KLOE: Status and recent results Commissione Scientifica Nazionale I Lecce, 22 September 2003 M. Moulson, LNF for the KLOE Collaboration New KLOE IR installed

KLOE: Status and recent resultsKLOE: Status and recent results

Commissione Scientifica Nazionale ILecce, 22 September 2003

M. Moulson, LNFfor the KLOE Collaboration

New KLOE IR installedJuly 2003

2KLOE: Status and recent results – M. Moulson – CSN1, Lecce, 22 September 2003

Days of running

Inte

grat

ed lu

min

osit

y (p

b1 )

2000: 25 pb1

80106 decays

First published results

2001: 176 pb1

550106 decays

2002: 296 pb1

920106 decays

Analysis in progress

KLOE data taking: 2000-2002

2002

2001

2000

2002 KLOE data taking

3 May—30 September Best value of Lpeak:

Best L dt in one day:

7.8 × 1031 cm2s1

4.5 pb1

KLOE: Status and recent results – M. Moulson – CSN1, Lecce, 22 September 2003 3

A major effort for KLOE in 2003 was to upgrade our capabilities for data processing, MC, etc., in order to exploit the ~450 pb1 of good data collected in 2001-2002. We are now using these tools to churn out physics results.

These tools will also allow us to capitalize on ~2 fb1 of new data to be taken in 2004.

The main focus of the collaboration is currently to ensure that these tools are kept in working order and are well used.


Summary of KLOE results 2002-2003

5 papers published in Phys. Lett. in 2002

11 KLOE measurements in PDG ’03 web update11 cases in which KLOE measurement is best

2 papers published in Phys. Lett. so far in 2003

+3 papers in preparation, contributed to EPS ‘03 and LP ’03

5 papers submitted to Nucl. Instrum. Meth., 2002-2003

>70 presentations at conferences and seminars 2002-2003


Milestones for 2003

Q1 1. Preparations for IR modifications

Q1-4 2. Maintenance of detector performance at 2002 levels

Q1 3. QCAL upgrade

Q1 4. Upgrades to raw-data filters

Q1 5. Reconstruction of all data

Q1 6. Preparation of DST’s for analysis

Q1-4 7. MC production for 2002 analyses

Q2-4 8. BR’s and cross section, KK

Q4 9. Repeat 2002 analyses with full data set


Detector open from January to July for installation of new IR

DC FEE maintenance: 60 channels repairedADC gates widenedWire tension measurementsGeometrical survey before/after new IR installed

EmC HV distribution boards repaired2 HV crates, 3 PMT’s replacedEmC fully operational

QCAL Complete FEE upgrade

Milestones: 1. Preparations for IR modifications (Q1)2. Maintenance of detector performance (Q1-4)


Current detector status

Milestones: 1. Preparations for IR modifications (Q1)2. Maintenance of detector performance (Q1-4)

July Water crisis impedes collection of cosmic-ray data with field on

Debugging of cryogenic system after KLOE/FINUDA compensators connected to main He circuit

Sept Machine operations starting upKLOE waiting to take test data with field on

Milestones satisfied 100%


EmC level

High level of high-frequency noise Correlated with machine RFBoth TDC and ADC affected

Ped

esta

l rm

s (A

DC

cou

nts)

channel

MIP: 200 cts(20 MeV): 50 cts

Milestone 3: QCAL upgrade (Q1)

Milestone satisfied 100%

• Worst 10 PMT’s replaced

• Work completed

• Waiting for field-on and beam for final testing

rece

iver

differential amplifier 8 increase in gain

10 cm coax

12 m twisted

PMT shielding

SDS

SDS

amplifier

60 cm coax12 m thin coax

• Old readout:

• New readout:


Milestone 4: Upgrades to raw-data filters (Q1)

Avg. L: 2.11031 cm2 s1

Avg. T2: 2000 Hz

500 Hz downscaled CR

2001 running

Avg. L: 5.41031 cm2 s1

Avg. T3: 1600 Hz

0 Hz downscaled CR

2002 running

Implementation of T3 for 2002 data taking eliminates need for downscaled CR sample and decreases DAQ load by 500 Hz

Status of offline software filters: • Currently reject ~60% of events before track reconstruction• Upgrades to increase rejection/reduce bias under study• Need restart of data taking for tuning of cuts, etc.



