doi:10.1016/j.nuclphysbps.2004.10.003

Results from the AMANDA neutrino telescope

C. de los Heros

, M. Ackermann

, J. Ahrens

, H. Albrecht

, X. Bai

, R. Bay

, M. Bartelt

S.W. Barwick

, T. Becka

, K.-H. Becker, J.-K. Becker, E. Bernardini

, D. Bertrand

D.J. Boersma

S. B¨oser

, O. Botner

, A. Bouchta

, O. Bouhali

, J. Braun

, C. Burgess

, T. Burgess

, T. Castermans

D. Chirkin

, B. Collin

, J. Conrad

, J. Cooley

, D.F. Cowen

, A. Davour

, C. De Clercq

, T. DeYoung

P. Desiati

, P. Ekstrom¨

, T. Feser

, T.K. Gaisser

, R. Ganugapati

, H. Geenen, L. Gerhardt

A. Goldschmidt

, A. Gross, A. Hallgren

, F. Halzen

, K. Hanson

, R. Hardtke

, T. Harenberg,

T. Hauschildt

, K. Helbing

, M. Hellwig

, P. Herquet

, G.C. Hill

, J. Hodges

, D. Hubert

, B. Hughey

P.O. Hulth

, K. Hultqvist

, S. Hundertmark

, J. Jacobsen

, K.H. Kampert, A. Karle

, J. Kelley

M. Kestel

, L. Kop¨ ke

, M. Kowalski

, M. Krasberg

, K. Kuehn

, H. Leich

, M. Leuthold

I. Liubarsky

, J. Madsen

, K. Mandli

, P. Marciniewski

, H.S. Matis

, C.P. McParland

, T. Messarius,

Y. Minaeva

, P. Mioˇcinovi´c

, R. Morse

, K. Munic¨ h, R. Nahnhauer

, J. W. Nam

, T. Neunh¨offer

P. Niessen

, D.R. Nygren

, H. Ogel¨ man

, Ph. Olbrecht

, A.C. Pohl

, R. Porrata

, P.B. Price

G.T. Przybylski

, K. Rawlins

, E. Resconi

, W. Rhode, M. Ribordy

, S. Richter

J. Rodr´ıguez Martino

, H.-G. Sander

, K. Schinarakis, S. Schlenstedt

, D. Schneider

, R. Schwarz

A. Silvestri

, M. Solarz

, G.M. Spiczak

, C. Spiering

, M. Stamatikos

, D. Steele

, P. Steffen

R.G. Stokstad

, K.-H. Sulanke

, I. Taboada

, L. Thollander

, S. Tilav

, W. Wagner, C. Walck

M. Walter

, Y.-R. Wang

, C.H. Wiebusch, R. Wischnewski

, H. Wissing

, K. Woschnagg

, G. Yodh

Division of High Energy Physics, Uppsala University, S-75121 Uppsala, Sweden

DESY-Zeuthen, D-15738 Zeuthen, Germany

Institute of Physics, University of Mainz, Staudinger Weg 7, D-55099 Mainz, Germany

Bartol Research Institute, University of Delaware, Newark, DE 19716, USA

Dept. of Physics, University of California, Berkeley, CA 94720, USA

Fachbereich 8 Physik, BUGH Wuppertal, D-42097 Wuppertal, Germany

Dept. of Physics and Astronomy, University of California, Irvine, CA 92697, USA

Universit´e Libre de Bruxelles, Science Faculty CP230, Bvd. du Triomphe, B-1050 Brussels, Belgium

Dept. of Physics, University of Wisconsin, Madison, WI 53706, USA

Dept. of Physics, Stockholm University, SE-10691 Stockholm, Sweden

University of Mons-Hainaut, 7000 Mons, Belgium

Dept. of Physics, Pennsylvania State University, University Park, PA 16802, USA

Vrije Universiteit Brussel, Dienst ELEM, B-1050 Brussels, Belgium

Dept. of Physics, University of Maryland, College Park, MD 20742, USA

Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA

Blackett Laboratory, Imperial College, London SW7 2BW, UK

Physics Dept., University of Wisconsin, River Falls, WI 54022, USA

Dept. of Chemistry and Biomedical Sciences, University of Kalmar, SE-39182 Kalmar, Sweden

