Neutrinos herald the birth of a neutron star in a supernova explosion. When a massive star consumes all of its thermonuclear fuel and becomes gravitationally unstable, it collapses until the degeneracy pressure exerted by protons and neutrons halt the collapse, creating a neutron star. This sudden halt converts the kinetic energy of the core into heat that is trapped inside the neutron star. This heat is radiated out of the core as neutrinos.
Despite the brilliance of a core-collapse supernova, the light we see is only a tiny fraction of the total energy released in the explosion; most of the energy escapes the exploding star as neutrinos. This energy is nearly invisible to us, because the very characteristic that allows the neutrinos to escape the supernova also prevents us from easily observing them. From virtually all supernovae, only electromagnetic radiation is observed. The sole exception is SN 1987A, a supernova observed in the Large Magellanic Cloud in the year 1987.[1] But this one event confirms the basic theory for core-collapse supernovae.
How does one observe neutrinos from a cosmic source? The same way that one observes gamma-rays: one looks for an interaction with an electron. With gamma-rays, such events occur easily, because electromagnetic radiation interacts strongly with charged particles. A gamma-ray enters a detector and collides with an electron; the detector measures the energy of the recoiling electron. A similar type of reaction occurs with neutrinos: a neutrino strikes an electron, giving up a large amount of energy to the electron. If one can recognize these events and measure the energy and direction of motion of the electron relative to the neutrino source, one can measure the energy carried by the neutrino.
The big difference between the gamma-ray and the neutrino in these interactions is that the electromagnetic interaction between a gamma-ray and an electron occurs readily, while the interaction between a neutrino and an electron occurs rarely. This difference dictates where the two types of detector are placed: the gamma-ray detector is placed above Earth's atmosphere, but the neutrino detector is placed deep underground. Gamma-rays cannot pass through Earth's atmosphere, so a gamma-ray detector must be above the atmosphere on a spacecraft or a high-altitude balloon to see cosmic sources. Most neutrinos, on the other hand, not only pass through the atmosphere unscathed, but also through the Earth. Because neutrinos interact so weakly with electrons, the detectors must be very large to see any interactions, and they must be heavily shielded against other types of radiation to suppress particle interactions that mimic the neutrino reactions. The most sensitive detectors therefore weight thousands of tons and sit deep under the Earth.
While there are many neutrino detectors in operation, only three detected SN 1987A; they are Kamiokande II under the Japanese Alps, IMB in a salt mine in Ohio in the United States, and Baksan under the Caucasus.[2] All three detectors work by observing the electrons scattered by neutrinos.
Kamiokande II and IMB are called water Cherenkov detectors. They use several thousand tons of water as a neutrino target, and they detect the interaction of a neutrino with an electron by detecting the light emitted by the scattered electron as it passes through the water. The reason light is emitted is that the electron is moving at nearly the speed of light, but its electromagnetic field, which interacts with the surrounding water, is moving at slightly less than the electron's speed. This drag on the electromagnetic field gives the field a cone-like appearance, much like the shock wave that forms around a supersonic jet. This electromagnetic cone is visible light, and it is called Cherenkov radiation; it is measured with an array of photomultiplier tubes in the neutrino detector. By measuring this radiation, physicists can derive the velocity of the electron. With this velocity and with a direction to the supernova that is established through optical observations, the energy of the neutrino can be derived.
The third detector, Baksan, is very much like a traditional gamma-ray detector; it relies on a scintillating medium rather than Cherenkov radiation in water for detecting the fast electron. As the electron passes through the medium, electrons bound to atoms in the medium move to excited energy levels. When the electrons fall back to the lowest unoccupied energy levels, they emit light, which is measured by a photomultiplier tube. This detector weighs several hundred tons.
The first supernova of 1987, SN 1987A, was noticed on February 24th in the Large Magellanic Cloud by observers in Chile using a ground-based telescope. After the announcement of the discovery, other observers found that they had seen the supernova the previous day. The neutrinos arrived at the neutrino detectors on February 23rd, just three hours before the supernova became visible to astronomers.
