Recent results from NESTOR a

L.K. Resvanis,NESTOR Institute, National Observatory of Athens and Physics Dept., University of Athens,

representing the NESTOR CollaborationG.Aggourasj , E.G.Anassontzisa, A.E.Balld, G.Bourlisf , W.Chinowskyh, E.Fahrung,

G.Grammatikakise, C.Greeng, P.Griederb, P.Katrivanosi, P.Koskeg, A.Leisosj,f , L.Ludvigh,E.Markopoulosj , P.Minkowskyc, D.Nygrenh, K.Papageorgiouj , G.Przybylskih, L.K.Resvanisa,j ,

I.Siotisi, J.Sohperh, T.Staverisj , V.Tsaglij , A.Tsirigotisj,f , V.A.Zhukovk

aUniversity of Athens, Physics Department, GreecebUniversity of Bern, Physikalisches Institut, Switzerland

cUniversity of Bern, Institute for Theoretical Physics, Switzerlandd CERN (European Organization for Nuclear Research), Geneva, Switzerland

eUniversity of Crete, Physics Department, GreecefHellenic Open University, School of Science and Technology, Patra, GreecegUniversity of Kiel, Institute of Experimental and Applied Physics, Germany

hLawrence Berkeley National Laboratory, Berkeley, CA, USAiNCSR ”Demokritos”, Athens, Greece

jNESTOR Institute for Deep Sea Research, Technology and Neutrino Astroparticle Physics, Pylos,Greece

kInstitute For Nuclear Research, Russian Academy of Sciences, Moscow, Russia

ABSTRACT

A module of the NESTOR underwater neutrino telescope, was deployed, inMarch 2003, at a depth of 3800m in order to test the overall detector performanceand particularly that of the data acquisition systems. A prolonged period ofrunning under stable operating conditions made it possible to measure the cosmicray muon flux, I0 · cosa(θ).We also present our plans for the near future.

1. Introduction

When high-energy neutrinos interact with matter, some of the time they producerelativistic muons that follow closely the direction of the incident neutrinos. Whensuch interactions occur in the sea water or bedrock close to the detector, these muonscan be observed by the Cherenkov light that they emit using arrays of sensitive opticaldetectors: from the arrival time and intensity of the light pulses detected, the directionof the muons, and hence those of the incident neutrinos, can be reconstructed.

The potential of such detectors for astronomy and cosmology has long been recog-nised. After the pioneering work by DUMAND 1) near Hawaii, detectors are currentlyoperating at Lake Baikal (Siberia)2) and in ice at the South Pole (AMANDA)3). Con-struction of a large array (ICECUBE 4)) is starting at the South Pole and the need for

aMost of the work reported here has been published in NIM A 552(2005)420-439 and AstroparticlePhysics Vol 23,377-392, 2005 and the reader is invited to refer to these publications for a completelist of references etc.

a complementary detector (∼ 1km3) in the northern hemisphere has led to a numberof projects in the Mediterranean 5,6,7,8).

