1. DPF 2002, Colonial Williamsburg, VA May 25, 2002 1 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 1 Leslie Groer Columbia University, New York DPF 2002, Colonial Williamsburg, VA May 25, 2002 Jet and Electron Identification in the Run 2 DØ Detector  Tevatron Run 2  DØ Detector upgrade  SMT  CFT  Preshower + ICD  Calorimeter  Jet ID  Algorithms  NADA  Trigger  Selection  Energy Scale  QCD Results  EM ID  Reconstruction  Trigger  Profile  Scale

2. DPF 2002, Colonial Williamsburg, VA May 25, 2002 2 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 2 DØ roll-in Run II start First Collisions Detector Commissioning; Timing in; Improve electronics,DAQ and offline Main Injector (new) Tevatron DØCDF Chicago  Booster 1.961.8  s (TeV) 36x366x6 #bunches Run 2aRun 1b Tevatron Run 2  p source 4.82.32.5 interactions/xing 1323963500 bunch xing (ns) 10517.33.2  Ldt (pb -1 /week) 5.2x10 32 8.6x10 31 1.6x10 30 typ L (cm -2 s -1 ) 1.96 140x103 Run 2b  New Main Injector and Recycler rings  Increased luminosity and energy  48 pb -1 delivered  15.2 pb -1 recorded physics events  L dt expected for 2002: 300 pb -1 Run 2a: 2 fb -1

3. DPF 2002, Colonial Williamsburg, VA May 25, 2002 3 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 3 Overview of Run 2a DØ Upgrade Muon, Calorimeter, Silicon fully commissioned and operational  Fiber tracker and preshowers fully instrumented. Central electronics complete, forward in a few weeks— commissioning this summer  Upgrade Calorimeter electronics readout and trigger  Add scintillator in muon for fast trigger and extended coverage for drift chambers  Replace inner tracking volume with Silicon and Fiber trackers with 2T solenoid magnetic field for central tracking and momentum measurement  Add preshower detectors and replace intercryostat detectors  Pipelined 3 Level trigger  Increase DAQ capability for 132 ns bunch crossings azimuthal angle  pseudorapidity  = -ln tan(  /2)

4. DPF 2002, Colonial Williamsburg, VA May 25, 2002 4 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 4 SS: single sided DS: double sided More in Harald Fox’s talk Silicon Microstrip Tracker 6 Barrels 12 F-Disks 4 H-Disks  Tracking up to |  | = 3  Provide good position resolution for vertexing  Innermost layer at r = 2.6 cm  Central region u 6 barrels, 4 layers, axial + 2 o /90 o stereo 12 cm long each, SS+DS u 12 F-disks (SS)  Forward region u 4 H-disks (SS)  793k channels  Radiation hard up to 1 Mrad  >90% channels operational  S:N > 10:1

5. DPF 2002, Colonial Williamsburg, VA May 25, 2002 5 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 5 Central Fiber Tracker 0 p.e. 1 p.e. 2 p.e. 3 p.e. CFT axial + stereo + SMT d~42  m pT > 3 GeV Beam spot ~28  m FPS d  Tracking out to |  | = 1.7  Good momentum resolution  20 cm < r < 51 cm, 1.8 / 2.6 m fibers  8 double layers (axial, stereo 3 o )  77,000 830  m fibers readout with VLPC  Operate at 9 K, 85% Q.E., good S/N  ~10 photons/m.i.p. get to the VLPC  Impact parameter resolution ~42  m for SMT+CFT tracks with p t > 3 GeV  No individual ladder or layer alignments yet  Beam spot size is about 28  m  Trackers shifted in z by 2.9 cm w.r.t calorimeter  shifts z o

6. DPF 2002, Colonial Williamsburg, VA May 25, 2002 6 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 6 Preshowers and Intercryostat Detector  Central and Forward Preshowers  Central mounted on solenoid (|  | < 1.2) u Forward on calorimeter endcaps (1.4 < |  | < 2.5) u CPS: 7,680 FPS: 14,000 channels u Extruded triangular scintillator strips with embedded WLS fibers and Pb absorber u Improve energy resolution measurements u Trigger on low-p T EM showers u Reduce overall electron trigger rate by x3-5 u Same readout electronics as CFT  Intercryostat Detector (ICD) u 384 scintillator tiles with WLS fiber to phototubes in low-B field region for readout u Improve coverage for the region 1.1 < |  | < 1.4 u Improves jet E T and missing-E T u Readout through Calorimeter electronics u LED pulsers used for PMT calibration u Relative yields measured > 20 p.e./m.i.p. ICD FPS CPS

