Aspects of neutrino astronomy
ASPECTS OF NEUTRINO ASTRONOMY A. M. BAKICH School of Physics, University of Sydney, Sydney 2006 N.S. IV., Australia (Received 18 August, 1988) Abstract. A broad overview of the current status of experimental neutrino astronomy is presented. Particular emphasis is given to the major recent developments that have occurred during the last few years. It is concluded that these developments and the next generation of experiments currently being installed signifies the coming of age of neutrino astronomy. I. Introduction The aim of this paper is to provide a review of various aspects of experimental neutrino astronomy. This field represents a dramatic merging of concepts and techniques of submicroscopic elementary particle physics with the astrophysical theories and phenomena on a supermacroscopic scale. The field is not new, as most ideas have been debated for some time (Ruderman, 1965; Chiu, 1966; Lande, 1979). Our motivation for this review stems from a number of very recent developments which appear to indicate that neutrino astronomy is gradually emerging from the phase of 'optimistic speculation and pilot experiments' into an active and exciting field of research. It is very likely that by the time review appears, new ideas and in particular new results will have been achieved. 1.1. SUMMARY OF RECENT DEVELOPMENTS We begin our discussion with a brief summary of recent progress with reference to a schematic neutrino energy spectrum shown in Figure 1. In the MeV energy range, there has been a long standing discrepancy between the predicted and the observed fluxes of the SB solar neutrinos, sometimes referred to as the 'solar neutrino problem'. Over the last couple of years, and for the first time, new data has been accumulating by a direct-counting electronic experiment. In addition, several other projects are in advanced stages of preparation, two of which will be measuring the flux of pp neutrinos (Section 2). The registration of supernova SN1987A by several neutrino detectors, although widely expected on theoretical grounds, came as a bonus to the experimenters. Since the frequency of observable supernovae explosions is known to be rather low, the obtained neutrino data, however meagre, is unique. These results and the significance of their interpretation are discussed in Section 3. At somewhat higher energies (,~ GeV) samples of 'fully contained' events have been recently accumulated by the nucleon decay detectors. These events can be directly attributed to the atmospheric (cosmic ray produced) neutrinos and represent the Space Science Reviews 49 (1989) 259-310. 9 1989 by Kluwer Academic Publishers. 260 A.M. BAKICH 15 10 r> -5 To0 O ~4 -~o J 2 -15 -20 -25 -30 i , i I I I I I I I i I I I I I I . ~ SN1987A A DIFFUSE 1MeV 1GeV 1TeV 1PeV ~ 1EeV ~ '~ I I | | I I | I I I I I I I 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 log (NEUTRINO ENERGY, eV) Fig. 1. A schematic representation of neutrino fluxes. This figure shows the expected extraterrestrial and atmospheric neutrino fluxes as a function of neutrino energy (after Koshiba, 1987; unpublished). background against which any extraterrestrial sources of neutrinos would have to be identified (Section 4). At still higher energies (,,~ TeV) the same underground detectors have been collecting data on upward and horizontal through-going muons. Since these muons are known to be a characteristic signature of neutrino interactions, this information could provide a means of establishing the existence of discrete point sources of high energy neutrinos (Section 5). One such (albeit still very controversial) identification attempt is the Cygnus X-3 pulsar. Finally, the long planned giant underwater experiments, intended to observe neutrino interactions at the ultra-high (~ PeV) energies, appear to be gradually moving past the preliminary prototyping stage. Interestingly, some estimates of the PeV neutrino flux limits have recently been obtained by a cosmic ray air shower detector (Section 6). ASPECTS OF NEUTRINO ASTRONOMY 261 The significance of these developments indicates that each one of them undoubtedly does deserve a separate and detailed review. However, our intention here is to provide a broad and global overview of the entire field of neutrino astronomy. It is the combined effect of all these recent developments that perhaps signifies the coming of age of neutrino astronomy. 1.2. NEUTRINO PROPERTIES AND INTERACTIONS It is perhaps surprising how much and yet how little is known of the neutrino itself, sometimes referred to as 'the most elusive particle in nature'. Neutrinos are point-like, neutral, spin -1 particles participating only in weak inter- actions. It is well established that at least two and almost certainly three flavours of neutrino (Ve, v,, and v~) exist in nature, in correspondence to their lepton partners, the electron, the muon, and the tan. However, whether the neutrino is its own antiparticle, or more specifically, what type of wave-function (two-component Majorana or four- component Dirac) actually describes the neutrino is not known and indeed is a subject of intensive theoretical and experimental investigations (Haxton and Stephenson, 1984). Similarly, the upper limits on neutrino masses have been progressively reduced by a series of painstaking and perhaps still controversial experiments. These limits are 17 eV < mve < 40 eV (tritium decay) mve < 18 eV (tritium decay) my, < 250 keV (g/~ decay) mv~ < 70MeV (zdecay) Boris et al., 1987, Fritschi et al., 1986, Abela et al., 1984, Albrecht et al., 1985. Hence, finite neutrino masses cannot be precluded, an uncertainty which has crucial implications for many theories and interpretation of experimental data (Vuilleumier, 1986). The question of neutrino types and masses is closely related to the possibility of neutrino oscillations (Pontecorvo, 1958). This hypothesis does not rely on an arbitrary (and often tacit) assumption of the massive neutrino eigenstates being aiso the eigenstates of the weak interaction; instead it relates the two eigenstate vectors by means of a flavour mixing matrix. The resulting oscillations between neutrino types are then determined by the neutrino masses and the mixing strength (or Am 2 and sin220 in the case of two neutrinos ve and vu). The above intrinsic or vacuum oscillations are modified on passage of neutrinos through matter (Wolfenstein, 1978) because the v,e (and the v,e) forward scattering amplitude is due to neutral current only, whereas vee does have an additional charged eurrent component. Recent realisation (Mikheev and Smimov, 1985, 1986) of the resonant character of this effect implies that even if the intrinsic mixing of neutrino types is very small, the oscillations can be dramatically enhanced under certain conditions, as determined by ~Am 2 cos 20 p- E r 262 A .M. BAK ICH where p is matter density, Ev is the neutrino energy and a is a medium and oscillation- type dependent parameter. This MSW effect has profound implications on the outcome of many neutrino experiments, as indicated throughout this review. In fact, because of the uniqueness of neutrino sources and the very large distances involved, neutrino astronomy experiments are particularly sensitive to these oscillation effects (Bilenky and Petcov, 1987). It is, therefore, widely expected that the developing field of neutrino astronomy, apart from providing information of purely astrophysical significance, will help to resolve some of the above fundamental problems. For the purposes of this review, the most important property of the neutrinos are the cross-sections of their interactions with other particles, such as elementary leptons and quarks as well as hadrons and nuclei. It is the characteristic relative weakness of these cross-sections and their dependence on energy that lead to uniqueness of neutrino physics and neutrino astronomy as a subject matter. On the one hand, it makes much of the astronomical universe effectively transparent to neutrinos providing information unobtainable by any other means. As an example, a 1 PeV neutrino traversing galactic density of -,, 1 nucl. cm-3 would have an interaction length of > 1015 light years, much greater than the radius of the Universe. On the other hand, it makes the very task of detecting these neutrinos extremely difficult, requiring vast and especially equiped detectors. This difficulty can be appreciated by considei-ing a 1 GeV neutrino to which the Earth diameter represents only 10-4 of an interaction length. The energy dependence of cross-sections of some of the relevant neutrino reactions are schematically depicted in Figure 2. The important feature of this plot is the linear energy dependence of the cross-sections, characteristic of a point-like fermion-fermion interaction G2M O'to t ,,~ - - E~, 7"C where G is the Fermi coupling constant and M is the target mass. Low-energy reactions on proton targets exhibit a quadratic energy dependence before these exclusive processes saturate due to the form factor dependence upon Q2. At very high energies the linear energy dependence is expected to flatten due to the IVB propogator effects and QCD evolution of structure functions. 1.3. UNDERGROUND NEUTRINO DETECTORS Throughout this review we will be referring to a number of underground detectors which have, in recent years, supplied the bulk of the data leading to our current understanding of neutrino astronomy. It is important to note that the initial and primary goal of most of these detectors has been to search for nucleon decay events; a task to which (perhaps ironically) atmospheric neutrino interactions themselves constitute a limiting background (Perkins, 1984; Meyer, 1986). These existing and operational detectors are listed in Table I. ASPECTS OF NEUTRINO ASTRONOMY 263 z 0 t- O iii o3 t.o 0 n- O v 0 -30 -31 -32 -33 -34 -35 -36 -37 -38 -39 -40 -41 -42 -43 -44 -45 -46 -47 -48 ' l l l l l l l l l l | l l l l / / J / ~ f f f / //J---/, v G a ~ v C I 1MeV 1GeV 1TeV , 1PeV 1EeV .... i l i I i ~3 111 ' = 1 1.1 1'6 log (NEUTRINO ENERGY, eV) Fig, 2. Energy dependence of neutrino interaction cross-sections. The medium and high energy regions have been measured in many experiments (e.g., Eisele, 1986). The extrapolation to ultra-high energies indicates the results of calculations by Quigg et al. (1986). It is clear that any neutrino detector should have a very large target mass in order to obtain significant event yields. An additional but equally important consideration is that these neutrino events have to be extracted and identified from a substantial and in most cases overwhelming background. This background is due to far more prolific particles, which are likely to either directly or indirectly mimic the genuine neutrino signal. Therefore, most neutrino detectors have to be heavily shielded and/or located deep underground, in addition to being well instrumented and quite sophisticated (and expensive) installations. Unfortunately the main aim of this review does not allow for a detailed description of either their design parameters or performance specifications, except in special cases where these considerations did have a significant effect on the interpretation of the obtained results. Similarly, although the experimental programs of most collaborations do include a diverse range of particle and/or cosmic-ray physics topics, we limit our discussion only to the issues and results directly relevant to neutrino astronomy. 264 A, M. BAKICH TABLE I Existing underground neutrino detectors Detector Depth Target Mass En~in AO Start, (location) (rowe) (tons) (MeV) (degrees) Upgrade HOMESTAKE 4400 chlorine 133 0.814 - 1968 (South Dakota) G2CI 4 KGF 7000 prop. tube 140 Oct. 1980 (India) calorimeter Dec. 1985 ASD 570 liquid 105 5 - 1978 (Ukraine) scintillator BAKSAN 850 modular 120 12 2 Aug. 1979 (Caucasus) scintillator (330) Jun. 1980 SOUDAN I 1800 prop. tube 31 1.4 Oct. 1981 (Minnesota) calorimeter NUSEX 5000 streamer 120 1.0 Jul. 1982 (Mont Blanc) calorimeter IMB 1570 water 3300 25 8 Aug. 1982 (Ohio) Cherenkov (6800) Jun. 1986 HPW 1450 water 900 3 Mar. 1983 (Utah) Cherenkov KAMIOKANDE 2700 water 680 8.5 2.7 Jul. 1983 (Japan) Cherenkov (2140) Jan. 1986 FREJUS 4400 flash tube 560 300 1.2 Mar. 1984 (Alps) calorimeter LSD 5200 modular 90 6 - Oct. 1984 (Mont Blanc) scintillator Specifically, we do not discuss such topics as nucleon decay, magnetic monopole searches, multiple muon bundles or neutrino geophysics. An indication of the activity and interest in the field of neutrino astronomy should be particularly apparent from the number of newly approved detectors and pending proposals. These second generation experiments, listed in Table II, are in various stages of preparation ranging from being presently installed to being tested as a preliminary prototype. Broadly speaking, and apart from the specialized radiochemical detectors, most of the neutrino detectors listed in Tables I and II can be classified into the following groups: (i) Water Cherenkov detectors (Figure 3), usually very massive (> 1000 tons) and capable of low-energy threshold levels (~ 10 MeV), The important parameter is the photosensitive area coverage, which determines both the spatial resolution (,-~ 1 m) and the angular resolution of a few degrees. ASPECTS OF NEUTRINO ASTRONOMY TABLE II New detectors and proposals 265 Detector Depth Target Mass /~min AO Status (location) (mwe) (tons) (MeV) (degrees) GALLEX 4000 Gallium 30 0.233 - setting up (Gran Sasso) GaC13 BAKSAN ~ 3250 Gallium 50 0.233 - setting up (Caucasus) metal LVD 4000 modular 1840 5 0.2 setting up (Gran Sasso) scintillator ICARUS 4000 liquid 6 500 5 prototype (Gran Sasso) argon SOUDAN II 2200 drift tube 1 100 setting up (Minnesota) calorimeter MACRO 4000 stream tube 700 0.2 setting up (Gran Sasso) scintillator SUPER-KAMIOKA 2700 water 22000 5 2 proposal (Japan) Cherenkov (45 000) SUDBURY SNO 6200 heavy water 1000 5 proposal (Ontario) Cherenkov SUNLAB 3300 water 250 6 - prototype (Australia) Cherenkov BAIKAL 1350 underwater - 0.5 prototype (Lake Baikal) Cherenkov DUMAND 4500 underwater 3 107 - 0.5 prototype (Hawaii) Cherenkov (ii) Liquid scintillator detectors (Figure 4), often of modular design and offering low-energy thresholds (few MeV) but without any signal directionality and limited angular resolution, unless supplemented with external tracking planes. (iii) Tracking calorimeters (Figure 5), typically 100 to 1000 tons with the bulk of the material in the form of iron plates sandwiched between crossed planes of either propor- tional, drift, streamer or flash tubes. The spatial resolution of these detectors is deter- mined by the tube size and is typically of the order of few mm to few cm. Although very good angular resolution of < 1 deg can be achieved for long tracks, the minimum energy threshold is usually quite high (> 100 MeV). We conclude this introductory section with a schematic representation of the under- ground location of various neutrino detectors, superimposed on the well known under- ground muon depth-intensity plot (Figure 6). 