Supernovae are exploding stars. Observations of nearby supernovae are conspicuous entries in the annals of Chinese imperial astrologers dating back to 185 C.E., and supernova observations are among the great works of the Renaissance astronomers Tycho Brahe and Johannes Kepler. However, the true nature of these "guest stars" was not understood until observations in the 1930s revealed the distances to these stellar disasters. Understanding supernova explosions depends on understanding particle physics: the properties of the very smallest components of the universe determine the properties of its most energetic events. Supernovae are important engines in transforming the microscopic properties of the universe. They fuse simple elements such as hydrogen and helium into complex ones such as iron, gold, and uranium, and they blast those products into the gas between stars to enrich the next generation of stars. Supernova debris can help form planets and makes up living things.
Because the brightest supernova explosions are about as bright as 4 billion suns, they can be detected at large distances. Careful measurements of supernovae provide the distance scale of the universe and help establish the 14-billion-year timescale of cosmic expansion. Light from supernovae that has traveled half the span of the observable universe shows that cosmic expansion, surprisingly, has been speeding up. This cosmic acceleration, first glimpsed in 1998, suggests a new property of empty space itself: space has an energy whose outward pressure is revealed only by the supernova data. If this picture is right, supernovae show that two-thirds of the universe resides in an enigmatic dark energy. Explaining this phenomenon in terms of fundamental physics will be an important challenge for the twenty-first century.
The sudden appearance of a new star is a surprise: the lifetime of a short-lived star is 100,000 times that of a very long-lived person, so humans think stars are permanent and unchanging. However, the universe is not constructed on the human scale in space or time. The sudden death of a star in a thermonuclear explosion or a gravitational collapse is rare in any single galaxy, such as the Milky Way, but common throughout the 100 billion galaxies of the observable universe. A single galaxy has a supernova explosion approximately every century, so there should be a billion supernova explosions every year in the observable universe—thirty events per second. In 2001, 254 supernovae were observed on Earth. Only a small fraction of all these events are actually seen because the entire sky is not observed every night and because the searches are not yet sensitive enough to reveal the most distant supernovae. There is room for improvement in the study of supernovae.
One key element in understanding supernovae is determining their distances: the apparent brightness of a supernova is not very conspicuous, with a few glorious exceptions such as Supernova 1987A in the Large Magellanic Cloud. When the observed flux is coupled with the distance, the intrinsic properties of supernovae become clear. In the early 1900s several new stars ("novae") were noted in spiral nebulae, which today are known as supernovae, in galaxies at distances of millions of light-years. At that time, however, it was more conservative to think that these were ordinary novae as seen in the Milky Way, which would imply that the spiral nebulae were part of the galaxy. In the 1920s the work of Edwin Hubble established that the spiral nebulae were outside the Milky Way, with the nearest of them at distances of millions of light-years. Hubble was puzzled by "that mysterious class of exceptional novae which attain luminosities which are respectable fractions of the systems in which they appear." (Hubble, 103). These were not ordinary novae, they were thousands of times brighter. Fritz Zwicky and Walter Baade dubbed this new class of exploding stars supernovae and set out to study them. Zwicky developed methods to search for supernovae: each time the moon was dark, he repeated a set of photographs of his target galaxies and compared the images by eye to find the new stars. Modern methods use the same approach he developed, except that the telescopes are automated, the detectors are giant electronic cameras with up to 100 million pixels, and the before and after comparison of gigabytes of data is carried out by computer algorithms.
In 1934 Zwicky and Baade proposed that supernova explosions come from the gravitational collapse of a star as it shrivels from 100-million-mile dimensions to "little spheres 14 miles thick." These dense clinkers were neutron stars—objects with the mass of the Sun but made of neutrons. Since neutrons had only been discovered in 1932, this was a remarkable extrapolation. In ordinary matter, electron clouds separate the massive nuclei from one another. Nuclei occupy only about 10-15 of the volume of ordinary matter, which is mostly the more or less empty space where the electrons orbit. In a neutron star, electrons and protons are compressed by gravity to form neutrons—a neutron star is a massive object made of 1057 particles whose density approaches the nuclear density of 1017 kg/m3.
This brilliant guess has been confirmed by modern work: neutron stars are real objects, and some supernovae do derive their energy from the gravitational collapse to a neutron star. A star with 8 solar masses or more fuses hydrogen to helium during most of its lifetime, which is measured in millions of years. Sub-sequent stages of nuclear burning, in which the ashes of each burning stage become the fuel for the next, lead to the accumulation of carbon, oxygen, silicon, and finally iron in the core of a massive star. Because iron is the most tightly bound nucleus, no further energy can be extracted by fusion. A star with a hot iron core has huge energy losses from neutrino emission but no energy source, and collapse is inevitable.
The actual moment of collapse is precipitated by energetic gamma rays in the hot, dense interior, which begin to break apart the iron nuclei, leading to a catastrophic loss of pressure. The core of the star collapses from a region about the size of the Earth (10,000 km) to the dimensions of a neutron star (100 km) in less than a second, with the inward velocity approaching one-third of the speed of light. This headlong implosion is halted with a violent snap when the core of the star approaches nuclear density. At that point, repulsive nuclear forces stiffen the forming neutron star, halting its collapse. As the material falling in smashes into the forming neutron star, computer simulations show that a powerful shock wave travels upstream, out through the star. This shock, refreshed by a blast of neutrinos emitted from the hot material raining down on the neutron star, cooks new elements from the iron just outside the forming neutron star and blows the star apart to create the visible explosion seen as a supernova. It is a strange picture: most of the gravitational energy of the collapse is emitted as massless, chargeless, and nearly undetectable neutrinos. Only about 1/10,000 of the energy is converted into the light by which supernova explosions are detected.