2. Main features of the NESTOR Detector, its site and infrastructure

A number of reports and papers have described in detail the elements of theNESTOR detector and the techniques used for its deployment and recovery 9,10,11,12,13,14).The main features are only briefly reviewed in this section. The prerequisites for thesite are great depth (several km), clear water, low underwater currents, very lowbioluminescent activity, minimal sedimentation and biofouling rates as well as closeproximity to support infrastructure on shore. The NESTOR site in the Ionian Seaoff the southwestern tip of the Peloponesse fulfils all these requirements. Extensivesurveys in 1989, 1991 & 19925,15) located a large flat plateau of 8 × 9 km2 in area ata mean depth of 4000 metres. Situated on the side of the Hellenic Trench that liesbetween the west coast of the Peloponnese and the submarine mountain chain of theEast Mediterranean Ridge, the site is well protected from major deep-water pertur-bations. The deepest water in the Mediterranean at 5200 metres is a few miles awayfrom the NESTOR site. Very deep water is essential in reducing the principal back-ground from muons produced by cosmic rays interacting in the Earth’s atmosphere.Also biological activity diminishes with depth. The locationb is 7.5 nautical milesfrom the island of Sapienza, where there are two small harbours, and 11 miles fromthe port of Methoni, while substantial port facilities are available 15 miles away in thetown of Pylos on the bay of Navarino. Regular measurements 16,17) of water qualityshow transmission lengths of 55 ± 10m at a wavelength of 460 nm, stable tempera-tures of 14.2 0C and water current velocities well below 10cm/s 18), light bursts of1-10 s duration, consistent with bioluminescent activity, represent around 1% of theactive time and there is little/no evidence of problems due to sedimentation or bio-fouling19). The sea bottom over the site has a clay deposit accumulated over sometens of thousands of years which provides for good anchoring20). The basic elementof the NESTOR detector is a hexagonal floor or star. Six arms, built from titaniumtubes to form a lightweight lattice girder, are attached to a central casing. Two op-tical modules are attached at the end of each of the arms, one facing upwards andthe other downwards. The electronics for the floor is housed in a one-meter diametertitanium sphere within the central casing. The nominal floor diameter at the opticalmodules is 32 metres. A full NESTOR tower would consist of 12 such floors stackedvertically with a spacing of 30 m between floors. This is tethered to a sea bottomunit (pyramid)14) that contains the anchor, the junction box, several environmentalsensors and the sea electrode that provides the electrical power return path to shore.The junction box houses the termination of the sea-end of the electrooptical cable,the fan-outs for optical fibres and power to the floors etc. as well as monitoring and

bSite coordinates: 36037.5’ N, 210 34.6’ E

protection of the electrical system. The optical module21) consists of a 15” diameterphotomultiplier tube (pmt) enclosed in a spherical glass housing which can withstandthe hydrostatic pressure up to 630 atmospheres. Other modules, above and beloweach floor, house LED flasher units that are used for calibration of the detector andthey are controlled and triggered from the floor electronics. Deployed14) equipmentis brought to the surface, together with the sea end of the electrooptical cable, bymeans of a recovery rope, released from the sea bottom by coded acoustic signal.Modifications or additions to the experimental package are made at the surface andall connections are made in the air with dry-mating connectors. The cable and exper-iment systems are then re-deployed and the recovery rope, with its acoustic releaselaid on the seabed. The NESTOR deployment ‘philosophy’ has always been to avoidthe need for specialised manned or unmanned underwater vehicles for deploymentand recovery operations that require the use of manipulators, wet-mating connectersand consequent high costs. All electrical and optical fiber connections are dry matedin the air. The objectives for the deployment reported in this paper were to test fullythe electrical supply and distribution systems, the monitoring and control systemsand the full data acquisition and transmission chain from the sea to the shore sta-tion (each NESTOR floor is independent from the others with respect to electricalpower supply and data acquisition and transmission). The titanium girder arms ofthe stars are made in standard modules of 5-meter length; for logistical reasons onthe deployment vessel, the star used for this experiment has an overall diameter of12 metres. In all other respects standard equipment was used. The detector staris located 80 metres above the sea bottom pyramid. The system was powered andmonitored during deployment while the pmts were switched on a few hours later whenthey had reached a quiescent state after brief exposure to daylight. The system wasoperated continuously for more than a month and several million triggers recorded.This has not only provided invaluable experience on the operation of the detector buthas initiated the development and testing of powerful tools for reconstruction andanalysis.

