Stato dell’analisi e della presa dati di KLOE

1 Graziano Venanzoni 7/07/2005 – CSN1/Trieste

Stato dell’analisi e della presa dati di KLOE

G. Venanzoni

LNF-Frascati

For the KLOE Collaboration


Outline

• Current physics results• Current data taking• The ultimate goal: physics with 2.5 fb-1

• Beyond the 2.5 fb-1: off-peak physics• Conclusions


Current Kaon Physics results

Major decays of KL

KL, e, +-0, 30 Paper in preparation

KL KL 30 PL B566 61 (2003)

Vus from KL and KS Paper in preparation

KL lifetime Paper in preparation

KS KS 00

PL B538 21 (2002)Update with ’01-’02 data in progress

KS ePL B535 37 (2002)Update with ’01-’02 data in progress

KS 000 PL B619 61 (2005) KS +0 In progress, uses also 2004 data

K0 mass KLOE Note 181, www.lnf.infn.it/kloe

Q.M. interference Preliminary results

Vus from K± Paper in preparation

K± K± K±

eK± lifetime

In progress

K+ +00 PL B597 139 (2004)


FISICA DEL KL

•Misura dei BR’s dei principali canali di decadimento•Misura della vita media del KL

•Misura dei fattori di forma semileptonici•Vus e test di unitarietà della CKM


Major KL decays: absolute BRs

BR(KLe) = 0.4049 0.0010stat 0.0031syst ~ 8105 events

BR(KL) = 0.2726 0.0008stat 0.0022syst ~ 5105 events

BR(KL 3) = 0.2018 0.0004stat 0.0026syst ~ 7105 events

BR(KL) = 0.1276 0.0006stat 0.0016syst ~ 2105 events

• DATA SAMPLE: 2001+2002 data sample: 400 pb-1 statistics, 50 x 106 tagged KL

- 3/4 of 2001-2002 data has been used for efficiency evaluation - 1/4 for BR measurement corresponding to 13106 tagged KL

•Absolute BRs results obtained with KL= (51.54 ± 0.44) ns in the acceptance

evaluation


KL BR’s comparison

KLOE

NA48

KTeV

PDG04

KLOE

KTeV

PDG04

KLOE

NA48KTeV

PDG04

KLOE

KTeV

PDG04

KL e KL

KL 30 KL


Major KL decays: absolute BRs

Imposing BR(KLall) =1

BR(KLe) = 0.4007 0.0006 0.0014(tag-trk)

BR(KL) = 0.2698 0.0006 0.0014(tag-trk)

BR(KL 3) = 0.1997 0.0005 0.0019(tag- -counting)

BR(KL) = 0.1263 0.0005 0.0011(tag-trk)

We measure BR(KL all) = 1.0104 0.0076

Including rare decays from PDG (0.36%)

KL)= (50.72

ns

~0.5% accuracy !


KL lifetime from KL30

+ data

Even

ts/0

.3 n

s

t*= LK/c (ns) 14 x 106 events

Fit region = 6 -26 ns (40% L)

KLOE direct

KLOE indirect

Vosburgh et al, PRD 6 (1972), 1834

KLOE average: τL = (50.81± 0.23) ns

L = (50.87 ± 0.16stat ± 0.26syst ) ns

(KLOE direct) L = (51.54 ± 0.44) ns

(Vosburgh et al.)

To be submitted to PLB


Semileptonic Form Factors Semileptonic FF describe the t- distribution of the decays Phase space integral depends on such a parametrization

Ko

e-, Koe+, Ko

-, Ko+ Dalitz plot analyses

• Tagging : clusters from KS decay associated to triggering sectors of the calorimeter

• Selection : 2 opposite-charge tracks with vertex : 35<Rt<120, | Z | <120

• Comparison between different mass hypotheses to identify Klong charged decays :

-

- semileptonic decays

)1)(0()(4

2"

2'

2

1

m

t

m

tt ff

2ppt K

)1)(0()(200

m

tt ff

Statistical Precision - 330 pb-1

’+ ’

o ’’+ ’’

o

Koe3 5% -- 10% --

Ko3 7% 15% -- --

Study of systematics in progress

Results expected for EPS conference


Vus from Kl3 decays and L

Prescription from hep-ph/0411097 (F. Mescia @ICHEP04):

1) Quadratic parametrization of the form factor :

from KTeV + ISTRA

2) KL lifetime from KLOE (average of the two measurements) :

KL

= (50.81 ns

(contribution to Vus ~ 0.1%)

3) BRs from KLOE (setting the sum =1):

4) Form factor f+K(0) from Leutwyler-Roos: 0.961(8)

BR(KLe) = 0.4007 0.0006

0.0014

BR(KL) = 0.2698 0.0006

0.0014

2

2

2 21)0()(

m

t

m

tftf

x 2 betterthan PDG !

|Vus| from neutral Kl3 partial decay widths

(K0 l) I(t) (1 + I(t,))(1 + )|Vus f+K(0) |2



KLOE results: |Vus

|f+

K(0) (KSe3

) = 0.2169 0.0017

|Vus

|f+

K(0) (KLe3

) = 0.2164 0.0007 |V

us|f

+K(0)(K

L3) = 0.2174 0.0009

From Unitarity: (1-|Vud

|2)1/2f+

K(0) = 0.2177 0.0028

|Vus| f+Kπ(0)


The grand summary on Vus with 0.5 fb-1

With charged KaonsWith neutral Kaons


FISICA DEL Ks

•BR del KS eVus, asimmetria di carica, test S=Q)•BR del KS 000

•BR del KS +-0


KS semileptonic decay

Sensitivity to CPT violating effects through charge asymmetry:

Sensitivity to CP violation in K0-K0 mixing:

AS = 2Re (CPT symmetry assumed)

_

AS never measured before

If CPT holds, AS=AL

ASAL signals CPT violation in mixing and/or decay with SQ

(KS,L -e+) (KS,L +e-)

(KS,L -e+) (KS,L +e-)

_

_AS,L =

Can extract |Vus| via measurement of BR(KS e)

Test of the S = Q rule, (KS e)/(KL e) = 1 + 4 Re(x)