Milestones: 5. Reconstruction of all data (Q1)6. Preparation of DST’s for analysis (Q1)

164 pb1 2001 data: Re-reconstructed with 2002 code in early 2002

288 pb1 2002 data: Reconstructed in “real time” during data taking

DST production following reconstruction for all streams except KK

KK stream re-reconstructed at DST stage (completed May 2003)

As of May, entire KLOE data set (452 pb1) available as DST’s

Total space: 4.3 TBMC DST’s add ~2.9 TB

DST disk cache: 5.7 TBMajority of DST’s disk-accessed

Milestones satisfied 100%


Ambitious program for MC development and production

Simulated event samples statistically comparable to data

all 452 pb1 at 1:5 scale ~300M events

KSKL 452 pb1 at 1:1 scale ~500M events

Comprehensive upgradesBoth MC executable and production procedure affected:

• State-of-the-art detector simulation• Inclusion of accidental activity from machine background• MC DST’s to provide transparent user interface

Each run in data set individually simulated s, p, x, background, dead wires, trigger thresholds…

Milestone 7: MC production for 2002 analyses (Q1-4)


Efficiency: from KS

pz (MeV)

pz (MeV)

da t

a/M

C d

a ta/

MC

New

Old

New

Old

OldNewData

Resolution: M(KS)

M (MeV)

MC-data comparison: trackingMilestone 7: MC production for 2002 analyses (Q1-4)

detail of radiative tail

log

scal

e


Background from e+e eventsAll DC hits (tracklets, etc.)All EmC clusters except isolated by r, t)

New reconstruction stream One background file per raw data file77,000 files, 8 GB total

Background hits (DC and EmC) inserted in MC events with timing preserved

dN/dE dN/d(cos ) Multiplicity

Selection of background clusters on EmCSelected clustersOut-of-time clusters in data

E (MeV) cos





GenerationEvents

requestedElapsed days (60 CPU’s)

Completed

all255M452 pb1, 1:5

26 100%

PHOKHARA 1.0

40M140 pb1, 5:1

6 100%

PHOKHARA 3.0

40M140 pb1, 5:1

6 100%

KSKL475M452 pb1, 1:1

43 (est.) 20%

Total output size: 17.6 TB + 2.9 TB DST at program completion

430M/810M events completed10M/day generation+reconstruction


Milestone 8: BR’s and cross section, KK (Q2-4)

DST’s for KK events newly available in early 2003

KK events re-reconstructed during DST production

• Dedicated energy-loss treatment (m = mK) in track fit

• Refined treatment of multiple-scattering correlation matrix

• Improved merging of split kaon tracks

• Dedicated algorithm for global t0 determination

• Kaon time-of-flight corrections to drift distances

• dE/dx reconstruction from DC ADC’s introduced


Milestone 8: BR’s and cross section, KK (Q2-4)

Software implementationdE/dx measurement

ADC gates aligned with respect to T1

33 cellsQ/Leff

Q/Lgeom

drift time (ns)

AD

C c

ount

s/cm

2Leff

Lgeom

1.8 s gate too shortCharge lost at large tdrift

Lengthened to 3.3 s

For 2002 data, calculate effective length for dE/dx

K

K 3-body

Bhabha

Tru

ncat

ed m

ean

(AD

C c

ount

s/cm

)

p (MeV)

K

Goal to use unambiguous K PID as tagKK DST’s a platform for dE/dx software


Milestone 8: BR’s and cross section, KK(Q2-4)


From N1 and N2 obtain sel and NKK

decay auto-triggers EmC: sel includes trig

Identification of 2-body decays

p* (MeV): m assumed

BR()/BR(): 281 pb1 ’02 data0.33000.00030.0025Systematics still under evaluation

K

KN1

K

K

N2

3 pts of ’02 scan give: = 210565 nb

m = 1019.450.11 MeV

= 4.070.21 MeV

Stat. error only

(nb)

s (MeV)


CPU available for data processing

Results of 3rd KLOE call for tenders (June 2003):

2002 2004 est.

