DUMAND A New Window on the Universe P.K.F. Grieder Physikalisches Institut, University of Bern, Switzerland For the DUMAND Collaboration*

The Deep Underwater Muon and Neutrino Installation in the ocean, and continuous Detector (DUMAND) is a completely new high-precision sonar tracking of the location concept, based on the detection of Cheren­ of each module had to be solved. kov radiation produced by relativistic muons This work, together with extensive theore­ particles as they move through a dielectric tical studies and the construction and suc­ medium (the ocean water) faster than the cessful operation of a prototype system, phase velocity of light in the medium. In DUMAND I, lead to the present scope of enabling the construction of a detector sys­ scientific activities and to today’s system tem of a size capable of identifying sources configuration for the DUMAND II detector. of high-energy neutrinos with a reasonable hope of success, DUMAND opens the era of Site and Environmental Aspects high-energy neutrino astronomy. It employs The site where the detector matrix will be the ocean’s water mass as a multi-purpose installed as well as the 25 km cable route to medium: as a shield against the unwanted the site have been carefully surveyed. Com­ cosmic-ray background, as a well-behaved position and topography of the sea floor are absorber for particles and radiation, as a tar­ well-known along with significant ocean get for muon and neutrino interactions, as a parameters around a depth of 4500 m such Cherenkov radiator for charged particles, as ocean currents (≤ 50 mm/s), water tem­ and as a dark room to prevent light from perature (≈ + 4.0°C), the optical attenuation interfering with faint Cherenkov radiation. length in the visible (≈ 40 m), salinity, etc. Comparatively small optical sensor mo­ The optical background In the ocean has dules constitute the detectors. They are dis­ been Investigated [1]. It is chiefly due to tributed in the form of a three-dimensional background light from seawater radioactivity matrix array throughout a well-defined vo­ (specifically, Cherenkov radiation from the lume at great depth in the ocean and sepa­ beta decay of 40K) and to bioluminescence. rated by a distance of the same order as the The former can be eliminated by noting local attenuation length of light in ocean water Fig. 1— DUMAND II array layout, showing coincidences between neighbouring modu­ (Fig. 1). The system is capable of detecting the 216 optical modules arranged in strings les. High levels of light were observed high-energy muon-neutrinos and their anti­ moored on the ocean floor some 4500 m during ship-tethered measurements owing particles of terrestrial and extraterrestrial ori­ deep. Systems for control, data handing, to wave motion inducing local turbulence gin that Interact within the detector matrix or auxiliaries, and support are not shown. causing bacteria to release light flashes. Its Immediate vicinity, producing detectable This stimulated bioluminescence whenever and reconstructable muon trajectories in the not confined to the size of an underground excessive turbulence is present is not array. Very high-energy cosmic ray muons cavity and the interpretation of data will not expected to be a major problem for originating from even more energetic inter­ be troubled by uncertainties in the density of DUMAND modules that are moored from actions of the primary cosmic radiation In the surrounding rock or in the complexity of the ocean floor, because ocean currents at the atmosphere are also monitored. the surface topography. The very large the site are weak. Sedimentation and bio- Distinction between cosmic-ray muons background rate found In detectors placed fouling have been studied and were found to and neutrino induced muons is achieved by in shallow lakes will also not be a problem. be of no concern. Lost equipment that we excluding from the neutrino analysis dow- have recovered some one and a half years ward-going muons within a zenith angle Prototype Experiment after immersion at a depth of nearly 5000 m range between 0° and 70°. Discrimination The main concern with this novel concept showed no signs of growth on its surface. against atmospheric neutrinos Is achieved is that completely new technologies have by statistical analysis and angular correla­ had to be developed; they must then be Prototype System and Experiment tion of arrival direction in conjunction with deployed in an unusual environment. A To deliver the ultimate proof of the feasibi­ energy cuts. DUMAND also has the capabi­ large amount of exploratory and pioneering lity and practicability of the DUMAND con­ lity