KS e

e11531 181

e11805 177

EMISS – cPMISS(MeV)

• KS tagged by KL interaction in EMC

• KS background (x103) suppressed by– vertex, 2-track mass

– e/ ID by TOF

– e+) = (24.10.1 0.2)%

– e-) = (23.60.1 0.2)%

• Radiative decays included in MC• Fit data to MC spectra of signal (missing

E - cP ~ 0) + background

• Normalize to PDG BR(KS to get BR(KS e)

BR(KS e) = (7.09 0.07stat 0.08syst) 10-4

AS = (2 9stat 5syst) 103

AL = (3.322 0.058 0.047) 103 [KTeV 2002]

AL = (3.317 0.070 0.072) 103 [NA48 2003]

410pb-1 of 2001/02 data


BR(KS )• Same motivations of KSe3, but more difficult

– Lower BR: expect 4 x 10-4

– Background events from KS : same PIDs of the signal– Troublesome charge identification for the signal

• Anyway, never done before. Here it is:

Cuts on P* + dpos + vtx

Cuts on P* + dpos + vtx

E (MeV)4040 0 20-20

2002 Data

KS + +

E (MeV)4040 0 20-20

2002 Data

KS + +

4686 4654


KS 30 test of CP and CPT

Observation of KS 30 signals CP violation in mixing and/or decay:

If CPT conserved: S000 = L000 |000|2 BR(KS 30) ~ 2 109

Best results: BR < 1.4 105 90% CL SND ’99BR < 7.4 107 90% CL NA48 ’04

Uncertainty on KS 30 amplitude currently limits precision on Im From unitarity (Bell-Steinberger relation):

Best results: Im = (2.4 5.0) 105 CPLEAR ’99Im = (1.2 3.0) 105 NA48 ’03 preliminary

describes CPV describes CPTV

f(1 + i tan SW) [Re i Im ] A*(KS f ) A(KL f )S

1

Compare tomK/mPlanck = 4 1020

K

KK

m

mm 00 < ~ 8 1019Im < ~ 2 105


BR(KS 0)Rarest decay studied by KLOE so far

Data sample: 0.5 fb-1 2001-2002 run– 37.8 x 106 (KL-crash tag + KS)

Require 6 prompt photons:

large background ~40K events

Kinematic fit, 20, 30 estimators (2, 3)

After all analysis cuts (3 = 24.4%)

– 2 candidate events found– Expected backgr. = 3.13 ± 0.82 ± 0.37

BR(KS) ≤ 1.2 x 10-7

@90% CL. Best limit

MC background(not scaled)

data Published in PLB619 61 (2005)


Search for KS +0 started

• Motivation– Present status

– BR(CPC)~ 3x10-7, BR(CPV)~1.2x10-9

– Direct measurement of CPC part possible with ultimate 2fb-1

– Measurement tests untested prediction of PT

• Data sample: 740 pb-1– 373pb-1 (2001/2 data) + 367 (2004 data !)

• Assuming BR=3x10-7, ~230 signal events produced


Reject background with kinematic fit

Constraints (8):• Global 4-momentum conservation• ’s from 0 have = 1• M() = m(0)

• M() = m(KS)

Require 2 < 30:• Cut efficiency = 48.5%• Overall signal efficiency = 3.3%• > 99% of background rejected

KS +0: Background rejection

From MC, 24060 events expected after preselection (of which ~16 signal)

2 from kinematic fit:

MC background

MC signal (L × ~100)

Preliminary results with 740 pb-1:

•Signal efficiency: ~ 1.5%, 6 candidates

Background (sidebands): ~ 3.5 events

•Statistical error: ~ 100%, systematics in progress


FISICA DEL K± •BR K±and Vus

•BR K±l

•Misura della vita media del K±


Tagging of charged kaons

• Tag performed selecting

K± ±, ±0 decays (85% of K±)

by measuring the charged particle

momentum in the K rest frame (P*)

• Trigger required on the tag side.


Measurement of BR(K++())

• Tag from K-- (self-triggering)• 2002 data 175 pb-1 (2/3 is used as efficiency sample)• Background events identified by the presence of a neutral pion.

e

P()*(MeV)

Particle momentum in K rest frameCombining the experimental value of

(K())/(())

with fK/f from lattice calculations

we can extract the ratio |Vus|/|Vud|

(Marciano hep-ph/0406324)

Selection


The result

BR(K+ +()) = 0.6366 0.0009stat. 0.0015syst.


Vus from K++()

fK /f =1.210±0.014 (MILC Coll. hep-lat/0407028)

Following the method from Marciano hep-ph/0406324 :

Vus=0.2223±0.0025KLOE preliminary

Vud=0.9740±0.0005 (superallowed -decays)

new unpublished Vnew unpublished Vudud value will shrink the error value will shrink the error band band


Semileptonic decays of K±

Absolute BR measurement. Tagging with K± , Kl3 selection:

K± vertex in DC Rejection of K2, K2 background0 in time Spectrum of charged daughter mass from TOF m2=p2 [(cT/L)2-1] Ratio of data and MC efficiency is used to correct MC acceptance.

KK

KK

ee

K++0 tagTag K+2 Tag K+2 Tag K-2 Tag K-2

NKe3 61 134 316 25 423 210 64 922 329 24 592 206

NK3 38 089 266 15 159 172 40 639 279 15 006 170

Result will be announced at EPS


Measurement of K± lifetime

we have two different methods to measure the charged kaon lifetime : (1) by K decay length and (2) by K decay time

the Particle Data Group values are questionable

KK = (12.385 = (12.385 0.025) 0.025) nsns


Measurement of K± lifetime: fit of the proper time slope

2 2 = 1.6= 1.6

ns

DCvxt resolution obtained smearing the MC DCvtx resolution comparing ((0 vtx) – (DC

vtx)) in DATA and MC

ns

t 400 ps

(0 vtx) – (DC vtx)

the proper time slope is fitted with an exponential function convoluted with the appropriate DCvtx resolution functions bin by bin (500 ps)(500 ps)


Kaon physics results with 0.5 fb-1 (2001-02)

• KLOE has performed a preliminary measurement of– Major KL BRs with 0.5% accuracy

– KL lifetime with 0.6% accuracy

– BR(KS e with 1% accuracy

– BR(K+ with 0.2% accuracy NEW

– All BRs are inclusive of the radiation– Final papers are under review by the collaboration

• KLOE has set the best upper limit on BR(KS3)

• A large number of K± semileptonic decays has been selected and allows a BR measurement with accuracy < 1%.

