0 fit update

18
0 0 fit update fit update P.Gauzzi

Upload: mahlah

Post on 07-Jan-2016

25 views

Category:

Documents


0 download

DESCRIPTION

 0  fit update. P.Gauzzi. Outline. Kaon Loop S ystematics on the fit parameters Fit with fixed VDM No Structure systematics fit with free VDM KL and NS fits with Adler zeros New version of NS model. KL. Correlation coefficients. M a 0 1.000 - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: 0  fit update

00 fit update fit update

P.Gauzzi

Page 2: 0  fit update

2

Outline

• Kaon Loop – Systematics on the fit parameters

– Fit with fixed VDM

• No Structure– systematics

– fit with free VDM

• KL and NS fits with Adler zeros

• New version of NS model

Page 3: 0  fit update

3

KLKL

Par. Fit uncert.

Ma0 (MeV) 982.5 1.5

gaK+K (GeV) 2.15 0.06

ga(GeV) 2.82 0.03

g a(GeV-1) 1.58 0.09(°) 222 12

Br(VDM) 0.92 0.40

R=Br( )/Br( 0) 1.70 0.04

Ra0=(ga0K+K─/ga0)2 0.58 0.03

2 157.6ndf 136

P(2) 9.9%

Ma0 1.000

ga0KK 0.931 1.000

ga0 0.584 0.550 1.000

0.277 0.236 0.156 1.000

Br(VDM) -0.252 -0.327 -0.513 0.130 1.000

R 0.004 -0.180 -0.111 0.008 -0.038 1.000

Correlation coefficients

Page 4: 0  fit update

4

KL Systematics KL Systematics

• Sensitivity to fixed parameters– ga0' = -1.13 ga0KK (4q) vs ga0' = 1.2 ga0KK (qq)

– gK+K─ = 4.49 0.07

• Normalization (N 1)

• Data-MC discrepancy for the wrong pairing fraction

11.5% (data) vs 14%(MC)

• Sensitivity to the starting values of the parameters

( 10% variations < 0.1%)

3/222K

KK/M4M1M

)KK(48πg

Page 5: 0  fit update

5

Systematics (KL)Systematics (KL)ga '= 1.2 gaK+K- N Wrong pairings g K+K─

Ma0 0.0% 0.0% 0.1% 0.0%

gaK+K 0.2% 1.5% 1.6% 0.2%

ga 0.3% 0.2% 1.2% 0.4%

g a -0.2% 0.5% 4.2% 0.7%

0.0% 0.1% -1.3% 0.3%

Br(VDM) 0.1% 2.4% -15.7% 0.0%

R=Br( )/Br( 0) 0.0% 0.2% 2.1% 0.0%

Ra0=(ga0K+K─/ga0)2 0.0% 2.4% 0.9% 0.4%

• Relative variations

Par. Fit uncert. systematic

Ma0 (MeV) 982.5 1.5 1.1

gaK+K (GeV) 2.15 0.06 0.08

ga(GeV) 2.82 0.03 0.04

g a(GeV-1) 1.58 0.09 0.16(°) 222 12 3

Br(VDM) 0.92 0.40 0.15

R=Br( )/Br( 0) 1.70 0.04 0.04

Ra0=(ga0K+K─/ga0)2 0.58 0.03 0.03

Page 6: 0  fit update

6

KL with fixed VDMKL with fixed VDM

• Br(VDM) = 410-6 (same as NS fit)

Par. Fit uncert. VDM fixed Fit uncert. variations

Ma0 (MeV) 982.5 1.5 980.0 0.5 -0.2%

gaK+K (GeV) 2.15 0.06 2.05 0.02 -4.8%

ga(GeV) 2.82 0.03 2.70 0.02 -4.2%

g a(GeV-1) 1.58 0.09 1.45 0.09 -8.2%(°) 222 12 236 8 6.4%

Br(VDM) 0.92 0.40

R=Br( )/Br( 0) 1.70 0.04 1.74 0.04 2.4%

Ra0=(ga0K+K─/ga0)2 0.58 0.03 0.57 0.01 -2.5%

2 157.6 179.7ndf 136 137

P(2) 9.9% 0.8%

• Worse 2 probability

Page 7: 0  fit update

7

NS Systematics NS Systematics

• Sensitivity to fixed parameters Ma0 = 982.5 MeV (KL fit) vs 985.1 MeV (PDG)

• Normalization (N 1)

• Data-MC discrepancy for the wrong pairing fraction

11.5% (data) vs 14%(MC)

• Sensitivity to the starting values of the parameters

(10% variations < 0.2%)

Page 8: 0  fit update

8

NSNS

Par. Fit uncert.