Avg. L (cm2 s1) 5.41031 2.01032

Avg T3 (Hz) 1600 Hz 4100

Reconstruction time (ms/ev) 19 25

CPU’s needed to match T3 30 100

2004 working estimate is:

Optimistic in terms of LPessimistic in terms of • MB rate• MB rejection capability Maximum reasonable load assumption

Must also add• MC generation + reconstruction ( all: 370 ms/ev, 1M events = 5 CPU days)• DST production

60 to 74 B80 CPU’s for reconstruction on current farm (reallocatable)

10 IBM 7028-6C4 p630 41.45 GHz POWER4 servers

Doubles CPU power of offline farm~70 150 B80 CPU equivalents

18.6 TB disk space, SAN/FC interface3 increase of offline disk spaceTo be added to DST cache and AFS cell


Tape library: status and upgrade plans

Raw 100.9

Reconstructed 51.2

DST 4.1

Old MC 4.9

New MC 9.5

MC DST 1.6

Total 172.3

Before reversal 200

Current 300

Current usage (TB)

Capacity (TB)

Space crisis averted by reversal to high-density cartridges (40 60 GB)

Final decision on new library by end of year

New IBM Magstar technology: 3494-L12 • 300 GB/cartridge (roadmap for 1 TB)

Proposed initial installation:• 1000 cartridges installed (300 TB)• Space for 3600 cartridges (1080 PB)• 2 robots, 6 drives (40 MB/s each) • Placed in LNF computer center• Remotely attached to the KLOE servers

via FiberChannel/SAN• 7 months of raw+reconstructed data at

200 b1/s with 1600 Hz MB


Milestone 9:Repeat 2002 analyses with full data set


PDG ’02 2.197 0.026 (avg.)

KLOE ’02 2.236 0.003 0.01517 pb-1 ’00 data, Phys. Lett. B538 21Error will be reduced to 0.1%

(KS ())/(KS )• First part of double ratio for Re • Provides information on EM isospin

breaking in K decays• Can extract 0 – 2 if effective E

cutoff known for channel

0 – 2 0 – 2

PDG widthsCirigliano et al. ’01

(568) (456) PT estimate Gasser & Meissner, ’91

KLOE ’02 value for

()/()(483) (47.71.5) scattering

Colangelo et al. ’01

E (MeV)

Current measurement fully inclusivedBR()/dE in progress


KS ee

6 7 8 104

CMD-2 ’99

PDG eval. from KL

KLOE ’02Phys. Lett. B537 21

KLOE ’03 prelim.

KLOE preliminary (170 pb1 ’01 data)

BR(e) = (3.46 0.09 0.06) 104

BR(e) = (3.33 0.08 0.05) 104

BR(e) = (6.81 0.12 0.10) 104

e2/d.o.f = 1.5N(e) = 3854 92

Emiss – pmiss

MeV

Data MC

MeV

e2/d.o.f = 1.1N(e) = 3863 88


KS e: CPT and S = Q

KTeV ’02 result for AL

AL = (3.322 0.058 0.047) 103

KLOE preliminaryAS = (19 17 6) 103

First measurement of AS!

AS AL 0 implies CPT viol.

A ee

ee

Re x1 BRS(e)/S – BRL(e)/L

2 BRS(e)/S + BRL(e)/L

KLOE preliminaryRe x = (3.3 5.2 3.5) 103

Compare to CPLEAR ’98 Re x = (1.8 4.1 4.5) 103

280 pb1 from ’02 running + KLOE measurements of L, BR(KL e) on the way


(KL) /(KL )• Dominated by long-distance contribution (, , )

PT calculation of BR sensitive to P

• Dominates long-distance contribution to KL

Exploits performance of EmC for reconstruction of photon vertex

KLOE: L = 51.6 ± 0.4 ns

PDG: L = 51.7 ± 0.4 ns

KL 3(1/10 of sample)

Regeneration on BP disabled in MC

NA48 ’02: (2.810.010.02)103

PDG ’02: (2.820.08)103

KLOE ’03: (2.7930.0220.024)103

362 pb1 ’01+’02 data, Phys. Lett. B566 61

KL


KL charged particles

Lesser of Pmiss Emiss in or hyp. (MeV)

Very preliminary78 pb1 ’02 data

KL e

KL

KL

KL

Errors are statistical only!(including MC statistics)

Systematic errors also ~1-2% but not yet fully evaluated

KL contribution fixed

BR(KL ), similar analysis:KLOE (2.04 0.04)103

PDG ‘02 (2.084 0.032)103

BR KLOE PDG ‘02

KL 0.1320.002 0.1260.002

KL 0.2710.002 0.2720.002

KL e 0.3840.002 0.3880.003

Very preliminary!