Kamiokande II observed 12 neutrinos, IMB observed 8 neutrinos, and Baksan observed 5 neutrinos, for a grand total of 25 neutrinos! I'll write that again—25 neutrinos!!! So out of 1058 neutrinos released from the supernova, and the 1015 neutrinos per square meter that passed through Earth, only 25 neutrinos were detected. But the observers running these detectors would have put many more exclamation marks after this sentence than I would ever dare, because the detectors didn't see 0 neutrinos.
These neutrinos were spread over several seconds. Of the 12 neutrinos seen by Kamiokande II, 9 are spread over 2 seconds, while the remaining 3 arrive 9 to 12 seconds after the first neutrino. Of the 8 neutrinos seen by IMB, 6 arrive over 2.5 seconds, and 2 arrive 5 seconds after the first. Of the 5 neutrinos seen by Baksan, 3 arrived over 2 seconds while 2 arrived 8 and 9 seconds after the first. So the most intense neutrino emission occurs over 2 seconds, followed by a less-intense emission several seconds later, perhaps with a short period of low neutrino emission between the two neutrino outbursts, although with only 25 events, this last conjecture is not adequately supported by the data.
With so few events, how believable is the association of these neutrinos with SN 1987A? The physicists who study neutrinos have great confidence in the observations by Kamiokande II and IMB; the probability of seeing so many events from the direction of the Magellanic cloud in only a dozen seconds is miniscule. The observations by Baksan, on the other hand, are less certain, because this detector is less sensitive than the water Cherenkov detectors; from the count rates seen in the first-two detectors, one expects an average of 2 neutrinos for Baksan, rather than the 5 they see. This may simply be a consequence of small-number probabilities—seeing 5 neutrinos is unlikely, but not highly unlikely—or it may indicate some misidentified events.
An now for the puzzle. A fourth detector measured a signal from the direction of the Large Magellanic Cloud, but the observation is out of line with the other-three detectors. The LVD detector underneath Mont Blanc, between France and Italy, saw 5 neutrino events 5 hours before those detected by the other three detectors. This detector is similar to the Baksan detector in design. The sensitivity of LVD is such that it should only see 1 neutrino from the supernova. The spurious time, the unexpectedly high count rate, and the inability of theory to easily account for such neutrinos cause most astronomers to doubt that this detector observed the supernova. The ambiguity generated by this measurement will persist until other supernovae are seen by neutrino detectors.
The big result is the measurement of neutrinos, which confirms the prediction of the core-collapse model of supernovae. Surveys of the Large Magellanic Cloud prior to the supernova show that the bomb was a blue supergiant star, so the first requirement of the theory—a massive star explodes—is satisfied. Second, the number of neutrinos detected from the supernova translates into an energy in neutrinos of 4×1053 ergs, which is 16% of the rest mass energy of a 1.4 solar mass core, which is consistent with the theory. A third important result is the duration of the neutrino burst. The duration is consistent with the neutrino cooling of a sphere of 27 km radius and 1.4 solar masses. The object is about 50% larger in radius than a neutron star, which is what one expects when core collapse is brought to a halt. As the newborn neutron star cools, its radius should shrink down to about 15km.
The next core-collapse supernova that is seen by neutrino detectors will likely be within our own Galaxy. Such a supernova should produce many times the neutrino flux of SN 1987A. When such a supernova is seen by the current or future generations of neutrino detector, the better statistics provided by high neutrino count rates should enable astronomers to get a better picture of the interior of a newborn neutron star.
[1]Arnett, W. David, Bahcall, John N., Kirshner, Robert P., and Woosley, Stanford E. “Supernova 1987A.” In Annual Review of Astronomy and Astrophysics, edited by G. Burbidge, D. Layzer, and J.G. Phillips, vol. 27. Palo Alto, California: Annual Reviews, 1989: 629–700.
[2]Bahcall, John N. Neutrino Astrophysics. Cambridge: Cambridge University Press, 1989.