Date post: | 18-Jan-2016 |
Category: |
Documents |
Upload: | katrina-mathews |
View: | 217 times |
Download: | 0 times |
0055
Outline • Event selection & analysis• Background rejection• Efficiencies• Mass spectrum• Comparison data-MC• Branching ratio evaluation• Systematics • Comparison with SND and CMD 2 • Conclusions
(Search for (Search for aa00))
P.Gauzzi
Workshop on KLOE Physics
La Biodola – May 23-25, 2001
0
• Signal: (nb) expected (=1) a00 0.10 1700 000 0.037 630
• Background: e+e-0 00 0.74 12500 00 00 0.10 1700 f000 0.37 6300 e+e-00 5.6x10-3 100 3 16.3 3500* 000 13.4 6000*
• Cross-sections from PDG-2000; a0, f0, and 0 from Novosibirsk data• Large errors on cross sections 15% — 25 %* reconstructed as 5 photon events (acc. to MC)
Initial sample
• Runs 15174—17033 Ldt 17 pb-1
• Events from radiative stream (selected by nrfilt)• “ “ substream neu_min_5g (no tracks and at least 5 prompt )
• First selection 5 prompt photons ( t.w.= min(5 t, , 2 ns) ) > 21o for each photon total energy > 900 MeV (to reject background)
• Initial sample : 36979 events
(tr.+nrfilt+ 5 pr. photons) signal 53 % (3%) 2084 expected evts 00 53 % 10865 3 3500 000 6000
Analysis scheme• First kinematic fit: 30 parameters (E,x,y,z,t for each photon + X,Y,Z of the I.P., Ee+, Ee-) 9 constraints (energy and momentum conservation + T-L/c=0 for each photon) 9 ndf cut: 2/ndf < 10• Best photon pairing in the following hypotheses: 1) 0 2) 00 3) 0 00 ( mass , E0 in the selection 2) 4) 3 ( mass , Erad in the selection 2)
rejection 3 rejection• Second kinematic fit : 30 parameters, 11 constraints( 9 + and 0 masses for 1) or two 0 masses for 2) 3) ) For each event this fit is performed three times hyp. 1) , 2) and 3) cut: 2/ndf (hyp. 1) < 10 • Final cuts:
00 rejection 7 rejection
MC samples
Samples used to study cuts and evaluate efficiencies : • a00 40000 evts • 0 (a0 “flat”) 45000 • e+e-0 00 80000• e+e-00 30000• 00 00 70000• 000 50000• f000 (old generation) 5000• 3 300000• 000 285000
• Cuts on the signal have been studied with 0 (a0 “flat”) in order to avoid introducing any bias on the unknown a0 shape
a0 “flat”
00 rejection
0 0
Second fit in the 000 hypothesis
= angle between the photon from and the 0 not coming from in the CoM system of (00)
0 0
00
f0 0 00
Sample composition
• Data– MC (without )
Agreement is not good, but expected number of events have large errors
signal = 48 % 824 expected evts 0+ f0 30 % 5719 00 37 % 635 000 3 x 10-3 736 ?
Sample composition
Fit the MC distributions to data: 1) 1(+) + 2() + 3(f0)2) (+) fixed + 1(7) + 2() + 3(f0)
• Data– fit
• Data– fit
1) 1 1.25 2 3 0.8
2) 1 0.95 2 3 0.8
Sample composition
Check with the other variable of the scatter plot (E)
• Data– MC
With parametersfrom 1st fit
without 000
With parametersfrom 2nd fitwith 000
• Sizeable content of 000 • decrease 0 and f0 cross-sections by 20% (see Simona’s results)
00 rejection
• Data– MC
• There are many events that have good probability to be both 00 and 0
2-2
000 rejection
• Cluster of maximum energy
• Data– MC
• invariant mass after the elliptic cut
2 cut
• Data– MC 7
Emax (MeV)
M /
Efficiencies
a00 27%
000 27%
e+e-00 27%
e+e-0 00 3.5 x 10-3
00 00 5%
f000 1.4 x 10-3 (depends on
the shape)
3 2 x 10-6
000 5 x 10-4
Mass spectrum
• Data: 666 events Backgrounds (MC)–– 00 (1293.7)–– 7 (11810)–– 00 (210.2)–– total (26810.7)
M (MeV)
cos of rad.