7. DPF 2002, Colonial Williamsburg, VA May 25, 2002 7 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 7 Calorimeter Overview L. Ar in gap 2.3 mm Ur absorber Cu pad readout on 0.5 mm G10 with resistive coat epoxy  Liquid argon sampling u Stable, uniform response, rad. hard, fine spatial seg. u LAr purity important  Uranium absorber (Cu (CC) or Steel (EC) for coarse hadronic) u Compensating e/   1, dense  compact  Uniform, hermetic with full coverage   < 4.2 (   2 o ),  int  total)  Single particle energy resolution  e:  E / E = 15% /  E  + 0.3%  :  E / E = 45% /  E + 4% Drift time 430 ns South End Cap North End Cap Central Cal.  50k readout cells (<0.1% bad)  Fine segmentation,  5000 semi-projective towers (0.1x0.1)  4 EM layers, shower-max (EM3): 0.05 x 0.05  4/5 Hadronic (FH + CH)  L1/L2 fast Trigger readout 0.2x0.2 towers ICD EM FH CH OH MH IH EM MG FPS

8. DPF 2002, Colonial Williamsburg, VA May 25, 2002 8 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 8 Calorimeter Electronics Calibration  Electronic readout  “live” sampled energy in L.Ar.  calibrated energy scale  Determine electronic calibration coefficients for absolute and channel-to-channel variations from pulser charge injection (DAC  ADC)  Dual gain readout with analog storage in switched capacitor arrays (SCA)  Non-linear behavior of SCA chip observed for low energies u ADC to GeV about 300 MeV underestimation per cell u Nonlinearity 1 GeV u Has significant effect in low energy region (jet widths and resolutions etc)  Can apply universal parametrized correction for all channels u Residuals after correction are better than  5 ADC counts on the whole range for both gains  Correct energy in cells before clustering  In calibration, correct for signal shape difference with simulation  Also correct for cell-to-cell gain (ADC/DAC) dispersion (5 to 10%)  Apply  intercalibration comparing slices in  -- flat within 2% after correction  Improves both Z mass mean and resolution Parameterized correction based on residuals compared to linear fit 1 ADC ~ 4 MeV ADC 8181 pulser ADC readout pulser shaper output dual gain ADC vs DAC

9. DPF 2002, Colonial Williamsburg, VA May 25, 2002 9 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 9 Jet Finding  Parton jet u Parton hard scattering and parton showers well described by pQCD  Higher cross-section expected in Run 2 for higher c.m.s  s=1.96TeV u x2  for p T > 400 GeV  Calorimeter jet u Jet is collection of towers with a given cone R u Cone direction maximizes the total ET of the jet u Various clustering algorithms  Particle jet u After hadronization u A spread of particles running roughly in the same direction as the parton u Correct for finite energy resolution u Subtract underlying event (modeled by minimum bias data) Jet inclusive pT spectrum

10. DPF 2002, Colonial Williamsburg, VA May 25, 2002 10 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 10 Run 2 Jet Algorithms  Run 1 Legacy Cone u Draw a cone of fixed size around a seed u Compute jet axis from E T -weighted mean and jet ET from  E T ’s u Draw a new cone around the new jet axis and recalculate axis and new E T u Iterate until stable u Algorithm is sensitive to soft radiation  Improved Run 2 cone u Use 4-vectors instead of E T u Add additional midpoint seeds between pairs of close jets u Split/merge after stable protojets found u Algorithm is infrared safe  kT-algorithm u Recombination algorithm based on relative momentum between ‘particles’ u Theoretically favored, no split-merge u To reduce computation time, start with 0.2 x 0.2 preclusters  Cell Nearest Neighbor u Floor-by-floor clustering starting with EM3 u Each local maximum starts a floor cluster then add in neighbors u Energy sharing according to transverse shape parameterization u Angular matching of floor clusters u Search for minima in longitudinal energy distribution to separate EM and hadronic showers  Energy Flow algorithm u use tracking information to better characterize the contributions from charged particles u In development Most results using simple cone for now