3. Data quality

In a typical example shown in Figure 1, the pulse height distribution has a shapecorresponding to a few (average 1.3) photoelectrons. The K40 background has beenused as a stable ‘standard candle’ in order to monitor the gain stability of the de-tector. The pmt pulse height distributions from each data file were compared to astandard shape defined at the beginning of the run and found to be extremely stablefor all of the pmts during the whole running period. However, there were periodsof time when the instantaneous counting rates of a group of pmts and the collectiontrigger rate show a large increase which lasts up to a few seconds. These phenomenalast typically from 1 to 10sec and represent a total 1.1% of the active experimental

Figure 1: The pulse height distribution of a pmt during operation in deep sea (main plot) and froma calibration run in the laboratory (insert plot). The solid line in the main plot is the result of afit to the data points using an exponential shape for the dark current (line a), as well as the onephotoelectron (line b) and the two photoelectrons (line c) pulse height spectra evaluated duringcalibration runs at the laboratory.

time. The effect is consistent with bioluminescent activity from marine organisms inthe detector vicinity. The pulse height distribution of the pmts during a period ofbioluminescence is very similar to the distribution due to the K40 decay. Biolumines-cence can be easily identified because of its characteristic time duration and thereforedoes not cause any background problem. In the analysis that follows, all events col-lected during periods of bioluminescence activity have been excluded. This representsa reduction of only 1.1% in the size of the data samplec. The average experimentaltrigger rate, corresponding to the coincidence of four or more pmt pulses above 30mVamplitude, (corresponding to 0.25 of a photoelectron), was 3.76Hz compared to anestimated rate of 3.79Hz derived from the Monte Carlo simulations. According tothe Monte Carlo estimation, only a small fraction (5.5%, 0.21Hz) of this trigger ratecorresponds to atmospheric muons passing close to the detector. When the pmtthresholds were set to 120mV, (corresponding to one photoelectron), the measuredtrigger rate was 0.29Hz, in agreement with the equivalent Monte Carlo estimate of0.30Hz. Furthermore, the measured coincidence rates, shown in Figure 2, are in verygood agreement with the Monte Carlo estimations for several levels of coincidence atdifferent pmt thresholds. In the same plots, we present the Monte Carlo estimated

cHigh levels of bioluminescence 23) can cause severe dead-time in data taking. Note that in otherMediterranean sites, periods with more than 30% of bioluminescence activity has been observed24).

Figure 2: Trigger rates as a function of the coincidence level, for two threshold settings. The pointsrepresent the data, the solid line the Monte Carlo estimation including background and the dashedline the Monte Carlo estimation for the contribution of the atmospheric muons.

contribution of the atmospheric muon flux to the triggers, showing that higher-levelcoincidences exclude the combinatorial background. A better rejection of the combi-natorial background is achieved at higher pmt threshold values.

4. Atmospheric muon studies

From the total data sample collected with a 4-fold or higher coincidence triggerand 30mV pmt threshold, a subset containing 45800 events has been selected thathave six or more pmt pulses (hits) within the 60 ns time window. These events havebeen analysed in order to reconstruct muon tracks. The arrival time of the digitizedpmt pulses was used to estimate the muon track parameters by means of a x2 fit whilstthe pmt pulse heights were used to reject ghost solutions and poorly reconstructedtracks. The results are summarized here. The zenith angular distribution of thereconstructed tracks is compared to the Monte Carlo prediction in Figure 3. Due tothe limited reconstruction resolution, the distributions extend to zenith angles higherthan 900.

5. Conclusions

In March 2003, the NESTOR collaboration successfully deployed a test floor of

Figure 3: Distribution of the Zenith angle (θ) of reconstructed tracks for the data (triangles) andMonte Carlo (solid points) event samples. The insert plot shows the same distributions on a linearscale.

