Performance of Jets and missing ET in ATLAS

1. 10th International Conference on Calorimetry in High Energy Physics Pasadena – March 2002 Performance of Jets and missing ET in ATLAS Monika Wielers TRIUMF, Vancouver on behalf of the ATLAS Collaboration M. Wielers, TRIUMF

2. Contents • Physics Goals for Jet/ETmiss Studies in ATLAS • The ATLAS Calorimeter System • Simulation of Detector Response and Pile-Up • Jet Reconstruction in ATLAS • Physics Examples for Jet Reconstruction • ETmiss Reconstruction • Conclusions I won’t have time to discuss  physics M. Wielers, TRIUMF

3. Physics Goals for Jet/ETmiss Studies • Jet Physics • QCD Studies • Measure jet cross section • Reconstruction of resonances • QCD: W  jet jet, Z bb, t  b W • Exotics: W’jet jet • SUSY: A  • Measure jet multiplicity in SUSY decay • Jet veto in central region to reject background • Jet tagging in forward region • Missing transverse energy measurement • SUSY and other New Physics signatures, e.g. HZZll • Reconstruct inv. mass in decays with ’s, e.g. A, tlb M. Wielers, TRIUMF

4. The ATLAS Calorimeter System EM Calorimeter • LArg technology • Coverage: ||<3.2 • High granularity up to ||=2.5 Hadronic Calorimeter • Fe-Scintillating tiles in ||<1.5 • LArg EC (||<4.9) • HEC, Cu/LAr : 1.5< ||<3.2 • FCAL: Cu/Tungsten/LAr 3.2<||<4.9 M. Wielers, TRIUMF

5. Simulation of Detector Response • Geant 3 still “standard”simulation program • Geant 4 to come  see various talks M. Wielers, TRIUMF

6. ~23 minimum bias events per bunch crossing at L=1034 cm-2 s-1 LArg uses bipolar shaper for read-out “history” from around 27 bunch crossings Noise peaks around zero GeV Tile uses mono polar shapers “history” from 4 bunch crossings Response ~ Gaussian with FWHM of 50 ns Simulation of Pile-up Noise M. Wielers, TRIUMF

7. Simulation of Pile-up Noise • ~700 minimum bias events need to be added to signal events at high luminosity ( 30 BC’s) • For given cell type weigh energy according to shaper response for given time • Read minimum bias events via secondary stream • On the fly via direct access file which is kept in memory • Add an additional simulation step (together with pile-up in all other detectors) • used to keep correlations between detectors • Keep big number of min. bias events in memory • Currently done in Fortran, work ongoing to get it running in C++ M. Wielers, TRIUMF

8. Experimental factors Electronic (and pile-up) noise Different response to charged and neutral particles (0) Non-linearities granularity Lateral shower size Dead material and cracks Longitudinal leakage Magnetic field effects Physics related factors Initial and final state radiation Fragmentation Underlying event Depending on luminosity: minimum bias events Factors influencing Jet Reconstruction No unique strategy for jet reconstruction, depends on • Jet reconstruction algorithm, luminosity, physics process M. Wielers, TRIUMF

9. Energy Resolution Calorimeter resolution at ||=0.3 /E = 52%/E  3.0 GeV/E  1.7% for R=0.7 at L=1033cm-2s-1 /E = 81%/E  3.9 GeV/E  1.7% for R=0.4 at L=1034cm-2s-1 • High luminosity pile-up, • + el noise •  el. noise ||=0.3 M. Wielers, TRIUMF

10. Physics Effects Particle level jet energy Parton energy Depends on physics process M. Wielers, TRIUMF

11. Algorithmic flow of jet reconstruction in software (in C++) (Sub) Detector Level cell signal reconstruction cell signal corrections Combined Detectors Jet Reconstruction tower finding cluster finding ProtoJet building Jet finding Jet reconstruction Physics Jet Jet classification Analysis • Get detector data • Calibrate to EM scale • Create towers and clusters (if input to jets) • Prepare input (cell, cluster, tower, jet…) for jet finder • Do jet finding • Reconstruct and calibrate jets • Do jet classification • Redo any step if desired! Detector Level Generic M. Wielers, TRIUMF

12. Forward Jet Tagging and Jet Veto • Important for background reduction for heavy Higgs production via Vector-Boson-Fusion ((VBF) ~ 1/5 (gg)) • H produced centrally • Accompanying jets in forward region • Small hadronic activity in central region • Veto low-pT jets  rejection of tt background 90 (80) % for ||<4 at low (high) luminosity (tt)~10% M. Wielers, TRIUMF

13. Simple Resonance Wjj at L=1034cm-2s-1 ET(W) ~ 150 GeV In resonance chain Hhhbbbb no electronic, no pile-up noise Use m(h) as mass constraint to reduce combinatorial background apply recalibration using m(h)=m(bb) requires b-jet tagging Resonance Reconstruction Bias in jet direction if jets overlap M. Wielers, TRIUMF

14. ETmiss Reconstruction • Important for new physics signatures (SUSY particle production and decay) • Invariant mass reconstruction for channels involving  • A, tlb, HZZll • ETmiss reconstructed from cell energies within ||<5 • Accurate calibration of calorimeters vital L=1033cm-2s-1 L=1034cm-2s-1 M. Wielers, TRIUMF

15. ETmiss Tails • ETmiss Reconstruction important for • H  ZZ  ll with m(H)~500-700GeV • Potential problem: background from Z+jet(s) events with badly measured jet pT(Z)>200GeV Requirement of rejection factor of 1000 for ETmiss > 200 GeV achieved! Jet reconstructed Assume jet is undetected M. Wielers, TRIUMF

16. Conclusions • Lot’s of interesting physics to be done using jets and ETmiss • Higgs, SUSY, the unknown,… • ATLAS calorimeter well designed for jet/ETmiss studies • We are ready to face the challenge Let’s see what 2007 brings us, when first collisions are expected ! M. Wielers, TRIUMF