A first look at longitudinal weights using DC2 MC

1
A first look at longitudinal weights using DC2 MC

• Karina Loureiro, Stathes Paganis
• University of Wisconsin - Madison
• LAr Week
• September 8th, 2004

2
Outline

• Results from longitudinal weights applied to
  electrons (8.0.58.5.0)
• Test of longitudinal weights applied
  (preliminary) to 50 GeV photons (8.0.58.3.0)
• Photon longitudinal weights obtained using 50 GeV
  photons (8.0.58.3.0) (all ?s)
• Photon longitudinal weights applied to
  (8.6.08.6.0) (most likely in the future)
• Plans

3
On Electrons ...
4
Electron Longitudinal Weights (Review)

• Following previous strategy, we define the
  reconstructed energy (EM barrel and end-cap) as
• Erec ? (b W0EpresE1E2W3E3)
• We extract the free parameters in 100 bins of ? (
  ? lt 0.025)
• Apply constants at the CBNT ntuple level (i.e.,
  E1 eg_e1, etc.)

5
Electron Resolution
6
EMC Longitudinal Weights By Calorimeter Regions
7
First look Linearity for hlt2.5
Without weights
With new weights
8
Performance 10GeV electrons
Without weights
With new weights
9
Performance 20GeV electrons
Without weights
With new weights
10
Performance 50GeV electrons
Without weights
With new weights
11
Performance 100GeV electrons
Without weights
With new weights
12
Electrons on Photons ...
13
Electron Longitudinal Weights Applied to Photons

• Effects on non-converted photons
• 8.0.5 official simulation and digitization.
  Reconstruction with 8.3.0 (used for converted and
  non-converted alike)
• DC2 longitudinal weights for electrons applied
  after reconstruction
• Effects on converted photons
• Conversions identified using the conversion block
  and the egtrkmatchnt_TP for lack of Geant4 truth
  info
• Applied DC2 longitudinal weights for electrons
  after reconstruction

14
Electron Longitudinal Weights Applied to
Non-converted Photons
15
Electron Longitudinal Weights Applied to
Non-converted Photons by ? Regions
16
Electron Longitudinal Weights Applied to
Non-converted Photons (cont.)

• Observe that the end-cap photons (red) need
  almost no corrections.
• There are other sweet regions where it would
  seem that the photons are doing just fine.
• Overcorrection (perhaps) due to fitting of tail.

17
Electron Longitudinal Weights Applied to
Converted Photons (5x5 cluster size)
50 GeV photons
After e- corrections
Before e- corrections
18
Electron Longitudinal Weights Applied to
Converted Photons by ? regions
19
Electron Longitudinal Weights Applied to
Converted Photons by ? regions (cont.)
20
Electron Longitudinal Weights Applied to
Converted Photons (3x7 cluster size)
After e- (scale) correction
Before e- (scale) correction
21
Photons on Photons ...
22
Photon Longitudinal Weights Applied to
Non-converted Photons
23
Photon Longitudinal Weights Applied to
Non-converted Photons
24
8.6.0 all the way

• Events generated, simulated and reconstructed
  using offline release 8.6.0 with electron
  longitudinal weights turned on
• Electron constants applied only to identified
  conversions
• Obtain new photon weights (non-converted) with
  electron based calibration formula

25
Results Plans

• Photon energies need to be corrected with photons
• Clustering choice has a strong effect on photon
  energy reconstruction (as expected)
• Need a range of photon energies to obtain the
  parameters to appropriately do the fit (in the
  works)
• Must integrate MCTruth information (a.k.a.
  Geant4) in order to tune the conversion block
• Will apply two different sets of constants for
  converted and non-converted photons (they will be
  integrated into Athena once tested)
• We might need a different way of calculating the
  corrected photon energies
• Will double check with results from the test beam
  runs

26
Motivation

• At the core of many analyses there are
  electrons and photons. The understanding of the
  calorimeter response and their correct weighting
  are paramount to the success of the analyses.