Effects of architectural and environmental issues on a km3 detector

1
Effects of architectural and environmental issues
on a km3 detector
R. Coniglione for the NEMO collaboration
Istituto Nazionale di Fisica Nucleare- Laboratori
Nazionali del Sud

• ? Effects of environmental parameters on a km3
  detector
• Optical background
• Absorption length
• ? Architectural effects on the detector muon
  effective areas
• ? Mass production of atmospheric muon background
  (work in progress)

2
The environmental effects
3
Simulation INPUTS

• ANTARES codes modified for km3 detectors LNS
  improvements
• surface m generation
• muons with E-1 spectrum - isotropic distribution
• can radius dmax/2 200 m
• Capo Passero absorption length (70 m _at_ 450 nm)
• Background 35, 60, 120 kHz
• Aart reconstruction with different triggers
  (LNS routine for triple local coincidences)

!!!! In order to get a fair comparison, quality
cuts on the reconstruction are applied to get
similar angular resolution (? 1 at 1 TeV)
4
The geometry

• Characteristics of the simulated km3 detector
• Square array - 140 m tower spaced
• - Dimension 1.14 1.14 0.68 0.88 Km3
• Tower characteristics
• total height 830 m
• instrumented 680 m
• number of bars 18
• number of PMTs per bar 4
• number of towers 81
• number of PMTs 5832
• bar length 20 m
• bar vertical distance 40 m
• PMT 10

5
Trigger for reconstruction

• The reconstruction procedure is based on a
  maximum likelihood method
• all the hits with amplitude lt 0.5 p.e. are
  rejected
• a causality filter with respect to the highest
  amplitude hit is applied
• a linear prefit is applied on a sub-set of hits
  that passed the following conditions
• (they are part of a local coincidence) OR (hit
  amplitude gt2.5 p.e.)
• Local coincidences hit with 20nsDt(L)
  difference in time in different PMT

Local coincidences
2/2 at least 2/4
at least3/4
6
Trigger and background rates
Aeff vs Em
Aeff vs Em
Aeff vs Em
35 kHz
60 kHz
120 kHz

• ? 2/4
• 3/4
• ? 2/2

• ? 2/4
• 3/4
• ? 2/2

• ? 1.5p.e.
• ? 2/2
• 3/4

Median vs Em
Median vs Em
Median vs Em

• Trigger effects are strongly dependent on the
  background rates
• Local coincidences (3/4) are relatively more
  efficient at higher background rate in particular
  at muon energy lower than 10 TeV

7
Effective muon area vs background rates
The best trigger for each rate value is applied

• ? 35 kHz - 2/4 local coinc.
• 60 kHz 3/4 local coinc.
• ? 120 kHz 3/4 local coinc

Even with local coincidence trigger (3/4) a
stronger and stronger effective area reduction is
observed with increasing background rates
8
Effects of labs
Aeff vs Em - 20 kHz
Aeff (70m)/Aeff (50m) vs Em
? l abs 70m _at_ 440 nm ? l abs 50m _at_ 440 nm
Aeff (labs 70m)/Aeff (labs 50 m)
At muon energy lower than 3TeV effective areas
higher than 20
9
The architectures
10
The architectural comparison
Compromise between performance and technical
feasibility

• Performance-gt Angular resolution and effective
  areas for
• up-going muons (Em ?102?108 GeV) in the whole
  angular range
• down-going muons (Em from 103 GeV) -gt moon
  shadow

• Studied architectures with
• about 100 structures
• about 5000-6000 OM
• tower height lt1000m

11
The geometries
String-dh_140_20
NEMO_140
String-d_125_16
5832 PMT 81 strings String height 680 m
(18floors/string) V 0.88 km3
5800 PMT 100 strings String height 912 m (58
PMT/string) V 1.15 km3
5832 PMT 81 towers String height 680 m
(18floors/tower) V 0.88 km3
12
The effective muon ares vs Em
Up-going muons
35 kHz background labs 70m_at_450nm