to detect electron-neutrinos and antineu­ work therefore had to be carried out during cept, and to test the reliability of our techno­ trinos, but this option will not be discussed. the early phases of the DUMAND project. logy, a one-dimensional prototype, called The DUMAND detector system is, in prin­ One of the foremost technical problems, the Short Prototype String (SPS), was ciple, freely expandable and allows the of course, was the development of a sui­ constructed and deployed for an experiment construction of a truly giant experiment. table detector module which met the many carried out in the Pacific Ocean about 30 km Unlike typical underground experiments, it is stringent requirements, keeping in mind the west of the island of Hawaii in November ever-present high pressure of the surroun­ 1987 [2]. The roughly 70 m long SPS string ding water masses. Major problems deman­ consisted of seven optical modules, two Peter Grieder graduated as an electronics ding great effort, ingenuity and much time calibration modules, an extensive environ­ engineer in Switzerland in 1951 and worked for four years in Industry before undertaking a were tackled during optimisation of the ove­ mental data recording system, wide-band Masters in accelerator physics at the IIT, rall response of the detector matrix, of its hydrophones, and a sonar system. Chicago, USA. He received his Ph.D. in experi­ sensitivity, efficiency and track reconstruc­ The primary purpose of this experiment mental astrophysics from the University of tion capability. Questions related to the opti­ was to demonstrate that muons could be Bern In 1961 where he was appointed cal background, high-speed long-distance detected In the deep ocean and their trajec­ Honorarprofessor in 1978. data transmission, system deployment and tories reconstructed. Because of the limited Europhys. News 23 (1992) 167 similar in principle to those employed for the SPS, and a very high precision sonar.

Capabilities The size and configuration of the array were chosen to yield a reasonably even rate and adequate sensitivity to detect neutrinos from extraterrestrial high-energy neutrino sources at their expected flux levels. Moreover, the array has been optimised for the detection of high-energy muons from neutrino interactions because calculations indicate that this is the best compromise for detecting the same sources. Finally, the Fig. 2 — a, left) schematic view of the Short Prototype system can has also detect contained neu­ String used for the first phase of DUMAND: b, upper) trino-induced cascades. In order to carry out diver’s view from a depth of 30 m of the top portion of useful astronomy, a high angular resolution the string during deployment to a depth of over 4000 (typically < 1°) is required. Cherenkov light m from the US Navy’s SSP Kaimalino, a highly stable from through-going particles will illuminate ship. The dark unit in the foreground with the protru­ on average a photomultipler in each of 13 ding rod is an optical calibration module. modules so the detector spacings ensure that muons with energies ≥ 50 GeV which pass through from outside will be detected with high efficiency and reconstructed in direction with a median accuracy of about 1°. The array will be able to resolve the direction of a neutrino to this accuracy if the neutrino energy is ≥ 1 TeV, when the scatte­ ring angle between a neutrino striking a nucleon and the resulting muon which passes into the detector will be ≤ 1°. Muons from atmospheric neutrinos will dominate the downward going muons above a zenith angle of about 80° (Fig. 5) leaving a Fig. 3 — Zenith angular distribution of solid angle of 2.357π steradians for neutrino muons at a depth on 4000 m in the observations. There will be about 3500 at­ ocean, measured with the DUMAND mospheric neutrino events detected each Short Prototype String. year. This gives a background rate of about one atmospheric neutrino per (2.80)2 per year. Thus, with the calculated resolution of 1° for muon direction reconstruction, we will size of the SPS (Fig. 2) and the very short Monte Carlo simulation. The simulation also be largely signal limited, rather than back­ time that was available for the experiment gave the effective detection area for muons ground limited, in the search for extraterres­ (11 days), the test was performed using as a function of the zenith angles. Multi­ trial point sources. high-energy muons produced by interac­ plying the events rates by the area gave the As in the case of the SPS, the array will tions of cosmic rays with the earth’s atmos­ total muon intensity as a function of depth. also contain wide-band hydrophones to