Transcript Kein Folientitel

The AMANDA-II Telescope
- Status and First Results Ralf Wischnewski / DESY-Zeuthen
for the
AMANDA Collaboration
TAUP2001, September 2001, LNGS/Italy
[email protected]
HE ’ Telescopes …
HE -Telescopes are still ”novel” instruments
proving sensitivity to their primary channels
m 
m
NmX
e N  e X
 from Astrophysical Sources ( Point / Diffuse )
 calibration possible only with atmospheric ‘s
Outline
1. AMANDA-II Detector
2. Calibration of the AMANDA-B10 detector with
Atmospheric ‘s
 Physics results from B10/1997, see A.Hallgren
3. Performance & Sensitivity of AMANDA-II
AMANDA-II Construction
1996 4 strings, coax transmission OMs
total 86 OMs
 AMANDA-B4
1997 +6 strings, twisted pair transmission OMs
total 302 OMs at 10 strings
 AMANDA-B10
1998 +3 strings, fiber transmission OMs
total 424 OMs
 AMANDA-B13
2000
+6 strings, fiber and digital transmission OMs
total 677 OMs at 19 strings
 AMANDA-II
2003-2008 ~80 strings with 60 OMs each, total 5000 OMs
 ICECUBE
AMANDA-II
(February 2000)
• 677 PMTs
• 19 Strings
• d = 200 m
• h  500 m
7 3
 Vgeo1.6 10 m
Trigger:
• Majority
>23 OM in <2.5 ms
• String Trigger
Atmospheric Muon Background
-Signature: Upgoing muon
m from CR
Background: Misreconstructed
atmospheric muons
m from atm.
Noise/Signal =
m
/
m (


atm
5…6
) = 10
Zenith of m‘s triggering AMANDA
Maximum Likelihood Reconstruction
Variation of track parameters until the spectrum
of arrival times has highest Likelihood.
Scattering results in a distance dependent timedelay relative to the Cherenkov cone.
Atmospheric -Analysis
• Two (largely) independent analyses were done for the
1997 data set of Amanda-B10 with respect to
- methodical aspects (Cut definitions, minim.functional)
- data cleaning (electronic noise ...)
• Used same signal & atmospheric muon Background MC
(BG-MC) and track model (ice optics and single muon)
Analysis A: Overview
• Conventional Likelihood description (no zenith weight)
• Multi-photon Likelihood and hit-probabilities
• Background rejected by specific cuts:
e.g. cascade fit for muon bremsstrahlung.
• Reject instrumental BG by hit-topology
• Cuts developed with emphasis on BG-MonteCarlo
• Deterministic method to define a small set of final Cut
parameters and their values (  CutEval)
Neutrino Analysis Chain (A)
Pre-processing :
Nevents bg
sig
S/N
1•109 0.95
0.95
1 : 2•105
5•107 5•10-2 0.37
1 : 3•104
4• 105 4•10-4 0.15
1 : 5•102
1• 104 1•10-5 0.07
1 : 30
223
10 : 1
Hit cleaning, Calibration ...
Level 1:
Fast track approximation (line fit)
Simple up/down cut
Level 2:
Time-likelihood reconstruction.
Simple cuts (zenith,L,Ndir)
Level 3:
First cut optimization.
Full likelihood reconstruction.
Level 4:
Final Neutrino Cuts (CutEval)
2•10-8 0.04
Analysis A: Neutrino Cuts & Sample Size
Find (minimum) set of CutVariables, which optimize
the signal to background
efficiency
Define „Sample Quality“:
Q = -log (NBG /NTL)
Data
BG-MC
Sig-MC
Sample Quality, Q
Neutrino Events – Statistics
Analysis A Analysis B Overlap
Experiment
Combined
223
378
204
265
102
119
325
524
-MC (new)
237
-
-
-
Background
10%
10%
-
-
-MC (stand)
130.1 days live time
Systematic Uncertainties
Calibration
Optical properties
Physics
Time
Geometry
Amplitude
OM sensitivity
Bulk Ice
Vertical structures
Hole Ice
m-prop./light yield
-flux
-oscillations
-5%
-5 %
<5%
-20 %
± 10 %
± 10 %
-50 %
± 20 %
± 30 %
-20 %
Nb. of trigger (Mill.)
Neutrino candidates
 - Luminosity versus day in 1997
Number of Neutrino events and total number of triggers in 1997
(analysis A). Total live time was 130.1 days.
Events /130d /0.1
Zenith angle distribution
Cos(zenith)
Analysis A
Energy distribution
E  ~ .07–3.4 TeV
(MC)
10 GeV
100 GeV
1000 GeV
Celestial distribution
Pointing:
3ª
-
4ª
No clustering
Combined sample
325 events
130.1 days live-time
Ratio Osci/No oscci
Events /130d /0.05
Oscillations:Zenith dependence
(MC)
Cos(zenith)
Cos(zenith)
Oscillations: Neutrino Energy
(MC)
Summary: B-10 atmosph. ’s
•The AMANDA-B data from 1997 (130.1 days live time) have been
analysed for neutrino induced events
• 2 independent analyses find a total of 325 neutrino events with
less than 10% background contamination
• Results are consistent with the MC expectation for background
(atmospheric muons) and signal (atmospheric neutrinos) within
(still relatively large) systematic uncertainties
• Improvement expected from
- better local ice-properties & OM sensitivity calibration
- improved MC light tracing & HE muon light yield
•AMANDA has reached „design luminosity“ !
