1 / 36

Missing E T in ATLAS (Emphasis of Object-Based MET)

Missing E T in ATLAS (Emphasis of Object-Based MET). X.Chen, L.Flores, B.Mellado, S.Padhi and Sau Lan Wu University of Wisconsin-Madison West Coast Physics Forum 08/22/06. Outline. Introduction Status of Cell-Based MET package (default) in Athena MET Tails and Fake MET studies

sileas
Download Presentation

Missing E T in ATLAS (Emphasis of Object-Based MET)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Missing ET in ATLAS(Emphasis of Object-Based MET) X.Chen, L.Flores, B.Mellado, S.Padhi and Sau Lan Wu University of Wisconsin-Madison West Coast Physics Forum 08/22/06

  2. Outline • Introduction • Status of Cell-Based MET package (default) in Athena • MET Tails and Fake MET studies • Status of Object-Based MET package in Athena • Looking at resolution of various objects • Towards the development of MET significance for ATLAS Bruce Mellado, West Coast Physics Forum, 08/22/06

  3. Loop over cells (cell-based) High PT Particle escaping detection Low PT particles • ATLAS is developing a robust effort to understand MET issues produced in physics events Jets, ’s Muons Loop over objects (object-based) Electrons, Photons Bruce Mellado, West Coast Physics Forum, 08/22/06

  4. Cell-Based Approach (default in ATHENA) D. Cavalli, S.Resconi (INFN/Milan) Bruce Mellado, West Coast Physics Forum, 08/22/06

  5. ATHENA MissingET : EtMiss Reconstruction and Calibration All Calorimeter cells in ||<5 |Ecell | > 2(noise) H1-like weights from jets OR OR OR Estimated energy loss in cryostat between LAr and Tile from reco jets (cone07) w*sqrt(EM3*TILE1) from MOORE mu ||<2.7 pT from external Spectrometer BestMatch /good quality MET_Final = MET_Calib + MET_Cryo + MET_ Muon Calorimeter cells in Topoclusters 4/2/0 H1-like weights from jets from MuonBoy muons in || < 2.7 from reco jets (kT) Calorimeter cells in Topoclusters 4/2/0 LocalHadronCalibration (under test) Topoclusters 4/2/0 LocalHadronCalibration (under test) Refinement of Calibration for cells in different objects (e/g, t, m ,jets...) a la TDR (implementation in 12.0.2) Bruce Mellado, West Coast Physics Forum, 08/22/06

  6. ATHENA MissingET: Performance Good EtMiss measurement crucial for mass reco in Z   and H   • Performance depends on: • Noise suppression, Calibration, Dead material correction … • Physics channel topology: presence of leptons, Eleptons, Njets, Ejets, energy outside phys objects, activity out of coverage... • No dependence on Montecarlo observed (Pythia QCD jets very similar to Herwig jets) EtMiss resolution vs SumET CSC MC CSC MC For CSC linearity within 5%, resolution slightly worse respect to TDR parametrisation (s(EtMiss)=0.46*sqrt(SumET)) Bruce Mellado, West Coast Physics Forum, 08/22/06

  7. Linearity vs EtMiss_Truth Bruce Mellado, West Coast Physics Forum, 08/22/06

  8. MET Tails and Fake MET studies Bruce Mellado, West Coast Physics Forum, 08/22/06

  9. ATHENA MissingET: detailed study of events in EtMiss tails CSC QCD J6 (560<pT<1120GeV), 50Kevts (129pb-1) reconstructed with 11.041 EtMiss_rec-EtMiss_true tipically < 100GeV, look at events with DEtMiss>250GeV F. Paige, Spephane Willoq EtMiss_rec-EtMiss_true (GeV) Shower leakage both from punchtrough and cracks important. Veto on muons chamber activity seems effective Fake muons also important: can reduce with more severe muon quality cuts For DEtMiss>100GeV less muon activity, cracks do not seem dominant Bruce Mellado, West Coast Physics Forum, 08/22/06