the detector tower, fully equipped with final electronics and associated environmen-tal sensors to a depth of 3800 m, situated 80 metres above the sea bottom station.The deployed detector was continuously operated for more than one month. Themonitored experimental parameters, operational and environmental, remained stablewithin the accepted tolerances whilst the readout and DAQ chain performed well andwith practically zero dead-time. The 1.1% of the total experimental time was lostdue to bioluminescent activity around the detector. This 1% dead time is consis-tent with previous measurements in the same site done with autonomous drops19).Events collected during such periods of activity were easily identified and rejected.Several studies have been made to ensure that the event selection trigger was un-biased and that the collected light on the pmts can be attributed to the expectednatural sources. The pmt pulse height distributions, the trigger rates and the to-tal number of photoelectrons inside the trigger window as functions of the signalthresholds and coincidence level settings as well as the arrival time distribution ofthe accumulated photoelectrons, agree very well with Monte Carlo predictions basedon the atmospheric muon flux parameterization of Okada22), on the natural K40 ra-dioactivity in the sea water and the pmt dark currents and after pulses. In parallel,calibration in the sea using the LED flasher units mounted above and below thedetector floor, provided a rigorous test on the time stability of the detector as wellas a measurement of the resolution of the arrival time of the pmt signals. A sub-

set of the accumulated data, consisting of events with six or more pmt pulses insidea 60ns time window, has been analysed and the trajectories of atmospheric muonshave been reconstructed. The distributions of the azimuth and zenith angles of thereconstructed muon tracks are found to be in a very good agreement with MonteCarlo predictions, based on the atmospheric muon model of Okada22). Finally, basedon previous measurements by the NESTOR collaboration15) concerning the shapeof the zenith angle distribution, we estimated the vertical atmospheric muon inten-sity at the deep-sea site. We parameterize the number of atmospheric muons (N)arriving at the detector depth per unit solid angle (Ω), per unit time (t) and perunit area (S) , dN/(dΩdtdS) , is usually parameterized as15,25,26) dN/(dΩdtdS) =I0cos

a(θ) where I0 is the vertical intensity. The measured vertical muon intensityand the index α, at a depth of 3800 m.w.e., are α = 4.7 ± 0.5(stat) ± 0.2(syst) andI0 = 9.0 × 10−9 ± 0.7 × 10−9(stat) ± 0.4 × 10−9(syst)cm−2s−1sr−1 are in very goodagreement with previous underwater measurements and with phenomenological ex-pectations. The objectives for this deployment of the NESTOR test detector were toperform a thorough test of the electrical supply and distribution systems, the moni-toring and control systems and the full data acquisition and transmission chain fromthe sea to the shore station. These objectives have been met successfully. In additionwe have been able to demonstrate the ability of the proposed neutrino telescope toreconstruct muon trajectories.

6. The future

Having demonstrated successfully the operation of a Neutrino Telescope at a depthof 4000m the obvious question is: How do you build a Neutrino Telescope big enoughto detect first and then to harvest the Physics? The NESTOR Collaboration issimultaneously following a two prong approach:

1. The recently funded by the EU three year Design Study of the MediterraneanCubic Kilometer Telescope (KM3NeT) together with twenty five European In-stitutions

2. In collaboration with colleagues from the US (mainly from the Lawrence Berke-ley National Laboratory-Space Sciences Lab., led by Hank Crawford) the NESTORCollaboration plans to build in the next three years a specialized Cubic Kilome-ter telescope to search for 100TeV neutrinos in coincidence with GRB (GammaRay Burst) emission. This is what we call NuBE-NESTOR (NuBE stands forNeutrino Burst Experiment).