• KLOE is now measuring:– K± lifetime

– Kl3 form factors

– BR(KS )

• Coming soon: (KS())/(KS) with few ‰ accuracy. Mature analysis


Current results on radiative decays

a0(980) 0 PL B536 209 (2002) – update in progress

f0(980) 00 PL B537 21 (2002) - update in progress

f0(980) +- New preliminary measurement

’ / PL B541 45 (2002) – update in progress

Upper limit BR( 3) PL B591 45 (2004)

0 Preliminary measurement

Dalitz plot 3 Preliminary measurement

Upper limit BR( +-) PL B606 276 (2005)

, +-0 PL B561 55 (2003)

0 In progress

leptonic widths PL B608 199-205 (2005)

Hadronic cross section PL B606 12-24 (2005)


f0 contribution to e+e +final state

DataEVA (ISR+FSR)EVA (ISR+FSR) + f0

(K-loop model)

Asy

mm

etr

y

M [MeV] M [MeV]

With the insertion of the f0(980) the MC now has the same shape of data

Forward-backward pion asymmetry


BR(). Tests O(p6) PT

signal + background after fit reproduces well the data

BR = ( 8.4 ± 2.7stat ± 1.4syst ) x 10

KLOE Preliminary

CrystalBall (2004)

GAMS (1984)

KLOE

NDATA = 735

Nbkg = 667

sig = 68


The Pion form factor, extracted frome+e + )

1.3% Error0.9% Error

CMD-2KLOE

0.4 0.5 0.6 0.7 0.8 0.9

45

40

35

30

25

20

15

10

5

0

s (GeV2)

KLOE vs. CMD-2

Fair agreement, but relatively large deviations: reasons unknown.

|F(s)|2

=

3s 2

3 |F(s)|2(ee )

Pion Form- factor

PUBLISHED

Relative Difference e+e vs.

s (GeV2)

e+e - and – data incompatible! Isospin breaking effects?!

– Data (ALEPH, OPAL, CLEO)

s (GeV2)0.2 0.4 0.6 0.8 1.0

20%

10%

0%

- 20%

-10%

A. Höcker @ ICHEP04


Our data used in a calculations

a 11 659 000 ∙ 10-10

Experiment E821

DEHZ’03 [e+e- based]

DEHZ’03 [based]

New Information:• New KLOE Measurement Phys. Lett. B606 (2005) 12

• New 4th order QED calculation (Kinoshita, Nio) Phys.Rev.D70 (2004) 113001

• New ‘Light-by-light’ calculation (Melnikov, Vainshtein) Phys.Rev.D70 (2004) 113006

A. Höcker @ ICHEP04: hep-ph/0410081

contains CMD-2and KLOE

Theory:DEHZ’04 [e+e-] –Data not considered

unsufficient understanding ofisospin breaking corrections!

a 11 659 000 ∙ 1010140 150 160 170 180 190 200 210 Theory (SM) - Experiment a

exp - atheo = ( 25.2 ± 9.2 ) ·10-10

2.7 “standard deviations”a 11 659 000 x 1010


And new e+e- data enter the game…

hep-ex/0506076(30 Giugno 2005)

SND data just releasedCMD-2 new data will be published soon…

A session at EPS devoted to hadronic

2


Stato della presa dati


INTEGRATED LUMINOSITY

2004 2005

Plot by P. Franzini

2001 2002

Ad oggi abbiamo ~1.7 fb-1 di dati raccolti su nastro


MONTH LUMINOSITY

Plot by P. Franzini

L’incremento in luminosita’ e’ stato accompagnato da una riduzione dei fondi macchina


DAILY LUMINOSITY

2004 2005

Plot by P. Franzini L/day

10.5 pb-1

7 pb-1

3.5 pb-1


Runs 28700 (9 May 04) to 37457 (30 Jun 05)1190 pb-1 of data on disk OK for reconstruction1176 pb-1 of data with full calibrations1107 pb-1 of data fully reconstructed

(94% eff)

Reconstruction in 2004-05

The reconstruction follows closely the data acquisition (each reconstruction job lasts 2h as maximum )

Tag=100

Luminosity (pb-1)

2001-02 440

2004 690

2005 (up to 30 Jun)

505

Total 1635

Data quality continuously monitored “online”!


Stability of beam energy at O(100 KeV)

May-04 Sep-04 Oct-04 Dec-04

L = 294pb-1 L = 397pb-12004

Feb-05 20-Apr-05 21-Apr-05 9-May-05

L = 248pb-1L = 85pb-12005

s mostly between 1019.3 and 1019.6: OK. Better than 2001


KS invariant mass stable within O(50 KeV)

Feb-05 20-Apr-05

21-Apr-05 9-May-05


Prospettive con 2.5 fb-1



- Contributions to the relative error on |Vus

|f+

K(0)

K0Le

K0L

K0Se

K0S

K± e K±

II

2Δ

BrBr

2Δ

2Δ

Further informationfrom KLOE

Black = KLOEBlue = PDG



K0Le

K0L

K0Se

K0S

K± e K±

II

2Δ

BrBr

2Δ

2Δ

Further information from KLOE

Contributions to the relative error on |Vus

|f+

K(0)


2.5 fb-1 prospects for rare KS decays

• Good opportunity to measure interesting rare KS decays, down to BR ~ few x 10-8

• However, need to exploit to the maximum KLOE and DANE

Decay channel Expected or measured BR

KS 0 2 x 10-9

≤ 1.2 x 10-7 KLOE

KS (4.9 ± 1.8) 10-8 NA48/1

KS +0 3.2+1.2-1.0 x 10-7 PDG04

KS (2.78 ± 0.06 ± 0.04) 10-6 NA48/1

KS 4 x 10-4

KS e (7.09 ± 0.07 ± 0.08) 10-4 KLOE


2.5 fb-1 prospects for KS 0

• Increased statistics: x 6.5 improvement– Luminosity x 5

– Add tagging by KL vertex in DC x 1.3

• Increased background rejection– Largest bkg source after all cuts is the splitting of e.m. clusters