phere that penetrate deep into the ocean. Vertical intensities obtained by multiplying The final DUMAND II array will in fact be by a correction factor agreed with previous Fig. 4 — Optical detector module for used to detect the much rarer muons indu­ estimates from underground experiments. DUMAND II with the upper hemisphere of the ced by the interactions of high-energy neu­ glass pressure housing removed. The coarse trinos with the surrounding ocean water. DUMAND II wire mesh is the magnetic shield. Each unit is The prototype system was deployed from After the successful completion of the pro­ fully monitored and computer controlled from a highly stable ship (it had a twin-hull design totype experiment, which delivered uncondi­ shore. For each string, data and command with a small wetted area) and was kept sus­ tional proof of the practicability of the transmissions are handled locally by a string pended for the duration of the experiment DUMAND principle and of the capability and centre controller, each of which has two-way using a cable with an adjustable tension to dependability of the technology we have communication with the shore via colour mul­ further reduce turbulence. A combined elec­ developed through the years, the DUMAND tiplexed optical fibre links operating at nearly tro-optical cable supported the full load, deli­ collaboration has designed an intermediate- 1 Gbd. vered power to the SPS, and carried the size detector, DUMAND II [3] which is now data stream to the control room aboard the under construction, as a next step towards a ship. The intensity of cosmic-ray muons was truly giant, deep ocean detector. The project measured at different depths using coinci­ is funded by the US Department of Energy, dence requirements within a well-defined the Swiss National Science Foundation and time window. This involved fitting an indivi­ Japanese agencies. dual trajectory to each downward-going The array (Fig. 1) will be moored from the muon event, employing both the arrival time ocean floor. The site is about 25 km west of of the conical wavefront of Cherenkov light Keahole Point on the Island of Hawaii at a at each detector module and the light inten­ depth of 4500 m — the same location as the sity for the reconstruction. Controllable light SPS trial. The array will consist of a total of emitting calibration modules were used to 216 optical modules (Fig. 4) similar to those calibrate timing and sensitivity of the detec­ used in the SPS trial, mounted on vertical tor modules and to measure the attenuation strings each with 24 modules, located at the length of light in the ocean. corners of an equilateral octagon, with the The observed angular distribution of the ninth string at the centre. The strings will be muons (Fig. 3) agreed with the results of a equipped with auxiliary and support systems 1 68 Europhys. News 23 (1992) Fig. 5 — Calculated counting rates per de­ gree and year for through-going cosmic-ray muons and atmospheric neutrino induced muons as a function of zenith angle. Fig. 6 — Extraterrestrial neutrino source detection capability of DUMAND II and comparison with the Irvine-Michigan-Brookhaven (IMB) detector, until recently the US’s largest operational explore the possibility of detecting acoustic muon and neutrino underground observatory. Shown is the number of muon detections from a shocks generated by ultra high-energy cos­ particular source versus source distance. Estimates are based on gamma-ray data. mic rays upon impact on the water. It has long been known that high-energy particles excite acoustic waves in water. As the atte­ nuation length for high frequency (20-50 DUMAND kHz) acoustic radiation is in the kilometre range (compared to 40 m for light) a truly Planning for Remote Operation gigantic array could be used for very high- The DUMAND neutrino telescope consists ever, there will still be a need for constant energy air showers. of 216 photomultiplier tubes in spherical monitoring and control. The operating para­ glass housings, 45 hydrophones for acousti­ meters of such a complicated system in the Scientific Goals cally surveying the 350 m tall, nine-string deep ocean have to be watched by a human High-energy neutrino astronomy and astro­ array, and 15 laser calibration units to moni­ observer who has to be able to adjust the physics tor the sensitivity of the optical detectors and mode of operation. This duty need not be Astrophysical sources of high-energy to calibrate their timings. The various mo­ performed at the shore station and no com­ neutrinos are the prime target for DUMAND dules in a string all communicate via fibre ponent or system at the