7. KM3NeT

The KM3NeT design must cover a wide band of energies and fluxes in Neutrino

Physics i.e. from the GeV SuperKamiokade sensitivity to the PeV region. Thus, theKM3NeT must be designed with a detector granularity and sensitive area, sufficientto bridge the gap in extraterrestrial neutrino detection from the few GeV to the PeVregion. These Physics requirements clearly argue for a variable density detector, withan increasing sensitive area as the neutrino energy gets higher and the flux is reducedcorrespondingly. On top of the Physics driven design considerations there are somecritical parameters which must be taken into account by the architectural designof a Deep Sea Neutrino Telescope, like the deployment-retrieval-maintenance of thetelescope elements. These considerations define to a large extend the connectivity ofthe telescope elements for: power, control command and data transfer between themand the shore station. All this should be done in a way to protect the telescope fromsingle failure points. The zeroth order approximation to solving the above problemis to follow the DUMAND1) architecture, which was later copied by Antares7), andpropose to build the telescope from strings which are equidistant (50m-60m) andare connected to an umbilical electrooptical cable using a manned submersible ora Remotely Operated vehicle (ROV). Assuming that one can afford the high costof either (ca 100k$/day from the moment the tender ship leaves its homeport untilshe returns back there) we do not believe that it is possible to connect the manyhundreds of strings which are necessary to deploy in order to construct KM3NeT. Theproblem is the safety and maneuverability inside, literally, a dense forest of hundredsof strings. Autonomous Underwater Vehicles (AUV) do not operate deeper than afew hundred meters. Manned submersibles are not used around ropes/cables/stringsbecause getting caught in one of these may prove to be lethal for the crew. ROVsusually operate from a ‘garage’ which the tender ship deploys on the ocean floor, theROV is connected to the garage with a thick and cumbersome umbilical cable whichsupplies power and transfers data and commands. So, if the ROV navigates in amaze of strings, the umbilical will be caught by the strings. These are the reasons forwhich we in NESTOR have been using the ‘tower’ structure. The NESTOR toweris composed of hexagons 32m in diameter with a pair of back, to back large (15”)pmts on the apexes and the center, one can then stack as many hexagonal floors ashe can in the vertical creating a tall tower (up to 12 in NESTOR, with 30m spacingbetween floors). Each tower has its own umbilical electrooptical cable to shore whichis raised to the surface whenever the tower is recovered. Clearly the hexagonal floorsprovide a rigid lattice onto which one can attach pmts at the necessary distances.One ought to think of a tower as a cluster of many strings with a single connectionpoint to the umbilical electrooptical cable. The NESTOR collaboration is developingthe necessary surface vessel infrastructure to enable the deployment of hexagons ofmuch larger than the present diameter i.e. up to 100m or even 120m especially if wedo not have to raise/deploy the umbilical cable every time the tower must come to thesurface. So, we are working to develop the concept of having on the ocean floor a pre-deployed electrooptical cable network (with at least two branches to the shore, in order

Figure 4: An 100m diameter hexagonal floor with pmts every 18.5m .

to avoid single failure points) with the necessary number of connection points (minijunction boxes) onto which an ROV connects a whole tower which now is composedof a cluster of 19 strings or more. Let me illustrate it with a possible example: an100m diameter hexagonal floor has a pair of back to back pmts every 18.5m alongits arm plus a pair at the center, Figure 4. A tower can then be created by stackingthese 100m diameter floors every 30m in the vertical. One can easily stack twelve ormore such floors. Such a tower would have a threshold sensitivity (50% reconstructionprobability) of about 5GeV for muons within an instrumented volume of about 15million cubic meters, (compare this to Superkamiokades 50 thousand tones) and amuon direction reconstruction of better than one degree. The tower will be deployedto the predetermined position from the surface and then later on when an ROV isavailable it will be connected to the pre-laid cable network. The way to proceedis obvious, taking this first tower as the center, one deploys six similar towers on ahexagon of radius 600m around it. The 600m space from tower to tower is comfortablefor an ROV to navigate safely and connect them to the electrooptical cable network.Any 100TeV or more neutrino that traverses the space between these towers will bereconstructed by at least two towers. The step to follow is again obvious, anothersix such towers deployed on a hexagon 1200m from the central tower, Figure 5. Thisarrangement of 13 long arm towers (or equivalently 13∗19 = 247 strings, with a totalof 5928 pmts) is a Cubic Kilometer Detector! This design provides variable densityin a modular and expandable fashion in order to harvest the 10Gev, 100GeV, TeV,10TeV, 100TeV and PeV Neutrino Physics and follow it, growing as necessary. It isrobust, retrievable, serviceable and does not suffer from major single failure points.

