May 2006 Hadronic Calibration Workshop Jet Session at Munich Jet Energy Calibration Using E/p of Single Charge Isolated Hadrons from Tau Decay J. Lu, D. M. Gingrich University of Alberta May 2006 Hadronic Calibration Workshop Jet Session at Munich 04/19/2006

Outline General Introduction: The single charged isolated hadrons from tau (W) decay: E/p=1. The precise measurement of their momentum (precision 0.5%) in the inner detector can be used to calibrate the energy measured in the calorimeters (precision 5-10%). Main issue is how to separate π+ and π+ π0 (ρ0) from tau decay. QCD background and Minimum bias events will affect E/p calibration. Data sample and event selections. Some results and outlook. 04/19/2006

Data sample TAUOLA BR (%) τ±→e±υeυτ 18.1 τ±→μ±υμυτ 17.6 τ±→π±υτ 11.1% Sample 1: Rome MC production: 4807 (sample A3), inclusive Z to tau tau tight filtered. 49K AOD /castor/cern.ch/grid/atlas/datafiles/rome/recov10/rome.004807.recov10.A3_Ztautau_tightfilter Sample 2: Rome MC production: 4124, 88.2 K AOD events (Z +jet, Z to tautau) /castor/cern.ch/grid/atlas/datafiles/rome/recov10NONT/rome.004124.recov10NONT.Sherpa_ZJ1tautau Sample 3: Run 4801 filtered, 50K (Z) pp →Z+ x and pp →Z+jet+X combined in the same luminosity Atlas release 10.0.1 TAUOLA BR (%) τ±→e±υeυτ 18.1 τ±→μ±υμυτ 17.6 τ±→π±υτ 11.1% τ±→K±υτ 0.7% τ±→ρ±υτ with ρ→ π± π0 24.2% τ±→a1±υτ with a1±→ π± π±/0 π-/+/0 13.0 τ±→K*±υτ 1.3 τ±→(mπ±+n π0) υτ 5.6 04/19/2006

Z to tautau, tau to single charge hadron event selection Trigger satisfy one of the following: Tau jet with ETmiss>30 GeV one lepton with PT>20 GeV, ETmiss>20 GeV ETmiss<1000 GeV Only one charge track in Tau Jet from AOD container PT>20 GeV, |η| < 2.5, ETmiss>10 GeV for electron trigger Tau likelihood greater than 4. ΔR between tau jet and electron >0.4, energy deposition in hadronic calorimeter >1GeV ΔR between tau jet and photon >0.4 Energy calibration factor in Rome 10.0.1 should be multiplied by a factor of 1.08 04/19/2006

E/P ratio without π0 separation Fit in the range 0.6-2.0 Mean: 1.079±0.008, σ: 0.172±0.008 Fit in the range 0.6-1.4 Mean: 1.010±0.010, σ: 0.093±0.012 1. Nucleon is a spin half particle, whose spin receives contribution from quark and orbital angular momentum. 2. HERMES studys nuclear spin structure by DIS, i.e deep inelastic scattering of electron or positron beam scattering from nuclear targets, which means exchange virtual photon has sufficient virtuallity Q2 to resolve the individual quark Q2>1 3. By using polarized beam, the interaction is ssensitive to the orientation of quark and nuclear spin, because quark can observe the virtual photon only if the quark spin is opposite the quark spin, because the hard interation conerve helicity. 4. Using this technique, it was found quark spin contribution Delta Sigma is small, leaving large room of contribution from gluon and orbital angular momentum Nucleon spin puzzle 6.(point to DVCS plot: two hard scattering vertex, the nuclear stays intact in the ground state) 5. The recent theretical development of GPDs (two quark correltion function) may help us Lq, while the DVCS can access GPDs. Large E/p tail, mean E/p depends on the fit region (Gaussian plus third-order polynomial fit) 04/19/2006