• Merging procedure removes bkg but leaves signal untouched• Candidates in data go from 2 to 0, in MC from 3.13 to 2

– Optimization of kinematic fit

– Overall better understanding of the known background

If we will be able to suppress the background to a ~negligible levelUL can be improved up to a factor of 10 (down to few 10-8)


2.5 fb-1 prospects for KS +0

• Signal efficiency ~ 1.5%• Very, very preliminary results with 0.74 fb-1

– Candidates: 6 events

– Background ~ 3.5 events (estimated from side bands in data)

– Observed events consistent with expectations within the statistical error (~100%)

– no systematic error, no error on background

• Scaling the values of signal and background to 2 fb-1 we expect– ~16 events, of which ~9 background

– ~60% statistical accuracy on BR(KS +0)

With further effort by KLOE to suppress background and 2.5 fb-1

by DANE, we can measure this BR with accuracy below 50%

competitive with other measurements (and the only direct search!)


2.5 fb-1 prospects for BR KS l

• Fractional accuracy <1% on the BR KS e

• Sensitivity at the level of 3 10-3 on the charge asymmetry AS

(ARe

• The first direct measurement of BR KS , accuracy <2%:

• Total uncertainty expected to be largely dominated by statistics

• Measurement of charge asymmetry much more complicated than KSe3


radiative anddalitz decays with 2.5 fb-1

• In addition to radiative, also Dalitz decays of the can be studied: N(2fb-1) ~ 7x109

• M(dilepton) spectra give transition FF(q2). FF(q2) with BR can test VMD and Lattice theoretical models– Predicted BR( ) = 5.3-6.8 x 10-6

• Then, can also study Dalitz and double Dalitz decays of , ’ (eeee, BR ~ 6.5 x 10-5, never observed)

• Reach the ultimate goal of understanding better the nature of ’, f0, a0 mesons (via Dalitz plot analysis)


Precision physics with 2.5 fb-1

M (MeV/c2)

Will improve by factor 2

Will improve by √(LNEW/LOLD)

(background limited)

Can use ee, to measure m() • With Pee from tracking and √s constraint can get few x

10 KeV accuracy, competitive w/NA48

Measure all major BR with permil accuracy• check that sum = 1


’physics with 2.5 fb-1


Beyond the 2.5 fb-1 goal

• Once the ultimate goal of the 2.5 fb-1 on tape has been achieved, we would like to perform a short program of off-peak physics– scan– Running at 1000 MeV c.m. energy

• Motivations for scan (4 points max, 5 MeV steps)

– Calibration of KLOE energy scale, line shape

– Studying the model dependence of the f0

BR leptonic widths, …

• Main motivations for running at 1000 MeV– Measurement of the () down to threshold

– Two-photon physics with KLOE• e+e- e+e-(,)


(e+e + ) at low energy1010

10

10

( ( aa

))st

atst

at

6

5

4

3

2

1

0

stat. error on ahad, is fully

dominated

by region (2m)2< s< 0.35 GeV2

+- signal+- 0 background (after fit)

high background there,

EVEN WITHlarge -angle

analysis,-tagging+kinematic

fit


(e+e + ) at low energy[nb]

s [MeV]

√s [MeV] (s)/(Peak)

1019.5 ~1

1010 ~0.1

1000* ~0.05

980* ~0.02

Running at √s=1000 MeV would allow a measurement of (e+e-+-) free from the resonant 3 backgroundAmount of integrated luminosity under evaluation

* from SND, Phys. Rev. D66 032001 (2002)

Channel


Conclusioni• Molte analisi pubblicate o in corso di pubblicazione:

– BR assoluti del KL e misura di Vus

– Fattori di forma del KL

– BR K+K±l3

– BR KS eKS KS

– Vita media del KL e K±

– Sezione d’urto adronica (a grande e a piccolo angolo)– Decadimenti radiativi della ( f0,a0)

• Presa dati attuale stabile anche se con qualche problema.• KLOE funziona egregiamente.• 1.7 fb-1 dal 2001 (~1.3fb-1 da maggio del 2004) di luminosita’

integrata.• Goal: 2 fb-1 integrati nel 2004-05 150 pb-1/mese x i prossimi 5

mesi.• Data taking del 2006 dedicato allo scan ed a ~200 pb-1 off-peak.• Le analisi dell’intero data set permetteranno di migliorare molte

delle misure fatte e di accedere al settore raro dei KS e allo studio del Dalitz plot dei decadimenti radiativi della .


SPARES/OLD


Absolute KL BRs

)(/)()()(

1

N

N i)BR(K

tag

iL allii tagtagLFVrec

Reconstruction efficiencies: KL , e (rec) 55% KL (rec) 40% KL (rec) 100%

Evaluated via MC simulation, Corrected by DATA/MC ratio

Integral over the fiducial volume: ε (FV, L) 26%, depends on L

FV51.7 ns = 0.0128

Tagging efficiency

Trigger required on the KS side


KLcharged

KS

2 tracks forming a vertex along the KL direction.Vertex and tracking efficiency 55% for Kl3 and 40% for 3

P-E() (MeV)

Events counted by fitting missing momentum minus missing energy in the pion-muon hypothesis



Events with 3 photons time-correlated 99.2% selection efficiencyResidual background (1.3%) subtracted.

KL +-0 used to determine :

EmC time-scale Vertex resolution Vertex reconstruction efficiency

π+π-π0 data:

LK (cm)

<σ (LK)> 2.5 cm

Time-correlated photons


KL30

KS

At least 3 neutral clusterNeutral vertex reconstructed along KL direction using TofEfficiency close to 100%, bkg 1%



• We measure a large fraction of KL decay length L/ 0.4 high statistical sensitivity to • KL momentum measured from KS+-

• KL30 efficiency >99% low sensitivity to efficiency variations along the KL path• Time scale calibration at 0.1% level enters directly in the lifetime measurement• KL+-0 as a control sample for the estimate of efficiency and resolution.