station, excluding II. The aim here is to search the universe for optic links to an electronics package at the storage devices, will normally require manual point sources of energetic neutrinos and bottom of each string. This unit contains an Intervention. determine the gross features of the neutrino application-specific digitising chip to record All operations will be controlled and moni­ spectrum, in an attempt to locate and study the arrival of Cherenkov light pulses with 1 ns tored through a main computer, with all func­ possible sources of very high-energy cos­ precision. The data and other control signals tions centralised at a single control worksta­ mic rays and other astrophysical aspects. from an entire string are then multiplexed tion. If possible, we want to be able to run the We expect detectable signals from a num­ onto a single-mode fibre link to the shore sta­ experiment sequentially from remote stations ber of point sources both inside and outside tion some 30 km distant. Ten data-carrying using teams of physicists on shift in Hawaii, the Galaxy and probably a more or less fibres working at 1300 nm (one for each Japan, Europe, and mainland USA. To such homogeneous flux of ultrahigh-energy neu­ string plus one for environmental data) trans­ an end, we have standardised on portable trinos from the recently proposed neutrino mit data at 625 Mb/s; filtering will reduce the computer code as well as vendor-indepen- sources In active galactic nuclei (AGN) for final data rate to 80 kb/s. dent platforms (e.g., C and OSF/MOTIF) in which no generally accepted model exists. The array will be lowered one string at a order to present the same interface to remote Fig. 6 shows a number of likely extraterres­ time and connected to a cable to shore by a operators. trial neutrino sources and their expected manned research submarine. The junction intensities in muon detections, based on box at the end of the cable will have instru­ Networked Videoconferencing gamma-ray intensities (see below). The mentation to monitor this activity. Power will We are currently investigating various net­ point source sensitivity of the array is be sent down the 15 mm diameter armoured working alternatives to link the shore station approximately 4-7 x 10-10 cm-2s-1 for neutri­ cable at high DC voltage, delivering 5.5 kW to universities around the world. Given the nos of energy > 1 TeV. Possible point to the array. The fibres also carry commands geographic spread, DUMAND could benefit sources such as the X-ray binary pulsar at 1550 nm out to the array which will contain from videoconferencing. However, the mini­ Cygnus-X3 in the Galaxy may produce 100 about 250 networked-linked computers to mum “near acceptable” bandwidth for this is counts/year in DUMAND II: one-tenth this adjust switches and many variables such as 1 Mb/s for the foreseeable future. Equipping voltages. The DUMAND array can thus be DUMAND to use this bandwidth for voice and level would be significant relative to the remotely tuned and even reprogrammed to data when video is not transmitted is expen­ expected noise from atmospheric neutrinos. cope with the unexpected, or to deal with any sive. Decreasing the size of images and their An important future aspect will be the failures. number to 2-3 per second and using com­ study of neutrino-gamma correlations, provi­ mercial videoconferencing communications ded that sources emitting simultaneously Remote Operation products (e.g., DECSpin) reduces the rate to high-energy neutrinos and gamma rays Traditionally in astronomy and high-energy 128 kb/s. But the local telephone operator exist and can be discovered. The production physics, physicists operate equipment direct­ only offers 56 kb/s lines so apart from dedica­ of energetic neutrinos is always accompa­ ly and manually, with around-the-clock ted lines for video, it is still early days for nied by the simultaneous gamma-ray pro­ watches in classical experiments. Direct ope­ videoconferencing via wide-area networks. A duction so both probe regions where par­ ration of DUMAND is not sensible. The expe­ solution being explored is to team up with the ticles are accelerated to high energies. We riment, once debugged and operating, has 64-mirror, CLUE gamma-ray detector plan­ can expect neutrino fluxes of anywhere from no moving components and no part can be ned for deployment next year on Maul Island equal to the gamma-ray flux to greater by a adjusted mechanically. The data rate after fil­ (CLUE Is a joint project between the Univer­ very large factor, depending on the thick­ tering and compression at the shore station is sity of Hawaii, three Italian universities