Stefano Di Falco & Achim Denig Universität Karlsruhe The new Montecarlo and Report on the luminosity measurement KLOE Workshop Conca Specchiulla, June 10 th - 12 th , 2002
Mar 26, 2015
Stefano Di Falco & Achim DenigUniversität Karlsruhe
The new Montecarlo
and
Report on the luminosity measurement
KLOE Workshop
Conca Specchiulla, June 10th- 12th, 2002
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
2
selection for small angle
• Trigger
• FILFO/par=5
• PPGTAG (tracks, vertex, MTRACK,…)
• Pion fiducial volume: 40o < < 140o New!!
• Photon fiducial volume: 0o < < 15o New!!
• E >10 MeV (in fact cut on E)
• Likelihood
• MTRACK
DATA…MC
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
3
HADR and Montecarlo generators
The hadronic cross section measurements needs MC help in more steps:
• acceptance: is determined using the angular distribution and ‘’
• correlations between selection efficiencies (mainly FILFO)
• MTRACK cut: cuts on the tail due to the radiation of one or more ISR
Furthermore:
•Final state radiation and ISR-FSR interference: needs to be know only with
rough approximation in the small angle analysis
• Radiation Function: used to extract the pion form factor F from the M2
• Luminosity determination (BABAYAGA)
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
4
Montecarlo generators
ISR FSR me precision
EVA*(J.Kühn et al.)
+coll.rad. pointlike 0 ~2%
PHOKHARA*o
(J.Kühn et al.) NO 0 ~0.5%
AFRODITE*(F.Jegerlehner et al.)
pointlike F
0 ?
KKMC (e+eff+n)(S.Jadach et al.)
CEEX NO 0 (<0.2%)
* Interfaced with Geanfi oModified to impose cuts on the pp system instead of the s
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
5
Comparison Eva-Phokhara:
EVA doesn’t give the correct cross section for < 0.1o:
min= 0o now available! Higher , lower background, no smearing problems !
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
6
Comparison Eva-Phokhara: Q2
EVA was generating too many events at small angle, nonetheless the Q2 spectrum shape was correct:
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
7
Comparison Eva-Phokhara: smearing
Nonetheless the cut >5o was causing troubles especially at large Q2
because of the resolution on :
+10%
-10%
The discrepancy was due to the EVA, with PHOKHARA it’s no more present:
EVA
PHOKHARA+10%
-10%
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
8
Comparison Eva-Phokhara: TOT
ISR in EVA doesn’t include the vacuum polarizationand O(2) terms
The global effecton for large
is ~2% (flat in Q2)
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
9
Contamination from FSR
ISR radiation (PHOKHARA)TOT= 14.3 nb
FSR radiation (EVA) TOT= 0.03 nb
(~0.2%)
FSR
ISR+FSR
M2
%
• ISR photons are mostly emitted collinear to the e+e- beams (small angle) while the FSR ones prefer the pion direction.
At leading order = and then the cut: < 15o or > 165o
suppresses FSR when 40o < < 140o
(also +-0 background is suppressed)
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
10
Summary & Outlook on generators
A new Montecarlo, PHOKHARA, with O(2) ISR, is available and has been adapted to our experimental cuts and interfaced with Geanfi. It allows to lower to 0 the lower cut on the missing momentum (‘’) resolving the smearing problems and decreasing the impact of FSR and background
The work to include FSR in Phokhara is going on.
Results are waited also from Afrodite (Berlin-Cracow group) and KKMC (work not yet started) to evaluate the systematics as discrepancies of the different theoretical calculations.
A realistic multiphoton generator including FSR will be very important to perform the measurement at large angle needed to cover the low M
2 region.