Departamento de F´ısica, Universidad Sim´on Bol´ıvar, Caracas, 1080, Venezuela

We review recent results from AMANDA on the search for cosmic point sources of neutrinos, both in the diffuse

and point-like channels. Assuming a E

−

spectral shape of the neutrino energy at the source, we derive limits on

the diffuse ν

flux as well as in the all-flavour diffuse flux from the cascade search. We report limits on selected

point sources as well as on GRB searches. We present results on primary cosmic CR composition in the range

∼

100 TeV-PeV obtained with the help of the SPASE air shower array run in coincidence with AMANDA.

Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

www.elsevierphysics.com

doi:10.1016/j.nuclphysbps.2004.10.003

1. INTRODUCTION

High energy neutrino astronomy is a natural

extension to traditional cosmic ray physics. They

share the interest on the same objects as poten-

tial accelerators of the cosmic rays observed at

Earth. Candidate cosmic accelerators include

supernova remnants, the accretion disk and jets

of Active Galactic Nuclei (AGN), or the violent

processes behind Gamma Ray Bursts (GRB).

Neutrinos are expected to be produced in these

violent environments through proton-proton or

proton-photon collisions and subsequent decay of

the created pions. Although these objects ’work’

at different energies, it is believed that in the

most violent environments protons can be accel-

erated to energies of the order of ? PeV and,

according to most models, one should therefore

expect neutrinos to be produced at TeV energies

and above.

While the original direction of the protons can

be randomized by intergalactic magnetic fields

and photons are easily absorbed in the infrared

background, neutrinos have the advantage of

traveling cosmic distances unabsorbed and un-

deflected. This turns into a disadvantage when

trying to detect them, requiring large detector

volumes to collect a decent rate. At extremely

high energies, O10

eV, where protons would

point to their sources, they are also absorbed

by the CMB. At such energies the detection of

cosmic rays suffers from the GZK suppression,

rendering the size of the visible Universe to ∼10

Mpc. A similar effect occurs for UHE gamma

rays.

The AMANDA detector has been built to ex-

plore the high energy universe in neutrinos, using

the advantages of neutrinos as cosmic messengers.

2. THE AMANDA DETECTOR

The AMANDA detector consists of an array of

optical modules (OM) buried deep in the ice at

the South Pole. The OMs consist of a Hama-

matsu 8-inch photomultiplier tube housed in a

glass sphere. They are connected to the sur-

face electronics by a copper cable which supplies

the high voltage to the photomultiplier tube and

transmits the signal to the surface. Some of the

modules were connected also through a fiber optic

cable in order to test the technique.

Due to the complexity of the construction

and the constraints set by the site, the detec-

tor was built is stages between 1996 and 2000.

In this paper we will mention results from the

10-string detector, AMANDA-B10, and from the

full AMANDA array, AMANDA-II. AMANDA-

B10 consisted of 302 OMs in 10 strings arranged

in two concentric circles, the outer one with a

diameter of 100 m. and was operational during

1997. The AMANDA-II detector was completed

in 2000 and it consists of an additional circle of

9 strings, increasing the total number of OMs to

677 and the detector diameter to 200 m. The

typical OM vertical separation is 10 m. and the

inter-string separation about 50 m. The instru-

mented height is 500 m, between 1500 m and

2000 m under the ice surface.

With the help of calibration devices deployed

with some strings and a YAG laser at the sur-

face, we have mapped the optical properties of

the ice at the depths of the detector [1]. The

antarctic ice is very transparent in the window

of Cherenkov wavelengths, with an absorption

length of ∼110 m at 400 nm. The scattering

length is quite independent of wavelength, but it

reflects the impurities of the ice with depth. It

is about 20 m at 400 nm with variations of more

than a factor of two for specific depths.