Check on time-scale calibration with KL+-0


Semileptonic Form Factors

• A unique t-binned analysis for all the decays of interest, selected by

• (e) Lesser of cPmiss-Emiss in the and (e and e) hypotheses

• Comparison between cPmiss-Emiss in the and (e and e) hypotheses used to determine the lepton charge

• The t behaviour of each component controlled by sub-samples with different fractions of Ko

e-, Koe+, Ko

-, Ko+, Background

• Control of DATA/MC agreement and systematics related to the corrections are in progress.



MC shapes used to fit data (input) Semileptonic slopes ’+,0 / ’’+,0 obtained contemporarly with the amount of each component

),()()()(),;()(20

1

",

', ijsmejjjjif trkefffsr

joo

Int

Study of the Dalitz plot shape MC precision - shape simulation One t-bin Fit Result - Behaviour for the different components

Koe- Ko

-Ke3K3

One bin in the variable to discriminate the different semileptonic componentsFit results - t behaviour



Statistical Precision - 330 pb-1

’+ ’

o ’’+ ’’

o

Koe3 5% -- 10% --

Ko3 7% 15% -- --

• Study of the systematics in progress• Further discriminant variables as PID from ECAL to control reliability of the MC distributions/ of the evaluation the different semileptonic components


KS semileptonic decay: Unitarity test of CKM matrix

Most precise test of unitarity possible at present comes from 1st row:

Can test if = 0 at few 10-3:from super-allowed 0+ 0+ Fermi transitions, n -decays: 2|Vud|Vud = 0.0010from semileptonic kaon decays (PDG 2002 fit):

|Vud|2 + |Vus|2 + |Vub|2 ~ |Vud|2 + |Vus|2 1 –

To extract |Vus| from K0e3 decays, have to include EM effects:

(K0 e) I(t) (1 + I(t,)) (1 + )|Vus f+K0-(0) |2

|Vus|

|Vus|

t)

t) f+K0-(0)

f+K0-(0)

= 0.5 0.5

2|Vus|Vus = 0.0011

0.3% 1%0.5%

Relative uncertainty:

Aiming at a 1% uncertainty for a BR ~ 7 × 10

Contribution from S to fractional uncertainty on Vus negligible (0.04%)


KS semileptonic decay: TOF PID

•After kinematic rejection, M < 490 MeV, background still dominated by KS () with decays in flight before entering the DC

•TOF identification: compare -e expected flight times, reject , bkg

e

e

Data MC e

1 2 4

0

1

2

3

31

4

5

1 0 5 1 2 4

0

1

2

3

31

4

5

1 0 5

dt,e(ns)

dt, e

(ns)

dt,ab = (tcl –L/c(a)) – (tcl –L/c(a)) (independent by T0 syst.)


KS semileptonic decay:Improved TOF

1000

()

Data— MC sig + bkg

500

1500

2000

0

()

200


400

600

800e

Background

02

24

2 204t, (ns)

6

6

t, e

(ns)

Emiss(e) cPmiss(MeV)20 200 40 60

Trigger signal is synchronized with machine rf, event T0 determined up to an integer multiple of the bunch crossing time (~2.7 ns)Improve TOF capabilities by using trial values for the global event T0Higher rejection power, as seen from Emisse cPmiss

Evt

s/M

eV

46

4 6

8

8

8 8


KS semileptonic decay:systematics

Dependence of corrections on charge state is crucial for the charge asymmetry

TOF efficiency responsible for the charge dependence:

= (4.3 ± 0.9stat ± 0.8syst)%

TOF difference arise from different hadronic interaction mechanisms for + and in EmC

Corrections studied as a function of time during data taking: the result is stable

Check fit stability and MC reliability by varying KL crash minimum energy

Correction ChargeFractional uncertainty

Statistical Systematic

Countinge

e

1.31%1.25%

0.5%0.5%

DC preselectione

e

0.2%0.2%

0.4%0.4%

TCAe

e

0.04%0.04%

0.4%<0.1%

Triggere

e

0.07%0.07%

0.5%<0.1%

TOFe

e

0.3%0.3%

0.1%<0.1%

Tag bias /e e

e

0.8%0.8%

0.1%0.1%

Efficiency for 0 0.3%

Totale

e

1.58%1.41%

1%0.6%


KS semileptonic decay results

Use BR(KL e±KTeV 04]:

Re(x) = (3.1 3.0stat 1.8syst) 10

Use BR(KL e±KLOE 05*]:

Re(x) = (.6 3.1stat 1.8syst) 10

Use average of KTeV and NA48 measurements of KS lifetime: S = (89.62 0.05) ps

Use new measurement of the KL lifetime: L = (50.81 0.23) ns [KLOE 05]

Compare (KS e) with (KL e): test of the S Q rule

Re(x) =1/4 [(KS e) / (KL e)1]

KTeV 04

KLOE 05*

PDG04

Most precise measurement of Re(x) (in CPT conserving transitions): compare w CPLEAR99, Re(x) = 6×103

*to be published

BR KLe3KLOE KS assuming S=Q

0.38

0.39

0.40

0.41


KS semileptonic decay:Analysis scheme

Analysis scheme:

1.Kcrash tagging

2.Two tracks from IP to EmC

3.Kinematic rejection of KS background, cut on M, Pmiss

4.Both tracks are geometrically associated to clusters in the calorimeter

5.Further rejection of KS background from TOF identification

• Compare cluster times with or electron expected flight times

• From TOF get a clean charge identification, too

6. Obtain number of signal events from a constrained likelihood fit of multiple data distributions

7.Normalize to KS count in the same data set

Use BR(KS ) to calculate BR(KS e)


KS semileptonic decay: event counting

Constrained binned-likelihood fit to multiple distributions of data

Likelihood accounts for statistical fluctuations of data and various MC sources

4 variables are used: Emiss(e) Pmiss,PCA= PCA1 – PCA2 are shown below

Data— MC fitsignalbad

bad

other

50 5000

100

200

300

400

500

600

700

Emiss(e) cPmiss(MeV)100150

PCA (cm)04 4 880

100

200

300

400

500

Evt

s/0.