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
11
Luminosity counters
Data
Trigger
EmCReconstruction
FILFO
LAB ( >45°)
Golden Bhabha ( > 25° including )
Trigger Luminosity(Bhabhas on Barrel)
DCReconstruction
GGL: > 45°)
DC - Events Selection: 2 Tracks from I.P. with opp. charge
First Hits > 50 cm separated 400 MeV < E+ - < 600 MeV
VLAB > 55°)
B and Energy calibrated
NO
NO
YES
YES
YES
EMC - Events Selection: 300 MeV < E+ - < 800 MeV, | TCL1 - TCL2 | < 4 ns, N Hits DC > 50,
cos( Acoll. (3d) ) < - 0.975
Effective(nb)
~ 800
~5700
~ 119
~ 800
~ 425
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
12
VLAB systematics
The luminosity is given as usual by: L = N / EFF where EFF is the theoretical cross
section evaluated from MC including the efficiencies for the selection cuts
MC-Data difference in:
1) Acceptance2) Efficiency3) Background contamination
Variations in time due to Running Conditions:
s , p , L-R asymmetries not const.,
Beam position not const. EmC - Calibration (DC?) Machine Background
Monitoring ToolHas been studied in the past: MC and Data distributions agree very well <1% level
The systematics of the measurement are given by :
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
13
Acceptance (checks)
TOTAL CROSS SECTION (CORNER REGION EFFECT? ): MC
MCDATA
o
o
o
o
o
o
N
N
N
N
N
N
40
55
40
55
40
55
= 0.2 0.5 %
%
e(degrees) e (degrees)
Relative difference
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
14
Efficiencies
Cluster Inefficiency:
Pmiss > 20 MeV| miss - 90o | < 35o
| Tcl | < 10 ns | Rcl | > 10 cm
additional Ncl = 0
“Tracking Inefficiency”:
Ncl=2Ntr=1
To be refined !
Data-MC: better than 1%
Data-MC: better than 1%
Monitoring…%
%
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
15
BackgroundLAB - Sample VLAB - Sample
LAB - Sample VLAB - Sample
Removed byE+- > 400 MeVCutAfter Likelihood
Before Likelihood
0.3 %
backgr.
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
16
Effect of run conditions: s
Suppose s = +1 MeV
Two effects:
• Change in Bhabha cross section:
-0.45 % L / L
• Change of the efficiency (95.65%) for
2 tracks with E + - >400 MeV
0.05 0.05 % L / L
L /
L %
Radiative Tail well described by Theory
BABAYAGA
GEANFI
Data
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
17
Effect of run conditions: beam position
Effect of a shift z=–1 cm of the interaction point on the cut on Polar Angle of Clusters
Corresponds to a left - right asymmetry = 1%
Variations in single bins reach ~2% but compensate each other…
VLAB(only barrel)
0.03 0.01 % L / L
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
18
GGLLAB
Additional infos on machine background
Mean value of not associated hits
Mean of Nclu
Mean of ETOT-E1-E2
Accidental rate inside+50 < T< +70 ns( T0= < t-l/c > )
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
19
Cross check VLAB – and GBHA
VLAB - GGL VLAB - GBH2002 dataRuns > 85nb-1
stat.error @ 100 nb-1: VLAB = 0.5% GGL = 1.0% GBH = 0.1%
= 0.3 %RMS = 1.0 %
= -2.8 %RMS = 0.6 %
S. Di Falco & A. Denig – Conca Specchiulla, June 10th -12th, 2002
20
Summary & Outlook (luminosity)
Error on Luminosity enters in our - cross section measurement as
absolute normalization Precision required: <= 1% ?!
VLAB sample is more stable and reliable than the previous LAB
Systematic error: - theoretical error estimated 0.5%
- no evidence for experimental effects larger than 1%
- agreement - VLAB ~0.3% (similar theoretical uncertainties)
In the latest version VLAB selection includes the monitoring of many variables
useful to determine the quality of the run (and eventually discard the bad ones)
• Works in progress:
• description of hard Bremsstrahlung, ….
• Cross check of Lumi-Bhabha-measurement by looking at - data