and ness and density of material surrounding a low enough (6 x 109 bytes/week) for optical the Italian INFN). source, the distance to the source, and the or magnetic storage, or for transmission G. Grajew nature of intervening matter that attenuates through networks to remote computers. How­ University of Hawaii, USA Europhys. News 23 (1992) 169 gamma rays (attenuation of gammas by Miscellaneous topics The shore building and the 20 m tunnel to photons in the neighbourhood of some Other topics of interest include: searches carry cables underwater are presently being sources may also occur). Knowing gamma- for ultrahigh-energy gamma-ray point built. ray fluxes would allow the computation of sources via TeV muon astronomy, assuming neutrino fluxes. that appropriate processes which enhance [1] Aoki T. et al., Nuovo Cimento C 9 (1986) muon production do exist; searches for par­ 642. [2] Babson J. et al., Phys. Rev. D 42 (1990) Cosmic ray physics ticular forms of dark matter; the monitoring of supernova explosions, etc. 3613; Matsuno S. et al., Nucl. Instr. Meth. A DUMAND will also study the properties of 276 (1989) 359. high-energy cosmic ray (atmospheric) Moreover, DUMAND and its support sys­ tem make it possible to carry out simulta­ [3] DUMAND Joint Proposal. The Inter­ muons, in order to deduce spectral features, national DUMAND Collaboration, University and other aspects of the parent primary neously research projects in other fields such as oceanography, ocean biology, geo­ of Hawaii, Manoa, Honolulu: HI.HDC-2-88 radiation at energies ≥ 10 TeV. Although (1988). optimised for up-coming muons from neutri­ physics, environmental sciences, and clima­ nos (i.e., for muons produced by neutrinos tology. A small number of carefully selected *DUMAND Collaboration: transversing the Earth), DUMAND will still experiments will be Incorporated. TH Aachen G; Bern Univ., CH; Boston Univ., have an effective area of 12 000 m2 for The time schedule for DUMAND II fore­ USA; Univ. of Hawaii, USA; Univ. of Kiel, G; Kobe sees the beginning of data taking with the Univ., JA; Kinki Univ., JA; Okayama Sci. Univ., downward-going muons, or a yield of 3.3 x JA; Scripps Institution of Oceanography, USA; 106 events per year, and has the capability central plus two other strings in 1993, and Tohoku Univ., JA; ICRR, Univ. of Tokyo, JA; KEK, of resolving the 2% of events which involve regular observation of the universe in “neu­ Tsukuba, JA; Vanderbilt Univ., USA; Univ. of multiple muons. By determining the muon trino light” with the complete array in 1994. Washington, USA; Univ. of Wisconsin, USA. energy spectrum from the angular distribu­ tion of down-going muons and measuring the spectrum of the electromagnetic bursts caused by the muons we shall be able to High Energy Neutrino Telescopes study the primary cosmic ray spectrum up through the ultrahigh-energy range. Multiple The most elusive of the established funda­ Neutrino detectors are classified according muons will be resolved for spacings > 10 m mental particles because of their extremely to the type of neutrinos involved, energy, tar­ and energy flows in tight muon bundles of a weak interaction with matter, neutrinos are get size, detection technique, location, and few metres in diameter so the composition unique tracers of phenomena occurring in angular resolution. Arrays designed to ob­ of primary cosmic rays can be determined in otherwise inaccessible regions of the Uni­ serve cosmic neutrinos expected from celes­ the disputed energy region between 100 verse. There are three well-recognised fami­ tial sources are called neutrino telescopes. and 10 000 TeV. lies of neutrino (electron-, muon- and tau- Neutrinos, being weakly interacting, are neutrino) of so far undetected but thought to difficult to detect. Nearly all neutrino tele­ Muon and neutrino physics be increasing mass. Being uncharged, neutri­ scopes are placed in neutrino observatories DUMAND will complement accelerator- nos maintain their directionality so point located deep underground or deep underwa­ based high-energy physics research since sources are identifiable. The lifetime of the ter in order to be shielded from atmospheri­ neutron is too short to survive great dis­ cally generated, charged-particle seconda­ there are no existing or planned facilities tances, even at the highest energies, so of ries of cosmic rays. There are some 20 neu­ that offer a neutrino beam above about the other known uncharged particles, only trino observatories around the world, and 600 GeV. There will be investigations of gammas are used for direct astrophysical new ones are being proposed all the time. muon and neutrino