Racha Cheaib Wednesday, January 30, 2013€¦ · 1/30/13 Racha Cheaib 5 € ε i = N i rec N i gen...

20
Racha Cheaib Wednesday, January 30, 2013

Transcript of Racha Cheaib Wednesday, January 30, 2013€¦ · 1/30/13 Racha Cheaib 5 € ε i = N i rec N i gen...

Racha Cheaib Wednesday, January 30, 2013

1/30/13 2 Racha Cheaib, McGill University

BK+νν BK0νν

BK*+νν BK*0νν

1/30/13 Racha Cheaib, McGill University 3

  Scan the mass spectrum of the recoiling neutrinos in BK(*)νν and look for bumps: new physics.

  Apply the same signal selection, background estimation, systematic errors, etc… ◦  Additional systematic errors will be introduced due to the signal

efficiency calculation.

ΒK(*)X Xνν

16/10/12 Racha Cheaib, McGill University 4

Size of sliding window was chosen to be +/-4σ, moving through the spectrum with 2σ steps.

σJ/ψ

ψ(2S)

Using signal MC.

B→K (*)νν

ψ(2S) J/ψ

o Determine the difference between the neutrino recoil mass and the neutrino truth mass. o Fit the profile with a Gaussian at various points of the 2D histogram and calculate the sigma. o Determine the function σ(m), as a function of the recoil mass.

1/30/13 Racha Cheaib 5

ε i =Ni

rec

Nigen

where Nrec is the number of events with a neutrino mass reconstructed in a specific window i and Ngen is the number of generated events with neutrino sum mass within the same window i.

(×10−2)

BK+νν BK+νν

1/30/13 Racha Cheaib 6

B0 →K*0νν

B+ →K*+νν For K* modes, a factor of 2 should be accounted for the signal efficiency.

1/30/13 Racha Cheaib, McGill University 7

• Use the J/ψ and ψ(2S) signal Monte Carlo. • Calculate the signal efficiency using this “sliding” window method. • Compare it to the signal efficiency from the actual J/ψνν signal MC. • Assign the difference as a systematic uncertainty on the K(*)νν efficiency.

Signal efficiency from the J/ψand ψ(2S) invisible analysis

Apply sliding window to the J/ψ and ψ(2S) signal Monte Carlo.

Sliding window results with K(*)νν MC at J/ψ and ψ(2S) masses.

1/30/13 Racha Cheaib, McGill University 8

• Use the J/ψ and ψ(2S) signal Monte Carlo. • Calculate the signal efficiency using this “sliding” window method. • Compare it to the signal efficiency from the actual J/ψνν signal MC. • Assign the difference as a systematic uncertainty on the K(*)νν efficiency.

Signal efficiency from the J/ψand ψ(2S) invisible analysis

Apply sliding window to the J/ψ and ψ(2S) signal Monte Carlo.

Sliding window results with K(*)νν MC at J/ψ and ψ(2S) masses.

Factor of 2 due to the helicity of the K*.

  Combinatorial background is estimated using the sideband data within a specific mass window scaled by the cumulative Ratio.

  Peaking Background estimated using B+B- or B0B0 Monte Carlo, which is corrected to match the peaking data.

  This is done on a bin by bin basis, where every bin corresponds to a 4 σ window.

1/30/13 Racha Cheaib 9

1/30/13 Racha Cheaib 10

Bi =Nobs

i − Nbkgi

NB B × ε

i

1/30/13 Racha Cheaib 11

Bi =Nobs

i − Nbkgi

NB B × ε

i

1/30/13 Racha Cheaib 12

Bi =Nobs

i − Nbkgi

NB B × ε

i

0 <M<1.461.31<M<1.75

1.85<M<2.142.21<M<2.44

2.49<M<2.682.73<M<2.89

2.93<M<3.073.11<M<3.24

3.27<M<3.393.42<M<3.53

3.56<M<3.663.68<M<3.78

3.80<M<3.903.92<M<4.00

4.03<M<4.114.13<M<4.21

4.23<M<4.304.32<M<4.39

4.41<M<4.484.49<M<4.56

4.58<M<4.64

-5 1

Cen

tral

Bra

nchi

ng F

ract

ion:

-2

0

2

4

6

8

i i A X, X0s KACentral Branching Fraction: B

1/30/13 Racha Cheaib 13

Bi =Nobs

i − Nbkgi

NB B × ε

i

0 <M<1.441.34<M<1.86

1.97<M<2.322.40<M<2.68

2.75<M<2.983.04<M<3.24

3.29<M<3.473.52<M<3.68

3.72<M<3.873.91<M<4.05

4.08<M<4.214.24<M<4.37

4.40<M<4.514.54<M<4.65

-5 1

Cen

tral

Bra

nchi

ng F

ract

ion:

-2

0

2

4

6

i i A X, X*+ KACentral Branching Fraction: B

1/30/13 Racha Cheaib 14

Bi =Nobs

i − Nbkgi

NB B × ε

i

0 <M<1.441.34<M<1.86

1.97<M<2.322.40<M<2.68

2.75<M<2.983.04<M<3.24

3.29<M<3.473.52<M<3.68

3.72<M<3.873.91<M<4.05

4.08<M<4.214.24<M<4.37

4.40<M<4.514.54<M<4.65

-5 1

Cen

tral

Bra

nchi

ng F

ract

ion:

-6

-4

-2

0

2

4

6

8i i A X, X*0 KACentral Branching Fraction: B

  Fit the data with a crystal ball function and generate a set of toy Monte Carlo samples.

  Test the sliding window method on these samples.

1/30/13 Racha Cheaib 15

1/30/13 Racha Cheaib 16

  Inject signal events (by adding a Gaussian) at a certain mass and test whether or not a signal is observed.

1/30/13 Racha Cheaib 17

1/30/13 Racha Cheaib 18

6 events at 2 GeV

30 events at 2 GeV

1/30/13 Racha Cheaib 19

6 events at 3 GeV

30 events at 3 GeV

1/30/13 Racha Cheaib, McGill University 20

• Sliding window is documented in my BAD # 2450. • Next project is the search for rare B-Λp-νν and BKττ.