Prospects for

16 Re / =BR(KL )

BR(KL )

BR(KS )

BR(KS )

Should scale down to 0.1% on full data set (~400 pb1)

KS

Statistical error: negligible

Systematic error:

Source Error (%)

Tagging 0.55 (’00 data) ~0.1 (’01-02

data) counting 0.20Trigger/t0 0.23

Tracking 0.26

Total error: 0.7% 0.4% Need at least 10 more data to reach the 104 regime

KL

Statistical error: ~1.5%Systematic error: ~2%, in progress

• Momentum reconstruction/resolution,

including DC occupancy effects• Effect of K3 decays on background

determination

• Separation of overlapping clusters• Regeneration


CP/CPT studies: longer term prospects

fKA

fKAe

S

Liff

f

With sufficient L, KLOE can use both absolute BR measurements and interferometry to measure many CP & CPT violation parameters of KSKL system

3 Im

f1

t2

t1e e

KL ,KS

KS ,KL

Example:f2

122

21 cos2

tme

t

Interference term in I(t |f1, f2)


KSKL )cos(2|)(| 2/|)|(||||2 tmeeetA ttt LSSL

A first glance at interference

KLOE preliminary340 pb1 ’01 + ’02 data

Fit with PDG values for S, L

2/d.o.f. = 43.7/47

m = (5.64 0.37)×1011 s1PDG ’02: (5.301 0.016)×1011 s1

First observation of quantum interference in relative decay-time distribution of KS, KL

No simultaneous events:same final state/ antisymmetric initial state

Peak position sensitive to m

Coherent KL regeneration on beam pipe

|t1 t2|/(KS)


• Asymmetries in K rates (~108) and Dalitz slopes (~105) signal direct CP viol.• Dalitz slopes give information on I = 1/2, 3/2 amplitudes for K 3 decays

K

BR(K)PDG ’02 fit (1.73 0.04)%

KLOE preliminary hep-ex/0307054441 pb-1 ’01+’02 data187 pb-1 counted as signal

(1.781 0.013 0.016)%

Preliminary fit to Dalitz plotF(X,Y) = 1 + gY + hY2 + kX2

Y T*( ) X T*()

PDGKLOE

0.01970.00540.00550.00260.0018k

0.0570.0180.0300.0100.013h

0.6520.0310.5860.0100.012g


Form factorF = F(p + p)

Ke (Ke4)

KLOE preliminary (441 pb-1 ’01 + ’02 data)BR(Ke4)= (2.43 0.20 0.22) × 10 –5

PDG fit: (2.1 0.4) 10-5

Best measurement: (2.54 0.89) × 10-5

Assuming I = ½:(K4

) = 2(K4) = ½(K40)

|F(K4)| = |F(K4)|

For K4, m = 0: = CF |F|2 |Vus|2

Angular distribution in , plane allows determination of 0

0 – 11

2 from kinematic fit

K4 K

K4 K K4

0 KL

Nobs = 40720Nbkg = 15218


Vus from K3 decays

KLOE will have BR(Ke3) soon

KLOE will measure all BR’s to 0.1% level and can also significantly improve , , L|Vus| = 0.2196 0.0019exp 0.0018th

For Ke3 modes:

0

0

20

1

2

1

BR

BR

2

1

f

f

V

V

us

us

exp th

Old data (Chiang ’72) or PDG fit valuesInclusiveness for Ke3?

|Vus| f(0)

Ke3 K

e3 K3 K

3

New E865 measurement, agrees better with current values of Vud

E865

KLOE preliminary KS e 170 pb1

Consistent with previous measurementsKLOE KS

KLOE preliminary KL 78 pb1

Also confirms previous measurementsKLOE KL


Parameters of the meson

Decay Method 0 (nb)

KK double tag 2080±42

KLKS KS, KL in EmC 1344±12

Kinematic selection 693±8

3 selection 58±1

Mass scale from CMD-2 resonance depolarization measurement

= 4.190.04 MeV

vis (

nb)

s (MeV)

2/Ndof

3/7

2002 scan: 3 points (m, m/2)~7 pb1 each

Preliminary– stat errors only

(ee) from AFB(ee)

s (MeV)