Ma0 (MeV) 982.5 FIXED

gaK+K (GeV) 1.59 0.22

ga(GeV) 2.21 0.09

g a(GeV-1) 1.63 0.11

R=Br( )/Br( 0) 1.70

a0 14.8 1.10 (deg) 42.8 1.5

a1 25.5 1.81 (deg) 61.5 3.5

Ra0=(ga0K+K─/ga0)2 0.52 0.11

2 148.3ndf 134

P(2) 18.8%

ga0KK 1.000

ga0 0.849 1.000

ga0 0.885 0.971 1.000

R -0.226 -0.261 -0.309 1.000

a0 0.858 0.793 0.905 -0.275 1.000

0 0.584 0.787 0.805 -0.283 0.768 1.000

a1 0.684 0.707 0.813 -0.253 0.933 0.836 1.000

1 0.807 0.912 0.936 -0.294 0.868 0.931 0.816 1.000

Correlation coefficients

• Large correlations

Page 9: 0  fit update

9

NSNS

Ma0=985.1 MeV N wrong pairings

gaK+K─ -20.7% -7.1% / -1% -3.2%

ga -3.9% -5.7% / -0.6% -4.5%

g a -3.5% -7.1% / +0.4% -3.9%

R=Br( )/Br( 0) -0.1% 0.3% 0.3%

a0 -0.2% -11.5% / -0.1% -7.9%0 1.0% 11.6% / 0.3% 11.4%

a1 -1.1% -14% / -0.3% -11.3%1 1.0% 19.5% / 0.3% 20.3%

Ra0=(ga0K+K─/ga0)2 -31.8% -3.1% / -0.8 % -1.3%

• Relative variations

• Large correlations also with the a0 mass

• Strange behaviour for N 1

Page 10: 0  fit update

10

NS with free VDMNS with free VDM• Br(VDM) as free parameter (as for KL)

Par. Fit uncert. VDM free Fit uncert. Variations

Ma0 (MeV) 982.5 FIXED 982.5 FIXED

gaK+K─ (GeV) 1.59 0.22 2.01 0.07 26.4%

ga(GeV) 2.21 0.09 2.46 0.08 11.3%

g a(GeV-1) 1.63 0.11 1.83 0.03 12.3%

Br(VDM)106 0.05 4

R=Br( )/Br( 0) 1.70 1.7 0.04 0.0%

a0 14.8 1.1 14.9 0.6 0.7%0 (deg) 42.8 1.5 38.3 1.1 -10.5%

a1 25.5 1.8 21.3 1.4 -16.5%1 (deg) 61.5 3.5 57.3 1.4 -6.8%

Ra0=(ga0K+K─/ga0)2 0.52 0.11 0.67 0.07 28.8%

2 148.3 140.6ndf 134 134

P(2) 18.8% 33.1%

• No sensitivity to VDM

• Better fit quality

• Couplings more similar to KL results (ga0 still compatible)

Page 11: 0  fit update

11

Fit with Adler zerosFit with Adler zeros

2KAA

22KKa

2ηAA

22ηπa

M2

1s ; smg

M 2

1-M s ; smg

0

0

• According to some theoreticians the Adler zero is needed as a consequence

of Chiral Symmetry (however there is no general consensus)

It forces the PP interaction to vanish close to threshold:

2K

2

2K

2

2KK

2KK

2

2

2ηπ

2ηπ

M21

M

M21

mgg

M21

MM

M21

Mmgg

0

00

0

00

a

aa

a

aa

• I tried to include them into the fit functions: in order not to change the

meaning of the parameters I used:

Page 12: 0  fit update

12

KL fit with Adler zerosKL fit with Adler zeros

Par. Uncert. Par. Uncert.

Ma0 (MeV) 982.5 1.5 1002 8

ga0KK (GeV) 2.15 0.06 3.09 0.02

ga0(GeV) 2.82 0.03 3.54 0.12

g a(GeV-1) 1.58 0.09 4.72 0.14(°) 222 12 165 13

Br(VDM) 0.92 0.40 1.44 0.61

Br( )/Br( 0) 1.70 0.04 1.70 0.04

R=(ga0KK/ga0)2 0.58 0.03 0.76 0.04

2 156.7 144.0ndf 136 136

P(2) 10.8% 30.3%

KL KL with Adler zero

— With A.z.

