Transcript kmu3_br_feb_14_2005_dabrowski

Update Kmu3 Branching Ratio
measurement
Anne Dabrowski
Northwestern University
Collaboration Meeting
22nd February 2005
A. Dabrowski, February 22 2005
Strategy: Measure Kmu3 Br normalised to
pipi0
Test 2 Particle ID muon strategies:
1) Muon Veto as Muon ID
• Check muon veto status 1 or 2
• Timing association of 2ns for track between muon veto
and hodoscope time
2) LKR and HAC as Muon ID
●
●
Use the mip signal in calorimeters:
LKR < 1.5 GeV and HAC < 5 GeV for the cluster
associated to the tracks..
Requirement for signal and normalisation:
●
1 track and 1 pi0
●
Kinematic cuts using LKR and DCH
A. Dabrowski, February 22 2005
Data Sample
●
●
Compact 7.2 & Database pass 5 Min bias 2003 (15745,15746 and 15747)
–
Alignment
–
E-baseline correction
–
Bad burst
–
Alphas and betas
–
Projectivity and Blue Field
MC Sample:
–
–
–
Ginsberg correction to Kmu3 – Evelina Marinova finalized
correction that Mengkei started
DCH resolution from Eddy, and latest official updates
Michal’s low energy correction to MC (not in presented Dec 2004
numbers)
A. Dabrowski, February 22 2005
Simple selection wanted....
Common for Kmu3 and pipi0
●
●
Track Section (no extra tracks allowed):
–
1 track after excluding Ghost-tracks
–
Hodoscope time window (-17. 20. ns)
–
Track quality > 0.8
–
x,y vertex (-1.8,1.8) cm ,
–
Blue Field correction applied
CDA < 2.5 , Beta, alpha corrections from database
z vertex (-500,8000)cm
Pi0 Selection (extra gammas allowed for both)
–
Energy of gamma (3, 65) GeV
–
Separation between gammas > 10 cm
–
Time difference between gammas (-5., 5.) ns
–
Pi0 mass cuts at 3 sigma and depends on pi0 energy
●
For this talk I use Michals cut, and cut 3 sigma, In dec meeting, I used my old 2004
pi0 cut for the official numbers
–
Projectivity correction
–
Latest Energy scale by Michal
A. Dabrowski, February 22 2005
Difference between Kmu3 and Pipi0 selection
●
Kaon Mass (assuming pi) <0.475 or >0.515 GeV
●
Mom (10, 40) GeV
●
PT track (0.0, 0.2) GeV
●
Nu mass (-0.01, 0.01) GeV2
●
Dist between track & gammas > 10 cm
●
Energy pi0 < 40 GeV
●
COM pi0 < 0.24 GeV
●
COM Track < 0.23 GeV
●
Mass of mu pi0 < 0.445 GeV
●
Particle ID for muons (2 methods used)
●
Kaon Mass (0.475,0.515) GeV
●
Mom (10, 50) GeV
●
PT track < 0.215
●
Nu mass (-0.0025, 0.001) GeV2
●
Distance between track & gammas >
35 cm
●
PT pi0 < 0.220
●
E/P < 0.95
●
Use muon rejection only when the
muon veto in used in the Kmu3
analysis.
A. Dabrowski, February 22 2005
Method 1:
Muon ID using the Muon Veto
•
Muon ID efficiency calculated
using Kμ2 sample from min bias
run;
– check status 1 or 2 and 2 ns
between hod time and muon
veto time
•
Kinematic cuts Momentum (10,40)
•
Event Timing and Fiducial cuts as in
Kmu3 Br analysis
Efficiency between 0.997 and
0.998
IN MC 6.4m decay volume, particle
decay not simulated – Apply a
correction to mc acceptance
•
•
– Banana PT vs P cut (Luca)
– Mass ν2 (-0.02;0.01) GeV2
– See Michal Talk Torino
A. Dabrowski, February 22 2005
Method 2:
Muon ID signals using the LKR and HAC
• Cuts chosen
– LKR < 1.5 GeV and HAC
< 5 GeV
• Muon sample using Kμ2
events from min bias run.
• Kinematic cuts
– Momentum (10,40)
– Banana PT vs P cut
– Mass ν2 (-0.02;0.01) GeV2
– Muon Veto requested
• Event Timing and Fiducial
cuts as in Kmu3 Br analysis
A. Dabrowski, February 22 2005
Method 2:
Muon ID efficiency using the LKR and HAC
• Muon ID requirement:
– LKR (cluster<1.5 GeV)
and HAC (cluster<5.0
GeV)
– Muon ID is energy
dependent with max
~0.987
– Analysis done bin by
bin in momentum
Method 1
eff at
0.998
Corrected binomial errors
A. Dabrowski, February 22 2005
Method 2:
Pion mis-identification as muons using the LKR and
HAC
•
Pions can be to mis-identified as
muons
– Need a pion mis-identification
probability, and background
subtraction.