interactions and other observations. topics of particle physics at energies far High-energy neutrinos and gammas Cosmic-Ray Physics beyond the range of present and planned coming from a point source interact at the top Cosmic rays are mostly relativistic protons accelerators (e.g., CERN’s LHC and the of the atmosphere to produce a cascade with a significant fraction of elections and SSC now under construction in Texas, containing muons that points back approxi­ heavier nuclei. There is a need to explore the USA), including testing the neutrino oscilla­ mately to the source since the trans­ tion hypothesis in hitherto untouched do­ verse momentum of the muons pro­ mains such as those provided by long-base­ duced is small. So high-energy par­ line experiments. These experiments in­ ticle astrophysics essentially relies volve using high-energy particles (especially on gamma-ray astronomy (using protons) produced by accelerators as a large extensive air-shower arrays of source. Owing to still unknown interactions, detectors), and neutrino and muon neutrinos may change their nature from say astronomy (using well-shielded a muon-neutrino to an electron-neutrino by detectors): Fig. 1 illustrates schema­ a process involving periodic changes in fla­ tically the various alternatives. vour (the MSW-effect). Theory predicts that the rate of change depends on the diffe­ Neutrino Observatories rence in neutrino masses Am and the over­ The wide range of detectable neu­ lap of the two states of which sin2(2θ) is a trino energies may be divided into measure, where θ is the mixing angle. In­ experimental ranges according to creasing the baseline simply extends the principal aims: < 40 MeV — e.g., solar neutrinos, time available for mixing. DUMAND will and neutrinos emitted within seconds make contributions in the range 0.01 ≤Δm2 of stellar collapse; < 100 eV2 for sin2(2θ) ≥0.1. >108 eV — solar flare neutrinos; The rising neutrino cross-section and the >1 GeV — “high energy neutrinos” large effective volume of DUMAND, which Is which may be: 108 m3 for 2 TeV neutrinos, offer the best - Cosmic-ray generated neutrinos available opportunity to study neutrino inter­ produced by the decay of mesons actions at energies above 1 TeV, employing produced in the nuclear interactions the dominant atmospheric muon-neutrinos. of cosmic-ray particles with the DUMAND will also be capable of setting Earth’s atmosphere (“atmospheric limits several orders of magnitude better neutrinos”) ; Fig. 1 — Muon and neutrino phenomenology. Shown than any previous detector in searching for - Astrophysical sources of three are sources of atmospheric and extraterrestrial high- exotic particles, rare phenomena such as basic types (point, diffuse emission energy neutrinos and of muons [from Grieder, P.K.F., WIMPS, dark matter, baryon decay by mag­ from the interstellar medium, and Proc 6th Int. Symp. on Very High Energy Cosmic Ray netic monopoles, etc. cosmological). Interactions (Editions Frontières) 1990].

170 Europhys. News 23 (1992) Setting up the NT pilot experiment on the frozen surface of Lake Inside the 8.103 m3 water-filled tank of the US's Irvine-Michigan- Baikal. Shown are the spherical detector modules and components of Brookhaven (IMB) underground detector. The walls, are covered with the support structure. Cherenkov counters. energy spectrum and elemental (chemical) dules with “intelligent” phototubes survived at the possibility of shallow water (50 m) detec­ composition of the primary radiation (before it 1070 m owing to leaky feedthroughs. This tor for the NESTOR collaboration in the Bay interacts with the atmosphere) at very high illustrates the difficulty of deep underwater of Pylos, Greece. energies. This is because indirect measure­ experiments. Preliminary data analysis gave ments indicate a change of slope in the pri­ 60% of the Monte Carlo predicted rate. The Underground Experiments mary spectrum around a few 100 TeV, and plan is to deploy NT-48 comprising four half Some large underground detectors also possibly a flattening beyond 106 TeV. Se­ strings of 12 detectors modules In 1993, and have the capability to carry out neutrino veral explanations have been proposed for the full 192 module NT-200 array of eight astronomy of a limited sort. The 400 m2 these phenomena strings of 24 detectors in 1994. Irvine-Michigan-Brookhaven (IMB) experi­ Muon and muon-neutrino physics are clo­ ment was original designed to study nucleon sely related and muon-neutrino experiments Shallow Underwater Observatories decay processes. It comprised 8.103 m of are, in principle, suitable for detecting muons A number of large-area, Cherenkov-type