Muons produced in charged-current ν

-nucleon

interactions near the array are detected by the

Cherenkov photons they emit. Electromagnetic

and hadronic cascades that result from neu-

tral current interactions of all three flavors or

from charged-current ν

and ν

are also detected

through the Cherenkov light of the produced

secondaries. AMANDA is therefore an all-flavor

neutrino detector. This is important in the con-

text of neutrino oscillations since neutrinos are

expected to be produced at the source with the

ratio ν

:ν

=1:2:0 (typical from pion decay),

but are expected at the Earth with a ratio 1:1:1.

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

Events are reconstructed by minimizing the

log-likelihood that the timing of the produced

light pattern is compatible with a track or cascade

hypothesis [2]. The current angular resolution

for muon tracks lies between 1.5

◦

and 2

◦

, while

the cascade angular resolution is ≈30

◦

. This

just reflects the longer lever arm for reconstruc-

tion of muon tracks, while cascades produce more

spherically-shaped light patterns. For this same

reason, the energy resolution (correlated to the

total light detected) is better for cascades, σ(log

E)≈ 15% than for muon tracks, σ(log E)≈ 40%.

3. PHYSICS TOPICS AND RESULTS

AMANDA can be used for different physics

topics, ranging from astrophysics to particle

physics. Due to the focus of this conference we

will restrict here to those topics more closely re-

lated to high energy astrophysics, rather than to

particle searches. We will present recently ob-

tained results on searches for a diffuse neutrino

flux using different signal channels and energy

ranges, results from the search for point sources

and results from cosmic ray composition studies

using downward-going muons. A recent summary

that includes results of other AMANDA topics

can be found in [3]

3.1. General analysis strategies

The analyses summarized in the next section

are all based in the following simulations. For

the background atmospheric muon flux we have

used the CORSIKA [4] air shower generator with

the QGSJET hadronic interaction model. The

average winter atmosphere at the South Pole has

been used. For the neutrino signal we have used

two AMANDA-grown generators, nusim and a

newly developed generator which treats all-flavor

generation in a more consistent manner, ANIS [5].

Both can be weighted to any desired spectrum, so

we produced events following an E

−1

spectrum,

that was later re-weighted to an atmospheric

spectrum, an E

−2

or to the spectra predicted by

the specific models probed.

The muons were propagated through the ice

including stochastic energy losses using the pro-

gram MMC [6].

A detailed simulation of the detector response is

performed for both signal and background events,

including individual OM response to light, trig-

ger simulation and pulse recording. A specific ice

model needs to be used at this stage to determine

the pulse timing. After the detector simulation,

data and simulated events are written in the same

format and the same reconstruction and event

selection procedure is done on both sets.

The main sources of systematic uncertainties

in the analyses mentioned below are the muon

propagation and the ice model parameters in the

detector simulation. For most of the analyses

presented in this summary, the systematic un-

certainties amount to ∼25%. In the case of the

search for UHE events, one should include the un-

certainty of the neutrino-nucleon cross sections at

these energies, which are just model-dependent

extrapolations of accelerator measurements at

lower energies [7]. There is an additional un-

certainty, common to all analyses which use the

atmospheric neutrino simulations as background

estimation, in the absolute normalization of the

atmospheric neutrino flux, only known to within

∼25% at energies ? 100 GeV, which is just a

translation of the uncertainties in the primary

cosmic ray flux.

We emphasize that AMANDA presents all its

limits following the ordering scheme described

in [8] and including all systematic uncertainties

in the calculation of the limit according to the

method derived in [9]. We should also mentioned

here that AMANDA follows the policy of per-

forming all analysis in a ’blind’ manner to guar-

antee statistical purity of the results. Cuts are

optimized on a subsample of the data set, which

is then not used to obtain the final result, or on a

time-scrambled data set, which is just unscram-

bled after cut values have been optimized and

frozen.

3.2. Diffuse search

Even if individual sources turned out to be too

weak to produce an unambiguous neutrino signal

from their direction in a detector of the size of

AMANDA, the summed neutrino output from all

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

sources could produce a detectable diffuse ’glow’.