AF

B


Fit parameters:m, For , ,

|Adir|, dir Amplitudes of direct,|A|, contributions

m = 775.8 0.5 0.3 MeV

= 143.9 1.3 1.1 MeV

m0 – m = 0.4 0.7 0.6 MeV

m+ m = 1.5 0.8 0.7 MeVRelative decay intensities from integration of |A|2 over Dalitz plot

I = 0.937

Idir = 0.0085

I = 0.0002

(ee) = 92 15 pb

Interference:I + Idir + I 1

KLOE ’03: Phys. Lett. B561 5517 pb1 ’00 data, 2M events

X = T* T* (MeV)

Y =

T*

(M

eV)

N(X,Y)|p p|2

F(X,Y) |p p|2 |A + Adir + A|2


Study of Dalitz plot gives u-d quark mass difference and sheds light on isospin-breaking mechanisms

Fit with a, b, d free parameters,c, d fixed to zero2/Ndof = 33/38

0.0460.031

0.0340.027

1.08 0.014

PDG

0.050.03

0.200.03

1.050.01

KLOE

d

b

a

Stat. errors only

When treated as free parameters:c = 0.0080.010

d = 0.010.03Will limit C violation in

Expansion of Dalitz plot:A(X,Y) 1 + aY + bY2 + cX + dX2 + eXY

Y + 1 T*

KLOE preliminary, ~100 pb1

352K events

X T* T*


5104 95% CL PDG ’02 (GAMS2000)1.8105 90% CL Crystal Ball ’02 prelim

Violates C

KLOE preliminary hep-ex/0307042410 pb-1 ’01 + ’02 data

BR(31.6105 95% CL

Require 4 with fiducial cuts on E, Reclustering to eliminate 3 background

Kinematic fit

m() veto eliminates ee and 5 background

Background est.from data

Fit result to whole spectrum

Emax (MeV)

4Erad = 363 MeV


ggZssYdduuX ''' 2

1

2

1)'(

BR

BRR

Allows determination of mixing angle (P) and gg content of (Z)

BR()KLOE ‘02 (6.10 0.61 0.43) × 105

PDG ’02 (6.7 1.5) × 105

9.16.18.41

P 9.16.19.12

P

09.006.0

2 06.0 Z

KLOE ’02: 17 pb1 ’00 data, Phys. Lett. B541 45

R = (4.70 0.47 0.31) × 103

R measured from 3 final state

M() (MeV)

KLOE ’0217 pb1

124 evts

KLOE prelim.100 pb1

700 evts.


(f0) and (a0)

BR( f0) BR( a0)

qq 5105 2105

qqqq 3104 2104

KK 105 105

KLOE ’0217 pb1 ’00 data

PDG ’02

BR() (1.090.030.05)104

Phys. Lett. B537 21(1.080.19)104

BR()

(8.510.510.57)105(7.960.600.40)105 Phys. Lett. B536 209

(8.91.4)105

Scalar mesons f0 and a0 not easily interpreted as qq states

Precise measurements of BR’s and d/dM may distinguish between various models


, : updated spectra

’00 data’01+’02 data

KLOE preliminary: ~400 pb1 ’01 + ’02 data

’00 data’01 data’02 data

5

M (MeV) M (MeV)


: Dalitz plot

Statistics of ’01-’02 data set allow interference between f, , contributions to be studied in a model-independent way


Search for f0

Completely dominated by ee with from initial state (ISR)

Fit to M spectrum with:

• continuum from both ISR and FSR

• Signal (f0)

• f0-FSR interference

KLOE preliminary340pb-1 ’01+’02 data

M(MeV)

Best results for destructive interference

Potentially study contribution from Analysis in progress


(eehadrons)

s (GeV2)

d(ee )/ds

H(s)

KLOE preliminary: 140 pb1 ’01 data, 1.5M events

bare(ee )

s (GeV2)

nb

nb/G

eV2

hep-ex/0307051

H(s)

d()/ds

F = 1

F

obtained fromPHOKHARA


(eehadrons)

Angular cuts ensure:• High statistics for ISR events• Very low contribution from FSR-only events• Reduced background contamination

ISR: M

2 s

FSR: M

2 s s

However, for a, inclusive cross section is needed: ee(), from FS

ISR+FSR events eliminated by kinematic cuts, must be resummedNow evaluating this correction with PHOKHARA v3.0This correction not yet included: currently our largest error


Preliminary results for ahad

Results are compatible given systematic errorsDefinitive comparison must await final FSR corrections

CMD-2 378.6 2.7stat 2.3syst

a ds (ee) K(s)

Status of a (Davier et al., Aug ’03) ~70% of ahad from s < m

ee-based estimates dominated by CMD-2 (revised analysis Aug ’03)

No comparable data set (until KLOE!)

