Aidan Robson Glasgow University for the CDF and D0 Collaborations EWK Cross-sections and Widths ICHEP, Philadelphia, 30 July 2008

Aidan Robson Glasgow University 1/24 CDF Z  ee (from Stirling, ICHEP04) 2004, using < 100 pb –1

Aidan Robson Glasgow University 2/24 A high-p T physics programme

Aidan Robson Glasgow University 3/24  More challenging  channels  Differential cross-sections  High-precision measurements of Standard Model parameters Now:

Aidan Robson Glasgow University 4/24  More challenging  channels  Differential cross-sections  High-precision measurements of Standard Model parameters  (pp  Z). Br(Z  ) d  / dy d  / dp T  W  Z (invisible) Outline

Aidan Robson Glasgow University 5/24 CDF and D0   = 1.0  = 0.6  = 2.0 muon chambers D0  =2  = m tracker had cal hadronic cal EM cal had cal solenoid pre-radiatorshower max silicon E M cal  =1 CDF Fibre tracker to |  |<1.8 Calorimeter to |  |<4 Muon system to |  |<2 Drift chamber to |  |<1 Further tracking from Si Calorimeter to |  |<3 Muon system to |  |<1.5

Aidan Robson Glasgow University 6/24 W and Z selection Electrons: good EM shower shape small hadronic energy isolated in calorimeter well-matching good track (except far forward) Muons: MIP in calorimeter isolated hits in muon chamber well-matching good track Z selection: 2 oppositely-charged electrons or muons invariant mass consistent with m Z W selection: exactly one electron or muon energy imbalance in reconstructed event, associated with neutrino

Aidan Robson Glasgow University 7/24 Taus at D0 Neural net separator trained on variables measuring: isolation shower shape calorimeter–track correlations Three D0 tau categories: Type 1: 1 track, no EM sub-cluster Type 2: 1 track, ≥1 EM sub-clusters Type 3: ≥2 tracks, ≥0 EM sub-clusters Type 1Type 2Type 3 0 NN output score 1 The elements:  calorimeter cluster (cone R<0.5)  energy concentrated in inner cone R<0.3  tracks in cone R<0.3, mass<1.8GeV  EM sub-clusters in finely segmented shower-max layer of calorimeter E T measurement: For E<100GeV, track p T & calorimeter E T are combined for Types 2&3, and single  ± energy corrections derived from special hadronic calorimeter simulation. Otherwise track p T used.

Aidan Robson Glasgow University 8/24 Z  Type 1Type 2Type 3 Select Z    h from inclusive muon trigger. Additionally:  p T  > 15 GeV  p T  > 15 GeV  opposite charge visible mass / GeV D0 1fb –1 preliminary  scalar sum p T of  tracks > 15GeV or 5 GeV (types 1&3/2)  NN > 0.9 or 0.95 (types 1&3/2) Backgrounds:  QCD (bb) data-driven from same-sign events  EWK backgrounds from MC corrected for SS–OS ratio in QCD-enhanced sample W+jets corrected for component accounted for in same-sign events m vis = √(P  + P  + P T ) 2

Aidan Robson Glasgow University 9/24 Z   (pp  Z). Br(Z  ) = 247 ± 8(stat) ± 13(sys) ± 15(lumi) pb Systematic sourceValue Tau Energy Scale1.0% NN2.4% Tau track reco1.5% QCD background0.7% W  background 0.5% Trigger3.6% Muon track match0.8% Muon identification0.4% Z p T reweighting1.6% PDF1.7% Total (excl lumi)5.4% Luminosity6.1% 2.5% 3.5% 2.3% Values from previous published analysis (PRD (2005)) Total 1527 events observed, ~20% background

Aidan Robson Glasgow University 10/24 Z rapidity rapidity y= ln E+pzE–pzE+pzE–pz 1212 closely related to parton x: Z production: m√sm√s x 1,2 = e ±y (LO) P1P1 P2P2 f q (x 1 ) f q (x 2 ) x1P1x1P1 x2P2x2P2  */Z e+e+ e–e– CDF e+e+ e–e– x y - Events / 0.1

