Trigger Bits

   Number of trigger bits, L1:  128                      1.6 kHz bandwidth
                           L2:  128 (See Note 1)         1.0 kHz
                           L3: approx 400 (See Note 2)  50-100 Hz

     Note 1: Each L2 trigger bit can be an "OR" of multiple L1 bits, so the
       effective number of bits is much larger
   
     Note 2: The L3 system does not allow direct "OR" combinations of L2 
       triggers.  Instead, each L2 trigger can feed many L3 bits providing the
       equivalent of an "OR" system
   

Simulation

 
    We use only Geant for mainstream simulation

    ttbar simulation time for Run IIb (with L0) takes about 
    3 minutes/event on a P4 3.2 GHz machine. Time in GEANT, Digitization and 
    Reconstruction steps is spent as follows:
 
      geant: ~ 2.5 min., digitization: ~10 sec and reconstruction: <~0.5 min.
   

Resolution

    Momentum resolution:  The momentum resolution from the TDR is 

         sigma_pt/pt = sqrt(0.015^2 + (0.0014 pT)^2)

    (with Pt in GeV/c).  We've measured this at one point, and the data
    is somewhat worse than this.
 

   Electron energy resolution

     The fractional energy resolution for electrons from Z -> e e at normal
     incidence on the CC [is_fiducial()] is 3.7%. This corresponds to an 
     average over some energy spectrum given by nature.  The spectrum spikes 
     at 46 GeV. As we stay in the CC, but go to eta = 0.7 , the fractional 
     energy resolution degrades to 4.8 %, despite the fact that the energy 
     spectrum now spikes at 52 GeV (which is higher than 46 GeV).

     The uncertainties on the resolutions are below 5 % (relative !).
 

Plots

(jpeg) (eps) Z->ee Loose EM ID efficiency vs eta, no track match requirement	(jpeg) (eps) Z->ee Loose EM ID effiicency vs eta, spatial track match required
(jpeg) (eps) Z->MuMu Loose muon ID efficiency vs. eta. No track match requirement	(jpeg) (eps) Data/MC muon ID efficiency ratio
(jpeg) (eps) Z->MuMu Tracking efficiency	(jpeg) (eps) Data/MC tracking efficiency ratio
(jpeg) (eps) Missing ET in Z->ee collider data events	(jpeg) (eps) Missing ET in Z->ee collider data events. Log scale
(jpeg) (eps) Tau ID efficiency and fake rate for 1-prong decays as a function of NN output value. The efficiencies are measured with respect to narrow jets having pT>15 GeV. The fake rate is measured relative to narrow jets in QCD events.	(jpeg) (eps) Tau ID efficiency and fake rate for 3-prong decays as a function of NN output value. The efficiencies are measured with respect to narrow jets having pT>15 GeV. The fake rate is measured relative to narrow jets in QCD events.
(jpeg) (eps) The b-tag probability (for b-flavored jets) at a function of jet ET for all jets within |eta|<2.4 for a neural network b-tagging algorithm. The red curve is for MC, and the green is for a background corrected data control sample. In analysis, the data curve is always applied, even for simulated events. This curve is shown for the loosest operating point used in analysis. The denominator for the efficiency is calorimeter jets with a minimal track requirement.	(jpeg) (eps) The b-tag probability (for b-flavored jets) at a function of jet eta for all jets within ET>15 GeV. The red curve is for MC, and the green is for a background corrected data control sample. In analysis, the data curve is always applied, even for simulated events. This curve is for the loosest operating point used in analysis.
(jpeg) (eps) The b-tag probability (for b-flavored jets) at a function of jet ET for all jets within |eta|<2.4 for a neural network b-tagging algorithm. The red curve is for MC, and the green is for a background corrected data control sample. In analysis, the data curve is always applied, even for simulated events. This curve is shown for the tightest operating point used in analysis	(jpeg) (eps) The b-tag probability (for b-flavored jets) at a function of jet eta for all jets within ET>15 GeV. The red curve is for MC, and the green is for a background corrected data control sample. In analysis, the data curve is always applied, even for simulated events. This curve is for the tightest operating point used in analysis.
(jpeg) (eps) The efficiency versus fake rate for the neural network (NN) and impact parameter (JLIP) based taggers for all jets ET>15 GeV and |eta|%lt;2.4. One sees clear improvement from the NN tagger.	(jpeg) (eps) The efficiency versus fake rate for the neural network (NN) and impact parameter (JLIP) based taggers for all jets ET>30 GeV and |eta|%lt;0.8. One sees clear improvement from the NN tagger.
Layer 0 Pictures