BhcK : Branching Fraction Studies

1
B?hcK Branching Fraction Studies
BaBar Week, La Biodola, Elba,Italy, 19-23 May,
2003

• Stefania Ricciardi (RHUL)
• for the hc group

hc group N.Barlow, P.Clark, G.Hamel de
Monchenault, F.Jackson, W.Kozanecki , C.Marker ,
S.R.
2
Outline

• Main differences wrt Moriond2002
• Data fits and Yields
• Toy MC studies
• Efficiency
• Systematic errors
• Plans

3
Main Differences wrt preliminary conference paper
(Moriond 2002)

• Data sample (Run1Run2)
• Selection (see September 2002 CM)
• New Skim
• Re-optimization of cuts
• Loose mX,mES cuts
• Events weighted to account for efficiency
  variations across Dalitz plot (shown at December
  2002 CM)
• ML fit in mX-mES plane

4
Phase Space Efficiency in ?C ? KK? Modes

Basic ideas more details in F.Jacksons talk
at Dalitz Workshop, SLAC, December 2002

• Signal MC contains
• non-resonant ?C ? KK? only.
• ?? ? ?C ? KK?
• indicated large ?C ? KK contribution.
• Now seen also in our exclusive B decay
• sample

L 80/fb
Need to account for efficiency variations across
the Dalitz Plot.
5
Alternative representation of D.P.
Dalitz Plot

• ?C ? K1K2? Dalitz plot phase space m2(K1?) vs.
  m2(K2?)
• Alternative representation helicity angle vs.
  mass of any 2-body pair from KK?
• Convenient rectangular Dalitz plot to
  parametrize the efficiency.

m2(K2?)
m2(K1?)
Phase Space Plot
cos?H(K1?)
m(K1?)
6
Efficiency Parametrization
mass slices
1 2 3 4 5 6 7 8 9 10

• Semi-continuous efficiency parametrization based
  on Legendre polynomials. (continuous in cos?H,
  binned in mass)
• Each event in data weighted with efficiency ratio
  (average/local) in corresponding D.P. position
• Efficiency variations not very pronounced.
• ?Impact on B.F. not great.

cos?H(K1?)
m(K1?)
Efficiency
7
2D Fit Description

• 2D ML Fit in mES and mX to subtract
  combinatorial and peaking
• PDF components
• hc Signal G1 in mES, BW ? G2 in mX
• J/y Signal G1 in mES, G2 in mX
• Combinatorial Argus in mES, linear in mX
• Peaking G1 in mES, linear in mX
• Free Parameters (10)
• Number of events of different components (4)
• s(G2), s(G1), slopes mX of background (2), Argus
  shape
• mean of G1 (1)
• Fixed Parameters
• 33.8 MeV (BaBar gg), J/y and hc mass (PDG),
  Argus endpoint

8
B0 ? hc Ks0 (hc ? Ks0Kp)
Combinatorial Peaking J/y
Number of hc events 78.8 12.7
9
B ? hc K hc ? Ks0Kp
Before D veto
10
hc ? KKp0
B
B
B N (hc) 134.3 17.4
B0
B0
B0 N (hc) 39.7 10.1
11
hc ? 4K
B
B N (hc) 26.1 8.3
B0
B0 N (hc) 3.9 3.5
12
hc ? ff
B
B N (hc) 18.9 4.9
B0
B0 N (hc) 2.9 1.7
13
Toy MC studies
Performed for KsKp and KKp0 modes so far Based on
700 fits/mode
No significant bias observed in Number of EtaC
14
KKp0K Toy MCPulls
15
Efficiency

• PID-killed, g-killed SP4
• Efficiency is further corrected for
• tracking (standard recipe)
• Ks (standard recipe)
• etac width (33.3 MeV gg?hc / 27.0 MeV SP4
  generated)
• lower cutoff in etac MC generation , m(hc) -150
  MeV ( lt 1)
• K PID B only ( 4.7)
• Various HV periods reweighted with data lumi

16
Efficiency systematic error
Cuts Estimated by DATA (control sample) vs MC
comparison
17
KID correction and systematic
Kfast selector NNVeryLoose
hc signal MC
D signal MC
Pid-killed MC underestimates our efficiency. We
correct for it and assign half of the
correction (4.7/2) as additional systematic.
18
Efficiency comparison in q bins
hc MC signal
D MC signal
Good agreement in bins of polar angle
19
Efficiency Summary of systematics (B)
Corrections
Cut Variations
20
Efficiency Summary of systematics (B0)
Corrections
Cut Variations
21
Yield Systematic error

• Components considered so far
• Toy bias (negligible)
• RF model (2D fit with RF fixed to MC double
  gaussian)
• hc width (1 sigma BaBar )
• hc mass (1 sigma PDG)
• Background model (mX , 0-2nd order polynomial)
• F. Dalitz weights (half-deviation from fit
  without weights)

22
Signal MC MES fit
Sizeable fraction (5-10) of self-combinatorial se
en in all modes
KKp0K
We have to correct for it. How? Efficiency
studies on full MC require high statistics. Plan
to mix SP4 signal with appropriate fraction of
Toy combinatorial and peaking to measure signal
extraction efficiency in realistic conditions.
23
Plans

• mES model introduce correction understand
  systematics
• Implement fit improvements simultaneous fit B0
  and B samples to determine common fraction of
  J/y , s(mES) , s(mX)
• Toy complete studies (4K,ff)
• Fit systematics complete for all modes
• Efficiency systematic errors
• - DE, f mass cut
• 4K separate non-resonant from resonant (ff)
  component
• Update BAD 579

Aiming to a conference paper for LP2003 followed
by a journal publication
24
(No Transcript)
25
Results

NOW (10-6) Moriond2002

stat syst
stat syst BF(B ? hc K) BF(hc? K0Kp)
42.5 3.4 3.1 52.8 7.9
7.3 BF(B0 ? hc K0) BF(hc? K0Kp) 31.6
5.1 2.9 36.8 11.6 6.0 BF(B ? hc
K) BF(hc? KKp) 11.3 1.5 0.9
15.5 3.6 2.5 BF(B0 ? hc K0) BF(hc? KKp)
9.1 2.3 0.8 11.3 5.1
2.4 BF(B ? hc K) BF(hc? 4K) 1.9
0.6 0.2 lt5.6 BF(B0 ? hc K0) BF(hc?
4K) 0.8 0.7 0.08
lt2.3 BF(B ? hc K) BF(hc? ff)
1.2 0.3 0.1 - BF(B0 ?hc K0)
BF(hc? ff) 0.5 0.3 0.05
-
Very Preliminary