We have performed such search with data col-

lected during 130 days of live-time in 1997 with

the AMANDA-B10 detector [10]. An analysis

with data taken during 2000 is under way. This

analysis is designed to retain high-energy events,

since the expected energy spectrum from shock

acceleration in cosmic sources (∝ E

−2

) is harder

than that of atmospheric neutrinos, ∝E

−3.7

. The

energy of the muon is correlated to the total light

detected by the array, higher energy muons pro-

ducing more light from catastrophic energy losses

along the track. We have used the hit-channel

multiplicity as the main variable to separate the

expected signal-like events from the atmospheric

neutrino background, and optimized to cut in

such variable for best sensitivity. After applying

all the quality cuts to the data and simulated at-

mospheric neutrino background, we obtain 3 data

events while we expect 3.1 atmospheric neutrinos.

Including the systematic uncertainties this leads

to a 90% CL flux limit on an E

−2

spectrum in the

region 3 TeV to 6 PeV of E

Φ(E

) = 8.4 × 10

−7

GeV cm

−2

−1

, which shown in figure 1 as

the full line labeled ’AMANDA-B10 E

−2

’.

Given specific models of neutrino production,

the cut optimization can be done to maximize

sensitivity to probe any chosen model. We have

done this for a sample of predictions resulting

in that three of the models in the market are

excluded at 90% CL ([11–13]). Figure 1 shows

the prediction and the AMANDA limit on one

of them, from reference [11]. This is an impor-

tant result since brings AMANDA to the level of

sensitivity to predictions of current neutrino pro-

duction models in AGN and allows the detector

to probe concrete physics predictions.

3.3. Diffuse cascade search

Another search for a diffuse neutrino glow can

be done in the cascade channel, being sensitive to

all neutrino flavors. The cascade analysis is sensi-

tive to the whole sky since down-going neutrinos

can also interact nearby the detector producing a

detectable cascade. A previous analysis with data

taken in 1997 with the AMANDA-B10 detector

has been published in [14]. Here we describe the

Figure 1. 90% CL limits on the neutrino flux from

the diffuse search and from the all-flavor cascade

search (dashed line in lighter shade) assuming a

−2

spectrum. We have end-marked both lines to

emphasize the different energy coverage of both

analyses. Shown in the figure is also the limit ob-

tained for an specific AGN model ([11]) which is

disfavored and the AMANDA-B10 limit on neu-

trinos from charm production in the atmosphere.

Limits from other experiments are shown for com-

parison.

more recent analysis performed with data taken

with AMANDA-II during 197 days of live-time

in 2000 [15].

Given the energy range we are looking at (50

TeV to ∼PeV), the contamination from atmo-

spheric muons is reduced to negligible levels, as

well as the atmospheric neutrino background. We

have simulated signal events for the three flavors

for an assumed spectrum of E

−2

and from all

directions. After all cuts are applied, 1 experi-

mental event remains, while 0.90

−

+00.

6943

events are

expected from atmospheric muons and 0.06

−

+00.

0904

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

from atmospheric neutrinos. This leads to a 90%

CL limit on the diffuse flux of neutrinos from

all flavors in the region 50 TeV to 5 PeV of

Φ(E) = 8.6 × 10

−7

GeV cm

−2

−1

. As

for the diffuse ν

search mentioned in section 3.2,

the cascade analysis can be optimized for given

model predictions. Figure 1 shows the E

−2

limit

compared to several models and to the result of

the diffuse search in the muon channel. Note that

the published cascade limit with AMANDA-B10

data in [14] was E

Φ(E)=8.9 × 10

−6

GeV

−2

−1

. Since the live-times of both anal-

yses are just about 50% different, the improve-

ment of an order of magnitude is mainy due to

the increased sensitivity of AMANDA-II due to

its larger size.

3.4. UHE neutrino search

Above a few PeV, the increasing neutrino cross

section with energy makes the Earth opaque.