11. DPF 2002, Colonial Williamsburg, VA May 25, 2002 11 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 11 NADA  NADA = New Anomalous Deposit Algorithm  Identify anomalous isolated energy deposits in the calorimeter = “Hot Cells” u Source: electronics, U noise, beam splash, cosmics etc  Improve object resolution and MET  Run 1: AIDA u Only examine neighbors in the same tower for E cell > 10 GeV u 99% efficient, BUT 5-10% misidentification rate  Not used for cells on boundaries of layers  FH1 and CH1 have more material E T thresold E T neighbour > 100 MeV or 0.02E cell  Examine all cells with > 1 GeV u Remove cells 500 GeV u E T 100 MeV u E T 2% E cell  High efficiency (90%) and low misidentification u E T > 1 GeV : ~0.5% u E T > 10 GeV : ~0%  On average about 0.8 cells / event

12. DPF 2002, Colonial Williamsburg, VA May 25, 2002 12 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 12 Jet Selection  Central jets (Run 2 cone, R=0.7)  Event Quality Cuts u Number of jets  1 u E total in the calorimeter  2 TeV u Missing E T  70% of the leading jet p T u Z vtx < 50 cm  Leading Jet Cuts u Jet p T > 8 GeV (offline cut) u 0.05  EMF  0.95 u CHF  0.4 (0.25 tight) u HotF  10 (5 tight) (HotF = E T 1st cell / E T 2nd cell ) u n90 > 1 (number of towers that contain 90% of jet E T )  Efficiencies from MC u Loose: ~100% Tight: ~ 98% u ~Flat in eta DØ Run 2 Preliminary CHF EMF n90 HotF Data — MC  Non-linearity of SCA included in MC

13. DPF 2002, Colonial Williamsburg, VA May 25, 2002 13 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 13  jet Jet Energy Scale  Correct Jet Energy back to the particle level  E offset energy offset from underlying event, pile-up, Uranium noise  determined from Min. Bias Events  R calo calorimeter response  Calibrate EM response on Z  ee mass peak  Measure from ET balance in  +jet events  R cone energy contained in jet cone  Correct for losses due to out-of-cone showering  Use MC-energy in cones around the jet axis Photon-jet Events Preliminary correction being applied with ~10% systematic uncertainty

14. DPF 2002, Colonial Williamsburg, VA May 25, 2002 14 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 14 Central Jet Triggers  L1 single jet efficiencies u Ask for one or two hadronic trigger towers (0.2x0.2) above threshold u Use muon trigger as unbiased reference for statistics to measure turn-ons u Ask for one and only one reconstructed jet in |  |<0.7 u L1 hadronic response about 40% low for current data set  L2 jet u Cluster 3x3 or 5x5 trigger towers around L1 seed towers  L3 jet u Simple cone or tower NN algo’s 0.1x0.1 towers u 3 single jet triggers (single tower): s JT_LO L1: 5 GeV, L3:10 GeV s JT_HI L1:10 GeV, L3:15 GeV s CJT40: L1:40 GeV u Efficiency s Standard jet selection, offline pT > 8 GeV s Very sharp turn on All L1 trigger towers at |  | <0.8 are instrumented, complete coverage coming soon L1 Trigger efficiency CJT(1,x) L1 Trigger efficiency CJT(2,x) Efficiency vs jet pT CJT(1,3) CJT(1,5) CJT(1,7) CJT(1,10)

15. DPF 2002, Colonial Williamsburg, VA May 25, 2002 15 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 15 First Run 2 QCD Physics  Central jets  Not fully corrected distributions: u Preliminary correction for jet energy scale (but no unsmearing or resolution effects) s 30-50% systematic error in cross-section u No trigger selection efficiency corrections Highest 3-jet event E T jet1 : 310 GeV E t jet2 : 240 GeV E T jet3 : 110 GeV E t miss : 8 GeV Only statistical errors Inclusive jet pT spectrum at 1.96 TeV Only statistical errors Dijet mass spectrum at 1.96 TeV  L dt = 1.9 ± 0.2 pb -1

16. DPF 2002, Colonial Williamsburg, VA May 25, 2002 16 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 16 u Tuned on MC in   bins of 0.1,  < 3.2 for different energies u HMx8 / HMx9 / Hmx41 s Energy fractions in each floor (PS), EM1, EM2, EM3, EM4 s ,  in EM3  grid (6,6) s log(E tot ) s Z/  z vertex EM ID and Reconstruction  Hmatrix u Measure compatibility of EM cluster with an electron shower   2 u Discriminate against hadronic (  ) decays that pass EM fraction and isolation cuts u Use longitudinal and transverse shower shapes to take into account correlations between energy in cells  Concentrate on high P T objects  Look for narrow isolated clusters with high EM fraction, track match for electrons, none for   Electron object reconstruction u P Tmin >1.5 GeV u EM fraction > 0.9 u Isolation u CC:  3x3 EM towers u EC: All cells in cone of 20 cm radius at EM3 around hottest channel u Track match p T > 1.5,  R<0.5  Preliminary fake rate calculated from 2 nd unbiased jet passing standard EM selection in jet triggers  0.6  0.1% e  HM41 Run 1 test beam+W  e log  2

