT: high-threshold hits
23/40
Number of pixel hitsNumber of pixel hits
Fraction TRT high-threshold hitsFraction TRT high-threshold hits
Transition Radiation Tracker (e/π separation)
Note: also tested using high-energy muons in the cosmic runs24/40
TRT: high-threshold probabilityTRT: high-threshold probability
Ze+e- candidate at √s=7 TeV
ET (e-) = 45 GeV ET (e+) = 40 GeV
Invariant mass = 89 GeV
25/40
Muons
Muons in top analyses
isolated & high-PT
- dominant trigger stream- reduce QCD multi-jet bckg. - handle on leptonic W-boson
Complications: - extrapolation from Zμ+μ- to top multi-jet environment - fake and non-prompt muons mainly from b-decays: identify using isolation
Muons in top analysesMuons in top analyses
27/40
Muon Spectrometer
Muon trigger and momentum resolution < 10% up to 1 TeVstandalone or combined with inner detector information
Barrel/endcap toroid B=0.5/1.0 T
Coverage: |η| < 2.7
28/40
Cosmic runs
Millions and millions of cosmics Transverse momentum resolution
Transverse momentum resolutionTransverse momentum resolution
Resolution close to nominalOngoing effort on alignmentand calibration
29/40
Muon trigger
L2 eff. (combined) w.r.t. offlineL2 eff. (combined) w.r.t. offlineL2 eff. (standalone) w.r.t. offlineL2 eff. (standalone) w.r.t. offline
PT track: EF versus offline PT track: EF versus offline Level 2 plots from events with: - L1 muon trigger - offline muon (PT > 2 GeV, P > 4 GeV)
- Nsilicon ≥ 6, match L1 ROI in dR<0.5
|η|< 1.05
30/40
Toroids are off
transverse plane
31/40
Muons in √s=900 GeV data
phiphi
etaeta
Monte Carlo expectation:Mainly π/K decays (~25% b/c decays)
MomentumMomentum
32/40
Muons in √s=7 TeV dataJ/Ψ peak in di-muon mass (opposite charge with p>3 GeV)
Mean = 3.06 ± 0.02 GeVResolution = 0.08 ± 0.02 GeV
Mean = 3.06 ± 0.02 GeVResolution = 0.08 ± 0.02 GeV
33/40
Muons in W- and Z-boson candidates in 7 TeV data
Wμν candidate Zμμ candidate
M(μ+μ-) = 87 GeV
34/40
Missing transverse energy
ET-miss in top analyses
- reduce QCD multi-jet bckg. ET>20 GeV: ε(tt) ~ 90%
QCD rejec. ~ 10 - handle on leptonic W-boson- tails important for new physics
Complications:- multi-jet topology (2 b-jets) - high-pT muons
ET-miss in top analysesET-miss in top analyses
36/40
Calorimeter responseEcell: EM EndcapEcell: EM Endcap
Ecell: Tile CalorimeterEcell: Tile Calorimeter
ET-miss= Calorimeter (+cryo+muons)
Electromagnetic calorimeter
Hadronic calorimeter37/40
Missing transverse energy and resolution
(Ex,Ey)-miss resolution (Ex,Ey)-miss resolution
Minimum bias events at 900 GeV and 2.36 TeV
RMS ~ 1.8 GeV38/40
Ex-missEx-miss
Ey-missEy-miss
Missing transverse energy: 7 TeV data
Note: - tails under control - cells from topo-clusters (EM-scale)
Collision data at √s = 7 TeV
Missing transverse energyMissing transverse energy
Ex-missEx-miss
Ey-missEy-miss
39/40
Summary:
- ATLAS collecting data at 7 TeV
- Good performance on reconstructing building-blocks for a top analysis.
shown here: leptons, ET-miss, b-tag
- Much more data & results available (at PLHC conference … next week)
Reconstruct many top quarks in 2010
Backup slides
Alignment tracking detectors
Pixel –x residualsPixel –x residuals
MC: 38 μmMC: 38 μm
SCT –x residualsSCT –x residuals
Data: 43 μmData: 43 μm
MC: 23 μmMC: 23 μm
Data: 28 μmData: 28 μm
Tracks with: PT > 2 GeV, Nsilicon ≥ 6, |d0|<10 mm
Barrel
Endcap
Pixel –x residualsPixel –x residuals SCT –x residualsSCT –x residuals
MC: 20 μmMC: 20 μm
Data: 23 μmData: 23 μm
MC: 40 μmMC: 40 μm
Data: 45 μmData: 45 μm
B-tag distributions in √s=7 TeV data
Flight length significance: L/σLFlight length significance: L/σL Jet probabillityJet probabillity