Above such energies the events should concen-

trate near the horizon, where the path length

is still not enough for complete absorption. We

have performed a search for events in the range

1 PeV-3 EeV near the horizon using data from

130 days of live-time in 1997. A new analysis

with data from 2000 is underway. At these en-

ergies the atmospheric neutrino background is

negligible, and the main source of background

arises from mis-reconstructed atmospheric muon

bundles. A somewhat different event selection

techniques were developed for this analysis. A

neural network based event selection was used,

where in addition of the hit channel multiplicity

and reconstruction quality variables, the num-

ber of hit channels with exactly one hit proved

to be a good discriminant variable. The upper

energy range available to this analysis is limited

by detector saturation effects: muons or cascades

from neutrinos in the EeV energy range emit

enough light to illuminate the whole detector,

and at about a few EeV, the event energy re-

construction based on the detected light becomes

unreliable.

Assuming an E

−2

spectral shape on the neu-

trino energy, and not having detected any excess

over the expected background, we have set a pre-

Figure 2. 90% CL limits in equatorial coordinates

in units of 10

−7

−2

−1

for an assumed E

−2

spectrum, integrated above 10 GeV. Systematic

uncertainties are not included in this plot.

liminary 90% CL limit of E

Φ(E

) < 1.5 × 10

−6

GeV cm

−2

−1

in the range 1 PeV < E

3 EeV, assuming a ratio ν

:ν

= 1:1:1 at the

Earth. This limit excludes as well the predictions

of the model of reference [11], in agreement with

the diffuse search mentioned in the previous sec-

tion, and sensitive to lower neutrino energies.

3.5. Point source search

The main aim of neutrino telescopes is the

detection of neutrinos from resolved cosmic ac-

celerators, whether known in the electromagnetic

spectrum or not. In this section we summarize

the results of a search for an statistically sig-

nificant excess of events from any spot in the

northern sky.

The extended size of AMANDA-II shows a

greatly improved sensitivity as a function of

zenith angle (see figure 3), down to practically

the horizon, as compared with a previous analysis

using data from 1997 with the AMANDA-B10

detector [16]. The integrated declination aver-

aged sensitivity above 10 GeV for an assumed

−2

spectrum is 2.3×10

−8

−2

−1

. Using data

collected during 197 days of live-time in 2000 we

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

Figure 3. AMANDA-II sensitivity in units of

−7

−2

−1

as a function of declination for

an assumed differential E

−2

spectrum.

have performed a grid search for an event excess

in the region 0

◦

< δ < 85

◦

[17] . The size of the

search bins has been optimized for each declina-

tion band with widths ranging from 6

◦

to 10

◦

This search benefits from the fact that we can

use off-source data as background estimation in

each declination band. Figure 2 shows the 90%

CL upper limit sky in units of 10

−7

−2

−1

for

an assumed E

−2

spectrum integrated above E

10 GeV.

We have also performed a search around a num-

ber of known blazars, microquasars, SN remnants

and other candidates, assuming an E

−2

spectrum

in each case. For this analysis a circular search

bin centered at the candidate position has been

used. The radius of the bin has been optimized

for each candidate. The number of background

events is taken from the off-source part of the dec-

lination band of the object. The non detection of

a signal for any of the candidates has allowed us

to set limits on the nuetrino flux from each of

them. In table 1 we show the limits obtained for

a few selected objects. See [17] for more details.

3.6. GRB search

A special kind of point source seach is the

search for neutrinos coincident with GRBs. In

this case we have the additional handle of the

timing of the event, obtained from the detect-

ing satellite. The detector stability at the time

of the burst is assesed by monitoring the noise

rate within one hour before and after the event.

The search for an statisticaly significant excess of

events in AMANDA is done in a time window of

±5 minutes around the GRB. We have used the

GRB sample collected by the BATSE instrument

on board of the CGRO satellite. AMANDA and

BATSE data taking periods overlapped between

1997 when AMANDA-B10 started taking data,

until 2000, when CGRO was decomissioned. We

have analyzed a total of 312 BATSE triggered

bursts from this period. No excess of events was

observed in coincidence with any of the bursts.