Astronomical observations confirm this physical picture as the source of some supernovae—most conspicuously, the observations of supernova 1987A in the Large Magellanic Cloud. There the presupernova star was observed: it was a 20-solar-mass star at an advanced stage of evolution. The supernova was discovered from its optical emission, but subsequent inspection of the records from underground neutrino detectors showed that hours before the supernova began to brighten, there was a brief flash of neutrinos, signaling the formation of a neutron star. Nuclear gamma rays from freshly synthesized
radioactive isotopes produced in the shock wave, especially 56Co, were seen in the months after the explosion. Although these observations confirm the basic picture, no neutron star has yet been found in the center of Supernova 1987A.
In the 1940s astronomers discovered there were two basic types of supernovae, as distinguished by their spectra. The original type had no hydrogen lines in the spectrum, but the new type, called Type II, did. Scientists have since discovered that Type II supernovae are core-collapse supernovae. Type Ia supernovae come from different types of stars and erupt by a completely different mechanism, but by coincidence, they emit a similar amount of light. Type Ib and Type Ic supernovae come from massive stars and are powered by gravitation but have no hydrogen in their atmospheres because they have exhaled it in a stellar wind before the explosion.
The modern picture of Type Ia supernovae is that they come from the thermonuclear explosion of white dwarf stars. Stars of less than about 8 solar masses, such as the Sun itself, fuse hydrogen to helium, and helium to carbon and oxygen, but they do not burn all the way up to iron. Nuclear fusion stops in these stars when the core of the star becomes degenerate—when the density becomes high enough so that the quantum mechanical properties of electrons themselves supply the pressure to support the star. For a star like the Sun, a carbon-oxygen white dwarf will be the endpoint of stellar burning about 5 billion years from now. The pressure in a degenerate star does not depend on its temperature, so a cooling white dwarf can be a stable object supported against gravity by its electrons. However, there is an upper limit, called the Chandrasekhar limit, to the mass of a star that can be supported by degeneracy pressure: about 1.4 solar masses. For stars in binary systems, where one star has become a white dwarf, the other star can transfer significant amounts of mass to the white dwarf. As the white dwarf accumulates matter and grows toward the Chandrasekhar limit, computations show that nuclear burning can begin again. In ordinary stars, the heat generated from fusion generates pressure that can make the star expand and cool slightly, decreasing the rate of energy production. This regulating effect ensures that ordinary stars will not explode.
Degenerate matter, on the other hand, is quite different—generating energy by fusing oxygen nuclei increases the rate of energy generation but does not make the star expand and cool. The result is a runaway thermonuclear explosion that rips through the entire white dwarf and destroys it as a Type Ia supernova. The burning wave turns much of a white dwarf into iron and blasts off the outer layers of the star at speeds above 10,000 km/s. Observations show that Type Ia supernovae have the chemistry of exploded white dwarfs: oxygen and carbon on the outside, and radioactive iron ashes inside. Radioactivity powers the light curve of Type Ia supernovae: they take about 20 days to reach their peak brightness, decline by about a factor of 2 in the first 2 weeks after maximum light, and then enter into a long exponential decline powered by the decay of freshly synthesized 56Co.
Because Type Ia supernovae come from a very well-defined physical situation, it is not too surprising to find that they have a well-defined peak energy output. It turns out to be about 4 × 109 solar luminosities. In the 1990s Type Ia supernova became the most powerful tools for measuring cosmic distances. The key improvement was to use the shape of the light curve, which is correlated with intrinsic brightness, to determine which Type Ia supernovae are brighter than the mean and which are dimmer. In 2002 the precision of the distance estimate to a single Type Ia supernova, after taking into account the light curve shape, was determined to be 8 percent, which makes them the best standard candles for judging cosmic distances.
The Hubble Space Telescope has been used to measure the distance to galaxies in which Type Ia supernovae have exploded by observing the brightness of Cepheid variable stars. This establishes the relation between cosmic redshift and cosmic distance, based on Type Ia supernovae. In 2002 the values of the Hubble Constant found from Type Ia supernovae ranged from about 60 to 75 km/s/mpc, with most of the uncertainty associated not with the supernovae but with lower rungs on the cosmic distance ladder.
Because Type Ia supernova are so bright, they can be detected at very large distances about halfway back to the Big Bang. This provides a way to study the history of cosmic expansion. The expansion of the universe during the time the light from a supernova is en route affects its apparent brightness. In a universe that is decelerating due to gravity, the light from a distant supernova travels a slightly shorter path from the explosion to a telescope. It would appear a little brighter than in a universe that is expanding at a constant rate. Observations reported in 1998 show the opposite: distant supernovae are about 25 percent dimmer than they would be in an empty universe. This surprising result, which implies that the expansion of the universe is speeding up over time, points to the presence of a significant amount of dark energy whose pressure produces the observed acceleration. When combined with information obtained from observing the fluctuations in the cosmic microwave background, the supernova results suggest a universe that is one-third dark matter and two-thirds dark energy. The intensive study of supernovae near and far has revealed two-thirds of the universe!
Goldsmith, D. The Runaway Universe (Perseus Books, Cambridge, MA, 2000).
Hubble, E. P. "A Spiral Nebula as a Stellar System, Messier 31." Astrophysical Journal 69 , 103–158 (1929).
Kirshner, R. P. The Extravagant Universe: Exploding Stars, Dark Energy, and the Accelerating Cosmos (Princeton University Press, Princeton, NJ, 2002).
Marschall, L. The Supernova Story (Princeton University Press, Princeton, NJ, 1994).
Woosley, S., and Weaver, T. "The Great Supernova of 1987" in Stars and Galaxies: Citizens of the Universe, edited by D.E. Osterbrock (W. H. Freeman, New York, 1990).
Robert P. Kirshner