266 A. M. BAKICH KAMIOKANDE -F- t ' J~ -~-~'~- - - J ~15.6m. - ~19m E J JUL ~U /UUL ~UL E LO c6 Fig. 3. Water Cherenkov detectors. BAKSAN LSD E Fig. 4. Liquid scintillator detectors. NUSEX Fig. 5. Tracking calorimeter de ASPECTS OF NEUTRINO ASTRONOMY 267 106 105 ~'E 104- x LL o 103_ 10 2 - 101 - 1AS D I ~ I I BAKSAN IMB SOUDAN I SOUDANII SUPER- -KAMIOKANDE 8UNLAB GALLEX "t \ LVD / MACRO / GRAN SASSO' ICARUS l KAMIOKANDE HOMESTAKE FREJUS NUSEX LSD SUDBURY I I I I I I 'i I I I I 0 2000 4000 6000 8000 Depth (mwe) Fig. 6. Underground location of various neutrino detectors. The curve represents the flux of underground cosmic-ray muons. The existing detectors are indicated by closed symbols labelled above the curve. The open symbols (labelled below the curve) represent new neutrino detectors, that are currently being assembled, and some of the proposed installations. 2. The Solar Neutr ino F lux The well-known discrepancy between the theoretically predicted flux of 8B solar neutrinos and the only experimental attempt that succeeded to measure this flux is commonly referred to as the 'solar neutrino problem'. A historical account of the issues involved can be found in BahcaU and Davis (1982). 268 A.M. BAKICH In this section we review the current status of this (until recently) stalemated problem, before concentrating on new results and new projects which promise, at long last, to provide some decisive information. 2.1. THEORETICAL FLUX PREDICT IONS The energy production in the Sun is understood to be due to a series of thermonuclear reactions originally formulated by Bethe (1939). These well-known reactions are listed in Table III. An essential by-product of some of these reactions is the emission of low-energy (~ MeV) neutrinos, thus constituting the solar neutrino flux. TABLE I I I Energy producing nuclear reactions in the Sun React ion Energy range (MeV) p + p ~ 2H + e + v 0-0.420 p + e - + p~ 2H + v 1.44 ZH + P ~ 3He + 7 3He + 3He .-~ 4He + p + p 3He + 4He ~ 7Be + ), 7Be + e - ~ 7Li + v 0.86 (90~o), 0.38 (10~o) 7Li + p ~ 4He + 4He 7Be + p ~ 8B + ~, 8B ~ 8Be + e + + v 0-14.1 8Be ~ 4He + 4He z2C +P ~13N +7 lSN ~13C +e + + v 0-1.20 lSC + P ~ laN + 7 14 N + p ~ 1sO + asO ~aSN +e + +v 0-1.73 15N + p ~ 12 C + 4He What is often overlooked is the fact that the observed discrepancy applies only to the 8B component of the solar neutrino flux, which represents a minute fraction (~ 0.01 ~o) of the emitted neutrinos. By far the most dominant (~ 90~) neutrino producing cycle is the fusion of four protons to form an alpha particle, which releases an energy of about 26.7 MeV 4p~+2e + +2v e. Due to the dominance of this process, corresponding to the first four reactions in Table III, a rough estimate of the total neutrino flux on Earth can be obtained directly from the solar luminosity (Bahcall, 1985) 2L| 1 - 6.5 101~ cm -2 s -1 . 26.7 MeV 4z~R~ ASPECTS OF NEUTRINO ASTRONOMY 269 Therefore, the task of any solar model calculation, apart from refining the above rough estimate, is to provide an estimate of the flux of the extremely rare higher energy SB neutrinos. Over the past quarter of a century many calculations have been performed in an attempt to predict the solar neutrino fluxes. One particular series of very detailed calculations has become known as the 'standard' solar model (see Bahcall and Ulrich, 1988, and references therein). In the following we provide a brief summary of the assumptions and the general procedure employed by this model. The basic premise of the model is that the energy production in the sum is primarily due to the thermonuclear reactions summarized in Table III. Several other (simplifying) assumptions are included such as: (i) hydrostatic equilibrium of pressure and gravitation, (ii) energy transport by radiation and convection, and (iii) a chemically homogeneous initial composition. In addition to the above assumptions, various other input data are also required, such as the (extrapolated) values of the cross-sections for all the energy producing reactions (reviewed by Filippone, 1986), the primordial elemental abundances and solar opacities. Starting with the initial set of parameters, the calculation itself proceeds by stepping through time the equations of state of the gradually evolving sum and re-calculating the observable characteristics of the sum. Thus the solar model can describe the long-term solar evolution with the aim of matching various fixed boundary conditions, such as the luminosity, mass, and radius. The resulting energy spectra of solar neutrinos calculated by numerical integration are presented in Figure 7. The total fluxes and their relative three-sigma uncertainties are listed in Table IV together with the estimated 37C1 capture rates. TABLE IV Solar neutrino fluxes predicted by the standard model (Bahcall, 1987) Neutrino Flux Error 37C1 Capture Rate source ( 106 cm-2 s - l ) (~/o) (SNU) pp 60 000 2 0.0 7Be 4700 15 1.1 13N 610 50 0.1 150 520 58 0.3 pep 140 5 0.2 8B 5.8 37 6.1 Total 7.8 + 2.6 Thus the 'standard' solar model prediction for the 37C1 experiment is 7.8 + 2.6 SNU. The estimated error should be treated with caution (Bahcall and Ulrich, 1988). Firstly, the error distributions of the input parameters on which this statistical uncertainty is based are not known sufficiently well. Secondly, since this calculation, however detailed, is based on what is essentially a very simple model (Roxburgh, 1985) 270 A.M. BAKICH 10 ~ 101~ 7Be _ ,-7 10 ~ 10 6 10 4 0.1 1.0 10 Neutrino Energy (MeV) Fig. 7. The energy spectra of solar neutrinos. This figure shows the energy dependence of solar neutrino fluxes, predicted by the 'standard' solar model (Bahcall and Ulrich, 1988). The neutrino fluxes are in units of (cm '-2 s - 1 MeV- 1) except for the monoenergetic sources (cm -2 s - 1). a possibility of systematic errors cannot be excluded. Finally, the extreme sensitivity of 8B neutrino flux on the central temperature should be noted, even though this tempera- ture is not an input parameter of the model. 2.2. SOLAR NEUTRINO FLUX MEASUREMENTS Early attempts at measuring the flux of solar neutrinos (reviewed by Reines, 1967) have not been succesful. Because of their small target mass and a limited ability to cope with low energy background, these detectors have only succeeded in egtablishing an upper limit of about 109 cm-2 s-1, and were subsequently abandoned. 2.2.1. Homestake Chlorine Experiment Until recently, the only experimental data on the solar neutrino flux has come from the original BNL experiment at Homestake. This radiochemical experiment is based on the reaction V e + 37C1 ~ e- + 3TAr, which has a threshold energy of 0.814 MeV (suitable for 8B neutrinos) and involves measurement of the production rate of 3TAr (half life of 35 days). The experimental procedure needs little introduction and only a brief outline is given here (see Rowley et aL, 1985). The neutrino target consists of 133 tons of 37C1 in the form of 610 tons of perchloroethylene (C2C14) located at 4400 mwe underground in the Homestake mine (South Dakota). At intervals of every 2 to 3 months the produced Ar is extracted by purging helium through the tank and collected into a small volume by ASPECTS OF NEUTRINO ASTRONOMY 271 a series of condensation, charcoal-trap cooling and purification steps. The sample is then transferred into a well-shielded proportional counter for 3TAr yield measurement. The counting procedure, lasting some 8 months, employs pulse rise-time and energy cuts to minimize the background. Finally, a maximum likelihood estimate of the 35-day half-fife decaying component above constant background is calculated. The well known results are presented in Figure 8, indicating an average production rate of about 0.46 + 0.04 Ar atoms per day. An allowance of 0.08 + 0.03 atoms per day 2.0 1.5 .,.9o 1.0 t~ r r t.-. o 0.5 0 n Fig. 8. f 1970 1980 8 I I I 1 1975 --t-- 1 1985 6 4 2 0 Experimental results from the Homestake solar neutrino detector. The measurements of Ar production rate in each individual run since 1970 are shown. The right-hand axis represents the 8B solar neutrino flux in SNU (1 solar neutrino unit = 1 capture per second per 1036 target atoms), with an offset of 0.08 atoms per day to allow for the cosmic ray muon induced background (Rowley et al., 1985). is made for the extensively studied muon-induced background, resulting in the net production rate of 0.38 + 0.05 Ar atoms per day, or 2.0 + 0.3 SNU (1 SNU = 1 neutrino capture per second per 1036 target atoms). It should noted that the Homestake technique has undergone many stringent tests and is now generally accepted as reliable, although a direct calibration test with an artificial source ofmonoenergetic electron neutrinos, such as 65Zn proposed by Alvarez (B ahcall and Davis, 1982) could not be performed. 2.2.2. KAMIOKANDE-H Experiment This direct counting experiment is based on the elastic scattering reaction v~+e-~v~+e- in which the scattered electron preferentially retains the direction of the incident neutrino, as illustrated in Figure 9. The effort originated in 1983 when it was realised that the original KAMIOKANDE-I nucleon decay detector did have the sensitivity to observe stopping # decay electrons down to energies of about 12 MeV. A major upgrade was undertaken to reduce the low-energy background by providing 1.5 m of active water shielding, new multihit ADC-TDC electronics, and an improved water filtration system. 272 A.M. BAKICH @ @ 9 0 Fig. 9. A response of the KAMIOKANDE-II detector to a low-energy electron. This exploded view of the detector shows the responding PM tubes (dot diameter proportional to the pulse amplitude) and the reconstructed Cherenkov ring (dashed line). The energy of this typical event has been estimated at 10 MeV (Cortez, 1986). The spread of photon hits is due to the multiple scattering of the electron, as illustrated by the Monte-Carlo simulated trajectories of twenty 10 MeV electrons. More recently (Totsuka, 1987) the top of the tank has been completely sealed off to prevent absorption of 222Rn from the open air. Despite all these efforts, the residual background trigger rate is too high to unambiguously identify electrons scattered by solar neutrinos. In fact, the off-line analysis employs three cuts (each reducing the background by a factor of about ten) specifically optimised to improve the signal to noise ratio (Totsuka, 1987). The final step of the data analysis relies on the directionality of the elastic scattering signal by estimating enhancement in the event rate away from the direction to the Sun, as shown in Figure 10. The results of the initial 128 day run with an energy threshold of 9.5 MeV produced a (conservative) 90 ~o confidence level limit of 3.2 106 cm -2 s - 1 (Hirata et al., 1987a). This upper limit is already substantially lower than the 'standard' solar model predicted flux and it can be expected to be further improved within the next couple of years. 2.3. IMPLICATIONS OF SOLAR NEUTRINO FLUX DISCREPANCY A first indication of a discrepancy between the predicted and the observed fluxes appeared after the two preliminary runs of the chlorine experiment in 1968. Virtually since that time many authors have attempted to provide some solution to this problem. These efforts are usually classified into two broad (but certainly not mutually exclusive) categories of either modifying the assumptions (and/or the parameters) of the 'standard' model or providing some other mechanism which could allow for a reduction of 8B neutrino flux. ASPECTS OF NEUTRINO ASTRONOMY 273 15 L i I I I r I I I L I I I t I I [ I ~s l o III "6 ~ ttt t I Itti -1.0 -0.5 0.0 ,0.5 1.0 cos(e) Fig. 10. Directional correlation of scattered electrons with the Sun. This plot shows the results of the first 128 days of operation of the KAMIOKANDE-II detector at energy threshold level of 10.5 MeV (Hirata et al., 1987a). The angular distribution of elastically scattered electrons is expected to be confined to the region of cos 0 > 0.75, as shown by the histogram corresponding to the 8B solar neutrino flux predicted by the standard solar model. No significant enhancement can as yet be seen in this region from this preliminary data. A. Alternative Solar Models Several 'non-standard' solar models that can account for reduced 8B neutrino flux have been suggested (see review by Rood, 1978, and a summary by Haxton, 1984). Details of these models are not considered here, except by listing the basic modifications to the 'standard' assumptions. These include: (i) low heavy element (Z > 2) initial core abundance, (ii) including mixing, either convective or turbulent diffusion, (iii) allowing for significant magnetic field effects. Some of these efforts can be described as either ad hoc, or too speculative, or perhaps not consistent with other currently accepted ideas. However, few of these 'non-standard' models have been studied sufficiently or calculated with precision of the standard solar model. This is particularly relevant now that the standard model itself appears to be inconsistent with the KAMIOKANDE-II results. B. Neutrino Propagation Neutrino properties that could explain the flux discrepancy include decay, electric or magnetic moment. Various 'exotic' solutions have also been proposed from time to time, such as the hypothetical weakly interacting massive particles (WIMPs) or 'cosmions' (Faulkner and Gilliland, 1985; Spergel and Press, 1985). A possibility that neutrino oscillations, suitably enhanced by the MSW effect, could be responsible for the flux discrepancy has been suggested (Mikheev and Smirnov, 1985, 274 A.M. BAKICH 1986) and subsequently re-examined by many authors. This hypothesis assumes that 8B (electron) neutrinos produced in the hot inner core of the Sun undergo enhanced oscillations on passage through the outer layers, some two-thirds of them emerging in transformed flavours undetectable by the chlorine experiment. As can be seen from Figure 11, the 8B flux deficiency can be explained if the oscillation parameters are within the range of Am2< 1 x 10-4(eV) 2 , sin220 X Am 2 > 3 x 10-8(eV) 2 , sin220< 8 x 10 -a . I I I I 16 3 16' 16 5 ~> 6 "" 16 E ASPECTS OF NEUTRINO ASTRONOMY 275 In addition to the above solar effects, the possibility of MSW regeneration &neutrino flux in the Earth, leading to day/night variation has been considered (Cribier et al., 1986). The plans to measure this day-night effect with the chlorine detector by a series of daily Ar extractions are now in progress (Cherry and Lande, 1988). An important conclusion from these results is that if neutrino oscillations do take place no single experiment could unambiguously resolve the solar neutrino problem. Conversely, a series of flux measurements with different targets could provide unique estimates not only of the solar neutrino flux but also of the intrinsic parameters Am 2 and sin 2 20. This intriguing possibility is one of the major reasons for the current proliferation of new solar neutrino projects and proposals. 