Assuming a broken power law Waxmann-Bahcall

type spectrum [18] with E

break

=100 TeV, we

obtain a 90% CL upper limit on the expected

neutrino flux at the Earth of 4.8×10

−8

GeV

−2

−1

There is a class of events that did not trigger

the BATSE detector but were found by a later

off-line nalysis on archived data [19]. This class

amounts to 26 events during 2000 in the north-

ern hemisphere during the up-time of AMANDA.

Since 2000 the only source of GRB detection is

the Interplanetary Network, a group of satellites

with GRB detectors on board which uses triangu-

lation to spacially locate the burst. We have used

their catalogue to add 44 new bursts to the 2000

set. The addition of these bursts to the analysis is

not straightforward since there are issues of sensi-

tivity of the different detectors to consider, as well

as the dependence of the model being tested on

information gathered from the triggered BATSE

bursts. Work to obtain a statistically consistent

limit with the addition of these new 70 bursts is

under way.

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

Table 1

90% CL upper limits on the muon-neutrino flux from several selected candidate sources in units of 10

−8

−2

−1

asuming a E

−2

spectrum and integrated above 10 GeV.

source

Dec.[

◦

]

RA[h]

obs

lim

SS433

5.0

19.2

2.38

0.7

M87

12.4

12.51

0.95

1.0

Cassiopeia A

58.8

23.39

1.01

1.2

Cygnus X1

35.2

19.97

1.34

2.5

Cygnus X3

41.0

20.54

1.69

3.5

Markarian501

39.8

16.90

1.57

1.8

Markarian421 (fm)

38.2

11.07

1.50

3.5

Crab

22.0

5.58

1.76

2.4

3.7. Cosmic-ray composition at the knee

The South Pole Air Shower Array (SPASE) [21]

located at the surface of the ice about 360 m from

the vertical of the AMANDA center, and provides

a unique tool for studing the composition of cos-

mic rays at energies above 10

eV by running

it in coincidence with AMANDA. SPASE is used

to reconstruct the position direction and electron

content of air showers, while AMANDA can mea-

sure the muon component. Track reconstruction

in AMANDA for coincident events benefits from

the knowledge of the shower core location at the

surface, reducing the number of degrees of free-

dom of the fit and giving an angular resolution

of less than half a degree.

The muon energy at the detector depth is esti-

mated from a a fit of the expected light intensity

of a muon bundle as a function of lateral dis-

tance to a given OM. This quantity correlates

with the total number of electrons in the shower

measured by SPASE. By means of CORSIKA

simulations these two observables can be related

to the energy and mass of the primary. The

resolution achieved in the determintion of the

primary energy is energy dependent itself, and

it is about 10% in log(E) in the range 1 PeV to

10 PeV. Figure 4 shows the result of plotting the

energy of the primary versus its mass, comparing

the results of AMANDA using data from 1998,

with other existing experiments. The error bars

for all experiments are statistical only. In the

case of AMANDA, systematic uncertainties arise

from the choice of the shower generation model

and muon propagation in the ice. The estimated

systematic uncertainties for this analysis are at

the level of 30%.

The AMANDA results are compatible with an

increase of the primary mass in the range 1 PeV to

6 PeV, and in agreement with HEGRA, Kaskade,

EAS-TOP/MACRO and Chacaltaya for example,

but in disagreement with BLANCA and DICE,

which seem to indicate a lighter primary compo-

nent at such energies. Details of this analysis can

be found in [20]

3.8. Summary and Outlook

AMANDA has reached the sensitivity to probe

specific physics models predicting neutrino pro-

duction in AGN and blazars, having disfavoured

a few of them at 90%Cl with the results of the

diffuse, cascade and UHE analyses. We have

also vastly improved our limits on known point

source candidates from previous analyses by using

the extended AMANDA-II detector, operational

since 2000.

Using events in coincidenc with the surface scin-

tillator array SPASE allowed us to study the com-

position of air showers at energies between 400

TeV and 6 PeV. The results indicate a mass de-

pendence with energy that is compatible with an

increase of primary mass in such energy range.

