1 / 31

Estimation of W+jets background using Z+jets as a control sample

Estimation of W+jets background using Z+jets as a control sample. Maria Ilaria Besana, Umberto De Sanctis, Tommaso Lari (Milano group). Contents :. Introduction Datasets and selection cuts W/Z as a function of jet multiplicity Background contamination as a function of jet multiplicity

ameliak
Download Presentation

Estimation of W+jets background using Z+jets as a control sample

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. Estimation of W+jets background using Z+jets as a control sample I. Besana T.Lari U.De Sanctis Maria Ilaria Besana, Umberto De Sanctis, Tommaso Lari (Milano group)

  2. Contents: • Introduction • Datasets and selection cuts • W/Z as a function of jet multiplicity • Background contamination as a function of jet multiplicity • Estimation of backgroud W+jets to top analysis • Conclusion and outlook I. Besana T.Lari U.De Sanctis

  3. Introduction: • Update ofpreviousanalyses: • http://indico.cern.ch/getFile.py/access?contribId=8&resId=0&materialId=slides&confId=25489 • http://indico.cern.ch/getFile.py/access?contribId=12&resId=0&materialId=slides&confId=25360 • http://indico.cern.ch/getFile.py/access?contribId=5&resId=1&materialId=slides&confId=52962 • Similar work doneby: • Marisa Sandhoff (Wuppertal) : http://indico.cern.ch/getFile.py/access?contribId=49&sessionId=6&resId=0&materialId=slides&confId=31799 • LBNL group : Atlas note ATL-PHYS-COM-2007-100 • Bonn group : http://indico.cern.ch/getFile.py/access?contribId=3&resId=0&materialId=slides&confId=45163 http://indico.cern.ch/getFile.py/access?contribId=6&resId=0&materialId=slides&confId=55855 I. Besana T.Lari U.De Sanctis

  4. Introduction: motivation • theoretical Xsection of W+jets is rather uncertain especially ratio: W+0j : W+1j : W+2j : W+3j : W+4j : W+5j • tool: Z+jets - Z is “easy ” to select with high purity (but σW~10*σZ) - idea: use Z to determine σW and properties of W+jets - assumption: σW+4jets = c *σZ+4jets (σW+0jets / σZ+0jets ) c = coef which takes into account mass difference for W and Z • distributions for W and Z should look similar I. Besana T.Lari U.De Sanctis

  5. Datasets: I. Besana T.Lari U.De Sanctis Datasets at 10 TeV rel 14: Alpgen datasets haven’any filter on jets.

  6. DATASETS: I. Besana T.Lari U.De Sanctis ! Not checked yet: QCD contamination in W and Z selected sample, how to estimate and subtract its contribution from data

  7. Selectioncuts: objectdefinition I. Besana T.Lari U.De Sanctis

  8. Selectioncuts: eventselection Z ee: • Trigger: e20i • 2 electrons with pT > 20 GeV, muon veto • 80 GeV < invariant mass < 100 GeV • max (|ETmissX|, |ETmissY|) < 3 σgauss W eν: • Trigger: e20i • 1 electron with pT > 20 GeV , muon veto • ETmiss > 20 GeV Top selection requires in addition: • at least 4 jets with pT >20 GeV & |η|<2.5 • at least 3 jets with pT >40 GeV & |η|<2.5 I. Besana T.Lari U.De Sanctis

  9. W/Z RATIO AS A FUNCTION OF NJETS MC level: Pythia vs Alpgen vs McAtNlo I. Besana T.Lari U.De Sanctis

  10. W/Z ratioas a functionofNjets: Reco level: Pythia vs Alpgen vs McAtNlo I. Besana T.Lari U.De Sanctis

  11. W/Z ratio as a function of Njets : Reco level (1jet norm.): Pythia vs Alpgen vs McAtNlo I. Besana T.Lari U.De Sanctis

  12. W/Z ratioas a functionofNjets: Reco level (2jet norm.): Pythia vs Alpgen vs McAtNlo I. Besana T.Lari U.De Sanctis

  13. I. Besana T.Lari U.De Sanctis

  14. I. Besana T.Lari U.De Sanctis

  15. EventSelectioncuts • e10i trigger chain (10**31 menu) • Electrons medium and isolated as before • W selection: • 1 electron pt >20 GeV, no muon pt>20GeV, • etmiss > 20 GeV, mT(e,EtMiss) cut. • Z selection: • 2 electrons pt>20 GeV, no muon pt>20GeV • max (|ETmissX|, |ETmissY|) < 3 σgauss • 80 GeV < invariant mass < 100 GeV I. Besana T.Lari U.De Sanctis

  16. Background contaminationfor Z: inclusive and low jet multiplicity sample Selection for ratio measurment [80GeV-100GeV] I. Besana T.Lari U.De Sanctis Z can be selected with high purity!