  10. Jet leakage from Tile/ExtTile crack, shower in muon system Bruce Mellado, West Coast Physics Forum, 08/22/06

  11. Fake EtMiss study A bad EtMiss mesurement could fake a non-zero reconstructed EtMiss in events with no true EtMiss dedicated meetings each 1-2 months • Event production (S.Asai, A. Gupta) • QCD jets (background for SUSY) - and • Z(ll)+jets events (background for Hllnn) • Filters • Events with true EtMiss>threshold (study true EtMiss distribution) • Events with a jet pointing to a crack (study fake EtMiss) (R. Pradhu) • Definition of crack regions • Dead material hits information • Contribution to EtMiss from DM regions will be calculated in MissingET • Study Instrumental effects: problems in electronic channels, crates, HV sectors… give fake EtMiss. Can correct and recover EtMiss? (R. McPherson et al) • Define Event quality variables: out-of-time cells, hot cells, number of muon segments… (D. Tovey et al) 100GeV  1TeV N. Kanaya Bruce Mellado, West Coast Physics Forum, 08/22/06

  12. Object-Based Approach (In Athena since 12.1.0/12.0.2) X.Cueh, L.Flores, B.Mellado, S.Padhi, Sau Lan Wu (Wisconsin) Bruce Mellado, West Coast Physics Forum, 08/22/06

  13. Object-Based Calibration Missing ET Reconstruction Object-Based Missing ET Final State Classification High ET Objects Low ET Objects Un-clustered depositions   0  e/ jets Bruce Mellado, West Coast Physics Forum, 08/22/06

  14. Object Based Calibration • The Object-Based approach is very FLEXIBLE. It accommodates any calibration strategy Use local hadronic calibration or in-situ calibration or a combination of both In-situ calibration Bruce Mellado, West Coast Physics Forum, 08/22/06

  15. OBMET in 12.0.2 • The following features are now available in 12.0.2 Bruce Mellado, West Coast Physics Forum, 08/22/06

  16. OBMET solved long standing shifts and degraded resolution in H, so important for low mass Higgs searches • Higgs mass reconstruction after the application of all analysis cuts with Hll ATLFAST Rome MC Samples =10.5 GeV FULLSIM RMS 19.8 GeV M=131.5 GeV =9.8 GeV Assumed in VBF SC Note RMS 19.6 GeV M=129.1 GeV M M Bruce Mellado, West Coast Physics Forum, 08/22/06

  17. Missing ET resolution after the VBF analysis cuts • Table with Gaussian mean, width and RMS before and after the implementation of low ET objects (in GeV) Bruce Mellado, West Coast Physics Forum, 08/22/06

  18. Resolution reported in TDR with A Bruce Mellado, West Coast Physics Forum, 08/22/06

  19. Resolution reported in TDR with A is 0.46ET Bruce Mellado, West Coast Physics Forum, 08/22/06

  20. Achieve ~1% linearity with W’ Bruce Mellado, West Coast Physics Forum, 08/22/06

  21. Achieve ~2% linearity with SUSY MET in SU2 is under study Bruce Mellado, West Coast Physics Forum, 08/22/06

  22. SUSY events with high PT leptons (e,) (OBMET) PT>20 GeV PTe>20 GeV Bruce Mellado, West Coast Physics Forum, 08/22/06

  23. Towards the Development of MET Significance in ATLAS Bruce Mellado, West Coast Physics Forum, 08/22/06

  24. There are a number of cases when the simple expression, EXYmissSumEt, breaks down • We cannot use Missing ET significance as • The Object-Based method offers the advantage of calculating the expected missing ET resolution based on the resolution of each of the objects. This also solves the issue of the event topology • Then Calculate missing ET significance (D0 experience) Bruce Mellado, West Coast Physics Forum, 08/22/06

  25. Towards MET Significance • For definition of MET significance see Ariel’s talk • It is crucial to understand the resolution of different objects as a function of || and Energy • ATLAS has cracks and significant amount of inactive material in front of the calorimeter Jet Linearity (EM scale) Bruce Mellado, West Coast Physics Forum, 08/22/06

  26. Shown are fractional deviation from the true energy for single pions with energy 100 GeV (hadronic scale) in “critical” regions Bruce Mellado, West Coast Physics Forum, 08/22/06