Amanda-II
•Larger effective area by 9 more strings (>double OMs)
•New technologies
- improved timing
- MultiHit resolution (muon bundles & HE events)
- reduced electronic noise
- Upgrade: full waveform sampling (FADCs) in 2002
•Triggering
- HE events by majority trigger
- E<100 GeV by string trigger
- EAS-Array (SPASE) as efficient veto for UHE events
•Improved angular sensitivity range; resolution ~ 2 degrees
Amanda-II
•Amanda-II analysis for 2000 started in spring 2001
(1.2 TB Data back from Spole)
•Neutrino Analysis profits substantially from Amanda-B10
analysis.
Tuning to new geometry and hardware is under way.
•Minimum bias and low neutrino-cut level data look ok vs. MC.
•MC-results given below for „final AM-II neutrino cuts“ (Aeff, ,...)
are still preliminary.
Amanda-II: Detector response
Number of hit Optical Modules
All Strings (1-19)
Inner Strings (1-10)
Outer Strings (11-19)
Data - dots
MC - line
Amanda-II: Detector response
Minimum bias Zenith & Azimuth acceptance
Event Rates: m and e
Trigger Level
ATM
m
e
Atmospheric
AGN
m - 11000 (CC)
-
130 (NC)
e -
160 (CC)
-
9 (NC)
AGN
(E-2 10-6 GeVcm- 2s- 1sr-1 )
m -
853 (CC)
e -
103 (CC)
Atmospheric m MC: Energy response
Trigger
Level
Aeff(E) much
improved
compared to
AMANDA-B10
After BG
rejection
Atmosph. m MC: Angular Sensitivity
Trigger
Level
~200 atm  per angular
bin & livetime-year
After B
rejectio
Nearly uniform angular
sensitivity to horizon
up
horizon
AMANDA-II:
a horizontal
m event
( preliminary Am-II neutrino - cuts )
Am-II: Effective Area vs zenith
200
E m = 10 TeV
Aeff depends
sensitively on the
physics objective!
A eff [ x10 3 m2]
150
Am-II Trigger
100
Am-II GRB Cuts
Point source
sensitivity is
uniform to near
horizon
50
Am-II Point Cuts
Am-B10 Point Cuts
0
-1
-0.8
-0.6
-0.4
Cos(theta)
-0.2
0
Amanda-II - Effective Volume
<V eff > (km3 )
0.5
Veff (m) is
0.3-0.5 km3
0.3
•Rm > 10 km
0.1
101
102
E (TeV)
103
(for Point Source Cuts)
Diffuse Flux m
-2
Convolved
energy
resolution
-2 -1 -1
[GeVcm s sr ]
10
Atm. 
10-4
AMANDA-B10 ('97)
10-6
E 2(dN/dE )
AGN Core
3C273
10-8
AMANDA-II (3 yr)
Anticipated
sensitivity
IceCube
AGN Core
-10
10
103
104
105
E (GeV)
106
107
New techniques
Upper Bounds on Diffuse Flux
Baikal NT-200
AMANDA-II
Anticipated
sensitivity
Point Sources
10-4
-2 -1
[GeVcm s ]
Atm. 
10-6
AMANDA-B10 ('97)
AMANDA-II (3 yr)
E 2(dN/dE )
Mk501 (=)
10-8
Atmospheric F +
energy resolution
3C273
IceCube
Crab
AGN Core
-10
10
102
103
104
E (GeV)
105
106
Point Sources
10-4
-2 -1
[GeVcm s ]
Atm. 
10-6
AMANDA-B10 ('97)
AMANDA-II (3 yr)
E 2(dN/dE )
Mk501 (=)
10-8
Focus on
Mk501 as
example
3C273
IceCube
Crab
AGN Core
-10
10
102
103
104
E (GeV)
105
106
UHE events
• Earth shadowing of ‘s becomes relevant for E > 1 PeV
•Horizontal and „downgoing“ events dominate
• New search strategies under development
- Total energy cuts & Energy flow
vs. track reconstruction
- EAS-Veto by surface Array SPASE
Downgoing  -induced MuonFlux
Angular distribution
Most events are
horizontal.
EeV sources cut off
very quickly below
horizon.
Direction provides
additional BG reject.
Up
Down
Summary
•AMANDA-B10
Final Analysis (1997) yields 325 HE neutrino events
The first high statistics -event sample for an
UWater/Uice Telescope
Proof of principle of operation in Antarctic scattering ice.
•AMANDA-II
5
2
>10 m Trigger Area - the Largest Muon and Neutrino
Telescope ever built.
Improved performance compared to B10-Telescope
Horizontal Angular acceptance
Event rates of 4-5 atm. ‘s per livetime day,
800-1000 ‘s for year-2000.