2.4. NEW SOLAR NEUTRINO DETECTORS AND PROPOSALS In this section we consider several new solar neutrino experiments that have been approved and are in advanced stages of preparation. There is little doubt that within the next few years these detectors will provide significant and perhaps decisive informa- tion on various aspects of solar neutrino physics. We also briefly mention a number of new projects that have either been recently proposed or are undergoing prototype testing. 2.4.1. Radiochemical Gallium Detectors (GALLEX and BAKSAN) Measuring the flux of the pp neutrinos has always been regarded as the crucial step in solving the solar neutrino problem. These lower energy (Ema x = 0.420 MeV) neutrinos constitute the bulk of the predicted flux and their flux is rather insensitive of specific model parameters. One particularly suitable reaction is neutrino capture in gallium v e + 71Ga---~ 71Ge + e - , Since the energy threshold of this reaction is only 0.233 MeV, a substantial fraction (5490) of the expected total capture rate of 132 SNU should be due to the pp neutrinos (Bahcall and Ulrich, 1988). Recently two projects based on the Ga reaction have been approved and are currently being installed. These are the European GALLEX experiment and the expansion of BAKSAN project by the Moscow INR group and their US collaborators, The GALLEX target is 30 tons of gallium in the form of concentrated GaC13 solution (Kirsten, 1986). The neutrino produced Ge would be extracted every two weeks (in the form of volatile GeC14) by a circulating stream of air or helium. Subsequent processing consists of absorption, re-extraction and conversion of GeC14 into GeH 4. After chromotographic purification, the gaseous sample is mixed in with proportional counting gas and loaded into a heavily shielded (Pb and Fe), minituarized (5 em 3) counting system. Finally the Ge yield is counted by means of a pulse-rise-time discrimi- nation technique in anti-coincidence with an active veto (NaI and plastic scintillator) enclosure. The INR (Baksan) target is 60 tons of gallium metal, which although less sensitive 276 A.M. BAKICH to background and more compact, does require a preliminary step of separating the produced Ge (Barabanov et al., 1985). This is achieved by dissolving out the Ge into an aqueous HC1 solution with increasing concentration until GeC14 can be swept out and processed as described above. Both extraction techniques have been tested in pilot experiments at up to 10~o of full scale, confirming the recovery efficiency of better than 95 ~o. However, both groups intend to use an artificial low-energy neutrino source not only to test the entire detector but also to provide direct data on the Ga neutrino capture cross-section. These calibra- tion runs would be performed on-line by inserting a 700 kCi source of s~Cr directly inside the Ga tank. 2.4.2. LANL 98Mo (Wolfsberg et aL, 1985) This geochemical experiment is aimed to test the hypothesis of possible temporal variations in 8B solar neutrino flux averaged over the past several million years. The basic reaction v e + 98Mo--+ 98Tc + e- calls for measuring the concentration of technetium (half-life of 4.2 million years) in a deeply buried, geologically stable deposit of molybdenum ore. This technically difficult project requires extraction and mass-spectrometric analysis of some 107 atoms of 98Tc from about 2600 tons of raw deposit ore. It is interesting to note that if the concentration of 97Tc produced via the reaction v e + 98Mo ~ 97Tc + e- + n could also be determined it might provide a limit on the flux of galactic neutrinos due to past supernova explosions (Haxton and Johnson, 1988). The experiment is in its initial stages and if the efficiency of the technique can be verified it should produce uniquely important data. The interpretation of these results does critically depend on the accuracy of the molybdenum capture cross-sections. 2.4.3. LVD (Bari etaL , 1988) The Large Volume Detector (LVD) which is being currently installed at the Gran Sasso Laboratory is based on the proven design of its Mont Blanc predecessor LSD. The detector consists of 1800 tons of modular liquid scintillator target interspersed with 20000 streamer tube tracking planes and 1800 tons of Fe shielding slabs. The very extensive experimental program includes a measurement of SB solar neutrino flux by means of the elastic scattering reaction ve+e-~ve+e- . With a proposed triggering threshold of only 3 MeV (which may be achievable in the well-shielded inner modules of the detector) and an energy resolution of 20 To, the LVD should have the capability to undertake a high statistics study of the 8B neutrino energy spectrum. ASPECTS OF NEUTRINO ASTRONOMY 277 2.4.4. ICARUS (e.g., Bahcall etal., 1986) An Imaging Cosmic and Rare Underground Signals (ICARUS) detector is a partly approved project consisting of 6500 tons of cryogenic liquid argon to be installed at the Gran Sasso Laboratory. In addition to the elastic scattering reaction, this detector will also be sensitive to v e capture in 4~ 2.4.5. Super-KAMIOKANDE (Suzuki, 1987) A proposal for a vast (see Table II) water Cherenkov detector to be installed near the existing KAMIOKANDE-II. With 40~o photosensitive coverage and an improved signal/noise ratio, the expected 8B neutrino counting rate is estimated at 46 events per day (at 5 MeV threshold) from the standard model flux. The experience gained by the Tokyo University group with the KAMIOKANDE-II detector will be very important for this project. 2.4.6. Sudbury SNO (Ewan et al., 1987) A proposal for a large heavy water Cherenkov detector to be located at 6200 rowe underground near Sudbury, Ontario. This experiment is aimed to measure the 8B (electron) neutrino flux via the charged current inverse beta reaction ve+d- - ,p+p+e- . One advantage of a DzO target is that provided the neutron background can be contained, a direct test of the MSW neutrino oscillations can be performed by measuring the total flux by means of the neutral current process vx+d~p+n+ Vx. 2.4.7. SUNLAB (Bakich and Peak, 1985) A proposal for a modular water Cherenkov detector to be built in a Broken Hill mine (Australia), featuring light collection with wave-length shifting panels and heavy lead shielding to reduce gamma background. The first prototype module of this detector is currently being tested. 2.4.8. Other Potential Projects Several other potentially promising solar neutrino targets have been considered at various times. These include 4He, VLi, 11B, SlBr, 115In, and 2~ In most cases, although considerable amount of preliminary work has been done, the technical feasibility of these experiments remains to be established and it is not clear whether sufficiently low levels of background can be achieved. We conclude our discussion of the solar neutrino fluxes by noting that the variety of these new experiments could hopefully lead to better understanding of solar physics and properties of neutrinos. 278 A.M. BAKICH 3. Supernova SN1987A Neutrinos The generally accepted scenario of gravitational stellar collapse (see recent review by Woosley and Weaver, 1986) assumes that stars of about 8 solar masses gradually evolve to an 'onion-shell' structure of successive layers of increasingly heavier nuclei. This evolution is sustained by thermonuclear fusion reactions which provide the radiation and particle pressure that maintain the star in a quasistatic equilibrium against the inward gravitational forces. As the central iron-nickel (endothermic) core, under increasing gravitational contraction, reaches the 1.4 solar mass limit, the lack of further thermonuclear fuel implies that the above equilibrium can no longer be maintained. It is at the time that the pressure support of the exhausted thermonuclear fuel is removed that the core inevitably begins to collapse under its own weight. The time scale of the collapse is dramatically short, as it is further accelerated by the nuclear disso- ciation and electron capture processes. When the central core density reaches the nuclear density of ,,~3 x 1014g cm -3 the infalling material is met with an over- compression bounce, resulting in a powerful outgoing shock wave that reverses and ejects most of the outer layers of the star. The remaining core contracts to form a neutron star. 3.1. SUPERNOVA NEUTRINO EMISSION To form a neutron star a binding energy of ~ 3 x 1053 ergs must be released. Since the total kinetic, optical and gravitational-wave energy release can account for only 1051 ergs, the bulk of the released energy (> 99~o) would be emitted in the form of an intense burst of neutrinos (Colgate and White, 1966). Various numerical calculations have been performed to obtain some quantitative predictions of the time structure, energy spectra and relative fluxes of neutrino types (e.g., Burrows and Lattimer, 1986; Mayle et al., 1987). These results depend on the assumed neutrino transport processes, the equations of state and zoning techniques, as discussed by Bruenn (1986). Two main stages of neutrino emission are usually distinguished (Figure 12): (i) The initial prompt (~ 10 -2 s) pulse of electron neutrinos due to the neutronization (or electron capture) by the rapidly dissociated protons e- +p- - .n + v e resulting in conversion of most of 89 A 1.4 M o ,,, 1057 electrons into a burst of about 1052 ergs of electron neutrinos. This neutrino emission is curtailed by the abruptly increasing core density, as the matter becomes opaque to neutrinos at around 3 1011 g cm -3 leading to neutrino 'trapping' by scattering and re-emission. (ii) The gradual (,-~ 10 s) thermal diffusion of pair-produced neutrinos of all flavours that continues after the formation of the neutrinosphere via several channels such as e +e- - -~v+ ~, plasmon --* v + ~, y+e~e+ v+u ASPECTS OF NEUTRINO ASTRONOMY 279 1 0 55 I I I I I I ico co 0 v cO 0 E .-J 10 54 1 0 sa 1 0 s2 % 10 51 i I L. ~ I 0 1 2 3 4 5 Time (s) Fig. 12. The predicted time structure of supernova neutrino emission. The very intense initial neutroni- zation burst of electron neutrinos is shown by the dashed line. The bulk of neutrinos of all flavours is emitted during the subsequent cooling stage (solid curve) continuing for some 10 s after the explosion. Adapted from Mayle et al. (1987). It is during this 'cooling' stage that the bulk (90%) of the neutrino flux is radiated. The dominance of the above pair-production processes means that roughly equal numbers of all neutrino flavours would be emitted. The typical calculated energy spectra of emitted neutrinos are shown in Figure 13. These results can be roughly approximated by a thermal Fermi distribution, modified by absorption and reemission (e.g., Nadezhin and Otroshchenko, 1980). We note that the average energy of v, and v~ neutrinos can be expected to be significantly higher since they are trapped in the neutrinosphere only by neutral current interactions. The above consideration indicate that neutrino interactions are a major form of energy transport that is directly related to the dynamics of a collapse. In fact, the neutrino emission processes play a crucial role as both the initial neutronization 'trigger' and the subsequent thermal cooling 'relief of the excess binding energy. It is for this reason that the features of neutrino bursts from collapsing supernovae and their theoreti- cal interpretation are of special interest to astrophysics and neutrino astronomy. 280 A.M. BAKICH i ~ I f ! I I I I I 1058 1056 ~ - Ve - .