2cm

TRK 2

TRK 1

Evt

s/M

eV

PCA2

PCA1

DC inner wall

IP


KS semileptonic decay: results

Dependence of efficiencies on charge mainly due to TOF effects, estimated using data control sample of KL e

AS = (2 9stat 5syst) 103

BR(KS e+) = (3.54 0.06stat 0.04syst) 104

BR(KS e-) = (3.55 0.05stat 0.02syst) 104

AL = (3.322 0.058 0.047) 103 [KTeV 2002]

AL = (3.317 0.070 0.072) 103 [NA48 2003]

Published result: (6.91 0.34stat 0.15syst) 104, KLOE 02

BR(KS e) = (7.09 0.08stat 0.05syst) 104

Uncertainty in charge asymmetry dominated by statistics

With data for an integrated luminosity of 2.5 fb, AS 3×10 ARe

BR(KS fundamental: a new update of our measurement of RS = BR(KS BRKS having 0.4% total error will be ready soon (it can be used to with phase shifts at s = mK)


20 200 04060 40


60 80

KS semileptonic decay

• Selected using TOF technique• Event counting obtained by fitting the E(e) - P distribution• The two charge modes are measured independently• Selected ~104 signal events per charge in the 2001-02 data

BR(KS e) = (7.09 0.07stat 0.08syst) 10-4

Emiss– Pmiss(MeV)


Search for KS 000

Preselection

• KS tagged by KL interaction in EMC

• 6 photons, no tracks from IP

• Kinematic fit in 20 & 30 hypothesis

Rejection of background

KS 0 + 2split/accidental clusters

• Define signal box in 23 vs. 2

2 plane:

“best” 3 photon pairs

“best” 2 pairs

Cut on E(KS) E


Search for KS 000 – 20 vs 30

Main background from KS + 2 fake ’s: Compare 3 vs 2 hypotheses:

Definition of the signal box obtained from UL optimization from analysis of a MC sample with a statistics equivalent to ~3data

DataMC KS 30

pairing of 6 clusters

with best 0 mass estimates

pairing of 4’s out of

6: 0 masses, E(KS), P(KS), c.m. angle between 0’s

00

20

40

80

60

4010 20 30


Search for KS 000 - sidebands

DataMC KS 30

00

20

40

80

60

4010 20 30

00

200

400

50 100

600

Data MC background

KS 30 decay switched on during MC production of 450 pb1 equivalent data, with BR equal to the SND upper limit


Search for KS 000 - sidebands

DataMC KS 30

00

20

40

80

60

4010 20 30

500

1000

1500

Data MC background

2000

00

10050 7525


Search for KS 000 – signal region

DataMC KS 30

00

20

40

80

60

4010 20 30

200

400

00

10050 7525

Data MC background


KS 000 – Final Results

KLOE: 450 pb1 ’01+’02 data

Nsel(data) = 2 events selected as signal

= 24.5%

Nsel(bkg) = 3.130.82stat0.37syst

bkg events expected from MC

N2BR(KS 000) = BR(KS 00) < 1.2 10-7

N3

Normalize signal counts to KS 00 count in the same data set:

Can state: N3 < 3.45 at 90% CL

PLB 619 61 (2005)

20

40

60

00

4 62


KS 000 – Final ResultsUsing the PDG value for L,S, and BR(KL 000) together with our limit:

@ 90% CL

|000| = < 1.8 10-2 A(KS 000)

A(KL 000)

Prospects for the analysis of 2 fb1:• Increased statistics: × 6.5 improvement• Increased background rejection, tighter recover-splitting technique• UL might decrease by an order of magnitude

BR(KS) ≤ 1.2 x 10-7

@90% CL. Best limit

This limit makes the 000 contribution to CPTV test from unitarity negligible. Such test is limited by +


Search for KS 000

Signal selection

KS’s tagged by means of Kcrash identification

6 photons (neutral clusters, TOF consistent with = 1)

No charged tracks from IP

• Impose KS mass and energy-momentum conservation, = 1 for each

• Estimate E, r, t, s, p

Rejection power of 2fit not sufficient to eliminate main background due

to KS + 2 fake ’s

Kinematic fit:


Using the PDG values and ourlimit we have:

Constraints for 000 and Im(CPTV)

KLOE, 0.45 fb-1: BR(KS 0) ≤ 1.2 107 @90% CL

000 =A KS 3 0 A KL 3 0

= L

S

B KS 3 0 B KL 3 0

|000| < 1.8 10-2 at 90% CL

This limit makes the 000 contribution to CPTV test from unitarity(Bell-Steinberger relation) negligible. Such test is limited by +


KS +0

Based on CPLEAR and E621 observations of time-dependent Dalitz-plot asymmetry from interference of KL, KS amplitudes to 0

KS 0 decay has never been observed directly

KS 0 decay dominated by CP-conserving I = 3/2 transition

Measurement of BR provides useful constraint for PT estimates of K 3 amplitudes

From PT and isospin analysis of existing K 3 data:

BR(KS 0) = (3.2+1.21.0) × 107 PDG’04

BR(KS 0) = (2.1–3.9) × 107

KLOE preliminary result based on analysis 740 pb1 of data from 2001-2004Assuming PDG value for BR, ~230 KS 0 events produced


After kinematic fit, expect 158 background events from MC:

~ 50% KK with K 00 at origin

~ 25% KS 0D0

(D) (0D ee)

~ 8% various other KK topologies

KS +0: Background rejection

KK Monochromatic ± from K 0 has p* = 206 MeVRequire p*(K ) < 120 MeV at ends of , tracks

KS 0D0

(D)Require at least one track to be associated to a clusterObtain e/ separation from TOF

Both typesVeto on prompt clusters not associated to tracks or 0 Cut on Efree = largest unassociated cluster energy

K+

+

“KL

crash”

Additional cuts to suppress background from:

