Update on Combinatorial Background for Partially Reconstructed Modes

1
Update on Combinatorial Background for Partially
Reconstructed Modes

• Amanda Deisher
• B Mixing and Lifetimes
• May 30, 2006

2
Outline

• Mass Distributions
• Mass shape of WS in partially reconstructed mass
  regions
• Possible physics backgrounds in WS
• Lifetime Distributions
• RS and WS in upper sideband
• Mass dependence of RS
• Mass dependence of WS
• Conclusions

3
Ingredients

• Data Sample B0 ?D- K ? -? - ?
• xbhd0h(0i)
• Wrong Sign (WS)
  D- K ? -? -? -
• Metric for Agreement
• Binned ?2 test
• Unbinned Kolmogorov test
• Lc veto Remove if 2.26 lt m(Lc ) lt 2.31 GeV/c2
• D veto Remove if m(Kpp) - m(Kp) lt 0.18 GeV/c2
• Lifetime

Definition Jeff Miles has been using
The definition Ive been using for this analysis
4
Mass Distributions
Analysis cuts only
Lc veto D veto

• DR lt 1.5
• p PT gt 1.2
• Lxy(B)/sLxy(B) gt 11.
• Lxy(B?D) gt -0.0300

• d0B lt 0.0110
• c2xy(B) lt 15.0
• c2xy(D) lt 15.0
• candPT gt 5.5

5
Mass Shape of WS in PR region

• Does the WS distribution in the partially
  reconstructed region agree with the RS
  combinatorial background extrapolation?
• Procedure
• Fit exponential flat background to RS upper
  sideband
• Extend curve to PR region compare

• Results depend on sideband definition
• Varied fit range (multiple curves)
  
• (5.6,6.38) ? (5.8,6.38)
• Note shape of combinatorial will be constrained
  with full mass fit (see CCKMP)

RSWS
6
WS Single B Physics Background

• Look at B ?DX and B0 ?DX Monte Carlo

25000
RS recod as D ? -
35
WS recod as D ?
Physics background in WS small and below region
of interest
7
WS and RS Agreement in Upper Sideband
5.4 lt m lt 5.7 GeV/c2 RS , WS
ct (cm)
6.0 lt m lt 6.3 GeV/c2 RS , WS
5.7 lt m lt 6.0 GeV/c2 RS , WS
ct (cm)
ct (cm)
8
WS and RS Agreement in Upper Sideband
?2 () ?2mPDG () KS () KSmPDG ()
5.4 lt m lt 5.7 GeV/c2 62.6 24.9 5.79 0.983
5.7lt m lt 6.0 GeV/c2 20.5 23.8 43.2 48.8
6.0 lt m lt 6.3 GeV/c2 22.1 52.0 19.3 12.9
Mass structure in (5.5,5.6) region not
understood. Try looking at agreement without
this contribution.
(5.4 , 5.5) ? (5.6, 5.7) 12.2 2.92

• WS and RS agree very well in the upper sideband,
  independent of ct definition

9
Mass Dependence of RS ct
Normalized to unit area
ct (cm)
10
Mass Dependence of RS ct
?2 () ?2mPDG () KS () KSmPDG ()
( 5.4 , 5.7 )?( 5.7 , 6.0 ) 0.132 7E-11 4E-4 3E-16
( 5.4 , 5.7 )?( 6.0 , 6.3 ) 5.9 6E-18 2E-3 1E-30
( 5.7 , 6.0 )?( 6.0 , 6.3 ) 89.7 30.8 14.1 8E-3

• Removing region (5.5,5.6) with unexplained mass
  structure

(5.4 , 5.5) ? (5.6, 5.7)?( 5.7 , 6.0) 6E-5 3E-14
(5.4 , 5.5) ? (5.6, 5.7)?( 6.0 , 6.3) 1E-4 7E-27

• What if we just start above 5.6?

(5.6,5.85)?(5.85,6.1) 9.1 6E-4
(5.6,5.85)?(6.1,6.35) 8.1 4E-10
(5.85,6.1 )?(6.1,6.35) 77.4 4.1

• Using the visible mass ? some mass dependence,
  significantly less than when using
  PDG mass

11
Mass Dependence of WS ct
Normalized to unit area
ct (cm)
12
Mass Dependence of WS ct
?2 () ?2mPDG () KS () KSmPDG ()
( 5.4 , 5.7 )?( 5.7 , 6.0 ) 65.7 8E-2 32.5 5E-5
( 5.4 , 5.7 )?( 6.0 , 6.3 ) 14.7 2E-13 6E-2 3E-23
( 5.7 , 6.0 )?( 6.0 , 6.3 ) 18.2 0.3 2.89 2E-6

• Removing region (5.5,5.6) with unexplained mass
  structure

(5.4 , 5.5) ? (5.6, 5.7)?( 5.7 , 6.0) 1.6 1E-5
(5.4 , 5.5) ? (5.6, 5.7)?( 6.0 , 6.3) 2E-4 3E-23

• What if we just start above 5.6?

(5.6 , 5.85)?(5.85 , 6.1) 4.1 6E-4
(5.6 , 5.85)?(6.1 , 6.35) 3E-3 2E-16
(5.85, 6.1 )?(6.1 , 6.35) 7.2 8E-3

• Hmm, the first and last mass bins dont agree at
  all

13
Mass Dependence of WS ct

• Looking WS ct in the signal region (5.1,5.4)

?2 () ?2mPDG () KS () KSmPDG ()
( 5.1 , 5.4 )?( 5.4 , 5.7 ) 0.3 9E-17 2E-4 3E-25
( 5.1 , 5.4 )?( 5.7 , 6.0 ) 2E-6 0 2E-8 0
( 5.1 , 5.4 )?( 6.0 , 6.3 ) 3E-10 0 5E-11 0

• Looking WS ct in the partially reconstructed
  region (4.8,5.1)

( 4.8 , 5.1 )?( 5.1 , 5.4 ) 2E-6 0 5E-12 0
( 4.8 , 5.1 )?( 5.4 , 5.7 ) 0 0 4E-20 0
( 4.8 , 5.1 )?( 5.7 , 6.0 ) 0 0 8E-20 0
( 4.8 , 5.1 )?( 6.0 , 6.3 ) 0 0 3E-26 0

• Definite mass dependence. Still need to work on
  optimizing cuts in this region.

14
Conclusions

• Full mass fit necessary to evaluate level of
  agreement between WS and current combinatorial
  background model
• Initial studies show little WS single B physics
  background in PR or signal region
• RS and WS ct distributions agree in the upper
  sideband and show same mass dependence
• Using visible mass in ct calculation greatly
  reduces mass dependence
• WS seems to track mass dependence of RS in the
  region where we can test it. While we will use
  the wrong sign as a proxy, we need to look at
  reducing the systematics by optimizing S/B in the
  PR region.

15
To-Do List

1. Perform full mass fit with templates for PR and
   then compare the shape of the extrapolated RS
   combinatorial bkgd and the WS combinatorial bkgd
   in the PR region
2. Evaluate systematic uncertainty on the B lifetime
   introduced by the uncertainties in the mass and
   lifetime dependence of the combinatorial bkgd
3. Re-optimize cuts to minimize this uncertainty