~ 10 ~4 10 52 10 50 0 20 40 60 80 O0 Neutrino Energy (MeV) Fig. 13. The energy spectra of supernova neutrinos. This plot represents results of calculations of Mayle et aL (1987), with the area under each curve representing the neutrino luminosity in MeV. Although the average energy of electron neutrinos and antineutrinos are comparable ( ~ 10 MeV), the energy ofmuon and tan neutrinos can be expected to be significantly higher, as discussed in the text. 3.2. NEUTRINO BURST DETECTION CONSIDERATIONS Because of their low-energy thresholds (typically of the order of 10 MeV) water Cherenkov and liquid scintillator detectors are considered especially suitable for detec- tion of neutrino signal from collapsing supernovae. For both types of detectors, the two main neutrino interactions for supernova flux registration are: (i) Ve capture Ve + P ---' n + e + ; (ii) elastic scattering v+e- - - - *v+e- Although the first reaction is applicable only to the Ve component of the supernova neutrino flux, it is by far the dominant process because of its higher cross-section (with quadratic energy dependence, see Figure 2), which more than compensates for the proton/electron target ratio. In this reaction, the isotropically emitted positron essen- tially retains the energy of the incident neutrino, as opposed to the directional (~ 15 ~ elastic scattering reaction. The sensitivity of any detector to a supernova neutrino burst is determined by several strongly interdependent factors: ASPECTS OF NEUTRINO ASTRONOMY 281 (a) Firstly, the target mass of the detector essentially defines its response in terms of the number of neutrino interactions that will take place in case of a burst. An approximate estimate of this response can be readily obtained by assuming an average interaction cross-section a corresponding to the mean neutrino energy Ev: Etot/6 a Number of events - - - fVNA, 4riD 2 Ev where Eto t is the total emitted energy, D is the distance (-~ 50 kpc for the LMC), f is the free proton ratio in the detector volume V, and NA is Avogadro's number. (b) Secondly, the energy threshold for both triggering and event reconstruction must be sufficiently low to register at least some of these neutrino interactions in the target. A useful parametrization (Krauss, 1987) of the triggering probability can be achieved by introducing a (detector dependent) efficiency parameterp: Triggering probability -- 1 - exp( - (Ev/Ethr)e). (c) Thirdly, the background event rate needs to be low enough to allow a genuine burst, occurring within a time interval, At, to be differentiated from normal statistical background fluctuations. Given the total background rate r, the rate of such random coincidences of multiplicity k can be calculated from Random rate - k(rAt)~k e-r~'. Atk! For any specific detector, the combined effect of these closely inter-related factors determines its sensitivity and efficiency of detecting neutrino signals from distant super- nova explosions. Obviously, a network of several detectors simultaneously observing a supernova neutrino burst would substantially improve the reliability of detection. Over the last decade several detectors have been continuously monitoring such bursts, with negative results, until the supernova SN1987A. 3.3. SN1987A NEUTRINO DATA Following the optical observation of supernova SN1987A on 24 February, 1987, four groups have reported the registration of a burst of neutrino events. Three of these signals (KAMIOKANDE-II, IMB, and BAKSAN) appear to be in coincidence, whereas the fourth (LSD) has apparently been observed some five hours earlier. Despite some controversial and variously interpreted discrepancies between these four registrations, it is now generally agreed that these bursts of low energy events do provide a compelling evidence of a genuine neutrino signal, directly attributable to the supernova SN1987A. This conclusion undoubtedly represents a milestone for neutrino astronomy. The following is a summary of the experimental data on the SN 1987A neutrino burst, as reported by the four groups, and devoid of any subjective or theoretical con- siderations. 282 A.M. BA~Cn KAMIOKANDE-H (Table V, Hirata et al., 1987b) A search for the SN1987A neutrino signal required a time-correlated (10 s) burst of low-energy (< 50 MeV), well-contained (outer response < 30 photoelectrons) events. One burst of 12 such events has been found at 07 : 35 : 35 UT (absolute error + 60 s). The accidental rate of occurrence of the burst due to a statistical fluctuation has been estimated at less than one event per 7 x 107 years or 1 x 105 years, depending on the multiplicity and energy selection criteria. Equally unlikely is the estimated probability (~ 10 -11) of the burst being induced by one of the preceeding muons. TABLE V KAMIOKANDE- I I SN1987A events (Hirata et al., 1987b) Event Event Number Electron Electron number time of PMT's energy angle (s) (Nhit) (MeV) (degrees) 1 0 58 20.0 + 2.9 18 + 18 2 0.107 36 13.5 + 3.2 15 + 27 3 0.303 25 7.5 + 2.0 108 + 32 4 0.324 26 9.2 _+ 2.7 70 + 30 5 0.507 39 12.8 + 2.9 135 + 23 6 0.686 16 6.3 + 1.7 68 + 77 7 1.541 83 35.4 + 8.0 32 + 16 8 1.728 54 21.0 + 4.2 30 + 18 9 1.915 51 19.8 + 3.2 38 + 22 10 9.219 21 8.6 + 2.7 122 + 30 11 10.433 37 13.0 + 2.6 49 + 26 12 12.439 24 8.9 + 1.9 91 + 39 Event time 0 corresponds to 07 : 35 : 35 UT. IMB (Table VI, Bionta et al., 1987) The search consisted of determining the number of (contained) events with fewer than 100 responding photomultipliers within non-overlapping 10 second intervals. A burst TABLE VI IMB SN1987A events (Bionta et al., 1987) Event Time Number Energy Angle number (UT) of PMT's (MeV) (degrees) 1 33162 07:35:41.37 47 38 74 2 33164 41.79 61 37 52 3 33167 42.02 49 40 56 4 33168 42.52 60 35 63 5 33170 42.94 52 29 40 6 33173 44.06 61 37 52 7 33179 46.38 44 20 39 8 33184 46.96 45 24 102 Event serial numbers are not sequential due to the 15 intervening muons. ASPECTS OF NEUTRINO ASTRONOMY 283 of 9 events has been located at a time of 07 " 35 9 41 UT (absolute error + 50 ms). One of the 9 events has been identified as a typical cosmic ray muon. BAKSAN (Table VII, Alekseev et al., 1987) Data analysis consisted of selecting events with > 4 signals within a 20 second sliding window, with a 50 MeV maximum energy cut. A burst of 6 signals starting at 07 9 36 " 06 UT (absolute error + 2 s) has been observed. Since the counting rate of individual pulses at 10 MeV level was about 0.033 s - 1 it was considered that the first (early) pulse should be attributed to background. We note that in a more recent paper (Alekseev et al., 1988) the BAKSAN group has reported that their absolute timing could have been late by up to 54 s. TABLE VII BAKSAN SN1987A events (Alekseev et al., 1987) Event Time Energy number (UT) (MeV) - 07 : 36 : 06.571 17.5 + 3.5 1 07 : 36 : 11.818 12.0 _ 2.4 2 12.253 18.0 3.6 3 13.528 23.3 4.7 4 19.505 17.0 + 3.0 5 20.917 20.1 + 4.0 The unnumbered first entry has been interpreted as unassociated background. LSD (Table VIII, Aglietta et al., 1987a) This burst has been registered and printed out in real time of occurence at 02 : 52 : 36 UT (absolute error _+ 2 ms). The burst consisted of 5 pulses (all in different counters, 3 of them within the internal fiducial volume) during 7 s. One of the 5 responding counters has also registered a low-energy (E = 1.2 MeV) pulse 278 gs after the main pulse, consistent with a neutron capture gamma signal. The frequency of a random fluctuation has been estimated at 0.7 events per year. In an accompanying paper TABLE VIII LSD SN1987A events (Aglietta et al., 1987) Event Time Energy number (UT) (MeV) 1 994 02:52:36.792 2 995 40.649 3 996 41.007 4 997 42.696 5 998 43.800 7 8 11, 1.2 (+ 278gs) 7 9 Note the second low-energy pulse in event 3. 284 A.M. BAKICH (Aglietta et aL, 1987b) the LSD collaboration has reported no evidence of an event burst at the time coincident with the other three detectors. Correspondingly, the other three detectors have no evidence of activity at the time of the LSD burst. We end this data summary with an illustration of the KAMIOKANDE-II response within a 400-s window around the time of the neutrino burst (Figure 14), and a graphic representation of one of the IMB SN1987A events (Figure 15). 100 , , , , , a ,m e-- Z 80 60 40 20 o 0 9 9 " . .,.. ': "9149 " "- 9 '" " 9 : t . ' ..7: 0 I I I I I I I -200 -100 0 100 200 Time (seconds) Fig. 14. The response of KAMIOKANDE-I I detector during the SN1987A explosion. This figure shows the sequence of events registered for a period of 200 s before and after the neutrino burst (Totsuka, 1987). The vertical axis represents the number of responding PM tubes in each event. Note that the nominal threshold level is Nhit = 20, which corresponds to electron energy of about 9 MeV. 3.4. INTERPRETATION OF SN1987A NEUTRINO SIGNAL The uniqueness of the SN1987A neutrino signal has triggered off an avalanche of preprints and papers attempting to extract all the possible physics content from the above experimental data. We subdivide these efforts into two categories; firstly, those concerned primarily with the astrophysics of neutrino emission from the collapsing star, and secondly, those concentrating on the properties of neutrinos themselves. 3.4.1. Neutrino Emission Firstly, the mere observation of neutrinos diffusing from the collapsed core over a period of several seconds means that no alternative cooling mechanism was involved. It confirms that neutrino emission is indeed the dominant process of releasing the binding energy of the collapsed core. ASPECTS OF NEUTRINO ASTRONOMY 285 4" I, / I t Fig. 15. A graphic representation of one of the IMB events due to SN1987A. This event is listed as number 3 in Table VI. The number of slashes at each point represents the pulse amplitude of the PM tube (Reines and Vandervelde, 1988). The mean neutrino energy has been estimated to be about 15 MeV. This estimate and the 10 s duration of the burst are certainly consistent with the predictions of trapped, lower energy neutrinos being emitted with a characteristic temperature of kT ,-~ 4 MeV. The lack of any pronounced directionality towards the LMC in the KAMIOKANDE-I I and IMB data (and the dominance of veP cross-section at these energies) implies that all of the observed events were due to electron antineutrinos. Although the first two KAMIOKANDE-I I events do appear to be directional, they cannot be attributed to Vee elastic scattering on both energy release and relative timing grounds (Sato and Suzuki, 1987). The total observed ~e luminosity has been calculated to be ~ 8 x 1052 ergs. Assuming roughly equal fluxes of neutrinos and antineutrinos of each flavour, the total energy release is (5 + 2) 10 s3 ergs. This estimate is consistent with a neutron star (rather then a black hole) being the final outcome of the collapse. Thus the neutrino data appears to agree remarkably well with the generally accepted standard scenario of a stellar collapse. It therefore can be seen as a dramatic and triumphant confirmation of the basic theory. 3.4.2. Neutrino Properties As mentioned above, the mere observation of a neutrino signal implies that no fluxes of other exotic light particles (such as axions), that couple to matter weaker then neutrinos were emitted (e.g., Raffelt and Seckel, 1988; Turner, 1988). 286 A.M. BAKICH Therefore, a limit on the number of neutrino species can be derived by comparing the maximum binding energy of a neutron star with the minimum estimated neutrino flux. This bound is Nv < 6-7 (e.g., Ellis and Olive, 1987). It has been long understood that, if finite, neutrino masses could be estimated directly from the arrival time delays (Zatsepin, 1968) according to m O(; 1) At= 2C E 2 ' 2 where D = 50 kpc and E is the neutrino energy. Various attempts at calculating neutrino masses from the arrival times and energies of the events have resulted in mass limits varying by a factor of about 10. The difficulty lies in the proper statistical treatment of a small data sample from an assumed source model that has to allow for an energy spectrum and delayed emission of neutrinos (Burrows, 1988). The consensus of opinion now appears to be that a very conservative 95 ~o confidence level upper limit can be assigned at about 15 eV. However, there is no convincing evidence that the observed electron (anti)neutrinos have non-zero mass. Bounds on other neutrino properties include the magnetic moment (less than 10 - ~2-10-13 of a Bohr magneton suggested by Goldman et al., 1988) and the electric charge (less than 10-16 of the electron charge suggested by Dar and Dado, 1987). Finally, no agreement exists on the significance of the 4 hour 43 minute time gap of the LSD burst. The hypothesis of two neutrino pulses due to the formation of a neutron star and its subsequent collapse into a black hole (Aglietta et al., 1987b) has not been generally accepted, although this scenario has been strongly supported by DeRujula (1987). We conclude that any neutrino properties derived from a very small sample of events, although useful, appear to be far less conclusive and categorical than the energy release balance arguments described above. 4. GeV Neutrino Atmospheric Background In the GeV energy range (0.1-100 GeV) substantial fluxes of atmospheric (cosmic-ray produced) neutrinos are known to be present. Although the detection of neutrino events in this energy range can be readily performed by the full containment cuts (as described below), the directionality of neutrino data is inherently poor. Therefore, the difficulty in searching for extraterrestrial neutrino fluxes lies in the fact that any such component must be recognized amongst a background of secondary, local neutrinos that are produced in the Earth's atmosphere. A detailed knowledge of the atmospheric neutrino fluxes is, therefore, absolutely essential. The general features of cosmic-ray interactions and subsequent air shower develop- ment are well understood (e.g., Hayakawa, 1969). These primary particles interact with air nuclei (proton mean free path ,-~80 g cm-2), transferring part of their energy (inelasticity ~0.5) into production of secondary hadrons, mostly ~ (and K) mesons. ASPECTS OF NEUTRINO ASTRONOMY 287 The neutral rc ~ mesons decay almost immediately (lifetime ~ 10 - 16 S) into gammas thus initiating the soft or electromagnetic cascade. The charged rc mesons can suffer further collisions contributing to the development of a hadronic (nuclear-active) component of the shower. Alternatively these mesons can decay into muons and neutrinos leading to a development of the penetrating shower component. The neutrino component is subsequently supplemented through muon decays. 4.1. CALCULATIONS OF ATMOSPHERIC NEUTRINO FLUXES The main meson decay modes that are expected to contribute to the neutrino flux are listed in Table IX. We note that the decay constant indirectly reflects the energy range TABLE IX Neutrino producing reactions Source Branching Lifetime Decay ratio (~o) (s) constant (GeV) ~--,# + v 100 2.6 10 -8 115 K~# + v 63.5 1.2 x 10 -8 850 ~t~e+ v+ v 100 2.2x 10 -6 1 K~ rcrc- ,##vv 68.7 8.9 10 -11 1172 K~ + rc + v 39,0 5.2 x 10 -8 202 --,#+ 7r+ v 27,1 K -+~e+Tr+v 4.8 1.2x 10-8 850 ~#+ 7r+ v 3,2 D -*K + 1 + v 4.8 9.2 10 -x3 4.3 10 7 ~K* + 1 + v 3.2 4.4 10 -13 9.0 107 A~ + ~A ~ + 1 + v 2.1 2.3 10 -13 2.1 x 108 ~A*+l+v 2.1 in which a given process should make a significant contribution to the total neutrino flux. Therefore, at lower energies o fE < 1 GeV the muon decay is by far the most dominant neutrino producing process. However, at higher energies o fE > 10 GeV and especially at ~ 100 GeV the zt and K meson decay contribution becomes more dominant, particu- larly in the case of muon neutrino flux. The prompt charmed decay modes are important only at much higher energies (see section 6). Because of the close relationship between neutrino and muon production, it is possible to infer the atmospheric neutrino fluxes from the well known energy spectrum of cosmic-ray muons (Allkofer et al . , 1971). However, more recent efforts do not rely on the muon flux measurements, except possibly for normalization. These calculations estimate the neutrino flux component directly by starting with the primary cosmic-ray spectrum and following the entire cascade shower development down through the atmosphere, by means of a combination of the diffusion equation and/or Monte-Carlo sampling approach. 