When a massive star at the end of its life collapses to a neutron star, it radiates almost all of its binding energy in the form of neutrinos, most of which have energies in the range 10-30 MeV. These neutrinos come in all flavors, and are emitted over a timescale of several tens of seconds. The neutrino luminosity of a gravitational collapse-driven supernova is typically 100 times its optical luminosity.
The neutrino signal emerges from the core of a star promptly after core collapse, whereas the photon signal may take hours or days to emerge from the stellar envelope. The neutrino signal can therefore give information about the very early stages of core collapse, which is inaccessible to other kinds of astronomy. In fact, an optical supernova display may never be seen at all for a given core collapse: some collapsing stars may never blow up into supernovae, or the star may live in an obscured region of the galaxy.
Neutrinos from gravitational collapse can be detected in various ways. A number of GC neutrino detectors are in operation or under construction around the world. They belong to various classes: scintillator, water Cerenkov, heavy water, radiochemical. The most important detection reaction for the most common detectors (in particular, scintillator and water Cerenkov detectors) is the absorption of electron antineutrinos on protons,
The positron from this reaction, which retains most of the energy of the incoming neutrino, is detected. In some cases the neutron can also be detected via a delayed fusion gamma ray.
Other reactions of importance for existing and planned detectors are neutral and charged current neutrino-electron scattering, excitation of C-12, deuterium breakup reactions, and interactions with other elements such as 40-Ar, 37-Cl and 71-Ga.
The famous supernova SN1987A, a gravitational collapse event in the Large Magellanic Cloud outside our Galaxy, was the first to have its neutrino signal detected. Two water Cerenkov detectors, Kamiokande II and IMB, detected 20 events between them. In addition the Baksan scintillator detector saw 5 events (another scintillator detector, LSD saw 5 events several hours early, but the significance of this signal is controversial.) The SN1987A neutrino data, although sparse, was sufficient to confirm the baseline model of gravitational collapse (and put some limits on neutrino properties, such as mass, as well). We await a more copious neutrino signal to be able to make distinctions between different theoretical models of core collapse and supernova explosions.
The MACRO detector is a large area underground scintillator/streamer tube array located at the Gran Sasso National Laboratory. MACRO's primary physics goal is the detection of magnetic monopoles; however the MACRO's 550 tons of liquid allow good sensitivity to gravitational collapse in our Galaxy.
Like most GC neutrino detectors, the liquid scintillator of MACRO is primarily sensitive to the antineutrino component of a gravitational collapse neutrino flux, via the positron in the charged current inverse beta decay reaction antineu_e+p->n+d. The secondary neutron can also be detected.
A burst of neutrinos from a GC at 10 kpc (near the center of the Galaxy) is expected to induce a total of approximately 200 events above 7 MeV in MACRO's scintillator within a few tens of seconds. Of these events, approximately half are expected to occur during the first second of the burst, with the rate of events dropping thereafter. The spectrum of neutrino energies emitted from the collapsing star softens with time, so that the average energies of detected events should decrease over the time scale of the burst. The average electron antineutrino energy is expected to be in the range 10-20 MeV.
The sensitivity of MACRO to a burst of GC neutrinos is determined by the rate of background events which produces ``fake'' bursts from Poissonian fluctuations. The background rate in MACRO is mainly due to natural radioactivity from the surrounding rock and concrete, and decreases rapidly with increasing energy threshold. The rate of background counts in one scintillation counter is approximately 1 Hz at a 4 MeV threshold, and 5 kHz at a 1 MeV threshold. A burst of neutrinos from beyond the opposite edge of our galaxy (but not quite out to the LMC) should stand out above background in the MACRO data.
In MACRO, two independent triggers provide sensitivity to GC neutrino events. In both cases, trigger decisions are made based on energy deposition in a tank, by making online corrections for light attenuation along the length of a tank. The PHRASE (Pulse Height Recorder and Synchronous Encoder), built by MACRO collaborators from the University of Pisa, provides a 7 MeV energy trigger by making an analog compensation for light attenuation. The ERP (Energy Reconstruction Processor), built at the University of Michigan, uses look-up tables to calculate the energies corresponding to the PMT pulse heights measured at each end of the tank, and stores events with energy deposition greater than 5-6 MeV in a buffer which is read out every several minutes.
Since the neutrino signal from a core collapse can arrive hours or days earlier than the optical signal, a prompt neutrino burst alarm signal can be very useful. (Unfortunately, no directional information is available from MACRO's signal).
For MACRO, special "spy" software looks at data coming in from the acquisition and does a preliminary analysis and searches for bursts of neutrino-like events with rate above background. If a sufficiently low probability burst of hits is found, an alarm message is generated, and relayed to an "on-call" physicist via a wireless communications system. There are two separate systems: one which looks at the Phrase data(with alarm messages sent within Italy) and one which looks at the ERP (with alarm messages sent to the U.S.) On the Italian side, Phrase alarms go to a special cellular phone with a modem link. On the US side, alarm messages for neutrino candidates go to a special email beeper rented from SkyTel. In either case, the on-call physicist is expected to log on immediately to check out the data in more detail, and raise an alert if the candidate burst looks sufficiently like a GC neutrino burst.
See the references for more information.