Back to my homepage

Jet energy corrections


29th Mar 2001 - talk on D0 RunII Qcd Meeting (Feb 15, 2001)
    - studies made on ntuples root_QCDpt20.0_mb*.root - QCD sample with ptmin=20GeV and ptmin=160GeV, 9000 events with 1.1 minimum bias in average, 9000 events with 4.0 minimum bias in average

Dependence on calorimeter tower energy cut for Cone R=0.7
  - Gaussian fit on fraction of energies on calorimeter and particle level for all bins in pT and pseudorapidity (see plots bin_dist-mb1.1-C7CB.ps, bin_dist-mb4.0-C7CB.ps)

  - low luminosity (minimum bias 1.1 on average)
low pT sample - as could be expected, algorithm with positive energy cut (C7CB) gives smallest energy corrections but there is almost no difference in energy resolution neither in central region, neither in forward region (ps)
high pT sample - the same situation as in low pT case (ps)

  - high luminosity (minimum bias 4.0 on average)
low pT sample - option with no energy cut gives the worst resolution especially in central region, (ps)

Dependence on splitting merging factor for Cone R=0.7
  - no difference can be observed within given statistics (ps)

Difference between Cone R=0.7 and R=0.5
  - energy corrections for larger cone are smaller
  - Cone R=0.5 gives a little bit better energy resolution, especially in forward region (mb1.1, mb4.0)
  - the differencies are smeared out with increasing pT (ps)

Dependence on calorimeter tower energy cut for kT D=0.7
  - the option with no energy cut gives the more energetic jets on calotimeter level, the jets are even more energetic than jets made with the option of positive energy cut, the difference is larger in forward region
  - again almost no difference in energy resolutions can be seen (pt20, mb1.1, pt20, mb4.0, pt160, mb1.1)

Comparison between Cone R=0.7 and KT R=0.7
  - Cone algorithm gives a little bit better resolution the KT (pt20, mb1.1, pt20, mb4.0, pt160, mb1.1)

Summary
  - option with no energy cut gives slightly worse resolution, especially in high luminosities
  - no difference between various splitting marging factors
  - smaller cone gives a little bit better resolution, especially in forward region
  - cone R=0.7 is better then KT R=0.7 in higher luminosities


Last update 29th Mar 2001 by AK.