We have recently finished a common processing of

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92

ln(A)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

5.25 5.5

5.75

6.25 6.5

6.75 7

7.25 7.5

log

prim

/GeV)

Figure 4. SPASE/AMANDA composition results

as a function of primary energy, compared with

other experiments.

four calendar years of data, from 2000 to 2003 and

results of different analyses from this extended

data set will be available soon. Since the begining

of 2004 AMANDA-II is fully equipped with a new

fully digital read out that improves single photo-

electron detection and will therefore improve en-

ergy resolution for the UHE analysis.

3.9. Acknowledgements

This research is supported by the following

agencies: National Science Foundation–Office of

Polar Programs, National Science Foundation–

Physics Division, University of Wisconsin Alumni

Research Foundation, Department of Energy, and

National Energy Research Scientific Computing

Center, UC-Irvine AENEAS Supercomputer Fa-

cility, USA; Swedish Research Council, Swedish

Polar Research Secretariat and Knut and Al-

ice Wallenberg Foundation, Sweden; German

Ministry for Education and Research, Deutsche

Forschungsgemeinschaft (DFG), Germany; Fund

for Scientific Research (FNRS-FWO), Flanders

Institute to encourage scientific and technological

research in industry (IWT) and Belgian Federal

Office for Scientific, Technical and Cultural af-

fairs (OSTC), Belgium. D.F.C. acknowledges the

support of the NSF CAREER program.

REFERENCES

1. K. Woschnagg et al. Proc. of the 26th Int. Cos-

mic Ray Conference, Salt Lake City, USA, 1999,

HE4.1.15.

2. J. Ahrens et al. Nucl. Inst. and Methods. A524

(2004), 169.

3. M. Ribordy et al. To appear in Proc. of the

XXXIX Rencontres de Moriond. hep-ex/0405035.

4. D. Heck et al. FZKA (1998) 6019. See also www-

ik3.fzk.de/h˜eck/corsika

5. M. Kowalski, A. Gazizov. Proc. of the 28th Int.

Cosmic Ray Conference, Tskuba, Japan, 2003,

1459.

6. D. Chirkin, W. Rhode. Proc. of the 27th Int. Cos-

mic Ray Conference, Hamburg, Germany, 2001,

HE2.02, 1017.

7. R. Gandhi, C. Quigg, M. H. Reno, I. Sarcevic.

Astropart. Phys. 5, (1996) 81.

8. G. J. Feldman, R. D. Cousins. Phys. Rev. D57

(1998), 3873

9. J. Conrad et al. Phys. Rev. D67 (2003), 012002.

10. J. Ahrens et al. Phys. Rev. Lett. Vol 90, No. 25

(2003), 251101.

11. F. W. Stecker, M. H. Salamon. Space Sci. Rev.

75 (1996), 341

12. A. P. Szabo, R. J. Protheroe. Proc. of the Work-

shop on High Energy Neutrino Astronomy, Uni-

vresity of Hawaii (World Scientific, Singapore,

1992), pg. 24.

13. Accretion Phenomena and Related Outflows,

IAU colloquium 163, ASP Conf. Series. 121

(1997), 585.

14. J. Ahrens et al.Phys. Rev. D67 (2003), 012003.

15. J. Ahrens et al. Accepted for publication in Phys.

Rev. D

16. J. Ahrens et al. Astrophys. J. 583, (2003), 1040.

17. J. Ahrens et al. Phys. Rev. Lett. Vol 92, No 7

(2004), 071102.

18. J.Bahcall, E. Waxmann. hep-ph/9807282 and

Phys. Rev. D64 (2001),023002.

19. B. Stern et al. A&AS 138 (1999), 413S

20. J. Ahrens et al. Accepted for publication in As-

tropart. Phys.

21. J. E. Dickinson et al. Nucl. Instr. Meth. A440,

(2000), 95.

C. de los Heros et al. / Nuclear Physics B (Proc. Suppl.) 136 (2004) 85–92