288 a.M. BAKICH Recent results by Gaisser and Stanev (1984), Mitsui etal. (1986), and Bugaev and Naumov (1987) are presented in Figure 16, together with an earlier calculation by Volkova (1980). I t I I 101 ,-~ 100 ~0~ ?E o T> 0~ 10 -~ x u_ o 10 -2 t- Z .2 10 -3 e" 0 E < 10 -4 10 -5 0,, /0~0~ ~'0% e/ 9 9 O. \ 0 Volkova, 1980 0 Gaisser and Stanev, 1984 rl Mitsui et al., 1986 Z~ Bugaev and Naumov, 1987 I I I 0.01 0.1 1 100 \ O \ \ , \ 10 Neutrino Energy (GeV) Fig. 16. Energy spectrum of atmospheric neutrinos. The graph shows the results of recent calculations of atmospheric neutrino flux. In the low-energy region, all spectra at solar minimum. The difference between neutrino fluxes at IMB (open symbols) and at KAMIOKANDE-I I (closed symbols) is due to the geomagnetic cutoff. The calculations of Gaisser and Stanev (1984) extend down to energies of 0.01 GeV where the linear cascade approximation that is usually employed is expected to lead to an overestimation of the neutrino flux by about a factor of two, as confirmed by Lee and Bludman (1988). At these low energies the atmospheric neutrino spectrum exhibits a peak due to the combined effect of steeply decreasing primary proton flux and the ASPECTS OF NEUTRINO ASTRONOMY 289 increasing neutrino production yield9 It is also at these energies that the effect of solar modulation and the geomagnetic cutoff on the primary protom spectrum is significant. However, at neutrino energies of > 1 GeV the agreement between different calcu- lations is generally believed to be quite good (N 25~o), We also note that the slope of the neutrino spectrum closely reflects the primary parent slope, indicating that at these GeV energies most mesons decay before interacting. 4.2. FULL CONTAINMENT NEUTRINO EVENTS In any underground detector, an energetic interaction vertex without a visible incoming track is an obvious candidate for a neutrino interaction. Since in many cases the direction of all vertex tracks cannot be unambiguously ascertained, the usual procedure of selecting neutrino events involves the full containment requirement. This means that not only the vertex itself but all the vertex associated tracks must be fully contained within the inner fiducial volume of the detector. Typical fully contained and partially contained events are shown in Figure 17. NUSEX lm lm o . jg ' " t ..? oo 9 9 s . , ; " ; " " . ~.- '~U 9 ' -~ . l rn lm FRE JUS Fig. 17. Examples of typical contained neutrino interactions. In each case only one projection of a small section of the total detector volume is shown. The fully contained NUSEX event corresponds to an inelastic v e (or v,) interaction with zenith angle of ~ 170 ~ and an estimated energy of 1.5 _+ 0.4 GeV. The Frejus event is only partially contained and because of the muon track escaping from the detector volume, only a lower limit of E~ > 10 GeV can be set on the incident neutrino energy. 290 A. M. BAKICH A useful feature of this method lies in the fact that since the event is fully contained, an energy estimate of the incident neutrino can be obtained. An additional bonus of the full containment condition is that it can provide an indication of the type of incident neutrino. Typically, an electron neutrino interaction would be recognised by an energetic (> 100 MeV) soft shower, whereas a muon neutrino event would be identified by the longest non-interacting track. However, if no obvious lepton candidate is present at the vertex, the event would be attributed to a neutral current interaction. Monte-Carlo studies indicate that up to 90 ~/o correct identification can be achieved by these proce- dures. The main shortcoming of the full containment method is that the finite size of the fiducial volume of the detector imposes a limit on the range of outgoing secondaries and, hence, on the maximum acceptable neutrino energy. For most existing underground detectors this limit is typically of the order of a few GeV but it can be exceeded if some well-identified secondaries are allowed to escape (partial or vertex-only containment, see Figure 17). This maximum energy restriction in turn implies that the direction of the incoming neutrino cannot be accurately determined because of the relatively large angle between the neutrino and the outgoing lepton at these lower energies. 4.3. COMPARISON WITH PREDICTED FLUXES As mentioned above, the underground detectors searching for nucleon decay have been gradually accumulating samples of fully and/or partially contained events over the past few years. A summary of the exposure times and event yields is given in Table X. TABLE X Fully contained event results Fully Partially Exposure Threshold Flux contained contained (kton yr) (GeV) KGF 15 20 NUSEX 31 - 0.40 0.250 152 IMB 401 - 3.80 132 KAMIOKANDE-1I 277 - 2.87 0.03, 0.205 FREJUS 88 39 1.00 Although most of these contained events are believed to be due to neutrino inter- actions, their analysis and classification is difficult for two reasons. Firstly, the details of neutrino interactions in the GeV energy range are not well understood because of various experimental difficulties associated with the many possi- ble final states. Only a few bubble chamber experiments have been dedicated to a study of quasielastic and single-pion production processes (Musset and Vialle, 1978). Very limited data is available on other final states such as double-pion and multi-pion production. Secondly, the response of an underground detector to GeV neutrinos cannot be easily tested nor directly calibrated in an accelerator beam exposure. Hence, extensive ASPECTS OF NEUTRINO ASTRONOMY 291 Monte-Carlo simulations are required to evaluate energy dependent detection threshold and triggering efficiencies. One exception is the NUSEX collaboration, who exposed one of their detector modules to a neutrino beam at CERN (Battistoni et al., 1984), collecting a sample of some 400 neutrino events at 90 o and 45 ~ to the calorimeter plates. A comparison of the observed contained events with the predicted atmospheric neutrino flux can be done (e.g., Gaisser and Stanev, 1985) by folding in the neutrino interaction cross-section and the detection efficiency: N(E)= ffd~a(E)f(E)dE, where dN/dE is the neutrino flux, a(E) is the production cross-section, and f(E) is the detection efficiency, and summing over all the possible final states. The resulting event histograms for the higher statistics IMB (Haines et al., 1986) and KAMIOKANDE-II (Hirata et al., 1988) data samples are shown in Figures 18 and 19. 40 > 30 O tD ,- 20 LIJ 10 0 0 I 0.5 1.0 1.5 Ec (GeV) Fig. 18, Atmospheric neutrino contained events (IMB). Visible neutrino energy distribution for the 401 contained events observed by the IMB detector (Haines et al., 1986), The curve represents the 12 year Monte-Carlo simulation, normalized to the 417 day run-time period. Considering the uncertainties of the atmospheric neutrino flux (> 25 ~o) and the total neutrino cross-section (~ 10~o) these data samples agree reasonably well with the predictions. This implies that no unexpected enhancement of the neutrino flux in the GeV energy range has been observed and the possibility of any extraterrestrial corn- 30 e- O) > Ill O 20 d: l E z 0 0 10 I I 1 I I I I I f 1 0.5 1.0 0 Momentum (GeV/c) I I 1 1 I ~ I I l 0.5 1.0 Momentum (GeV/c) ! l ! l l l l l l l ! ve 40 I I ! 1 ! 1 1 1 1 1 1 292 A. M. BAKICH Fig. 19. Atmospheric neutrino contained events (KAMIOKANDE-II). Momentum distributions of the fully contained single ring electron-like and muon-like events obtained by the KAMIOKANDE-II detector (Hirata et al., 1988). The curves represent the distributions expected from atmospheric neutrino inter- actions. ponent can be ruled out at the level of few percent of the atmospheric flux (LoSecco, 1986). Both of these data samples have been used to test the neutrino oscillation hypothesis. The IMB data imposes a limit of Am 2 < 2.2 x 10 -5 eV z at maximum mixing by compar- ing the fluxes of downward (0 < 53 ~ ) and upward (0 > 127 ~ ) events with a mean energy of about 0.92 GeV (LoSecco et al., 1985). The KAMIOKANDE-II apparent deficit of v, single ring events has been interpreted as possible evidence of MSW oscillations (Learned et aL, 1988). However, the significance of these neutrino oscillation tests is limited by poor statistics and by the uncertainties of the predicted fluxes, particularly in the low energy range of 0.1 to 0.7 GeV. We conclude that the prospects of searching for any astrophysical flux component in the GeV energy range do not appear promising. In fact, the presence of atmospheric background is so overwhelming that the advantages of a neutrino detector on the Moon have been considered (Reines, as quoted in the following reference). This suggestion, recently re-examined by Shapiro and Silberberg (1985), indicates that background reduction of about 10 3 can be expected in the 1 to 100 GeV energy range. 5. Search for High-Energy (TeV) Neutrino Point Sources High-energy neutrinos have long been considered as potentially important messengers of astrophysical information (Markov, 1960; Greisen, 1960). The energy range of ASPECTS OF NEUTRINO ASTRONOMY 293 0.1-1000 TeV is believed to be particularly suitable for detection of fluxes of extra- terrestrial neutrinos from discrete point sources. At these energies neutrino detection is achieved by the induced muon method with ever increasing directional accuracy and in the presence of steeply diminishing atmospheric neutrino background. Recent pro- gress (reviewed by Weekes, 1988) in the field of TeV gamma-ray astronomy has provided further incentive for these efforts. A common feature of all the potential astrophysical point sources of high energy neutrinos is the process by which these neutrinos are believed to be produced. This basic process is very high energy pp (and pN) collisions leading to neutrino production via rc (and K) meson decays, p + p~XTr(K)~/~ + v. The two essential ingredients for neutrino emission from any source are: (i) An intense, localized flux of high-energy protons, due to some source dependent acceleration mechanism. Several such acceleration mechanisms have been proposed, such as the Fermi shock acceleration models (Eichler and Vestrand, 1985) and the accretion powered unipolar induction mechanism (Chanmugan and Brecher, 1985), however, details of these acceleration mechanisms fie beyond the scope of this review. For most neutrino astronomy calculations the proton luminosity at the source and the exponent of the power law spectrum are usually the only important parameters. (ii) A presence of low-density gaseous matter in the vicinity of the acceleration region. This effective 'target' should provide sufficient interaction path length (z = S P dl) for mesons to be produced, yet be tenuous enough to allow these mesons to decay before interacting. The optimum conditions for neutrino emission have been evaluated at p < 10-8 g cm-3 and z > 102 g cm-2 (Stenger, 1984). These two general requirements can be employed as a 'standard source scenario' for prediction of various features of neutrino fluxes for a wide variety of specific source models (Berezinski et al., 1985). 5.1. POINT SOURCE CANDIDATES AND MODELS The close relationship between gamma-ray astronomy and neutrino astronomy, mentioned above, is due to the well-known inherent connection between neutrino and gamma-ray emission. Since in any hadronic process neutrinos are products of charged rc meson decays, the accompanying gamma-ray emission should inevitably be present due to ~z ~ decays. In fact if the gamma-ray flux is not accompanied by neutrinos, one must assume an electromagnetic rather than a hadronic production process. Therefore neutrino and gamma-ray fluxes can be related by (e.g., Kolb et al., 1985) Fv(>E ) = 2EnF~(>E), where n is the exponent of the primary spectrum and 2 is the source dependent enhancement factor. Because of the above considerations, most sources of very high-energy gamma-rays should be considered as potential candidates for neutrino emission. Some of the 294 A.M. BAKICH better-established gamma-ray sources are listed in Table XI (Weekes, 1988). This table represents a list of confirmed sources of high-energy X-rays and gamma-rays, except for fast (ms) pulsars which are believed to emit gamma-rays of electromagnetic origin. TABLE XI Sources of very high-energy gamma-rays (Weekes, 1988) Source Period Maximum Flux Distance Luminosity energy (cm -2 s - 1 ) (kpc) (ergs s - 1) Cygnus X-3 4.8 hr 1 PeV 2 x 10 -14 > 11.4 6 x 1036 Hercules X-1 1.24 s 0.5 PeV 3 x 10 -12 5 2 1037 4U0115+63 3.61 s 1 TeV 7 10 - l l 5 6 x 1035 Vela X-1 8.96 d 3 PeV 9 10 -15 1.4 2 1034 LMC X-4 1.41 d 10PeV 5 x 10 -15 50 1 x 1038 Crab Nebula variable 1 PeV 1 10 -13 2 2 x 1035 Cen A steady 1 TeV 4 x 10 -12 4400 3 1040 The question of whether any of these point source candidates could produce a detectable neutrino flux has been addressed by various authors by considering specific source models and configurations. 