Use of data involves corrections for isospin breaking, etc.

374.1 1.1stat 5.2syst 3.0th (+7.5)

FSR

(0KLOE preliminary

a 1010 (0.37 < s < 0.92 GeV2):a

1.4% < 1% 2.0% < 1%In progress:


Preliminary results for ahad: ee vs.

KLOE data confirm discrepancy between ee and data for s > m2

Point-by-point comparison must await final FSR corrections

s (GeV2) a1010 KLOE a

1010 CMD-2

0.37-0.60 256.2 4.1 (+5.1-0) FSR 256.4 2.5

0.60-0.92 117.9 2.1 (+2.3-0) FSR 123.3 1.8

For s > m2, |F|2 from data

systematically 10-15% higher

|F|2(ee) |F|2()|F|2()

(Davier et al., Aug ’03)

s (GeV2)


Milestones for 2003: reprise

Q1 1. Preparations for IR modifications 100%

Q1-4 2. Maintenance of detector performance at 2002 levels 100%

Q1 3. QCAL upgrade 100%

Q1 4. Upgrades to raw-data filters 80%

Q1 5. Reconstruction of all data 100%

Q1 6. Preparation of DST’s for analysis 100%

Q1-4 7. MC production for 2002 analyses 90%

Q2-4 8. BR’s and cross section, KK 70%

Q4 9. Repeat 2002 analyses with full data set 80%


2004 run plans

Expected DANE performance• Luminosity up to 21032 cm2 s1

• Lifetime > 1 hr (was 0.6 hr in 2002)• 10 pb1/day, 200 pb1/month

DANE run plans (from KLOE perspective)• Start FINUDA data taking A.S.A.P• 250 pb1 for FINUDA in ~2 months• Realistically restart KLOE towards end of year• Run KLOE for ~1 year with goal of collecting 2 fb1


Milestones for 2004

Q1-4 KLOE data taking and maintenance

Q1-4 Online reconstruction, all data

Q1-4 DST production quasi-online

Q2 Installation of new tape library

Q3 Introduction of level-4 filter

Q1-4 MC production for 2003/2004 data set

Q3-4 BR(Ke3) and Vus

Q3-4 Hadronic cross section down to threshold

Q4 KS e BR and semileptonic asymmetry to 0.4%

Q4 BR’s for , decays


Additional information


Operations: January-June 2003

January February

Detector opened DC wire tension measurementsDC FEE work: 60 channels repaired Geometrical surveyHV boards for EmC repaired

MarchIR dismountedAssembly of new IR started

April DAQ on for tests of DC electronics

MayGas flowing in DC DC ADC gates widened 2 EmC HV crates fail and are replaced

June

QCAL upgrade completeEmC: 3 PMT’s replacedNew IR installedDC + IR resurveyed


Operations: July 2003-present

July

Detector closedField ramp cycles to test for deformations with new IRFinal status of DC tested: 86 dead channels (preamps near IR)EmC fully operational: ~1 hr of field-on cosmic-ray data takenWater crisis impedes collection of cosmic-ray data with field onDebugging of cryogenic system after KLOE/FINUDA compensators connected to main He loopNo running in sight: DC gas flow interrupted

SeptemberMachine operations starting upKLOE still waiting to take test data with field on


Offline upgrades for 2003

• 12 upgrade kits for Magstar drives in tape library

40 60 GB/cartridge

Total capacity 208 312 TB (5200 cartridges installed)

• 3.5 TB FiberChannel disk space for DST-access cache

• 0.5 TB FiberChannel disk space for AFS cell (analysis)

• 2 GB RAM for DB2 server (run/file database)


