1. Progress in angular acceptance study
Yuehong Xie
University of Edinburgh
12 March 2009

2. Reminder of previous talk
Talk in bs meeting 22 January 2009
Proposed a event-by-event method to include effects on angular distributions from
detector geometry
kinematic cuts
particle efficiencies
Discussed the possibility to calibrate particle efficiencies using control channels

3. This talk How well does our model work with fully simulated data?
Predict Bs→J/yf average acceptance functions
Develop methods to calibrate particle efficiencies

4. Full simulation test
Test if the following factors can explain the observed angular acceptance
detector geometry
cuts on particle P and/or Pt
dependence of particle reconstruction efficiency on kinematics (P and Pt)
Use B+→J/yK+
cosqm in J/y rest frame w.r.t. B direction in this frame
cosqK in B rest frame w.r.t. (0,0,1)

5. Detector geometry
Only explicitly require charged tracks are inside Velo acceptance
Pt/Pz >0.02
Pt/Pz <0.4
Effects of other detector geometrical acceptance enter the momentum-dependent particle efficiency

6. Explicit cuts Use standard DC06 selection
Kaons: Pt > 1.3 GeV, P > 10 GeV
Muons: Pt > 500 MeV
J/y: Pt > 1 GeV

7. Particle efficiency e(P)
Obtain efficiency as a function of P for muons and kaons respectively from MC truth
muon
kaon
P (GeV)
P (GeV)

8. Test method Get reconstructed distribution of an angle
Follow steps in my previous talk to construct conditional pdf for each event
Use known theoretical distributions
Use per event angular acceptance function
Sum of conditional pdf should match the reconstructed distribution if the considered factors in the per event angular acceptance model are adequate

9. cosqm known theoretical distribution: 1- cos2qm cosqm reconstructed
sum of conditional pdf
cosqm

10. Average efficiency of cosqm
Relative difference
from MC truth
from conditional pdf
cosqm
cosqm

11. cosqK known theoretical distribution: flat cosqK reconstructed
sum of conditional pdf
cosqK

12. Average efficiency of cosqK
from MC truth
Relative difference
from conditional pdf
cosqK
cosqK

13. Part I summary
Our model describes the basic trend of the angular acceptance functions
has prediction power
Some fine tunings of the model are needed to reduce the ~10% relative difference
e.g. consider particle efficiency as a function of P and Pt

14. Average acceptance in Bs →J/yf
~ 20% variation
cosqtr
cosy
ftr

15. Effects of Pt/Pz>0.02
~ 5%
relative variation
ftr
cosqtr
cosy

16. Effects of Pt/Pz<0.4
~ 5%
relative variation
cosqtr
cosy
ftr

17. Effects of particle efficiencies
< 5%
relative variation
cosqtr
cosy
ftr

18. Effects of Kinematic cuts
~20%
relative variation
cosqtr
cosy
ftr

19. Part II summary
Shapes of angular acceptance functions in Bs→J/yf roadmap document are confirmed
Decreasing order of importance
Kinematic cuts
Velo geometry
Particle efficiencies

20. Obtain particle efficiency e(P) from B+→J/yK+
Assume an initial form of e(P)
The P distribution of a certain particle can be predicted and compared with the reconstructed P distribution
If the predicted and reconstructed P distributions don’t match, update e(P)
Iterate until a perfect match

21. Iteration 0: start with em(P)=1
reconstructed
predicted with em(P)=1
updated eff.
muon P distributions in GeV
em(P) for muons

22. Part III summary
Have found a simple way to check consistency of assumed e(P) with data
Aim to obtain e(P)
w/o using MC truth
w/o any assumption of B momentum spectrum
A few steps away from success
Still have some technical problems to get e(P) converged to the right form
Non-trivial to deal with background