Talk for Collaboration Meeting

1

Branching Fraction and Direct CP Violation in
B0??K 0 and B ???h ?
Alfio Lazzaro Dipartimento di Fisica and INFN,
Milan on behalf of the BaBar Collaboration America
n Physical Society 2003, Philadelphia
2
Overview

• Physics Motivation
• Analysis Strategy Samples, Event Selection, ML
  fits
• Results on Branching Fractions and Charge
  asymmetries
• Conclusion

3
Physics Motivation

• Measurements of Branching Fractions allow
  constraints on phenomenological models
• Charge asymmetries (Ach) expected large (gt20) in
  B?h K (Soni, PRL 43, 242 (1979)) and B?h p
  (Gronau, PRL 79, 4333 (1997))

Previous Measured Branching Fractions (?10?6)
Need precise measurements

4
Data Samples and Sub-channels
We have used 81.9 fb-1 on-resonance (88.9
millions of BB pairs)
_
We considered 3 channels

Reconstructed in
(23)
(39)
5
Event selection

• Beam energy to constraint B mass and energy

• Event shapes to distinguish B events from
  continuum udsc
• Fisher discriminant combining 4 variables
• direction of B thrust axis wrt beam direction
• direction of B momentum wrt beam direction
• 2 Legendre moments of the energy flow of rest of
  event wrt B thrust axis
• cos (qthrust) lt 0.9

• Loose cuts based on shapes variables, PID and
  kinematical quantities of secondary daughters,
  Cherenkov angle cut for h? selection.
• We prepare the input to Maximum Likelihood Fit

6
Observables

• Two types of background
• qq (continuum ) dominant
• bb (non continuum ) 1 2
• Observables ?E, mES , Fisher, h mass

_
_
We use Unbinned Extended Maximum Likelihood (ML)
Fit to extract signal yields and charge
asymmetries
7
Results for hK0 and hh?
Preliminary
First observation
8
Negative logLikelihood

Dashed blue line Dotted red line Solid black
line
9
mES and DE projections

Shaded histograms represent h ? gg
10
Systematics on BF and Ach

• Sources
  Error on yields
• Systematics due to ML fit efficiency and bias
• Daughter Branching fractions
• Reconstruction efficiencies
  (h, p0, KS,
  charged tracks)
• Other sources BB Number, MC statistics,
• PID,

?0.3 - ?0.5 ?0.03 - ?0.4 ?0.3 - ?2.0 ?0.03 - ?0.6
Total systematic error on BF
7.1 (hh )
7.7 (hK 0)
Charge Asymmetries Systematic on charge
dependence bias
1.1
11
h?K0, h?K?

• Branching Fraction for B?h?K larger than
  initially expected by theory
• Tree diagrams are Cabibbo suppressed
• Interference between two penguin diagrams and the
  known h h ? mixing angle conspire to
  greatly enhance B?h ?K and suppress B?h K
  (Lipkin, Phys. Lett. B 254, 247 (1991))

12
Results for h?K0 and h?K?

BF
Preliminary
Ach
See also talk by Fred Blanc, Session T11
13
mES and DE Projections

Shaded histograms represent h? ? hpp
14
Conclusion
We have presented the following preliminary
results

First observation
Suppressed B?h K Enhanced B?h ?K
2.5s
15
Altre Analisi in Corso
Bgt h'?

• BAD-462 Supporting Document
• BAD-605 Physics Note

s
g
?
_
_
t
s
_
b
_
d
??
B0
W
d
d
Risultati di CLEO
16
Bgt h'?

• Analisi finita da tempo
• Yields negativi
• Quindi abbiamo fatto una analisi
• di tipo Bayesiano
• Si sta ancora discutendo con
• i referee (Bob Cahn) ed altri
• esperti sul modo piu appropriato
• di presentare i risultati

17
Bgt hKL

• Supporting document BAD-591
  
• Scopo Misura di BR e Time-Dependent
  CP-Violating Asymmetries ( analogo a quanto
  abbiamo fatto con ???S)
• Attualmente usiamo il sotto-decadimento ??? ???
  ma in un prossimo futuro verra aggiunto ???
  ??
• Statistica usata RUN 1 RUN 2

18
Bgt hKL

• Funzioni di likelihood per la identificazione
  delle KL (Antimo)
• Constraint su MB e sulle direzioni per la
  ricostruzione cinematica
• Analisi ML con le variabili Mse , M?? , M?
  , Fisher
• Forte presenza di fondo comporta tagli su cos
  qT ? bassa ? ( 6 )

19
Bgt hKL

• Ora stiamo cercando di aumentare lefficienza di
  ricostruzione sostituendo alcuni tagli
  e il discriminante di Fisher con una rete
  neurale (NN). Uscita della rete usata come
  taglio per input ML

DE , M?? , M?

• Analisi ML utilizzando le variabili

( NN , ?? )

• Analisi cut count nel piano

• Control Sample ee- ? ?? ( ? ? ?S ?L )

20
Bgt hKS

• Stiamo misurando il BR di questo canale nel
  sottodecadimento
• h?? h p p e h? 3 p KS? ? ?-

• Stiamo aggiornando lanalisi a tutta la
  statistica (Run1 Run2)

• Useremo questi dati nella prossima analisi
  CP-time dependent

• Supporting Document BAD-488

21
Bgt h?? , h?h , h?h? , hh

• Stiamo misurando il BR di B???? nei
  sottodecadimenti
• con h???? e h?? h p p (h? ??)

• Per i canali hh, hh?, h?h ? stiamo usando le
  varie combinazioni dei sottodecadimenti con h???
  e h?3?, h???? e ?????? . In queste analisi
  siamo ancora agli inizi!