Offline computing resources: 2003

Tape library: 312 TB5200 60GB slots12 Magstar drives14 MB/sec each

RECONSTRUCTION, MC & DST PRODUCTION

32 Sun + 76 IBM CPU’s

Managed disk space: 6.5 TB0.8 TB SSA: I/O staging

I/O stagingReconstruction/MC/DST

2.2 TB SSA + 3.5 TB FC:DST’s maintained on disk

afs cell: 2.3 TBUser areasAnalysis/working groups

ONLINE FARMDAQ and calibration1.4 TB local disk used as reconstruction staging

ANALYSIS8 Sun + 16 IBM CPU’s

File servers

IBM B80 4375 MHzSun E450 4400 MHz

Catalyst6000

AFS serverDB2 server


2001-2002 trigger rates

Avg. L: 2.11031 cm2 s1

Avg. T2: 2000 Hz• 700 Hz cosmic rays• 500 Hz downscaled CR• 200 Hz physics• 600 Hz mach. bkg.

2001 running

Avg. L: 5.41031 cm2 s1

Avg. T3: 1600 Hz• 700 Hz cosmic rays• 0 Hz downscaled CR• 500 Hz physics• 400 Hz mach. bkg.

2002 running


Reconstruction load scenariosAssumptions: Standard B80 CPU’s

5 ms/ev overhead+41 ms/ev to track /Bhabha events+17 ms/ev to track 50% of MB events

2002:

Avg. L: 5.41031 cm2 s1

Avg. T3: 1600 Hz• 700 Hz cosmic rays• 500 Hz physics• 400 Hz MB

Reconstruction time 19 ms/ev

2004 working estimate:

Avg. L: 2.01032 cm2 s1

Avg. T3: 4100 Hz• 700 Hz cosmic rays• 1800 Hz physics• 1600 Hz mach. bkg.

Reconstruction time 25 ms/evLoad = 68%

3.9 kHz T338 ms/ev

2002

Saturation of 150 CPU offline farm2004estimate

7.7 kHz T319 ms/ev

MB rate (Hz)

Lum

inos

ity (b

1/s

)

Saturation of installed farm

Optimistic for luminosityPessimistic for MB, rejection…


Tape library scenarios

Assumptions:Current T3 configurationCurrent event sizes

• Raw 2.6 KB• Streamed decay 12.3 KB• Streamed Bhabha 9.2 KB

Current event classification• All events streamed• No CR/MB events streamed

Except:1:10 for Bhabhas < 45

Raw and reconstructed dataDuration refers to 100% uptime300+50 TB new+old libraries80 TB reserved for MC, DST’s…

Optimistic luminosityPessimistic MB, rejection…

Duration of 350 TB library space, months

MB rate (Hz)

Ave

rgae

lum

inos

ity (b

1/s

)

100 b1/s, 550 Hz MB13 months

200 b1/s, 1600 Hz MB7 months


DC Geometry revisedRealistic wire sags from survey implementedNew s-t relations from simulated cosmic raysDead wire simulation (map a function of run number)

EmC Geometry revised’s from events in data used to calibrate:• sampling fraction• energy response and resolution• time response and resolution• position dependence of aboveDead zones at module interfaces simulatedVariations in fiber attenuation length in endcaps simulated

Trigger Thresholds adjusted in terms of reconstructed EmC/DC hitsThresholds a function of run number

MC upgrades: detector simulation


New decay/ISR generatorEffective s sampled before decay channel chosen BR from up-to-date cross-section measurements vs. s

KL KS regeneration updated

Regeneration on Al in DC wall enabled, nuclear recoil treated, angular distributions modified

, f0, a0 generators updated

a0 and f0 mass spectra from fits to KLOE 2000 data

New generators for KS/KL andNo cutoff at low E

MC upgrades: event generators


Level-3 DAQ filter

T3 filter installed before recorder to review and enforce cosmic-veto decision

• Fast clusteringVeto enforced if t12 = cosmic ray TOF

• Fast tracking on coarse lattice of DC wiresVeto enforced if no activity near IR

• Runs on online farm Can handle 1.9 KHz cosmic-veto rate

event builder

monitoringprocesses

T3 filter recorder Eve

nts

cos

Recovery of ee

ee, barrel

Eve

nts

dN/d(cos ) 1 + cos2

cos


Data-summary tapes

DST typeInput size

(GB)Output

size (GB)Total time

(days)