K 00; K 0


Selection of KS +0

• KL-crash tag, 3-body selection cuts

• Kinematic fit

KL-crash

TOF

p*

UnassociatedEnergy

DATA

Standard BackgroundMC

25130 eventsin the data


Backgrounds for KS +0

• Backgrounds after cut

on 2 from kinematic

fit (98.8% rejection)

• ’ background

Dedicated MC, TCA cut

Dedicated MC, cut onunassociated energy

K+

+


2 fb-1 prospects for KS +0

• Signal efficiency ~ 1.5%• Very, very preliminary results with 0.74 fb-1

– Candidates: 6 events

– Background ~ 3.5 events (estimated from side bands in data)

– Observed events consistent with expectations within the statistical error (~100%)

– no systematic error, no error on background

• Scaling the values of signal and background to 2 fb-1 we expect– ~16 events, of which ~9 background

– ~60% statistical accuracy on BR(KS +0)

With further effort by KLOE to suppress background and 2 fb-1

by DANE, we can measure this BR with accuracy below 50%competitive with other measurements (and the only direct search!)


Vus from K± decays at KLOE

i(K l) |Vus f+K(0) |2 SewIi(+,0,0) (1 + i

em+Ii)

BR(Kl3)/K

(exp)Kl3 form factors (exp) PT, lattice rad.

corr.

alternative method to extract Vus from K± ±() given

fK/fcomputed on lattice (K ()) ( ())

Vus

Vud

fK2

f2

2

absolute BR(Kabsolute BR(K++ ++(()))) donedone

absolute BR(Kabsolute BR(Kl3l3)) close to be close to be

completedcompleted

KK lifetime lifetime close to be close to be completedcompleted

KKl3l3 form factors form factors in linein line

(Marciano hep-ph/0406324)


Measurement of BR(K++())

P()*(MeV)

Event counting performed byfitting the P* distribution with signal and background shapes obtained from data control samples: bkg-sample selected by looking for 2 photons from neutral pion; control sample for signal shape selected by identifying the muon cluster in the Emc. The shapes are properly corrected using MC simulation

Systematics are dominated by efficiency estimate

BR(K+ + = 0.6366 ± 0.0009 (stat.) ± 0.0012 (syst.)

Chiang = 0.6324 PDG fit = 0.6343

~2.6 Millionsignal events


K region

The method

TBDCTAG

)K

ε1

ε1

N

N))(μBR(K (γν

the signal is given by K+ decay and is selected asking for a vertex in the FV (40 cm xy 150 cm) of the DC ( 60 pb-1)

the background is mainly due by events with a 0 in the final state :

K-> 0 K0eK

tag bias estimated from MC

DC = DATA corrMC

DC trk+vtx GEO

CTB= 1.01650.0002

P. de Simone 14/06/2005 – Kaon 2005 12

Entries 2176449

CTB

MC includes radiative process

P (MeV)


scan: BR()


2.5 fb-1 prospects for BR(KS

• Selection algorithm provides few% statistical accuracy

• Total uncertainty expected to be largely dominated by statistics

• Measurement of charge asymmetry much more complicated than KSe3

• Timescale for a preliminary measurement: few months


DAILY LUMINOSITY

2004 2005

Plot by P. Franzini


(e+e + ) at low energyAnother background to the hadronic +- spectrum:

f0 +.

M [MeV]

√s=1020 MeV

(IS

R+

FS

R+

f0)

(IS

R+

FS

R)

√s=1000 MeV

Isidori-Maiani (no-structure) model

Kaon-loop model

M [MeV]


scan: f0 cross section

Use scan of 4 points, 5 MeV steps, with O(10)pb-1/point to establish parameters of scalar amplitude

(plus decay BR and forward-backward asymmetry of pion tracks)

2002 data compared to extrapolations from on-peak data:K-Loop fit green curvesIsidori-Maiani NS fit red curve

7 pb-1

5 pb-1


Quantum interferometry: KSKL

Fit including t resolution andefficiency effects + regenerationS, L fixed from PDGKLOE preliminary0.41 fb1 (2001 + 2002 data)

m = (5.34 0.34) 10 s1

PDG04m =(5.290 0.016) 10 s1

Window on decoherence (D, ) violation of QM

KLOE preliminary:

CPLEAR:

DHit ,038.0035.0032.0

LS KK 520.0

18.0 1016.000

KK

16.013.0, LS KK 7.04.000 ,

KK

D related to CPTV. CPTV also beyond QMV. Many models to test with 2fb-1

Coherent KL regeneration on beam pipe

t/S

I , ;t

EPR. First observation ever ofcoherence in two Kaon system


Q. Interferometry: KSKL

Toy MC with from data andgaussian vertex res = 6mm

With 2.5 fb-1 KLOE could reach

the statistical precisionIm ’) ~ 0.015

PDG 2004 fit:

00 0.20.4

(00 ) 2.6

2 fb-1

(Im ’/)

Int. Lum(pb-1)


2.5 fb-1 prospects for Kaon semileptonic decays

• BRs, lifetime, form factors are all measured at the level of few permil currently

• With 2.5 fb-1 they should all get to 0.1% or below


Paving the road for KS BR differs from CHPT O(p4) by 30%,

useful to fix one O(p6) counterterm

Projections based on– 150 pb-1 of 2001 background MC

– 10K events of signal MC

With 2+0.5 fb-1 we expect– 500x106 KS events tagged by Klcrash

– N(KS, tagged) = 500x106 x 2.8x10-6 = 1400 events

– acceptance 0.4

– Nsig = 560 events

2 fb-1: with good background rejection ~ 4% statistical error

BR = (2.78 ± 0.06 ± 0.04) 10-6

3% accuracy, NA48, PLB 551 2003


Paving the road for KS

• Strategy of analysis

– No recover splitting and large

angular acceptance

– Kinematic fit to exploit two

body kinematics

• Distribution of kinematic

variables after fit

• Background separation looks

promising

bkg signalA.U

.

A.U

.

bkg signal

MC distributions, no data yetN. of events in A.U.

After fit

Reco

ns.


f0 KK decays with 2.5 fb-1 ?