highly purified water in tank located 1600 arising from interactions of primary cosmic shallow water (< 1000 m) neutrino and muon mwe (metres, water-equivalent) underground rays in the atmosphere. Existing experiments observatories have been, or are being, ope­ with 2048 photo-multipliers to detect Cheren­ to measure the cosmic-ray muon spectrum rated including LENA, Lake Issuk-Kul, and kov light mounted on six sides of the tank. are too small to acquire reliable statistics and AMANDA: proposed are GRANDE, PAN, IMB, which operated from 1982 to 1991, estimates of the energies are poor. Obser­ NET, and a shallow water NESTOR. The detected along with the Japanese vations of multi-muon events gives data of backgrounds associated with such arrays are Kamiokande experiment, the neutrino burst the primary chemical composition because about 105-106 times higher than for the deep from supernova SN1987A that heralded the the multiplicity distribution depends on the underwater experiments so it is difficult to era of neutrino astronomy. Upgrading of the elemental mass composition of primaries. convince people that they can be used effec­ Kamiokande water Cherenkov detector some But such events may possibly be confused tively for detecting astrophysical neutrino 2400 mwe underground from 2.103 to 5.104 with sequences of events unless the angular point sources. m3 by 1996 will extend possibilities. Mean­ resolution is good. Hence the importance in The Japanese 4.104 m2 LENA project was while, the 3.103 m2 MACRO and the 3.103 cosmic-ray physics of large muon-neutrino abandoned in 1991 after a pilot phase. Pre­ m3 LVD liquid scintillator detectors in the detectors with good angular resolution. liminary data taking has been reported this Gran Sasso lab some 3200 wme under­ summer for a 2.106 m3, 256-module Cheren­ ground will continue searches for neutrino Underwater Telescopes kov array operating at 600 m in Lake Issuk- point sources begun in 1989 involving the The main deep underwater, large-area Kul, CIS. AMANDA, sited on Antarctic Ice, observation of energetic muons produced by detectors primarily Intended for searching for offers some Intriguing advantages but faces interactions with the Earth's atmosphere. astrophysical point sources are NT-200 (pilot horrendous logistics. Funded by the US Although having good background rejection in operation; 2.103 m2 planned; Lake Baikal, National Science Foundation, the first phase and good pointing ability, MACRO and LVD CIS); DUMAND (2.104 m2 ; under construc­ calls for two strings of five detectors spaced 5 may be either too small or too marginal to tion off Hawaii, see p. 167); and NESTOR m apart to look for muons produced by rare effectively observe astrophysical point (3.103 m2 pilot under construction; 105 m2 neutrino interactions in the ice; gradual sources of neutrinos. proposed; in the Mediterranean off Pylos, extensions to 105 m2 are foreseen. Relatively see page 172). All are based on the concept small upward looking detectors have already A model of the MACRO detector in the Gran suggested In 1969 of detecting charged cur­ been sunk to detect energetic muons asso­ Sasso underground laboratory. Each of the rent interactions of neutrinos by observing ciated with extensive air showers. Measure­ six 12 x 12 x 5 m3 supermodules comprises the Cherenkov radiation of the produced ments of the transmission length are needed streamer tubes interleaved with passive muons as they move through highly transpa­ to see if trapped air bubbles are a problem as detector material. rent water. the length sets the detector scale (and cost). The NT (Neutrino Telescope) collaboration The 3.104 m2, 50 m deep GRANDE artifi­ involving three Russian institutes and cial lake project proposed for a quarry in DESY's Zeuthen lab in Germany saw the Arkansas, USA, was discontinued in favour deployment during the 1991 /2 winter from the of DUMAND. The Swedish PAN collaboration frozen surface of Lake Baikal of the basic for a 104-105 m2 array in a northern Swedish heptagonal “umbrella” mechanical structure lake has recently made some hydrological to 1360 m In deep, clear water. A two-string measurements and is presently exploring the mini-array of 14 detector modules was sus­ possibility of joining another collaboration. pended for long-term tests and to make a The Gran Sasso lab in Italy has proposed space reconstruction of atmospheric muons. NET, a 105 m2 detector placed in a purified However, only four of the 14 Russian-desi­ water filled, artificial lake built above the exis­ gned upward and downward-looking mo­ ting underground facilities. Finally, there Is

Europhys. News 23 (1992) 171