5.1.1. Binary Systems Many high-energy gamma-ray sources are believed to consist of a compact object (such as a neutron star or an active pulsar), orbiting around a massive companion star. In these binary systems, the neutrinos are expected to be produced via the nadronic process either in the atmosphere behind the companion or in the accretion disc. Hence, both the periodicity and the enhancement of the neutrino signal are determined by the transparancy of the companion star, as can be seen from Table III. Several calculations TABLE XlI Neutrino transparency along the diameter of stellar objects (Berezinsky et aL, 1985) Object Mass Radius Column density E v Ev (M o ) (R o ) (g cm -2) (TeV) (TeV) White dwarf 1 3 10-2 2 x 10 TM 0.07 1.5 Sun 1 1 2 x 10 ll 0.8 1.6 Main sequence 14 7 5 x 101~ 3 6 Giant 5 102 1 108 transparent Supergiant 20 5 102 2 107 transparent of neutrino induced muon fluxes have been performed (e.g., Berezinski etaL, 1985; Gaisser and Stanev, 1985). The results of these calculations indicate that a cosmic-ray luminosity of at least 1043 ergs s- 1 is required to detect a source at a distance of about 10 kpc with an underground detector of 100 m 2 fiduciai area. ASPECTS OF NEUTRINO ASTRONOMY 295 5.1.2. Young Supernova Shells Supernovae are believed to be capable of accelerating protons to very high energies. It has been suggested that for an initial period of a few months to several years after the explosion proton energies of > PeV can be produced, resulting in a delayed emission of TeV neutrinos (Berezinsky and Prilutsky, 1978). These ideas have been recently re-examined by several authors in connection with the SN1987A supernova. In these calculations, the important parameter is the column density of the expanding shell. Once this density decreases sufficiently to allow the produced pions to decay (rather than interact), high-energy neutrinos would be copiously emitted. This neutrino emission can be expected to continue for a period of about 107 s (4 months) until gradually subsiding due to continuing expansion. These models are of particular interest in connection with the SN1987A. In addition to the above gamma-ray sources, an enhancement of high-energy neutrino emission from the centre of our Galaxy has been proposed (Stecker, 1979). These neutrinos would be produced in nuclear collisions of cosmic-ray protons with the interstellar proton gas (galactic pp neutrinos). Since the observed cosmic-ray spectrum is well known, the neutrino flux is largely dependent on the interstellar hydrogen column density. We note that this enhancement can be estimated to be some three orders of magnitude above the diffuse neutrino flux from normal galaxies. 5.2. NEUTRINO INDUCED MUONS An underground detector can register neutrino fluxes indirectly, by observing the muons produced in neutrino interactions outside the fiducial detector volume. The basic reaction involved is the charged current process v~, + N~X + #- . In this detection method, it is the surrounding rock that constitutes the neutrino interaction target, while the underground detector itself merely records the passage of secondary muon through its cross-sectional fiducial area. This method of neutrino detection becomes particularly important with increasing neutrino energy. Since the energy (and, hence, the range) of the muon increases with increasing neutrino energy R --- 2.5 x 10 5 ln(1 + 2E,(TeV)) g cm -2 , so does the rock 'target' mass, which essentially is the product of the detector area and the muon range. Therefore, the effective 'target' mass can be several orders of magnitude larger then the mass of the detector itself. Even more important is the improvement in the directionality of the technique due to the dependence of the kinematics of muon production. A useful approximation of the r.m.s, neutrino-muon angle can be obtained from 0 --- (E(TeV)) 1/2 deg. 296 A.M. BAKICH The reduced angular spread is crucially important not only for unambiguous identifi- cation of potential point sources but also for extraction of the source signal from the atmospheric muon and atmosperic neutrino background. Although the muon multiple scattering in the rock is also a decreasing function of energy, its contribution to the directional uncertainty does not improve significantly as it remains to be determined by the detection energy threshold (Nakatsuka et al., 1987). The method is obviously restricted to detection of muon neutrinos and since the interaction vertex is never directly observed no neutrino energy estimate can be obtained. Hence, only the integral flux above the detector threshold can be determined from neutrino induced muon fata. 5.3. EXPERIMENTAL RESULTS In this section we summarize the attempts to identify neutrino fluxes from the various discrete point sources discussed above. 5.3.1. Searches for Point Sources Since, as illustrated in Figure 20, the downward-going background flux of secondary cosmic-ray muons is prohibitively high, only the near-horizontal or upward-going 1 0 0 0 , J , , , ~ i I L ' I ' ~r 100 ~/~ATMOSPHERIC t.1 ~ LLI " - ), 0 .Q z 10 / / / / / / / / t11 ~L ,\ 1 , ~ , i ~ i, I l i i i% r i i I , 0 30 60 90 Zenith Angle (degrees) Fig. 20. Zenith angle distribution of underground through-going muons. The graph shows the KGF data collected over a run-time period of 3.6 years with an effective area of ~20 m 2 at a depth of 7000 rowe (Krishnaswamy et aL, 1986). The neutrino induced component is predominant in the horizontal direction for zenith angles of > 60 deg. Note that the sharp cut-off at about 85 deg is due to trigger restrictions. ASPECTS OF NEUTRINO ASTRONOMY 297 muons can be accepted for directional point source searches. Even at these large zenith angles the potential source candidates have to be identified against the (roughly is 9 background due to the atmospheric neutrino induced muons. The results of such arrival direction studies performed by the IMB and KGF collaborations are presented in Figures 21 and 22. 90 60 r 9 30 ~" 0 ._o o -30 a -60 I I I I I I | I '1 I [ \ 9 9 9 ~o \ ; . \ . . o ~ Q 9 '~ 9 \ 9 : 9 : . . 9 eo 9 ~ 9 9 9 9 # LMC X-4 El. -90 I I 0 4 f _ J / 9 9 HER X-1 YG X-3 o / . . . . 9 . ' l co* e / 9 9 9 9 9 9 9 9 9 9 / 9 9 9 9 9 4 9 9 G .C . /@' 9 9 ~ . 9 9 \ ' " .4". . " . 9 . . . . -~- - j~_ .~ / . I I I I I i I I 8 12 16 20 24 Right Ascension (hours) Fig. 21. Arrival direction search for point sources of neutrinos (IMB). The scatter plot indicates the directions of a sample of 172 upward-going (0 > 90 ~ ) muons collected by the IMB detector over a run-time period of 344 days (Svoboda et al., 1987). Neither plot exhibits any significant clustering, indicating that most of the observed events are due to the atmospheric neutrinos. This conclusion is in agreement with the KAMIOKANDE-I I results. The estimated 90~o confidence limits on upward-going muon fluxes in the direction of a number of potential source candidates are presented in Table XIII. These muon flux limits can be translated into proton luminosities at the source of approximately L < 104o ergs s - 1 and do not represent any stringent restriction on the predicted fluxes. 5.3.2. Young Supernova SN1987A The SNI987A has provided a rare opportunity to test the hypothesis of TeV neutrino emission during the first few months after the collapse. Several groups have reported monitoring induced muons from the direction of SN1987A. 298 A, M. BAK ICH 90 I f I I I I 9 I i I I ! ~n ~- 9 9 l j +,, j 9 \ " , 9 . " / . . : \ \ . : . . . ; .. cY+x : a) 30 9 " : \ : 9 HE x-+ 9 ~) 9 9 o 9 N =\ / . (1) 9 9 CRAB o~ 9 " / "0 9 9 oe" 0 " \ . 9 . o . - / . . 9 \ 9 Q r - 9 \ 9 9 9 9 G.C. / 0 " ,~ m -30 " 9 : " \ " 9 o . . . . 9 9 e e e e oC ' ]CENA / 9 9 " \ . ~/ 9 9 -60 . LMC X-4 I-1 -go 9 I I I I I I I l I I I 0 4 8 12 16 20 24 Right Ascens ion (hours) Fig. 22. Arrival direction search for point sources of neutrinos (KGF). Arrival directions of through-going neutrino induced muons obtained by the KGF collaboration (Krishnaswamy et al., 1986). TABLE XIII Upward muon flux limits (9070 C.L.) Source IMB KAMIOKANDE-II (cm-Z s -1 ) (cm-2 s -1 ) Cygnus X-3 2.3 10 -13 4.6 x 10 -13 Hercules X-1 1.1 10 -13 6.5 x 10 -13 Vela X-1 3.3 X 10 -14 1.3 10 -13 LMC X-4 3.0 x 10 -14 1.1 x 10 -13 Crab 1.2 x 10 -13 2.8 x 10 -13 Cen A 3.7 x 10-14 _ Geminga 7.8 X 10-t4 -- SS433 6.2 X 10 -14 2.3 X 10 -13 3C273 4.1 X 10 -14 - Galact ic centre 5.1 x 10 -14 1.6 x 10 -13 No enhancement of upward muon flux has been observed by the FRE JUS Collabo- ration (Berger, 1988) during the first 70 days after the supernova explosion. Similarly, KAMIOKANDE- I I (Oyama et al., 1987) reported no upward muon events from SN1987A in the first 6 months after the collapse. The 90~o configence level limit on the muon flux has been calculated to be < 1.2 x 10 - 13 cm -z s - 1 for E~ > 1.7 GeV. One explanation of the negative result is a low accelerated proton luminosity at the ASPECTS OF NEUTRINO ASTRONOMY 299 source. However, other possibilities of either initially delayed or containment-prolonged emission cannot be excluded (Stanev, 1988). 5.3.2. Underground Muons from Cygnus X-3 Two groups have reported detecting time-modulated underground muons from Cygnus X-3 by means of the very technique that is employed for registration of neutrino events (Marshak et al., 1985, Battistoni et al., 1985). Figure 23 represents the fluxes l d 1~ 10 -~ 1612 _ ii~ 13 I I I I l \ \ \ \ aAKSAN \ HOMESTAKEV \ \ ~'~" __ __ - -~ ~ - - . . ~ _ . _NUSE: \ KAMIOKANDE ~- FREJUS~ I I I I 1 0 2000 4000 6000 Depth (mwe) Fig. 23. The underground muon flux from Cygnus X-3. This figure shows the in-phase correlated excess fluxes of underground muons observed by SOUDAN-I and NUSEX detectors in the direction of Cygnus X-3. Also shown are the upper flux limits reported by several other collaborations. observed by NUSEX and SOUDAN-I together with the upper limits reported by several other collaborations. Although there appears to be an obvious disagreement between these results, direct comparison is difficult because of the different observation periods and the well-known variability of Cygnus X-3. 300 A.M. BAKICH The controversy of the result lies in the nature of the primary particle. For any known particle to qualify as the progenitor of the in-phase underground muons its energy would have to be extremely high (> 1 EeV). The resulting output from Cygnus X-3 would then exceed the total cosmic-ray flux measured by the extensive air shower arrays. If, however, a reasonable assumption about the primary energy is made (~ 10 TeV) all known particles are eliminated by very simple arguments. (i) Charged particles would be scattered by the galactic magnetic fields. (ii) The neutron lifetime of ~ 900 s is too short for the ~ 12 kpc flight path. (iii) Photons are very inefficient muon producers. (iv) Neutrinos should have an isotropic zenith angle distribution. Therefore either a new (exotic) particle and/or new type of interaction is required to explain the underground muon data. The properties of such a particle can be inferred directly from the observed signal (Halzen, 1987). It should be neutral, with mass < few GeV, lifetime > 108 s, and have a cross-section of 10 pb < a< 1 rob. Although several such suggestions have been proposed, the unconfirmed status of the signal does not warrant their discussion here. 5.4. NEW EXPERIMENTS AND PROPOSALS The relatively poor statistics of the experimental data collected to date by the existing underground detectors is due the fact that their design has not been specifically optimized for the induced muon technique. Some of the new detectors listed in Table II should be more suitable for detection of astrophysical neutrino point sources. One typical example is the MACRO (Monopole, Astrophysics, and Cosmic Ray Observatory) installation being at Gran Sasso (Calicchio et aL, 1988). It consists of three 72 12 m 2 liquid scintillator planes interspersed with 18 layers of streamer tubes and supplemented with a layer of track etch detectors. For observing neutrino point sources this detector is offering a total surface area of 3240 m 2 (isotropic acceptance 104 m 2 sr) and an angular resolution of about 0.2 deg with up/down time-of-flight rejection capability. In addition to the above, planning is underway for the next generation detectors with a specific primary goal of observing the extra-terrestrial point sources of TeV neutrinos. The two important parameters involved are: (i) a largest possible sensitive detector area (rather then volume or mass), (ii) a high angular resgJution for through-going muon tracks. The basic idea is to employ a very large area (but shallow depth) water Cherenkov detector for detection of upward-going induced muons. One proposed design (Koshiba et aL, 1986) involves a lake 150 m diameter and 30 m deep, surrounded by an array of 50 cm photomultipliers on a 5 m grid and including an outer layer of 5 5 5 m 3 water Cherenkov modules (Figure 24). Another proposal (Gajewski et al., 1987) employs of three 250 250 m 2 planes each eqttipped with 2500 photomultiplier tubes on a 5 m grid. The planes would be separated by non-transparent plastic sheeting and positioned 10 m apart some 30 m below the surface of a shallow clear water lake. This type of configuration would easily reject ASPECTS OF NEUTRINO ASTRONOMY 301 9 , , , . . . . . . . . Fig. 24. The proposed LENA detector. This configuration of Lake Experiment on Neutrino Astronomy (LENA) detector is specifically designed for detection of upward-going neutrino induced muons Koshiba (1986). down-going and horizontal muons while providing an aperture of about 1.3 r~ x 62 500 m 2 sr for upward-going tracks. Both of these large area configurations should be able to monitor the upward-going (i.e., neutrino induced) muon fluxes of some 4000 events per year, with an angular resolution of about 1 deg. In conclusion, it is interesting to note that these 'shallow lake' water Cherenkov detectors would be significantly cheaper and easier to deploy then the giant deep underwater detectors described in the next chapter. 6. Ultra-High Energy (PeV) Neutrino Fluxes It is well known that the cosmic-ray energy spectrum does extend to at least 10 z~ eV (Hillas, 1984). Interactions of these particles with interstellar and intergalactic photons should lead to production of ultra-high energy PeV neutrinos. The atmospheric neutrino background, so troublesome at lower energies, is decreasing very steeply in this energy range, and becomes almost negligible at Ev > 1016 eV. It, hence, becomes possible that at these ultra-high energies one may be able to detect the diffuse extraterrestrial neutrino component (Berezinski and Zatsepin, 1977). Detection of these neutrinos is of great interest for two reasons: Firstly, establishing the existence of ultra-high energy neutrino fluxes could provide information of unique astrophysical significance, such as the cosmic-ray acceleration processes in extragalactic sources and the enhancement of their activity during the very early stages of the Universe. Secondly, it offers an opportunity of studying neutrino physics at energies significantly higher than available at accelerators and well beyond the intermediate vector boson threshold. Obviously this possibility presents itself only if substantial fluxes of PeV neutrinos are detectable. 