Measurements of Electron Charge Misidentification Rate and Electron

Measurements of Electron Charge Mis-identification Rate and Electron Efficiency within SUSY analysis framework 1 Zhichao Zhan SDU Shan Jin IHEP Xueyao Zhang SDU Emmanuel Monnier CPPM FCPPL April 7, 2011 Jinan, China

outline Introduction Tag-and-Probe Method Electron Charge Misidentification Rate Electron Reconstruction Efficiency Conclusion and Prospect 2

Introduction ATLAS SUSY group has published 2 paper 0 lepton channel and 2 lepton channel. Charge mis-identification is the very important part for SUSY 2 same sign lepton final states analysis, which is also an important issue for the physics channel that has lepton charge requirement (SS/OS). Several constituents for charge misidentification Bremsstrahlung; Conversion events; fake electron from jet We try to get a precise charge misidentification rate. 3

Tag-and-probe Method 1. Good Events(pls see the Appendix) 2. 2 good electrons, one tight tag electron; one probe electron. 3. Missing. Et<20 Ge. V 4. dphi(e 1, e 2)>pi*3/4 Data control sample is fitted by Breit Weigner function convoluted with Gaussian distribution function (Line shape); The background is fitted by Chebyshev polynomials. Definition: Working progress 4 4 Line shape

Object and Bakground check: Electron: Author=1||3; pt>20 Ge. V fabs(ηcalo)<2. 47; etcone 20<10 Ge. V; ptcone 30/pt<0. 2 Rubust. Medium; Electron eta Electron pt Working progress Data and MC check Electron phi 5 There is no significant discrepancy between MC and data.

Electron charge mis-ID rate data~34. 3 pb-1 @7 Te. V Robust. Medium Robuster. Tight DATA Left: opposite sign events fitted by line shape 6 right: same sign events fitted by line shape

Robust. Medium Robuster. Tight MC Apply the same fit process to MC Zee sample 7

Electron charge mis-ID rate summary |DATA-MC|(%) 8 Robust. Medium Robuster. Tight 0. 09± 0. 24 0. 33± 0. 18

Electron efficiency using the same technique Robust. Loose Robust. Medium Robuster. Tight 9 9 Left: the probe electron fulfill three level selection criteria Right: the probe electron does not pass three level selection criteria

Working progress 10 10 Working progress

Positron efficiency Robust. Loose Robust. Medium Robuster. Tight 11 11 Left: the probe positron fulfill three level selection criteria Right: the probe positron does not fulfill three level selection criteria

Working progress 12 12 Working progress No difference between positron and electron efficiency

Conclusion&Prospect Electron charge misidentification rate and electron reconstruction efficiency from Tag-and-Probe method sounds reasonable. Using Tag-and-probe method, we can get a charge misidentification rate with respect to bin-by-bin pt distribution, when we have enough data statistics. The discrepancy of electron charge mis. ID rate between real data and MC is under investigation Thanks a lot! Merci ~ 13

Appendix I Lepton Electron : Author=1||3; pt>20 Ge. V fabs(η 2 calo)<2. 47; etcone 20<10 Ge. V; ptcone 30/pt<0. 2 Rubust. Medium&&Expect. Hit. BLayer; Muon: Staco combined; 0<=match. Chi 2<=100 fabs(η)<2. 5; pt>20 Ge. V; etcone 20<10 Ge. V; Jet&MET l Jet: l Anti. Kt 4 Topo; EMJES calibration; l p. T>20 Ge. V, fabs(η)<2. 5 l MET: MET_Loc. Had. Topo 14

Jet Cleaning: remove events with any jes with pt(EM)>10 Ge. V, satifying: emf>0. 95 && fabs(quality)>0. 8 n 90<=5 && hec. F>0. 8 |t|>=50 ns hec. F >= 1 - fabs(quality) Step 1: GRL Step 2: Trigger Step 3: Jet Cleaning Step 4: Vertex requirment(Ntrack>4) Step 5: Veto events with crack electrons Step 7: OR for jets/electrons(remove jets within d. R=0. 2 of an electron) Step 8: Veto events with electron in bad OTX regions Step 9: Veto events with muon having|Z 0 -Zpv|>10 mm Step 10: Remove electrons/muons within d. R=0. 4 of a jet Note: use electron cluster(eta, phi)for OTX veto use electron object(eta, phi)for OTX veto 15

Appendix II 16 16

Appendix III Robust. Medium

tight

Truth tag mis-charge rate. Strategy： 1. Selecte an Electron, which fulfill with pt >20 Ge. V, |eta|<2. 47, is. EM_tight, 2. author==1||author==3 3. Tag a truth Electron fufill match(d. R<0. 1) 4. Calculate its mischarge rate. Working progress 19 Working progress The charge mis-identification rate from reconstruction is around 0. 8%, in the truth tag method measurement 19