Unitarity test of CKM matrix: Vus

IlK() phase space integral, Sew short distance corrections (1.0232)

|Vus| from neutral Kl3 partial decay widths

f+K0-(t = 0) form factor at zero momentum transfer: pure theory calculation (PT, lattice)

emKl electromagnetic correction (amplitude and phase space)

slopes (momentum dependence of the vector and scalar form factors)

Most precise test of unitarity possible at present comes from 1st row:

= 0.0042 0.0019 was the situation as of PDG02: ~ 2.2 deviation2|Vud|dVud = 0.0015 from super-allowed 0+ 0+ Fermi transitions, n -decays

2|Vus|dVus = 0.0011 from semileptonic kaon decays (PDG02 fit)

|Vud|2 + |Vus|2 + |Vub|2 ~ |Vud|2 + |Vus|2 1 –

(K0 l) I(t) (1 + I(t,))(1 + )|Vus f+K(0) |2

|Vus|

|Vus|

t)

t) f+K(0)

f+K(0)= 0.5 0.5

Relative uncertainty:

~(0.5% 0.3% 1%)


Preselection criteria:

• Vertex at origin with zero net charge

• 2 prompt neutral clusters

• Each pair of clusters is a 0 candidate. For each:

- Close 3-body kinematics using m(0), m(KS)

- Set t0 using pair of clusters to establish KL-crash timescale

- Use p(KS) and p() to search for KL-crash cluster in 20° cone

• Choose 0 pair giving best agreement between KS, KL momentum

-

KL

KS +0: Event selection

Preselection efficiency ~7%Small acceptance for low-pT tracks


Preliminary results with 740 pb-1:• Signal efficiency: ~ 1.5%• Candidates: 6 events• Background (sidebands): ~ 3.5 events• Number of events observed consistent with expectation• Statistical error: ~ 100%• Evaluation of systematic error in progress

Scaling the values of signal and background to 2 fb1 we expect:• 16 events, of which 9 background

• Measurement of BR(KS) with 60% error

About the same precision as interference-based measurements • First observation of decay in a direct search

KS +0: Current status


Semileptonic decays of K±

K++0 tag

ml2=pTRK

2 [(c(tl-t)/LTRK )2-1]

Tag K+2 Tag K+2 Tag K-2 Tag K-2

NKe3 61 134 316 25 423 210 64 922 329 24 592 206

NK3 38 089 266 15 159 172 40 639 279 15 006 170


Measurement of K± lifetime

K tag

K

Kvtx

Li

Signal selectionSignal selection

self-triggering tag K track on the signal side decay vertex in FV

signal K extrapolated to the IP dE/dx correction applied along the path

LLii = step length = step length

global efficiency global efficiency trk+vtxtrk+vtx and DC and DCvtxvtx resolution functions resolution functions needed wrt the needed wrt the KK of the Kaon of the KaonP. de Simone 14/06/2005 – Kaon 2005 25


Measurement of K± lifetime: fit of the proper time slope

2 2 = 1.6= 1.6

ns

DCvxt resolution obtained smearing the MC DCvtx resolution comparing ((0 vtx) – (DC vtx)) in DATA and MC

ns

t 400 ps

(0 vtx) – (DC vtx)

the proper time slope is fitted with an exponential function convoluted with the appropriate DCvtx resolution functions bin by bin (500 ps)(500 ps)


0

200

400

600

800

1000

1200

1400

May-05

1.5 fb

Estimate of Tape library usage (TB):Long term estimate incl. all MC and reprocessing

raw

recon

DST

MC04(05)all

2001/02IncludingMC

2.2 fb

Presentcapacitypb reproc

2001/02

reproc2004

We need additional tapes (1000)

TBTotalcapacity


CPU and Disk resources:10 IBM p630 servers: 10×4 POWER4+ 1.45 GHz (equivalent to 100 (40x2.5) B80 CPU )

23 IBM B80 servers: 92 CPU’s

“Online” data Reconstruction and calibrationDST and MC productionReprocessing

CPU allocation flexible. Simply redefine queues with LoadLeveler

Current recalled areas

Production 0.7 TB

User recalls 2.1 TB

DST cache12.9 TB

(10.2 TB added in April 04)

3.2 TB added to AFS cell in Nov 04

2001/02 2004 2005

Data DSTs

(TB)

4.5 6.2 4.6

MCDSTs

(TB)

8.2 - -


“New

” mac

hine

s (p

630

Serv

er)

fibm Users MC Physmon Datarec DST Reproc13 4 114 4 115 4 117 4 118 4 119 4 120 4 121 4 122 4 123 4 1 1 2 224 4 1 1 2 225 4 1 1 2 226 4 1 2 227 4 1 2 228 4 1 2 229 4 1 1 330 4 1 1 331 4 1 1 332 4 1 1 333 4 1 1 334 4 1 1 335 2 436 2 437 2 438 2 439 2 440 2 441 2 442 2 443 2 444

Machine Configuration on April-May05

TOT 84 30 12 36 18 30

“Old

” mac

hine

s (B

80 S

erve

r)


Consideration on CPU needs

16 ± 2 months on 150 B80 CPUs

Task B80 days

Online reconstruction 19300

Reprocessing ’01-’02 8000

MC04/05 28000

Reprocessing ‘04 15900

Re-reconstruction MC ’01-’02 (full sample) 1900

Total 73100

More CPUs are needed if we don’t want to delay the analyses!


Remarks from LNFSC and CSN1

Offline status/priorities discussed at closed session of LNFSC (24 May)– Need for prompt MC04/05 production emphasized by LNFSC members– Requests for additional CPU and disk space supported

Requests for computing upgrades discussed with KLOE CSN1 computing referees yesterday (8 Jun)

– 6 new 4-way 1.5 GHz Power5

1 Power5 = 8 B80 new machines equivalent to 192 B80s

Total KLOE computing power 200 392 B80s– 20 TB of additional disk space for DST staging

Requests will be discussed at the next CSN1 meeting in Trieste 6-7 Jul

One strong possibility purchase 2 of 6 boxes now– No gara needed, machines could arrive in less than 1 month

Stato dell’analisi e della presa dati di KLOE

Documents

Transcript of Stato dell’analisi e della presa dati di KLOE