• ? string-dh_140_20
• string-d_125_16
• ? NEMO_140

13
Angular resolution vs qm
Up-going muons
Em 102 ? 103 GeV
Em 103 ? 104 GeV
Q. Cut applied

• ? string-dh_140_20
• string-d_125_16
• ? NEMO_140

Horizontal
Vertical
The string detectors show a worst angular
resolution for near vertical muons
14
Down-going muons
Down-going muons
35 kHz background labs 70m_at_440nm

• ? string-dh_140_20
• string-d_125_16
• ? NEMO_140

The string-d_125_16 detector shows a worst
angular resolution and Aeff for down-going muons
15
Atmospheric muon simulations with CORSIKA
16
Atmospheric muons at the sea level and CORSIKA
A mass production of atmospheric muons down to
the sea level has been undertaken in
collaboration with ANTARES (J.Brunner).
Full atmospheric shower simulation with CORSIKA
(version 6.200)

• Adronic interaction model -gt QGSJET and GHEISHA
• curved version for horizontal showers and
  flat for vertical showers
• Angular primary distribution
• Primary ions -gt p, He, N, Mg, Fe
• Primary energy -gt 10-105 TeV/nucleon
• Primary zenith angles gt 0? 85
• Energy threshold for muons at sea level -gt 0.5
  TeV for ions between 0 and 60 and 1 TeV for
  ions between 60 and 85
• Slope primary spectrum ? E-2

D.Heck et al.,Report FZKA 6019(1998).
Forschungzentrum Karlsruhe http//www-ik.fzk.de/c
orsika/physics_description/corsika_phys.html
17
Propagation to the detector

• Muons produced at the sea level are propagated to
  the detector (NEMO detector 81 tower 140m spaced,
  depth 3400m) with MUSIC (can size changes as a
  function of the Rbundle )
• Muon light and reconstruction with the ANTARES
  software (20kHz background, labs 70m_at_440nm)
• Only one track reconstructed per event
• Events are weighted in order to reproduce the
  observed atmospheric muon flux (parameterization
  by J.R. Horandel Astr. Phys. 19(2003)193)

Propagation of few files already done (work in
progress) -gt preliminary results
18
Propagation on a Km3 detector 3400 m depth
Comparison between different simulation codes
Muon spectra at the detector can
Can dimension H 956m R 947m -gt V2.7 km3
Number of m per second
Black CORSIKA Red HEMAS Green OKADA
log10Em(GeV)
Em gt 300 GeV
Number of m per second
cos(qm)
19
Muon shower multiplicity
Black all at the can Red recon. with q. cut
7.8 Green ric cut-7.8 up-going
20
The statistics and life time
jobs run in Catania jobs run in Lyon
1 year 8760 h
21
CPU times and GRID
CPU time for all the statistics 17.2 years Disk
space (zipped files) 576 G
For the jobs run in Italy ? GRID
GRID in Italy

• Tutorial on GILDA https//gilda.ct.infn.it/
• GILDA is a virtual laboratory to
  demonstrate/disseminate the strong capabilities
  of GRID computing
• Mass production on INFN-GRID
• http//grid-it.cnaf.infn.it/

22
Summary

• A study on the possible trigger conditions on a
  tower based detector has been undertaken -gt
  triple local coincidences are more efficient with
  increasing background rates.
• Stronger and stronger reduction on effective
  areas with increasing background rates up to Em
  104 GeV.
• Comparison of effective muon areas as a function
  of labs shows an improvement higher than 20 for
  muon energy lt3TeV
• Architectural study on tower vs string based
  detector performance shows that the detectors
  based on string have a bad angular resolution
  for Em lt 104 GeV up-going muons with qmlt35.
• A full shower simulation is necessary to evaluate
  the detector response to high order muon
  multiplicity events. Full atmospheric muon
  simulation with very large statistics has been
  undertaken and the analysis is in progress.