302 A. M. BAKICH 6.1. ORIGINS OF DIFFUSE EXTRATERRESTRIAL NEUTRINOS The two sources of astrophysical neutrino production are the proton-proton (pp) and proton-gamma (p 7) collisions. The pp process is a very inefficient source of ultra-high energy neutrinos because of the low interstellar and intergalactic gas densities. However, the P7 collisions can produce neutrinos via meson photoproduction reactions such as p + ~N*- - rN I r~NI~v~N+ e + vz + v~ + v e. Because in the center-of-mass system these processes correspond to the extensively studied low-energy (,-, GeV) 7P reactions, their cross-sections are well known (,~ 300 #b at resonance). The threshold of these reactions is given by 89 with the average neutrino energy being about 59/0 of Ep. 6.1.1. p 7 Neutrino Flux It has been suggested by Greisen (1966) and by Kuzmin and Zatsepin (1966) that the interactions of ultra-high energy cosmic rays with the 2.7 K universal microwave background should lead to a cutoffofthe extragalactic cosmic-ray spectrum at energies of > 3 1019 eV. It is these p 7 colfisions that should also result in the p 7 neutrino flux. The remnant photon 'target' is described by the photon density of 390 era-3 and mean photon energy of 6.3 x 10 -4 eV. The main uncertainty in calculating the extragalactic P7 neutrino flux lies in the exact shape of the cosmic-ray spectrum at ~ 1019 eV. In recent calculations of Berezinski et al. (1986) a two-component spectrum was assumed with a change in exponent from - 3.08 to - 2.5 at E ~ 1019 eV. The resulting p 7 fluxes are shown in Figure 25. 6.1.2. Bright Phase Neutrinos The above flux estimates are based on the presently observed cosmic-ray energy spectrum and the present energy density of the remnant photons. In some cosmological models of galactic evolution the e .arly epoch of the Universe (the 'bright phase', Peebles, 1967) is believed to be characterized by intensified cosmic-ray production and hotter remnant photon radiation. This 'bright phase' of galactic evolution, continuing for some 107 years, has been hypothesized as a possible origin of an enhanced PeV neutrino flux. Since this bright phase period corresponds to large red shifts (z) both the remnant density and energy would be scaled up as (1 + z) 3 and (1 + z), respectively. Although the calculations of bright phase neutrino flux are critically dependent on the assumed shape of the cosmic ray energy spectrum, a considerable enhancement can be expected as shown in Figure 25 (Hill and Schramm, 1983, 1985). 6.1.3. Active Galactic Nuclei The question of the above bright-phase models can be related to the active galactic nuclei, or extragalactic objects with emission substantially in excess of normal stellar processes. These include strong extragalactic sources such as Seyfert galaxies, various types of quasars and some radiogalaxies. The vast energy output (luminosities of ASPECTS OF NEUTRINO ASTRONOMY 303 10-11 -12 10 -13 10 1614 l d 1~ 10-16 10 -17 ld 18 l l j 19 lO 4 [ I I I I I [ % ,% z=6 " , \ z=4 '-~-\C"- \ , \ \ \ Z=2 ~\ \ \ k \ \ \ z=o,, \ \ \ \ p' o Pz " FLY'S EYE \ 9 \ \ \ \ \ \ \ \ 1015 10 6 1017 ld 8 1019 10 0 101 Neutrino Energy (eV) Fig. 25. The integral energy spectra of ultra-high energy neutrinos. The solid lines show the steeply decreasing atmospheric neutrino spectra from calculations by Mitsui et al. (1986) and Bugaev et aI, (1987). The dashed lines show the extragalactic diffuse p ? neutrino fluxes predicted by Berezinsky et al. (1986). The 'bright phase' neutrino fluxes (at red shift z) represent calculations by Hill and Schramm (1983, 1985). Also shown are the upper flux limits obtained by the Fly's Eye detector (Baltrusaitis et al., 1988). > 1045 ergs s -1) from these objects is one of the unanswered questions in astrophysics. Some of this energy could be emitted in the form of neutrinos if the column density in the inner regions is sufficiently high. However, the expected fluxes of > PeV neutrinos cannot be reliably calculated due to the uncertainties in the assumptions of the cosmo- logical models and the lack of experimental data. 304 A.M. BAKICH 6.2. GIANT UNDERWATER EXPERIMENTS The task of detecting ultra-high energy PeV neutrinos is frustrated by the extremely low expected fluxes, requiring experiments on an entirely different scale to that previously attempted. One possibility of setting up a giant underwater detector array has been extensively studied over the last ten years. The water simultaneously constitutes a target for neutrino and muon interactions, a detection medium for Cherenkov light, and a shielding against cosmic rays and sunlight. High energy neutrino interactions (inside or outside the array) produce a through- going muon track, registered by 'hits' in many modules from which both energy and direction can be estimated. It is important to note that this type of detector should also provide a high statistics data on all of the point sources discussed in the previous chapter. 6.2.1. DUMAND The DUMAND (Deep Underwater Muon and Neutrino Detector) proposal aims for a 250 x 250 x 500 m 3 array of 756 detector modules to be located at a depth of 4500 m in the Pacific Ocean near Hawaii (Figure 26). This represents the enclosed target mass is 30 Mtons and an effective detector area of about 105 m 2. The most important DUMAND 3Okm \ 4.5kin Detector Location Detector Array ~4Ocm ~ Optical' Module Fig. 26. Proposed configuration of the DUMAND detector. This expanded schematic diagram shows the underwater location of the detector, the full array of 36 strings of optical sensors, and a single PM tube module. ASPECTS OF NEUTRINO ASTRONOMY 305 parameter of the detector is its angular resolution estimated at 15 to 45 mrads, depend- ing on the muon energy (Peterson, 1984). The proposal, recently reviewed by Grieder (1986) has been subdivided into four stages of a short prototype string, a full prototype string, a complete plane (6 strings) and the full array. Stage I of the project, a 60 m string of 7 modules submerged at a depth of 4 kin, is a thorough engineering test of the optical sensors and data handling. This work has recently been completed, recording some 2.4 x 104 atmospheric muons. 6.2.2. Lake BAIKAL A similar proposal (total aperture 4 x 105 m 2 sr, with ,-, 103 optical sensors) has been put forward by Domogatsky and his collaborators for lake Baikal. Although limited to a shallower depth of about 1300 m, this location has the advantages of high water transparency, lack of underwater currents, reduced level of bio-luminescence and a low level of 4~ radioactive background. Practical implementation is facilitated by an ice cover during the winter months of operation. A series of preliminary experiments with a single string of 9 modules (36 photo- multipliers total) has also been completed. Although both DUMAND and BAIKAL proposals face strong competition from the latest large area proposals, both collaborations are proceeding with the second stage involving the deployment of several strings. 6.3. PeV NEUTRINO FLUX L IM ITS The Fly's Eye installation has been designed to study ultra-high energy (E > 1017 eV) cosmic-ray air showers (Cassiday, 1985). This optical detector observes the nitrogen fluorescence excitation produced in the atmosphere by the developing cascades. Located in Utah, the detector consists of two stations (67 mirrors with 880 photomultipliers and 36 mirrors with 464 photomultipliers) separated by 3.4 km. One of the important advantages of this detector is in the energy dependence of its effective sensitive area, which can extend up to 20 km 2 for sufficiently energetic showers. However, the optical observation periods are severly limited, with an average duty factor being only about 5Yo. For each visible shower the pulse integral and arrival times are recorded and the shower track geometry reconstructed, followed by estimation of longitudinal develop- ment profile and total shower energy. The accuracy of event reconstruction depends on the number of photomultipliers that view the shower and, hence, on its energy and in particular on the perpendicular distance from the station(s) to the shower axis (the impact parameter). Typical reconstruction errors are 20~o for energy and 3 deg for zenith angle. Since air showers can be attributed to high energy neutrinos, a search for these events has been performed, as part of normal Fly's Eye operation. The obtained preliminary upper limits on diffuse neutrino flux are shown in Figure 25. It is interesting to note that these results are within a factor of 30 from some of the extraterrestrial flux predictions, and indicate that the red shift of maximum bright phase activity must be less than 10. 306 A. M. BAKICH 7. Conclusion We conclude this review of the current status of neutrino astronomy with a summary of what progress can be expected within the next few years. It is clear that the existing KAMIOKANDE-II detector should produce a significant result on the 8B solar neutrino flux, possibly within a couple years. At that time, some preliminary data from several new experiments should also become available. In particu- lar, the pp neutrino flux measurements will provide a fascinating opportunity of explor- ing the possibility of neutrino oscillations. For the next supernova burst, these new neutrino detectors will have larger targets, lower energy thresholds and much better timing accuracy. In the GeV energy range the prospects for neutrino astronomy do not appear to be very promising. It is unlikely that even the next generation of detectors will have the event statistics to overcome the poor directionality of the signal in the presence of high atmospheric neutrino background. At higher energies, some of the large new detectors could provide a first neutrino point source identification within the next few years. However, a systematic study of these sources will probably have to await the next generation of detectors with sensitive areas of at least 10 4 m z, whether shallow lake or deep underwater. Finally, in the ultra-high energy region some further improvement of the Fly's Eye upper flux limits can be expected. Acknowledgements The author is grateful to L. S. Peak for useful comments and encouragement and acknowledges the support of the School of Physics, University of Sydney, where this review was written. References Abela, R., Daum, M., Eaton, G. H., Frosch, R., Jost, B., Kettle, P. R., and Steiner, E.: 1984, Phys, Letters 146B, 431. Aglietta, M., Badino, G., Bologna, G., Castagnoli, C., Castellina, A., Dadykin, V. L., Fulgione, W., Galeotti, P., Kalchukov, F.F., Kortchaguin, B., Kortchaguin, P.V., Malguin, P.V., Malguin, A. S., Ryassny, V. G., Ryazhskaya, O. G., Saavedra, O., Talochkin, V. P., Trinchero, G., Vernetto, S., Zatsepin, G. T., and Yakushev, V. F.: 1987a, Europhys. Letters 3, 1315. Aglietta, M., Badino, G., Bologna, G., Castagnoli, C., Castellina, A., Dadykin, V. L., Fulgione, W., Galeotti, P., Kalchukov, F.F., Kortchagnin, B., Kortchaguin, P.V., Malguin, A. S., Ryassny, V.G., Ryazhskaya, O. G., Saavedra, O., Talochkin, V. P., Trinchero, G., Vernetto, S., Zatsepin, G.T., and Yakushev, V. F.: 1987b, Europhys. Letters 3, 1321. Albrecht, H., Binder, U., Harder, G., Philipp, A., Schmidt-Parzefall, W., Schroder, H., Scbultz, H. D., Wurth, R., Drescher, A., Grawe, B., Matthiesen, U., Scheck, H., Spengler, J., Wegener, D., Schubert, K. R., Stiewe, J., Waldi, R., Weseler, S., Brown, N. N., Edwards, K. W., Frisken, W. R., Fukunaga, Ch., Gilkinson, D.J., Gingrich, D.M., Goddard, M., Kim, P. C. H., Kutschke, R., Macfarlane, D.B., McKenna, J. A., McLean, K. W., Nilsson, A.W., Orr, R. S,, Padley, P., Patel, P. M., Prentice, J.D., Seywerd, H. C. J., Stacey, B. J., Yoon, T. S., Yun, J. C., Ammar, R., Coppage, D., Davis, R., Kanekal, S., Kwak, N., Kernel, G,, Plesko, M., Jonsson, L., Oku, Y., Babaev, A., Danilov, M., Golutvin, A., ASPECTS OF NEUTRINO ASTRONOMY 307 Lubimov, V., Matveev, V., Nagovitsin, V., Ryltsov, V., Semenov, A., Shevchenko, V., Soloshenko, V., Sopov, V., Tichomirov, I., Zaitsev, Yu., Childers, R., Darden, C. W., and Gennow, H.: 1985, Phys. Letters 163B, 404. Alekseev, E. N., Alekseeva, L. N., Volchenko, V. I., and Krivosheina, I. V.: 1987, J. Eksperim. Theor. Phys. Letters 45, 589. Alexeyev, E. N., Alexeyeva, L. N., Krivosheina, I. V., and Volchenko, V. I.: 1988, Phys. Letters 205B, 209. Allkofer, O. C., Carstensen, K., and Dan, W. D.: 1971, Proc. 12th Cosmic Ray Conf. Hobart 4, 1314. BahcaU, J. N.: 1979, Space Sci. Rev. 24, 227. Bahcall, J. N.: 1985, Solar Phys. 100, 53. Bahcall, J. N. and Davis, R., Jr.: 1982, Essays Nucl. Astrophys. 12, 243. Bahcall, J. N. and Ulrich, R. K.: 1988, Rev. Mod. Phys. 60, 297. Bahcall, J. N., Baldo-Ceolin, M., Cline, D. B., and Rubbia, C.: 1986, Phys. Letters B178, 324. Bahcall, J. N., Huebner, W. F., Lubow, W. H., Parker, P. D., and Ulrich, R. K.: 1982, Rev. Mod. Phys. 54, 767. Bakich, A. M. and Peak, L. S.: 1985, AlP Conf. Proc. 126, 238. Baltrusaitis, R. M., Cassiday, G. L., Cooper, R., Dawson, B. R., Elbert, J. W., Fick, B. E., Liebing, D. F., Loh, E. C., Sokolsky, P., and Steck, D.: 1988, Nucl. Instr. Meth. A264, 87. Barabanov, I. R., Veretenkin, E. P., Gavrin, V. N., Danshin, S. N., Eroshkina, L. A., Zatsepin, G. T., Zakharov, Yu. I., Klimova, S. A., Klimov, Yu. B., Knodel, T. V., Kopylov, A. V., Orekhov, I. V., Tikhonov, A. A., and Churmaeva, M. I.: 1985, AlP Conf. Proc. 126, 175. Bari, C., Basile, M., Bruni, G., Cara Romeo, G., Castelvetri, A., Cifarelfi, L., Contin, A., Del Para, C., Giusti, P., Iacobucci, G., Maccarrone, G., Massam, T., Nania, R., O'Shea, V., Palmonari, F., Perotto, E., Prisco, G., Sartorelli, G., Willutzky, M., Chincellato, J.A., Dobrigkeit Chincellato, C., Fauth, A.C., Turtelli, A., De, K., Shapiro, A. M., Widgoff, M., Rohrbach, F., Zichichi, A., Caputi, P., Susino, G. L., Barbagli, G., Confronto, G., Landi, G., Pelfer, P., Bianco, S., Anzivino, G., Casaccia, R., Cindolo, F., Endorini, M., Fabbri, F., Laakso, I., Qian, S., Rindi, A., Spallone, A., Votano, L., Zallo, A., Lau, K., Lipps, F., Mayes, B., Mo, G. H., Pinsky, L., Pyrlik, J., Sanders, D., Sheldon, W.R., Weinstein, R., Dai, Y., Din, L., Jing, C., Jing, G., Lu, Z., Shen, P., Zhu, Q., Alyea, D., Di Sciascio, G., Scrimaglio, R., RoteUi, P., Kocharov, G. E., Deutsch, M., Hafen, E. S., Haridas, P., Jeckelmann, B., Ji, G., Kuang, H. H., Pitas, A., Pless, I.A., Yuan, Y., Zhao, C.Z., Berezinski, V. S., Dadykin, V.L., Khalchukov, F.F., Korolkova, E. V., Kortchaguin, P. V., Kortchaguin, V. B., Kudryavtsev, V. A., Malguin, A. S., Markov, M.A., Ryassny, V.G., Ryazhskaya, O.G., Talochkin, V.P., Yakushev, V.F., Zatsepin, G.T., Moromisato, J., Saletan, E., Shambroom, D., Von Goeler, E., D'Ali, G., De Pasquale, S., Alpat, B., Artemi, F., Cappelletti, C., Diodati, P., Salvadori, P., Aglietta, C., Badino, G., Bergamasco, L., Castagnoli, C., Castellina, A., Cini, G., Dardo, M., Fulgione, W., Galeotti, P., Morello, C., Navarra, G., Periale, L., Picchi, P., Saavedra, O., Trinchero, G.C., Vallania, P., Vernetto, S., Grianti, F., and Vetrano, F.: 1988, Nucl. Inst. Meth. A264, 5. Battistoni, G., Bellotti, E., Bologna, G., Campana, P., Castagnoli, C., Chiarella, V., Cundy, D. C., D'Ettore Piazzoli, B., Fiorini, E., Iarocci, E., Mannocchi, G., Murtas, G. P., Negri, P., Nicoletti, G., Periale, L., Picchi, P., Price, M., Pullia, A., Ragazzi, S., Rollier, M., Saavedra, O., Trasatti, L., and Zanotti, L.: 1984, Nucl. Inst. Meth. 219, 300. Battistoni, G., Bellotti, E., Bloise, C., Bologna, G., Cami~ana, P., Castagnoli, C., Castellina, A., Chiarella, V., Ciocio, A., Cundy, D., D'Ettore-Piazzoli, B., Fiorini, E., Galeotti, P., larocci, E., Liguori, C., Mannocchi, G., Murtas, G., Negri, P., Nicoletti, G., Picchi, P., Price, M., Pullia, A., Ragazzi, S., Rollier, M., Saavedra, O., Satta, L., Serri, P., Vernetto, S., and Zanotti, L.: 1985, Phys. Letters 155B, 465. Berezinsky, V. S. and Prilutsky, O. F.: 1978, Astron. Astrophys. 