Calibration without π0 separation If in the first month LHC running we collect 100 fb-1, around 2000 tau jet events passing E/p event selections without π0 separation . The statistics uncertainty for E/p mean is around 0.01. Statistics is not enough for further π0 separation. It is possible to calibrate to E/p bias without π0 separation. E/p bias induced by π0 is (1±1)% by only fit in the narrow region around peak. E/p bias induced by QCD background and pile up is small, less than 1%. These conclusions will be cross-checked by CSC samples. 04/19/2006

Ratio of energy deposited in EM calo and hadronic calo ----- for π0 separation Same total entries Good separation of h± and h±+nπ0 for EM/Had<0.05. Try other variables to separate h± and h±+nπ0 for EM/Had>0.05 EM/Hadronic <0.02 <0.03 <0.05 <0.1 <0.2 Eff (e2, %) h+ 4.68 7.16 10.46 17.08 21.76 Eff(e1, %) h+ + nπ0 0.15 0.29 2.06 7.64 e1/e2 0.021 0.028 0.12 0.35 1. Nucleon is a spin half particle, whose spin receives contribution from quark and orbital angular momentum. 2. HERMES studys nuclear spin structure by DIS, i.e deep inelastic scattering of electron or positron beam scattering from nuclear targets, which means exchange virtual photon has sufficient virtuallity Q2 to resolve the individual quark Q2>1 3. By using polarized beam, the interaction is ssensitive to the orientation of quark and nuclear spin, because quark can observe the virtual photon only if the quark spin is opposite the quark spin, because the hard interation conerve helicity. 4. Using this technique, it was found quark spin contribution Delta Sigma is small, leaving large room of contribution from gluon and orbital angular momentum Nucleon spin puzzle 6.(point to DVCS plot: two hard scattering vertex, the nuclear stays intact in the ground state) 5. The recent theretical development of GPDs (two quark correltion function) may help us Lq, while the DVCS can access GPDs. 04/19/2006

Variables to separate τ→h±υ and τ→(h±+ nπ0)υ EM/Had >0.05 Width of energy deposition in strips >0.005 Fraction of energy deposited in ΔR (0.1-0.2) >0.08 EM radius of τ > 0.07 After this 3 variables cuts: efficiency h±: 12.0%, h± +nπ0: 4.7%, not strong separation Another variable: the number of strips in the strip layer of the EM calo with energy deposition above a threshold. But not in the “TauJetCollection” AOD container. 04/19/2006

E/P ratio with π0 separation EM/Had>0.05 and other 3 variable cuts Mean: 1.030±0.020, σ: 0.044±0.027 EM/Had<0.05 is applied Mean: 1.012±0.019, σ: 0.023±0.024 1. Nucleon is a spin half particle, whose spin receives contribution from quark and orbital angular momentum. 2. HERMES studys nuclear spin structure by DIS, i.e deep inelastic scattering of electron or positron beam scattering from nuclear targets, which means exchange virtual photon has sufficient virtuallity Q2 to resolve the individual quark Q2>1 3. By using polarized beam, the interaction is ssensitive to the orientation of quark and nuclear spin, because quark can observe the virtual photon only if the quark spin is opposite the quark spin, because the hard interation conerve helicity. 4. Using this technique, it was found quark spin contribution Delta Sigma is small, leaving large room of contribution from gluon and orbital angular momentum Nucleon spin puzzle 6.(point to DVCS plot: two hard scattering vertex, the nuclear stays intact in the ground state) 5. The recent theretical development of GPDs (two quark correltion function) may help us Lq, while the DVCS can access GPDs. Gaussian plus third-order polynomial fit 04/19/2006

Calibration with π0 separation Using ratio of energy deposition in EM calorimeter and hadronic calorimeter <0.05 can greatly reduce π0 contamination, the lower the cut, the better. In this case, the hadronic calorimeter can be calibrated. If in the first month LHC running we collect 1 fb-1, around 1000 tau jet events passing E/p event selections without π0 separation . The statistics uncertainty for E/p mean is around 0.005. The long tail from π0 contribution is highly reduced. E/p bias induced by QCD background, π0 and pile up is around (2±2)%. These conclusions will be cross-checked by CSC samples. A note using Rome sample: ATL-COM-PHYS-2006-023. Another ATLAS note with CSC sample will be produced. 04/19/2006