•
Sample used for calculating the
mis-identification probability
– Pions from my standard pipi0
selection, with the muon Veto
requirement.
– Plus a tighter Kaon mass cut
for this sample (0.485, 0.505
GeV).
– Event Timing and Fiducial cuts
as in Kmu3 Br analysis
Corrected binomial errors
A. Dabrowski, February 22 2005
●
Main difference between this and the dec
meeting:
–
1. the pipi0 acceptance is defined as just
those events for which the pi+ does not decay
.. So pipi0dk is a background to pipi0.
–
For all analysis, the Number events =
Ns(1+Ns/Nb) is used.
Comparison in result between 2
methods K+
No
No no
Backgrou
nd
Backgrou
nd
525248
532838
Muon
3186589
Veto
529125
Kmu3
#
Events
Data
Raw Acc
MC
Acc * Particle ID
(muon veto or
E/P < 0.95)
Backgrounds
55905
0.1051
0.1049 ±0.0002
Pipi0dk 0.0133+0.0007
Pipi0pi0dk 0.00111+3x10-5
3186911
pipi0
488334
0.1541
0.1532±0.0003
Ke3 0.00010+-6x10-6
Correction due to
pion decaying after
LKR (0.993)
536558
Kmu3
54623
0.1051
0.1018 ± 0.0002
Pipi0dk 0.0129+0.0003
Pipi0pi0dk 0.00112+3.3x10-5
LKR
HAC
3192325
324662
pipi0
497464
0.1541
0.1532 ±0.0003
Pipi0dk 0.013+0.0062
Kmu3 0.00329+7.6x10-5
A. Dabrowski, February 22 2005
Comparison in result between 2
methods K-
Muon
Veto
No
No no
Back’grnd
Back’grnd
291274
295517
Kmu3
#
Events
Data
Raw Acc
MC
Acc * Particle ID
(muon veto or
E/P < 0.95)
Backgrounds
33635
0.1053
0.1051 ±0.0002
Pipi0dk 0.0135+0.0007
Pipi0pi0dk 0.00111+3x10-5
1771466
1771633
pipi0
271027
0.1540
0.1530±0.0003
Ke3 9.28x10-5+-6x106
Correction due to pion
decaying after LKR
(0.993)
293394
LKR
HAC
297397
Kmu3
30331
0.1053
0.1020 ± 0.0002
Pipi0dk 0.0125+0.0007
Pipi0pi0dk 0.0011+3.3x10-5
1773866
1804720
pipi0
276147
0.1539
0.1530 ±0.0003
Pipi0dk 0.014+-0.0062
Kmu3 0.00330+7.6x10-5
A. Dabrowski, February 22 2005
Br Result
Br Ratio
Muon
Veto
LKR
HAC
•
•
Br(Kmu3)+
0.0346 ± 0.0003
Br(Kmu3)-
0.0345 ± 0.0003
Br(Kmu3)+
0.0350 ± 0.0002
Br(Kmu3)-
0.0350 ± 0.0002
The error statistical and includes:
– Data sample
– MC statistics
– Errors in particle ID efficiency
No systematic errors have been included
–
–
I must still propagate the errors due to the background
And the “not decay after lkr error”
A. Dabrowski, February 22 2005
My to do – and questions
My own pi0 mass cut as a function of energy (right now using Michals –
possible source for decrease of events in data? Have to check, will
check changes in data base / compact why events have decreased
since dec meeting)
The Br as a function of momentum
When kmu3 is a source of background – see the effect of the kmu3 Br
from PDG used in the analysis (higher order todo)
Vertex in data/mc problem
The factor to correct for particles not decaying after lkr in MC – need
to checked if a small correction should be applied to LKR/HAC case –
for the region between lkr and hac … for the pipi0dk and pipi0pi0dk
backgrounds. Right now not corrected.
Re-measure the correction for decay not simulated in MC as a check.
Check sensitivity to vertex and pt cuts
A. Dabrowski, February 22 2005
Pion ID efficiency E/P < 0.95
(common to both analysis methods)
• Pion ID efficiency
calculated using pipi0
sample from min bias
run.
• Kinematic cuts (as in my
selection)
– Muon veto requirement
to reject muons
– But have a tighter Kaon
mass cut for this sample
(0.485, 0.505 GeV).
• Event Timing and
Fiducial cuts as in Kmu3
Br analysis
Corrected binomial errors
A. Dabrowski, February 22 2005