66, 325. Berezinskii, V. S. and Zatsepin, G. T.: 1977, Soviet. Phys. Usp. 20, 361. Berezinsky, V. S., Castagnoli, C., and Galeotti, P.: 1985, Nuovo Cimento C8, 185. Berezinskii, V. S., Gazizov, A. Z., Zatsepin, G. T., and Rozental, I. L.: 1986, Soviet. J. Nucl. Phys. 43, 406. Berger, C. (FREJUS Collaboration): 1986, Phys. Letters 174B, 118. Berger, Ch., Hofmann, A., Raupach, F., Schleper, P., Schmitz, G., Tutas, J., Voigtlander, B., Arpesella, C., Benadjal, Y., Deuzet, G., Dudelzak, B., Eschtruth, P., Jullian, S., Lalanne, D., Laplanche, F., Longuemare, C., Paulot, C., Roy, Ph., Szklarz, G., Behr, L., Bland, R.W., Degrange, B., Nguyen- Khac, U., Serri, P., Tisserant, S., Tripp, R., Bareyre, P., Barloutaud, R., Chardin, G., Di Ciaccio, L., Edmunds, D.L., Ernwein, J., Gerbier, G., Jabiol, M.A., Kohon, W., Mosca, L., Moscoso, L., Pietrzyk, B., Becker, K. H., Daum, H.J., Dernski, S., Hinners, R., Kohrs, W., Kuznik, B., Mayer, R., 308 A.M. BAKICH Meyer, H., Ortmann, D., Peters, J., Schubnell, M., Thierjung, J., Wei, Y., and Wintgen, P.: 1988, Nucl. Inst. Meth. A264, 24. Bethe, H. A.: 1939, Phys. Rev. 55, 434. Bilenky, S. M. and Petcov, S. T.: 1987, Rev. Mod. Phys. 59, 671. Bionta, R. M., Blewitt, G., Bratton, C. B., Casper, D., Cioeio, A., Claus, R., Cortez, B., Crouch, M., Dye, S. T., Errede, S., Foster, G. W., Gajewski, W., Ganezer, K. S., Goldhaber, M., Haines, T. J., Jones, T. W., Kielczewska, D., Kropp, W. R., Learned, J. G., LoSecco, J. M., Matthews, J., Miller, R., Mudan, M. S., Park, H. S., Price, L. R., Reines, F., Schultz, J., Seidel S., Shumard, E., Sinclair, D., Sobel, H. W., Stone, J. L, Sulak, L. R., Svoboda, R., Thornton, G., van der Velde, J. C., and Wuest, C.: 1987, Phys. Rev. Letters 58, 1494. Boris, S., Golutvin, A., Laptin, L, Lubimov, V., Nagovizin, V., Nozik, V., Novikov, E., Soloshenko, V., Tihomirov, I., Tretjakov, E., and Myasoedov, N.: 1987, Phys. Rev. Letters 58, 2019. Bruenn, S. W.: 1986, Astrophys. J. Suppl. 62, 331. Bugaev, E. V. and Naumov, V. A.: 1987, Proc. 20th Cosmic Ray Conf., 6, 196. Bugaev, E. V., Zaslavskaya, E. S., Naumov, V. A., and Sinegovsky, S. I.: 1987, Proc. 20th Cosmic Ray Conf., 6, 305. Burrows, A.: 1988, Astrophys. J. 328, L51. Burrows, A. and Lattimer, J. M.: 1986, Astrophys. 9". 307, 178. Calicchio, M., Case, G., Demarzo, C., Erriquez, O., Favuzzi, C., Giglietto, N., Nappi, E., Posa, F., Spinelli, P., Baldetti, F., Cecchini, S., Giacomelli, G., Grianti, F., Mandrioli, G., Margiotta, A., Patrizii, L., Sanzani, G., Serra, P., Spurio, Ahlen, S., Ciocio, A., Feleini, M., Fieenec, D., Incandela, J., Marin, A., Stone, J., Sulak, L., Worstell, W., Barish, B., Lane, C., Liu, G., Peck, C., Poulard, G., Sletten, H., Cohen, S., Ide, N., Manka, A., Steinberg, R., Battistoni, G., Bilokon, H., Bloise, C., Campana, P., Chiarella, V., Grillo, A., Iarocci, E., Marini, A., Reynoldson, J., Rindi, A., Ronga, F., Satta, L., Spinetti, M., Valente, V., Heinz, R., Mufson, S., Petrakis, J., Monaeelli, P., Reale, A., Longo, M., Musser, J., Smith, C., Tarle, G., Ambrosin, M., Barbarino, B.C., Grancagnolo, F., Onnembo, A., Palladino, V., Angelini, C., Baldini, A., Bemporad, C., Flaminio, V., Giannini, G., Pazzi, R., Auriemma, G., De Vincenzi, M., Lamanna, E., Martellotti, G., Palamara, O., Petrera, S., Petrillo, L., Pistilli, P., Rosa, G., Sciubba, A., Severi, M., Green, P., Webb, R., Bisi, V, Giubellino, P, Marzari Chiesa, A., Ramello, L., Soli, D., and Trower, P.: 1988, Nucl. Inst. Meth. A2(ut, 18. Cassiday, G. L.: 1985, Ann. Rev. Nucl. Part. Sci. 35, 321. Chanmugan, G. and Brecher, K.: 1985, Nature 313, 767. Cherry, M. L. and Lande, K.: 1988, Phys. Rev. D36, 3571. Chin, H. Y.: 1966, Ann. Rev. Nucl. Sci. 16, 591. Colgate, S. A. and White, R. M.: 1966, Astrophys. J. 143, 626. Cortez, B. (KAMIOKADE-II Collaboration): 1986, AlP Conf Proc. 150, 1087. Cribier, M., Hampel, W., Rich, J., and Vignaud, D.: 1986, Phys. Letters B182, 89. Dadykin, V. L., Zatsepin, G. T., Korchagin, V. B., Korchagin, P. V., Malgin, A. S., Ryazhskaya, O. G., Ryasnyi, V.G., Talochkin, V.P., Khalchukov, F.F., Yakushev, V.F., Aglietta, M., Badino, G., Bologna, G., Castagnoli, C., Castellina, A., Fulgione, W., Galeotti, P., Saavedra, O., Trincero, J., and Vernetto, S.: 1987, J. Eksperim. Theor. Fis, Letters 45, 593. Dar, A. and Dado, S.: 1987, Phys. Rev. Letters 59, 2368. DeRujula, A.: 1987, Phys. Letters 193B, 514. Eichler, D. and Vestrand, W. T.: 1985, Nature 318, 345. Eisele, F.: 1986, Rep. Prog. Phys. 49, 233. Ellis, J. and Olive, K. A.: 1987, Phys. Letters 193B, 525. Ewan, G. T., Evans, H. C., Lee, H. W., Leslie, J. R., Mak, H. B., MeLatchie, W., Robertson, B.C., Skensved, P., Alien, R. C., Buhler, G., Chen, H. H., Doe, P. J., Sinclair, D., Tanner, N. W., Anglin, J. D., Bercovitch, M., Davidson, W. F., Hargrove, C. K., Mes, H., Storey, R. S., Earle, E. D., Milton, G. M., Jagam, P., Simpson, J. J., McDonald, A. B., Hallman, E. D., Carter, A. L., and Kessler, D.: 1987, Sudbury Neutrino Observatory Proposal (unpublished). Fanlkner, J. and Gilliland, R. L.: 1985, Astrophys. J. 299, 994. Filippone, B. V.: 1986, Ann. Rev. Nucl. Part. Sci. 36, 717. Fritschi, M., Holzschuh, E., Kundig, W., Petersen, J. W., Pixley, R. E., and Stussi, H.: 1986, Phys. Letters B173, 485. Gaisser, T. K. and Grillo, A. F.: 1987, Phys. Rev. D36, 2752. ASPECTS OF NEUTRINO ASTRONOMY 309 Gaisser, T. K. and Stanev, T.: 1984, Proc. Neutrino-84 Conf. Dortmund 370. Gaisser, T. K. and Stanev, T.: 1985, Proc. 19th Cosmic Ray Conf., 8, 156. Gajewski, W., Haines, T., Kropp, W., Learned, J., Potter, M., Price, L., Reines, F., Schultz, J., Sobel, H., and Svoboda, R.: 1987, UCI87-4, preprint. Goldman, I., Aharonov, Y., Alexander, G., and Nussinov, S.: 1988, Phys. Rev. Letters 60, 1789. Greisen, K.: 1960, Proc. Int. Conf. Inst. in HEP (lnterscience) 209. Greisen, K.: 1966, Phys. Rev. Letters 16, 748. Grieder, P. K. F.: 1986, Nuovo Cimento 9C, 222. Haines, T. J., Bionta, R. M., Blewitt, G., Bratton, C. B., Casper, D., Claus, R., Cortez, B. G., Errede, S., Foster, G.W., Gajewski, W., Ganezer, K. S., Goldhaber, M., Jones, T.W., Kielczewska, D., Kropp, W. R., Learned, J. G., Lehmann, E., LoSecco, J. M., Matthews, J., Park, H. S., Price, L. R., Reines, F., Schultz, J., Seidel, S., Shumard, E., Sinclair, D., Sobel, H. W., Stone, J. L., Sulak, L., Svoboda, R., van der Velde, J. C., and Wuest, C.: 1986, Phys. Rev. Letters 57, 1986. Haines, T., Bratton, C. B., Casper, D., Ciocio, A., Claus, R., Crouch, M., Dye, S. T., Errede, S., Gajewski, W., Goldhaber, M., Haines, T. J., Jones, T. W., Kielczewska, D., Kropp, W. R., Learned, J. G., LoSecco, J. M., Matthews, J., Miller, R., Mudan, M. S., Price, L. R., Reines, F., Schultz, J., Seidel, S., Shumard, E., Sinclair, D., Sobel, H. W., Sulak, L. R., Svoboda, R., Thornton, G., and van der Velde, J. C.: 1988, Nucl. Inst. Meth. A264, 28. Halzen, F.: 1987, 2nd Aspen Part. Phys. Conf. 237. Haxton, W. C.: 1984, Proc. Neutrino-84 Conf., Dortmund 214. Haxton, W. C. and Johnson, C. W.: 1988, Nature 333, 325. Haxton, W. C. and Stephenson, G. J.: 1984, Prog. Part. Nucl. Phys. 12, 409. Hayakawa, S.: 1969, Cosmic Ray Physics, John Willey, New York. Hill, C. T. and Sehramm, D. N.: 1983, Phys. Letters 131B, 247. Hill, C. T. and Schramm, D. N.: 1985, Phys. Rev. D31, 564. Hill, C. T., Schramm, D. N., and Walker, T. P.: 1986, Phys. Rev. D34, 1622. Hillas, A. M.: 1984, Ann. Rev. Astron. Astrophys. 22, 425. Hirata, K., Kajita, T., Koshiba, M., Nakahata, M., Oyama, Y., Sato, N., Suzuki, A., Takita, M., Totsuka, Y., Kifune, T., Suda, T., Takahashi, K., Tanimori, T., Miyano, K., Yamada, M., Beier, E. W., Feldscher, L. R., Kim, S. B., Mann, A. K., Newcomer, F. M., Van Berg, R., Zhang, W., and Cortez, B. G.: 1987a, UT-ICEPP-87-04. Hirata, K., Kajita, T., Koshiba, M., Nakahata, M., Oyama, Y., Sato, N., Suzuki, A., Takita, M., Totsuka, Y., Kifune, T., Suda, T., Takahashi, K., Tanimori, T., Miyano, K., Yamada, M., Beier, E. W., Feldscher, L. R., Kim, S. B., Mann, A. K., Newcomer, F. M., Van Berg, R,, Zhang, W., and Cortez, B. G.: 1987b, Phys. Rev. Letters 58, 1490. Hirata, K. S., Kajita, T., Koshiba, M., Nakahata, M., Ohara, S., Oyama, Y., Sato, N., Suzuki, A., Takita, M., Totsuka, Y., Kifune, T., Suda, T., Nakamura, K., Takahashi, K., Tanimori, T., Miyano, K., Yamada, M., Beier, E. W., Feldscher, L. R., Frank, E. D., Frati, W., Kim, S. B., Mann, A. K., Newcomer, F. M,, Van Berg, R., Zhang, W., and Cortez, B. G.: 1988, Phys. Letters B205, 416. Kirsten, T.: 1986, Proc. Neutrino-86 Conf., Sendal 317. Kolb, E. W., Turner, M. S., and Walker, T. D.: 1985, Phys. Rev. D32, 1145. Koshiba, M.: 1987, Phys. Today 38. Koshiba, M., Orito, S., and Kawagoe, K.: 1986, Non-Accel. Phys. Workshop, KEK. Kranss, L. M.: 1987, Nature 329, 689. Krishnaswamy, M. R., Menon, M. G. K., Mondal, N. K., Narasimham, V. S., Sreekantan, B. V., Hayashi, Y., Ito, N., Kawakami, S., and Miyake, S.: 1986, Proc. Asia-Pacific Physics Conf., Bangalore 1, 424. Kuzmin, V. A. and Zatsepin, G. T.: 1966, J. Eksperim. Theor. Phys. Letters 4, 78. Lande, K.: 1979, Ann. Rev. Nuel. Part. Sci. 29, 395. Learned, J. G., Pakvasa, S., and Weiler, T. J.: 1988, Phys. Letters B207, 79. Lee, H. and Bludman, S. A.: 1988, Phys. Rev. D37, 122. LoSecco, J. M.: 1986, New Frontiers in Particle Physics, World Scientific, Singapore, p. 376. LoSeceo, J. M., Bionta, R. M., Blewitt, G., Bratton, C. B., Casper, D., Chrysicopoulou, P., Claus, R., Cortez, B. G., Errede, S., Foster, G.W., Gajewski, W., Ganezer, K. S., Goldhaber, M., Haines, T. J., Jones, T. W., Kielczewska, D., Kropp, W. R., Learned, J. G., Lehmann, E., Park, H. S., Reines, F., Schultz, J., Seidel, S., Shumard, E., Sinclair, D., Sobel, H. W., Stone, J. L., Sulak, L., Svoboda, R., Van der Velde, J. C., and Wuest, C.: 1985, Phys. Rev. Letters 54, 2299. 310 A.M. BAKICH Markov, M. A.: 1960, Proc. lnt, Conf on HEP, Rochester 578. Marshak, M. L., Bartelt, J., Courant, H., Heller, K., Joyce, T., Peterson, E. A., Ruddick, K., Shupe, M., Ayres, D. S., Dawson, J., Fields, T., May, E. N., Price, L. E., and Sivaprasad, K.: 1985, Phys. Rev. Letters 54, 2079. Mayle, R., Wilson, J. R., and Schramm, D. N.: 1987, Astrophys. 3". 318, 288. Meyer, H.: 1986, Proc. Neutrino-86 Conf., Sendai 674. Mikheev, S. P. and Smirnov, A. Yu.: 1985, Soviet. Nucl. Phys. 42, 913. Mikheev, S. P. and Smirnov, A. Yu.: 1986, Nuovo Cimento C9, 17. Mitsui, K., Minorikawa, Y., and Komori, H.: 1986, Nuovo Cimento 9C, 995. Musset; P. and Vialle, J. P.: 1978, Phys. Rep. 39, 1. Nadezhin, D. K. and Otroshchenko, I. V.: 1980, Soviet Astron. 24, 47. Nakatsuka, T., Kobayakawa, K. and Kitamura, T.: 1987, Proc. 20th Cosmic Ray Conf. 6, 261. Oyama, Y., Arisaka, K., Kajita, T., Koshiba, M., Nakahata, M., Suzuki, A., Totsuka, Y., Kifune, T., Suda, T., Sato, N., Takahashi, K., and Miyano, K.: 1986, Phys. Rev. Letters 56, 991. Oyama, Y., Hirahata, K., Kajita, T., Koshiba, M., Nakahata, M., Kifune, T., Suda, T., Nakamura, K., Takahashi, K., Tanimori, T., Miyano, K., Yamada, M., Beier, E. W., Feldscher, L. R., Kim, S. B., Mann, A. K., and Cortez, B. G.: 1987, Phys. Rev. D36, 3537. Quigg, C., Reno, M. H., and Walker, T. P.: 1986, Phys. Rev. Letters 57, 774. Peebles, P. J. E.: 1967, Astrophys. J. 147, 868. Perkins, D. H.: 1984, Ann. Rev. Nucl. Part. Sci. 34, 1. Peterson, V.: 1984, Proc. Neutrino-84 Conf., Dortmund 543. Pontecorvo, B. M.: 1958, Soviet Phys. JETP. 34, 247. Raffelt, G. and Seckel, D.: 1988, Phys. Rev. Letters 60, 1793. Reines, F.: 1967, Proc. Roy. Soc. A301, 159. Reines, F. and Vandervelde, J.: 1988, Phys. Rep. 163, 137. Reno, M. H. and Quigg, C.: 1988, Phys. Rev. D37, 657. Rood, R. T.: 1978, Proc. Conf. Solar Neutn'no Research, Brookhaven 1, 175. Rowley, J. K., Cleveland, B. T., and Davis, R.: 1985, AlP Conf. Proc. 126, 1. Roxburgh, I. W.: 1985, Solar Phys. 100, 21. Ruderman, M. A.: 1965, Rep. Prog. Phys. 38, 411. Samorski, M. and Starnm, W.: 1983, Astrophys. J. 268, LI7. Sato, K. and Suzuki, H.: 1987, Phys. Rev. Letters 58, 2722. Schramm, D. N., Mayle, R., and Wilson, J. R.: 1986, Nuovo Cimento 9C, 443. Shapiro, M. M. and Silberberg, R.: 1985, Proc. 19th Cosmic Ray Conf., La Jolla 8, 160. Spergel, D. N. and Press, W. H.: 1985, Astrophys. J. 294, 663. Stanev, T.: 1988, Nucl. Inst. Meth. A264, 32. Stecker, F. W.: 1979, Proc. Neutrino-79 Conf., Bergen 1,475. Stenger, V. J.: 1984, Astrophys. aT. 284, 810. Suzuki, A.: 1987, UT-ICEPP-87-06, preprint. Svoboda, R., Bionta, R. M., Blewitt, G., Bratton, C. B., Casper, D., Chrysicopoulou, P., Ciocio, A., Claus, R., Cortez, B., Dye, S. T., Errede, S., Foster, G. W., Gajewski, W., Ganezer, K. S., Goldhaber, M., Haines, T. J., Jones, T. W., Kielczewska, D., Kropp, W. R., Learned, J. G., LoSeeco, J. M., Matthews, J., Park, H. S., Reines, F., Schultz, J., Seidel, S., Shumard, E., Sinclair, D., Sobel, H. W., Stone, J. L., Sulak, L., Thornton, G., van der Velde, J. C., and Wuest, C.: 1987, Astrophys. J. 315, 420. Totsuka, Y.: 1987, UT-ICEPP-87-02, preprint. Turner, M. S.: 1988, Phys. Rev. Letters 60, 1797. Volkova, L. V.: 1980, Soviet J. NucL Phys. 31,784. Vuilleumier, J. L.: 1986, Rep. Prog. Phys. 49, 1293. Weekes, T. C.: 1988, Phys. Rep. 160, 1. Wolfenstein, L.: 1978, Phys. Rev. D17, 2369. Wolfsberg, K., Cowan, G. A., Bryant, E. A., Daniels, K. S., Downey, S. W., Haxton, W. C., Niesen, V. G., Nogar, N. S., Miller, C. M., and Rokop, D. J.: 1985, AlP Conf. Proc. 126, 196. Woosley, S. E. and Weaver, T. A.: 1986, Ann. Rev. Astron. Astrophys. 24, 205. Zatsepin, G. T.: 1968, J. Eksperim. Theor. Phys. Letters 8, 205.