• Data: 666 events –– background (268 events)–– signal (MC) + background
Comparison data-MC
2 of the first fit, ndf =9(without mass contraints)
2 of the second fit, ndf =11(with mass contraints)
• Data– MC
• Data– MC
Comparison data-MC
• Data– MC
• Data– MC
M /
M /
• Pion and eta mass (before the second fit)
Branching ratio
39% γγ)Br(η
produced 105.38Lσ ; 27%ε
10)0.050.71(γγ)Br(ηεLσ
BNγ)ηπBr(
7
4stat
0
M (MeV)
a00• If interference in (a0+0) 0 is negligible (Achasov-Gubin: Phys.Rev.Lett.D63,094007)
• N(0) = 171 60(from cross-sect.) events
-400 10))0.14(syst.0.39(γ)ηπγaBr(
a00
M (MeV)
Efficiency
M (MeV)
bineach for )(MN
)(MN)ε(M
ηπreconstr.
ηπreconstr.
ηπ
cuts before
cutsafter
• Not corrected for photon detection efficiency
Efficiency
M (MeV)
dBr/dM (MeV-1)
0.69ε
BN
γγ)Br(ηLσ
1γ)ηπBr(
ii
ii0
)B(NεN NεN jobsjj
1ijij
genjij
reci
• Next step:
M (MeV)
QCAL veto
• Veto = hit in QCAL on time with the event• <(0)> does not change• ( 7) 3 x 10-4 (40% less)
564 events selected (102 rej.)
M (MeV)
Systematics
γγ)Br(ηεN
BNγ)ηπBr( 0
• Background subtraction (main contribution): B = 118() + 150(others)
20% uncertainty on cross-sections Br/Br = 8%
uncertainty on to be evaluated
• Efficiency : uncertainty to be evaluated
• N : uncertainty on Ldt
“ on : 5% (from Kloe memo 234)
Comparison with SND
• SND: 39 events with <> = 2.1% 2x107 (Phys. Lett. B 279 (2000),53) Br=(0.880.140.09)x10-
4 S/B 0.6• KLOE bckg subtracted and corrected for efficiencyS/B = 1.5
M (MeV)
dBr/dM (MeV-1)
• SND: from tab.1 of their paper
• KLOE spectrum before bckg subtraction, not corrected for eff.
(No normaliz.)
(Normalized)
M (MeV)
Comparison with CMD-2
• CMD-2: after bckg subtraction; <> = 6% 1.9 x107 Br = (0.90 0.24 0.10) x 10-4 (Phys.Lett.B 462 (1999), 380)
• KLOE spectrum bckg subtracted
(Normalized)
Conclusions
• Ldt 17 pb-1 from 2000 data analyzed• 666 events survive the selection • 268 background events estimated from MC• 398 signal events• Br(0) = (0.710.05(stat)) x 10-4
• good agreement between data and MC
• Efficiency as a function of M: need a better evaluation
• Systematics: bckgd. subtraction is dominant( 8%)
uncertainty on efficiencies to be evaluated
• Fit to the spectrum has to be performed in order to get the a0 parameters
• Comparison with the Novosibirsk results: our Br is 20% less, BUT: we have much more statistics, greater efficiency, better S/N ratio.
Efficiencies
a00 27% 26.8%
000 27% 26.8%
e+e-00 26% 25.7%
e+e-0 00 3.5 x 10-3 3.4 x 10-3
00 00 5% 5%
f000 1.4 x 10-3 1.4 x 10-3
000 5.3 x 10-4 4.1 x 10-4
2/ndf<10 2/ndf<5
2/ndf<5
000 63 (not bckg.)
e+e-00 8
e+e-0 00 37
00 00 83
f000 7
000 91
Expected events
Total. 226 9
1.297% ηγ)Br( ; 170284N(
10)0.05(0.68
ηγ)Br(N(
ε(ηγ)
ε
BNγ)ηπBr(
4stat
0
ηγ)
ηγ)
From data : N = 599 events
000
Cut
e+e-0 00 26.8%
f000 2.7% 166
f000(f0g_neu4) 5.1% 314
00 00 8.0% 133
0 2.5% 37
5x10-4 226
From data : 3443 events
Total : 562 – 710
nb 0.650.61γ)ππσ(ωπ 000