QE Analysis Update Chris Smith

1
QE Analysis UpdateChris Smith

• Contents
• Introduction
• Event Selection
• Energy Reconstruction
• Results
• Summary

2
Introduction

• Analysis of QE events can provide a valuable
  cross-check of results from full analysis
• Different systematics to full analysis
• Independent of energy calibration
• Lower statistics
• Using reconstructed MDC files to
• Evaluate reco performance wrt QE events
• Find PDFs for likelihood event selection
• Perform oscillation parameter fits to obtain
  confidence limits

3
Event Selection

• Looked at a number of variables to discriminate
  QE, CC, NC
• Initial cut demand 1 reconstructed event and 1
  reconstructed track
• Efficiencies QE 88, CC 70, NC 15

4
Event Selection previously...

• Previously, for QE selection
• Variables ave dEdx, chg ratio, (prange-pcurv)/p
  for NC/CC discrim
• Cut on Etrack/Etotal for QE selection

5
Event Selection previously...

• This resulted in the following energy spectrum
• Selection efficiencies QE 40 CC 20 NC 5
• Sample fractions QE 30 CC 65 NC 5

6
Event Selection

• Want to remove dependency on reconstructed shower
  energy
• Look at hit fraction in track
• NC-like CC events typically have higher Y and
  QE-like have lower Y as expected

7
Event Selection

• Use Track Hit Fraction, Charge Ratio, Average
  dEdx distributions in a likelihood calculation
• Construct PID Parameters
• PIDParQECC -log(ProbCC)1/2 -
  -log(ProbQE)1/2
• PIDParQENC -log(ProbNC)1/2 - -log(ProbQE)1/2

Cut on PIDPars in 2D to select QE-like events
8
Event Selection
9
Event Selection

• True neutrino energy spectrum for selected events
• Selection efficiency by event type

10
Energy Determination muon

• (True-Reco)/True when fitting only to central
  peak
• Curvature mean -0.027 sigma 0.070
• Range mean 0.016 sigma 0.056

11
Energy Determination neutrino

• Estimate neutrino energy from muon momentum and
  angle using relation
• E? (mN E? - m?2/2) /(mN - E? P?cos??)
• For true QE events, mean 2.2 - related to
  momentum offset?

12
Fits

• Carried out single experiment for 7.4, 16e20 POT
  and sin22theta1, dm20.002
• Including systematic errors at 2 level for
  normalisation and energy shift

13
Resolution Vs Statistics

• By changing PID parameter which discriminates
  between QE and CC, relative contribution of CC
  events can be varied

14
Resolution Vs Statistics

• Find minimum contour size by playing off energy
  resolution vs stats
• A cut of -0.2 on the PID parameter gives you the
  best measurement
• Analysis does not currently include systematics
  related to the DIS event fraction present how
  much will this change the contours?

15
Summary

• Reconstruction finds tracks in 90 of QE snarls
• Muon energy reco gt neutrino energy reco good to
  2
• Event Selection
• QE selection efficiency 70 level
• QE sample purity 40
• NC background at a very low level lt1
• CC DIS dominates energy spectrum
• Energy resolution vs statistics vs systematics
  needs to be understood
• Closer links with NearDet Physics group after
  this meeting will help to accomplish this