KK 5270 2070 50

KSKL 8900 1350 15

Neutral radiative2920

13010

Charged radiative 640

(80 pb1) 250 60 1

As of May, entire KLOE data set (452 pb1) available as DST’s

Total space: 4.3 TBMC DST’s add ~1.6 TB

DST disk cache: 5.7 TBVast majority of DST’s disk-accessed


KS ()KS

Phys. Lett. B538 21 (2002)

KS ePhys. Lett. B537 21 (2002)Preliminary update with 2001 data

KS mass KLOE Note 181 (http://www.lnf.infn.it/kloe)

KL KL Phys. Lett. B566 61 (2003)

KL charged Preliminary results

CP violation & interference

Prospects & preliminary results

K hep-ex/0307054, submitted to EPS’03, LP’03

K e Preliminary results

Vus Prospects & preliminary results

Kaon physics at KLOE


meson parameters Preliminary results

Phys. Lett. B561 55 (2003)

Preliminary results

hep-ex/0307042 submitted to EPS’03, LP’03

Phys. Lett. B541 45 (2002)

Preliminary update with 2001 data

f, aPhys. Lett. B536 209 (2002), B537 21 (2002)

Preliminary update with 2001 data

(ee hadrons) hep-ex/0307051 submitted to EPS’03, LP’03

Hadronic physics at KLOE


Tagged KL and KS “beams”

KL tagged by KS vertex at IPEfficiency ~ 70% (mainly geometrical)KL angular resolution: ~ 1°KL momentum resolution: ~ 2 MeV

KKSS

KKLL 2 2

KS tagged by KL interaction in EmCEfficiency ~ 30% (largely geometrical)KS angular resolution: ~ 1° (0.3 in )KS momentum resolution: ~ 2 MeV

KKLL “crash”“crash”= 0.22 (TOF)= 0.22 (TOF)

KKSS ee


Extraction of 0 – 2

• Extraction of K amplitudes from measured widths must take into account effective cutoff for processes with in final state

• Including isospin-breaking EM effects:

II iII

iI ee AAA

III I = EM phase shift

K decays actually measure 0 – 2

For 0 – 2, need theoretical input (0 – 2)

2

20

20

20

2000

20

43

32

31

31

32

i

ii

ii

e

ee

ee

AA

AAA

AAA

Conventional extraction of strong phase shifts from KS, K decays:

0 – 2 0 – 2

PDG widthsCirigliano et al. ’01

(568) (456) PT estimate Gasser & Meissner, ’91

KLOE ’02 value for

()/()(483) (47.71.5) scattering

Colangelo et al. ’01


m(KS) KLOE preliminary497.583 0.005 0.020 MeVKLOE Note 181

2001 peak scan: 29 pts, 0.5 pb1

W (MeV)

(e

e

KSK

L)

(b)

KSKL KS

Measurement of KS mass

Momentum scale calibrated to CMD-2 ’01m() = 1019.4830.0110.025 MeV

W = M(ee P = p() p()

m(KS) = W2/4 P2

m/m 0.05(P/P)

CMD-2 NA48 KLOE

MeV


Tagging via K or K6105 tags/pb1 for measurement of absolute BR’s

KK

KK

ee

m2 (MeV2)

Ke3 event selection by charged daughter mass• p, L from DC track• TOF from EmC with

tdecay from 0 clustersKe3

K3

Charged kaon decays


, : mass spectra

f0 parameters compatible with 4q model a0 parameters not well described

Kaon-loop model for S contribution (VMD)Pointlike contribution for f0

f0

f0- interf.a0

dBR

/dM1

08 (

MeV

1)

dBR

/dM1

06 (

MeV

1)

M (MeV) M (MeV)


Theory 0.8%Radiator function (H) 0.5%Vacuum polarization 0.1%Luminosity 0.6%

FSR corrections 2.0%

Experiment 1.4%Acceptance 0.3%Trigger efficiency 0.2%Tracking efficiency 0.3%Vertexing efficiency 1.0%/e identification 0.1%Track mass cut 0.2%Background subtraction 0.5%Unfolding 0.6%

Systematic errors for ahad

Documents

KLOE: Status and recent results Commissione Scientifica Nazionale I Lecce, 22 September 2003 M. Moulson, LNF for the KLOE Collaboration New KLOE IR installed