— Without A.z.

M (MeV)

• Fit quality improves

• All relevant parameters become larger

Page 13: 0  fit update

13

NS with Adler zerosNS with Adler zeros• The fit convergence improves: no need to fix the a0 mass

Par. Fit uncert. with A.z. Fit uncert.

Ma0 (MeV) 982.5 FIXED 971.9 2.6

gaK+K─ (GeV) 1.59 0.22 2.70 0.06

ga(GeV) 2.21 0.09 2.60 0.04

g a(GeV-1) 1.63 0.11 1.97 0.02

R=Br( )/Br( 0) 1.70 0.04 1.69 0.04

a0 14.8 1.1 16.6 0.50 (deg) 42.8 1.5 35.2 1.0

a1 25.5 1.8 26.5 0.91 (deg) 61.5 3.5 59.0 1.8

Ra0=(ga0K+K─/ga0)2 0.52 0.11 0.58 0.03

2 148.3 146.3ndf 134 133

P(2) 18.8% 20.3%

• However according to Gino no Adler zeros should be put in the model, but…

Page 14: 0  fit update

14

New modelNew model

• t’Hooft, Isidori, Maiani …. “A theory of Scalar Mesons” arXiv:0801.2288

• Scalars are “tetraquarks”

• “Instanton” induced transitions to explain decays like f0

• All couplings are written in terms of 2 parameters: cf and cI (|cI|<<| cf |)

• Pseudoscalars appear in the effective lagrangian with derivative couplings:

2

P22P1

2SP1P221SP1P2SP1P2 MMm

2

1g)p(pgg

• From the technical point of view of the fit, the effect is very similar

to the Adler zero

Page 15: 0  fit update

15

New NS fitNew NS fit

2P

2P

2S

2P

2P

2

PSP(fit)

PSP MMM

MMmgg

Par. Fit uncert. new NS Fit uncert.

Ma0 (MeV) 982.5 FIXED 983.6 1.3

gaK+K─ (GeV) 1.59 0.22 1.40 0.05

ga(GeV) 2.21 0.09 1.94 0.04

g a(GeV-1) 1.63 0.11 1.38 0.02

R=Br( )/Br( 0) 1.70 0.04 1.70 0.03

a0 14.8 1.1 11.9 0.60 (deg) 42.8 1.5 37.6 0.9

a1 25.5 1.8 18.8 1.81 (deg) 61.5 3.5 60.4 0.6

Ra0=(ga0K+K─/ga0)2 0.52 0.11 0.52 0.02

2 148.3 148.9ndf 134 133

P(2) 18.8% 16.4%

The mass is in agreement with KL and with PDG cf=6.6 GeV-1 cI=1.6 GeV-1

(cf= 22 GeV-1 cI= -2.6 GeV-1 best fit by t’Hooft et al.)

Page 16: 0  fit update

16

ConclusionsConclusions

• We should decide which fit show in the paper

• KL fit is stable

• NS has some problems; what systematics should we quote ?

• NS with VDM free goes in the same direction as KL– Very small VDM contribution

• Adler zeros improve the convergence; are they really needed ?

• New NS model: we should discuss the results with the authors (we are in contact with A.Polosa in Rome)

Page 17: 0  fit update

17

Systematics from photon pairingSystematics from photon pairing

C(a0)

C(a0) is the difference between the

first and the second photon combination

for the a0 hypothesis

• Two component (right and wrong pairing)

fit to the C(a0) distribution of the data

(final sample)

• Right and wrong pairing shapes from MC

wrong pairings = (11.5 ± 0.70) %

(from MC 14 %)

• Data

─ Right p.

─ Wrong p.

Page 18: 0  fit update

18

Mgen (MeV)

Mre

c (M

eV)

Systematics from photon pairing Systematics from photon pairing

• From C(a0) between the first and

the second best combination

Wrong pairings: data (11.5 ± 0.70) %

MC 14 %

• Check done by scaling the off-diagonal

part of the efficiency matrix by 0.115/0.14

and the diagonal region accordingly

to conserve normalization