Aidan Robson Glasgow University 11/24 Z rapidity: shapes Fraction Acceptance x efficiency Si tracking efficiency energy scaling factor |  ’| |  |< <|  |<0.8 |  |<0.8All  10 E T 60 Calorimetry Tracking Total

Aidan Robson Glasgow University 12/24 Z rapidity Isolation E Events Systematics: Detector material modeling Background estimates Electron identification efficiencies Silicon tracking efficiency Acceptance

Aidan Robson Glasgow University 13/24 Z rapidity fractional systematic uncertainty (%)

Aidan Robson Glasgow University 14/24 Z p T p T (Z) pQCD reliable resummation / parton shower with non-perturbative model Measurement of Z p T tests QCD predictions for initial state gluon radiation resummation required multiple soft gluon radiation 30 0 RESBOS event generator implements NLO QCD and CSS resummation (with BNLY form-factor) 3-parameter function, global data fit Recent global fits suggest extra small-x form-factor

Aidan Robson Glasgow University 15/24 Z p T p T (Z) / GeV d  /dp T / nb/GeV all y |y| > 2 pp  Z 0 X, Tevatron :√s=1.96TeV d  /dp T / pb/GeV  Recent global fits suggest extra small-x form-factor – implies p T (Z) broadened at high y D0 e+e+ e–e– p T (Z) / GeV d  /dp T / nb/GeV d  /dp T / pb/GeV pp  Z 0 X  e + e – X, LHC :√s=14TeV pp  W + X  e +, LHC :√s=14TeV |y| < 2.5 p T e > 25 GeV |y e | < 2.5 p T e > 25 GeV E T > 25 GeV p T (Z) / GeV ATLAS e+e+ e–e– p T (W) / GeV pp  Z 0 X, Tevatron :√s=1.96TeV No small-x broadening (BLNY) With small-x broadening No small-x broadening (BLNY) With small-x broadening

Aidan Robson Glasgow University 16/24 Z p T Regularized unfolding p T (e 1 ), p T (e 2 ) > 25 GeV |  |<1.1 or 1.5<|  |< < m ee < 110 Selection: 64k events, > 5k for |y Z |>2 Backgrounds: fit m ee with templates 1.3–8.5% Unfolding method (input MC) Smearing parameters PDF Unfolding method (parameters) QCD background p T dependence of eff. x acc.

Aidan Robson Glasgow University 17/24 Z p T without small-x effect:  2 /dof = 11.1/11 including small-x effect:  2 /dof = 31.9/11 (curves normalised) PRL (2008) Data does not favour small-x broadening (but no refit of usual 3 parameters has been done)

Aidan Robson Glasgow University 18/24 Z p T PRL (2008) The complete spectrum: Compare 4 models: Resbos with default parameters Resbos with additional NLO–NNLO K-factor NNLO (Melnikov and Petriello) NNLO rescaled at to data at 30GeV/c

Aidan Robson Glasgow University 19/24 W width  W predicted in Standard Model:  W SM = 2091±2 MeV (PDG) Experimentally have access to transverse quantities:

Aidan Robson Glasgow University 20/24 W width  Generator: LO MC matched with Resbos (QCD ISR) and Berends/Kleiss (QED FSR)  Fast simulation for templates: electron conversions + showering muon energy loss parametric model of recoil energy (QCD, underlying event + brem)  Tracking scale/resn  Calorimeter scale/resn m  (GeV) m ee (GeV)  Backgrounds m T (GeV)  21 MeV, 31 MeV  17 MeV, 26 MeV  32 MeV  33 MeV  54 MeV (ele), 49 MeV (mu)  2 /dof = 27.1/22  2 /dof =18/22