  17. Background contaminationfor W: inclusive and low jet multiplicity sample Selection for ratio measurement: [40 GeV, 90 GeV] I. Besana T.Lari U.De Sanctis

  18. Selection of W, Z candidates • All numbers normalized to 10 pb-1 • W signal (for purity) is sum of en and tn decays • Background in Z selection < 0.01% (but QCD not included yet) • We need to correct the number of W candidates for the efficiency e of mT cut • We need to correct for the background to W selection • Systematics on the two corrections need to be studied. • In the following, assume : - background known to 20% • - D(1-e)/(1-e) = 50% I. Besana T.Lari U.De Sanctis

  19. W/Z ratio • (Very preliminar) overall error on W/Z normalization factor: • 2.1% using 0 jet selection (0.6% with 100 pb-1) • 4.2% using 1 jet selection (1.5% with 100 pb-1) • 10.4% using 2 jet selection (5.9% with 100 pb-1) • QCD background not included yet • Error from MC statistics on ratios is small (0.04 to 0.12) • As long as efficiencies ratios, acceptance ratios, JES, etc. are the same as in • to the top signal region (3-4 high pt jets) we do not care! These effects need to be • included on systematic error on the extrapolation (later). I. Besana T.Lari U.De Sanctis

  20. Ratio of ratios • Call Rn = (W+njets)/(Z+njets) • Rn depends on n differently according to different generators. This is a systematics in extrapolating to top signal region (large n) • We can use measured values of R2/R0 and R1/R0 to test/tune/constrain generator predictions R1/R0 = 0.791 ± 0.039 ± 0.026 R2/R0 = 0.726 ± 0.076 ± 0.044 First error is statistics for 10 pb-1. Second is systematics, dominated by the efficiency term. Statistical error on MC prediction is <0.01 (sarebbe carino confontare col valore a livello di verita’ MC) I. Besana T.Lari U.De Sanctis

  21. Background contaminationfor Z: top-antitop signalregion Mee invariant mass (200 pb-1) NO cut on ETmiss I. Besana T.Lari U.De Sanctis Mee invariant mass (200 pb-1) Cut on ETmiss

  22. I. Besana T.Lari U.De Sanctis

  23. Evaluationof background tosemileptonic top-antitop sample (200 pb-1) I. Besana T.Lari U.De Sanctis 22 % difference between Alpgen and Pyhia

  24. Three-jetinvariant mass distribution (200 pb-1) Alpgen I. Besana T.Lari U.De Sanctis Good agreement !

  25. Conclusion & Outlook W+jets background can be estimated from Z+jets: • with a statistical error of …% (according to Alpgen) at 200 pb-1 , • difference between Alpgen and Pythia of 22%, • using R1/R0 and R2/R0 (RN=W+Njet/Z+Njet) to constraint the spread of Montecarlo predictions and reduce associated systematics looks promising. Next steps: • include QCD background, • study systematics: constrain generator dependence with first data. I. Besana T.Lari U.De Sanctis

  26. Backup I. Besana T.Lari U.De Sanctis

  27. Pythiasystematicerror at MC level (rel 12) There is a distortion in the pT shape of the most energetic jet wich comes from doing an η cut on the 2nd electron in Z ee, but not on the ν of W eν I. Besana T.Lari U.De Sanctis

  28. Pythiasystematicerror at MC level (rel 12) I. Besana T.Lari U.De Sanctis

  29. Etmissdistribution: I. Besana T.Lari U.De Sanctis Selection cut: max (|ETmissX|, |ETmissY|) < 3 σgauss

  30. Selectioncuts: efficiency W selection Pythia I. Besana T.Lari U.De Sanctis Alpgen

  31. Selectioncuts:efficiency Z selection Pythia Alpgen I. Besana T.Lari U.De Sanctis Ratio (signal/background) > 200

More Related