We are testing various materials and hexagonal structure configurations in orderto build these 100m-120m diameter ‘floors’. These will be followed by extensivedeployment tests under various payload dimensions in different sea conditions.

8. NuBe - NESTOR

The discovery of Gamma Ray Bursts signifies the detection of the most violent,

Figure 5: Bird’s eye view of a proposed Cubic Kilometer size telescope made out of 13 towers.

Figure 6: Schematic of NuBE - NESTOR.

truly cataclysmic phaenomena in the Universe. In order to constraint the models ofGRB production it is of paramount importance to determine whether neutrinos areproduced contemporaneously with the GRBs (300-400 per annum reported ). Therelativistic fire ball model by Waxman and Bahcall 27) predict a measurable flux of∼ 100TeV neutrinos accompanying the gamma rays to Earth. According to thismodel, a detector of ∼ 2km2 effective area, will observe 40-200 neutrino inducedmuon showers per annum in coincidence with GRBs. Other models of astrophysicalprocesses also demand production of high energy neutrinos, including other burstmodels, AGN models and topological string models. We plan to build a ∼ 2km2 inorder to determine whether such neutrinos arrive in coincidence with GRB.

This will be a sparse detector since we want to detect 100TeV events in a narrowcoincidence window (within ∼ 100s) with a GRB. At these energies the muon rangein water exceeds 5km and the muon suffers from catastrophic bremstrahlung leadingto many showers all along its track, each producing a lot of Cherenkov light whichcan be detected with high efficiency at perpendicular distances >300m . The cascadeevents (i.e. the nonmuon neutrino interactions) leave a short (∼ 10m) but extremelyintense shower which produces Cherenkov light which can be detected at distances inexcess of 400m in the NESTOR site (transmission length 55m 17).

Figure 6 shows the planned configuration which approximates a sphere of diameter∼ 1000m, creating an effective area > 2km2 for high energy events. The detectorconsists of four autonomous strings placed at the corners of a square having ∼ 600m

Figure 7: Schematic of a string.

diagonals, as shown in Figure 6. The center of the array is a NESTOR tower, initiallycomposed of four floors. Each string, Figure 7, is composed of two independent photondetector nodes (with 16 pmts each) separated by ∼ 300m along the string.

The strings are built with extremely low power electronics (e.g. the pmts haveCockroft -Walton bases) and can run on batteries for one year. Data is stored locallywhile selected triggers are transmitted to the NESTOR tower acoustically and thento the shore station via the electrooptical cable. A high energy neutrino event willlead to local triggers in more than one node (or floor), with the coincidence amongmany nodes determined by comparing local clock values off-line. Local node clocksare periodically synchronized using bright flashes of blue emitting LEDs on the Tower.The Tower synchronization can also be accomplished by direct signal from the shoretraveling along the coax connecting the floors.

A signal of a high energy event consists of a local trigger in any node/floor oc-curring within ∼ 3µs of a local trigger in any other node. The luminous travel time

across the array is ∼ 3µs. The coincidence that signals the high energy events isdetermined off-line, after the strings are retrieved. Cascade events which hit three ormore nodes provide the incident direction to within as little as FWHM ∼ 25 degrees.The requirement that these coincidences occur to within a 100s window to a GRBobserved by a satellite reduces to practically zero the probability of a random coinci-dence. If the tower is hit by a muon then its higher pmt density may allow angularreconstruction to a FWHM ∼ 5 degrees. This angular resolution capability providesrobust verification of the correlation of a GRB based on time coincidence, because itwill allow us to directly compare to the position determined by the satellite.

9. References

1) DUMAND Proposal, Report No. HDC-2-88, 1988 (unpublished).2) Balkanov V. A. et al [Baikal Collaboration], Nucl.Phys.Proc.Suppl. 87, pp405-407 (2000).

3) Andres E. et al, Astropart.Phys. 13, 1-20 (2000).4) Ahrens J. et al [Icecube Collaboration], Nucl.Phys.Proc.Suppl 118,388-395(2003).