Aidan Robson Glasgow University 21/24 W width  W = 2032 ± 73 (stat+sys) MeV  W SM = 2091 ± 2 MeV) PRL (2008) Compare to CDF indirect measurement: NNLO calcFrom LEP SM value  W (indirect) = 2092 ± 42 MeV J Phys G World most precise single measurement

Aidan Robson Glasgow University 22/24 Z invisible width  Z (invisible) measured very precisely indirectly from LEP:  Z (invis) = ± 2.6 MeV However combined direct LEP measurement:  Z (invis) = 503 ± 16 MeV Tevatron measurement uncorrelated, relies less strongly on theoretical inputs  Z (inv)  Z (ll)  (Z+1jet). Br (Z  inv)  (Z+1jet). Br (Z  ll) = N obs – N bck  (Z  ll +1jet). L = Must be corrected for different acceptance of ‘1-jet’ selection in Z  versus Z  ll (applied after lepton removal) Select single-jet events (E T trigger) Independently measure  (Z+1jet). Br (Z  ll) from high-p T lepton trigger Selection:  E T > 80 GeV  E T jet > 80 GeV second jet E T <30 allowed but ≥ 3 jets E T >20 rejected

Aidan Robson Glasgow University 23/24 Z invisible width EWK ‘background’ includes Z  prediction WW 2010 ± 69 WW 1570 ± 54 W  e 824 ± 28 Z  ll 87 ± 3 QCD708 ± 146  +jet 209 ± 41 non-collision52 ± 52 Z  3203 ± 137 Total predicted8663 ± 332 Data observed8449 Measured  ( Z  ll+1jet) = ± pb  Z (inv)  Z (ll) = ±  Z (inv) = 466 ± 42 MeV ~ equal contributions from EWK bck, QCD bck and  (Z  ll+1jet) N = 2.79 ± 0.25

Aidan Robson Glasgow University 24/24 Summary W and Z cross-section measurements underpin the Tevatron high-p T physics programme Dedicated measurements are harnessing the high statistics datasets:  improving tau identification  testing higher-order calculations and PDFs and probing QCD  making precision measurements of SM parameters

Aidan Robson Glasgow University 25/24 Backup

Aidan Robson Glasgow University 26/24 longer Q 2 extrapolation smaller x (W James Stirling)

Aidan Robson Glasgow University 27/24 D0 Z rapidity

Aidan Robson Glasgow University 28/24 CDF d  /dy

Aidan Robson Glasgow University 29/24 Z p T p T (Z) / GeV d  /dp T / nb/GeV all y|y| > 2 pp  Z 0 X, Tevatron :√s=1.96TeV d  /dp T / pb/GeV CDF e+e+ e–e– Nadolsky et al:  global fits to HERA and fixed-target data suggest increased intrinsic p T carried by proton constituents, for interactions involving only a small fraction of proton’s momentum  they insert extra factor in differential cross-section ; p T (Z) broadened at high y

Aidan Robson Glasgow University 30/24 Z p T p T (Z) / GeV d  /dp T / nb/GeV d  /dp T / pb/GeV pp  Z 0 X  e + e – X, LHC :√s=14TeV pp  W + X  e +, LHC :√s=14TeV |y e | < 2.5 p T e > 25 GeV |y e | < 2.5 p T e > 25 GeV E T > 25 GeV ATLAS e+e+ e–e–  LHC: beam energies ~7x higher than Tevatron  Probing new part of phase space  Tevatron forward detectors map on to LHC central detectors Problem! W/Z production is a benchmark

Aidan Robson Glasgow University 31/24 Z p T C-P Yuan

Aidan Robson Glasgow University 32/24 Indirect  W

Aidan Robson Glasgow University 33/24

Aidan Robson Glasgow University 34/24 without small-x effect: p T < 5GeV/c:  2 /dof = 0.8/1 p T <30GeV/c:  2 /dof = 11.1/11 including small-x effect: p T < 5GeV/c:  2 /dof = 5.7/1 p T <30GeV/c:  2 /dof = 31.9/11

Aidan Robson Glasgow University 35/24 W±W± pp l±l± d u u u u d