17. DPF 2002, Colonial Williamsburg, VA May 25, 2002 17 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 17 Triggering on electrons  L1 EM Trigger u Look for single EM trigger tower (0.2 x 0.2) over threshold u Scale calibrated ~10% u No hadronic veto u Use “bootstrap” method to calculate efficiencies  L2 EM Trigger u Use 3x3 NN algorithm with 1 GeV seed  L3 EM Trigger u To measure the trigger efficiency, select good EM objects:  EM frac > 0.9, isolation < 0.2, HM41 < 200, |  | < 0.8 u L3EM(1,15,emfr) rejection 5.1 u Add shower shape  can drop energy threshold L3EM(1,12,emfr,shape) rejection 4.2 L1 TT vs Offline L1 Trigger effic. CEM(1,x) L3 Trigger effic.  L3EM(1,15)  L3EM(1,12,shape)  L3EM(2,10)

18. DPF 2002, Colonial Williamsburg, VA May 25, 2002 18 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 18 Reconstructed EM profiles EM4 EM3 EM2 EM1 Energy Fractions  pT>20 GeV from EM_HI trigger — QCD MC EM1 EM2 EM3 EM4  Energy Fractions  good EM candidates that reconstruct to Z mass — Zee MC Efficiency vs.  Efficiency from 2 nd e in Zee sample (pT>20GeV, with track match) HMx9 < 100 : 94% HMx9 < 25 : 82% DØ Run 2 Preliminary

19. DPF 2002, Colonial Williamsburg, VA May 25, 2002 19 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 19 Energy Scale from Z  ee CC+EC  Not applied phi-intercalibration, pulser corrections etc. so calculate energy correction for each cryostat region to restore Z-peak to its expected value  Gives correction < few %  Work underway to add tracking information, calibration for individual cryostat quadrants  Compare data and Zee MC Mass distributions to get absolute energy scale  Use standard EM selection with geometrical corrections (phi cracks, eta dependence etc) u  2 EM objects, ET > 20 GeV u isolation < 0.1 u 0.95 < EM fraction u HMx8 < 100  Paramaterize E true = E(1 +  )  Fit for Z mass with Breit-Wigner and find  which maximizes a likelihood

20. DPF 2002, Colonial Williamsburg, VA May 25, 2002 20 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 20 Check Energy Scale with W  e EM cluster with SMT track  em objects  with track match  el-id criteria W selection  EM Object u ET>25 GeV in |  | < 0.8 u -0.05 < isolation < 0.1 u 0.95 < emf < 1.05 u HM8<50, HM41<200  MET > 25 GeV Search for cluster-global track match in EM sample (scale corrected)  Global tracks u 10 < Nhits < 16 u p T > 5 GeV  Track to EM cluster match u  < 0.05,  < 0.2 E/p ~ 1 Fit the electron p spectrum

21. DPF 2002, Colonial Williamsburg, VA May 25, 2002 21 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 21 Summary  Tevatron Run 2 well underway  DØ detector performing extremely well but many new systems coming online u Complete readout and integration of tracking and preshowers u L1: extend coverage in eta; track triggers u L2: calorimeter and track triggers u New EM/jet algorithms (e.g. did not discuss identification of softer electrons in jets, especially useful for semileptonic b-decays – use road method)  Expect rich physics program from large statistics for high p T events u Improve knowledge of QCD, proton structure functions u Measurements of heavy flavor and Electroweak physics u Searches for new phenomena, quark compositeness, extra dimensions, W’, Z’… u The elusive Higgs boson  D0 detector poised to take full advantage of the higher instantaneous and integrated luminosities

22. DPF 2002, Colonial Williamsburg, VA May 25, 2002 22 Leslie Groer Columbia UniversityJet and Electron Identification in the Run 2 DØ Detector 22 EM geometric corrections and resolution  EM Geometric Correction u Energy corrections for geometric effects (e.g. phi cracks, eta dependence due to dead material in front of calorimeters) u Single electron MC  EM Resolution u Single electron MC u Calorimeter info only (no preshower) u Correcting for phi cracks and eta correction Calculate  from cluster position in EM3 Eta correction factor sampling noise constant EM resolution E (GeV) Eta correction factor 5 GeV electrons  E (GeV)