  27. Fractional deviation from the true energy for single electrons with different energies in bins of 0.625<||<1.25 in steps of 0.025 Bruce Mellado, West Coast Physics Forum, 08/22/06

  28. Fractional deviation from the true energy for single electrons with different energies in bins of 1.25<||<1.875 in steps of 0.025 Bruce Mellado, West Coast Physics Forum, 08/22/06

  29. Outlook and Conclusions • ATLAS has embarked in a vigorous collective effort to develop robust algorithms for Missing ET reconstruction • Cell-Based and Object-Based algorithms available in ATHENA releases • MET performance is being evaluated and studies in a large variety of final states • Shifts in the module of MET are now significantly reduced. Efforts made to improve resolution • Study of various sources of tails and fake MET underway • Regular meetings set up and new tools are being developed • Moving towards development of MET significance • At present compiling resolution of different objects as a function of Energy and || Bruce Mellado, West Coast Physics Forum, 08/22/06

  30. Additional Slides Bruce Mellado, West Coast Physics Forum, 08/22/06

  31. All Calorimeter Cells with |Ecell | > 2 ( noise ) SumET EtMiss Resol -- Rec -- Truth <> = 528.2 <> =560.1 <> ~ 3.1  ~ 16.0 Calorimeter Cells in TopoClusters (4/2/0) EtMiss Resol SumET -- Rec -- Truth <> = 538.3 <> =560.1 <> ~ 1.5  ~ 15.2 • Figure shows A   sample with mass of A = 800 GeV • Resolution and scale improves with the use of noise suppression using CaloTopoCluster Bruce Mellado, West Coast Physics Forum, 08/22/06

  32. EtMiss with early data:resolution with minimum bias and Wjets Minimum bias: Possible to test EtMiss resolution up to SumET=300GeV N. Kanaya Can use W+njets to evaluate EtMiss resolution up to SumET ~ 1TeV. • Need a way to distinguish effect of resolution and offset. Bruce Mellado, West Coast Physics Forum, 08/22/06

  33. Montecarlo comparison s(Ex(y)miss) vs SumET in CSC Jets data (Pythia andHerwig) Bruce Mellado, West Coast Physics Forum, 08/22/06

  34. EtMiss with early data: in situ scale determination with Z   Rec  mass Signal Z   Inclusive W e Inclusive W  top Rome data. Applied cuts : pt(lep) > 15 GeV, ||<2.5 pt(jet) > 15 GeV, ||<2.5 isEM & 0x7FF) ==0 , lep isolation: Etcone30<5GeV 1.< < 2.7 or 3.6< <5.3 mT(lept-EtMiss)<50GeV -likelihood > 8 (-eff ~ 30%) 66<rec mtt<116 GeV Expected in 100pb-1 ~ 300 evts with ~ 20% backgd Possible to loosen cuts to increase statistics? Or more severe cuts necessary to reduce bb backgd? In TDR :EtMiss>20 GeV mT(lept-EtMiss)<25GeV D. Cavalli <> ~ 90  ~ 16 Rec  mass vs EtMiss scale +3% -3% - 10 % +10 % Results still preliminary due to low statistics Need to have also a bb sample Trigger-aware analysis and Cuts tuning Bruce Mellado, West Coast Physics Forum, 08/22/06

  35. Expect O(10K) per 1pb-1 • With 100 pb-1 of data can check EtMiss in most of useful range for low mass Higgs physics and with different jet topologies • The ratio pT(n)/pT(lep) can be calculated with MC. It depends on experimental cuts 100pb-1 R calculated using truth +3% +2% No cuts X.Cueh, L.Flores, B.Mellado, S.Padhi, Sau Lan Wu (Wisconsin) PTL>20 GeV, |l|<2.5 Method is sensitive to EtMiss scale. Less sensitive to EtMiss resolution (use di-jets). Addressing systematic errors and combination with Zll events Bruce Mellado, West Coast Physics Forum, 08/22/06

  36. PTL>20 GeV, |l|<2.5 ATLFAST +5% shift 1pb-1 +3% shift Nominal Bruce Mellado, West Coast Physics Forum, 08/22/06

More Related