5) L. K. Resvanis for NESTOR coll., NESTOR: A Neutrino Particle AstrophysicsUnderwater Laboratory for the Mediterranean, Proceedings of the High EnergyNeutrino Astrophysics Workshop, V. J. Stenger, J. G. Learned, S. Paksava andX. Tata editors Hawaii, 325, (1992), Word Scientific, Singapore.

6) L.K.Resvanis, Proceedings of the 2nd NESTOR International Workshop, L.K. Resvanis editor, 1-19 (1992); website http://www.nestor.org.gr .L.K. Resvanis, Proceedings of the 3nd NESTOR International Workshop, L.K. Resvanis editor, 1-17 (1993); website http://www.nestor.org.gr .

7) ANTARES: Aslanides E. et al [ANTARES Collaboration], ANTARES Pro-posal, astro-ph/9907432 (1999); website http://antares.in2p3.fr .

8) NEMO: Migneco E. et al [NEMO Collaboration], The NEMO Project, Pro-ceedings of the VLVnT Workshop pp 5-10,NIKHEF, Amsterdam 5-8 October2003; website http://nemoweb.lns.infn.it .

9) Ball A. E. et al [NESTOR Collaboration], CERN Courier 43N9,23-25, 2003.10) Peter K. F. Grieder [NESTOR Collaboration], NESTOR neutrino telescope

status report, Proceedings of the 28th International Cosmic Ray Conference,1377-1380 (ICRC 2003).

11) Tsirigotis A. G [NESTOR Collaboration], Proceedings of the International Eu-rophysics Conference on High Energy Physics (EPS 2003, Aachen Germany),The European Physics Journal C, ISSN,1434-6044.

12) Tzamarias S. E. [NESTOR Collaboration], Proceedings of the VLVnT Work-shop pp 11-16, NIKHEF, Amsterdam 5-8 October 2003.

13) Resvanis L. K., NESTOR, First Results, Proceedings of the 8th Interna-

tional Workshop on Topics in Astroparticle and Underground Physics, 187-190, (TAUP 2003) September 5 - 9, 2003, University of Washington, Seattle,Washington.

14) Anassontzis E. G. et al, Sea Technology 44-10, 2003.15) L. K. Resvanis, Proceedings of the 2nd NESTOR International Workshop, pp

1-19, L. K. Resvanis editor (1992), website http://www.nestor.org.gr .16) S. A. Khanaev et al, Proceedings of the 2nd NESTOR International Workshop,

page 253, L. K. Resvanis editor (1992), website http://www.nestor.org.gr .17) E. G. Anassontzis et al, Nuclear Instruments and Methods A349,242-246

(1994).18) T.A. Demidova et al, Proceedings of the 2nd NESTOR International Work-

shop, page 284, L.K.Resvanis editor (1992), website http://www.nestor.org.gr.19) Sofoklis Sotiriou, PhD thesis, University of Athens, website

http://www.nestor.org.gr .20) E. Trimonis et al, Proceedings of the 2nd NESTOR International Workshop,

page 321, L. K. Resvanis editor (1992), website http://www.nestor.org.gr .21) E. G. Anassontzis, et al, Nuclear Instruments and Methods A479, pp 439-455

(2002).22) A. Okada, Astroparticle Physics 2, 393 (1994).23) E. J. V. Battle et al, Proceedings of the VLVnT Workshop pp 41-43, NIKHEF,

Amsterdam 5-8 October 2003.24) U.F. Katz, for the ANTARES Collaboration, Status of the ANTARES project,

Eur Phys J C 33, s01, s971s974 (2004).25) P.K.F. Grieder, Cosmic Rays at Earth, Elsevier, Amsterdam (2001), Chapter

4.26) J. Babson et al, Physical Review D42, 3613 (1990).27) Waxman and Bahcall, Phys. Rev. Letters 78,2292-2295 (1997)