Astronomy › Section Q · The death of stars › Topic Q.3

Supernova observations

Supernovae are rare in any one galaxy, yet their light, their debris and even their neutrinos have been caught. From medieval guest stars to the nearby blast of 1987, observations have tested the theory of how massive stars die.

Guest stars

Before telescopes, a star that suddenly appeared where none had been seen was taken to be new, and the Latin word nova, meaning new, stuck. Twentieth-century astronomers named the most luminous, far more violent outbursts supernovae. Chinese astronomers called them guest stars, because they arrived, stayed for months and left. The Milky Way probably has a supernova roughly every 30 to 100 years, yet none has been seen in it since 1604. Most were likely hidden by the interstellar dust that dims light across our galaxy's disc.

Supernovae in history

In 1006 a supernova in the southern constellation Lupus became the brightest star-like object ever recorded, bright enough to cast shadows. In 1054 Chinese astronomers recorded a guest star in Taurus, visible by day for weeks; its expanding debris is today's Crab Nebula. A fainter one followed in 1181. In 1572 the young Tycho Brahe measured a new star and showed that it did not shift against the background stars, so it lay far beyond the Moon, overturning the old belief that the heavens never change. Kepler studied another in 1604. Modern study of their remnants suggests the 1006, 1572 and 1604 events were white dwarf explosions, while the 1054 one was a core collapse.

How bright and how fast

At its peak a supernova can emit roughly a billion to ten billion times the Sun's luminosity, where one solar luminosity, \(L_\odot\), is about \(3.83 \times 10^{26}\) W. For a few weeks one star can outshine the billions of stars in its host galaxy. The light then fades over months to years. Spectra show the outer layers flying outwards at about 10,000 km/s, sometimes twice that. Spectra also sort supernovae into types. Type I spectra show no hydrogen lines; Type II spectra show strong ones. Type II, and the Type Ib and Ic subtypes, are core collapses of massive stars, while Type Ia are exploding white dwarfs, covered in "Binary star evolution and gamma-ray bursts".

Worked example: a growing remnant

Debris from a supernova leaves at about 10,000 km/s. How far does it travel in one year, and how big could a 1,000-year-old remnant be? Take one year as \(3.156 \times 10^{7}\) s, one astronomical unit (AU, the Earth-Sun distance) as \(1.496 \times 10^{8}\) km and one light-year as \(9.46 \times 10^{12}\) km. 1. Distance in a year: \(10^{4} \times 3.156 \times 10^{7} = 3.16 \times 10^{11}\) km. 2. In AU: \(\frac{3.16 \times 10^{11}}{1.496 \times 10^{8}} = 2110\) AU. 3. In light-years: \(\frac{3.16 \times 10^{11}}{9.46 \times 10^{12}} = 0.0334\) ly. 4. After 1,000 years at that speed: about 33 ly from the centre. 5. Check: real debris slows as it sweeps up interstellar gas, so a 1,000-year-old remnant should be somewhat smaller, a few tens of light-years across, as observed.

Supernova 1987A

On 24 February 1987, Ian Shelton, working at Las Campanas Observatory in Chile, found a new star on a photograph of the Large Magellanic Cloud, a dwarf galaxy orbiting the Milky Way about 160,000 light-years away. Observatory staff member Oscar Duhalde had glimpsed it by eye the same night, and Albert Jones in New Zealand spotted it independently. It was the nearest supernova seen since 1604 and still visible without a telescope. For the first time, the star that exploded could be identified in older photographs: a blue supergiant of about 20 solar masses. Theory had expected a red supergiant, so the surprise showed that mass loss and composition change how a dying massive star looks.

Neutrinos arrive first

The strongest test of core-collapse theory came from underground. Hours before the supernova's light was first recorded, three detectors, Kamiokande-II in Japan, IMB in the United States and Baksan in the Soviet Union, registered a burst of about two dozen neutrinos within about 13 seconds. Neutrinos escape the collapsing core at once, whereas light must wait for the shock to reach the star's surface, which explains the head start. Their number and energies matched predictions: about \(10^{46}\) J carried off by roughly \(10^{58}\) neutrinos, confirming that the core had collapsed to a neutron star's size.

Radioactive light

After its first few days SN 1987A did not simply fade as the blast cooled: it brightened for about three months, then dimmed at a steady rate. Radioactive nuclei made in the explosion were powering it. Nickel-56 turns into cobalt-56 (half-life roughly 6 days), which in turn becomes stable iron-56 (half-life roughly 77 days). Each decay releases gamma rays, which the expanding gas absorbs and re-emits as visible light. For months the supernova faded at almost exactly the rate set by cobalt-56, and orbiting and balloon-borne telescopes detected the gamma-ray lines of cobalt-56 directly. This showed that supernovae really do forge new elements, including much of the iron in the universe.

Worked example: fading with cobalt

In a supernova, all the cobalt-56 is in place a few weeks after the explosion. What fraction is left 308 days later, and how much fainter should the radioactive glow be? 1. Number of half-lives: \(\frac{308}{77} = 4\). 2. Fraction left: \(\left(\frac{1}{2}\right)^4 = \frac{1}{16} = 0.0625\), about 6%. 3. The glow is powered by decays, so if the gas still traps all the gamma rays, the brightness drops in proportion, to about a sixteenth. 4. Check: after 154 days, two half-lives, a quarter should remain, halfway down on a logarithmic scale. In practice the gas thins out and lets some gamma rays escape, so real supernovae later fade a little faster than this.

The remnant of SN 1987A

Hubble images since the 1990s show a ring of gas that the star shed thousands of years before it died. Debris from the explosion began hitting the ring in the late 1990s, lighting it up with bright spots. For decades no neutron star could be seen in the dusty centre. In 2024 a team using the James Webb Space Telescope reported argon and sulfur atoms near the centre stripped of several electrons, which needs high-energy radiation that a young, hot neutron star would supply. That is the strongest evidence yet that SN 1987A left a neutron star, as the neutrinos implied.

Watching for the next one

Robotic surveys now find thousands of supernovae a year in other galaxies, many within days of the explosion. The Vera C. Rubin Observatory in Chile began issuing alerts in February 2026 and started its ten-year sky survey in mid-2026, and it is expected to find many more. If one goes off in the Milky Way, neutrino detectors will catch the burst first and send an automatic warning to telescopes.

Common mistakes to avoid

First, thinking a supernova is a newly born star: it is a dying one. Second, thinking a nova and a supernova are the same: a supernova is vastly more luminous and destroys or transforms the star. Third, assuming all supernovae come from massive stars: Type Ia are white dwarf explosions. Fourth, thinking the light carried most of SN 1987A's energy: neutrinos did. Fifth, assuming none have occurred in the Milky Way since 1604: dust probably hid several.

Key takeaways

Historical supernovae in 1006, 1054, 1181, 1572 and 1604 left remnants we still study, such as the Crab Nebula. Supernovae peak at about a billion to ten billion solar luminosities and eject gas at about 10,000 km/s. Type II spectra show hydrogen; Type I do not. SN 1987A in the Large Magellanic Cloud came from a blue supergiant of about 20 solar masses. About two dozen of its neutrinos were caught hours before its light, and the decay of nickel-56 to cobalt-56 to iron-56 powered its later glow. Webb data in 2024 point to a neutron star at its centre.

Flashcards

Flashcard

Guest star

Old Chinese name for a star that appeared and faded

Many were supernovae.

Test your understanding

Question 1 of 5

Why did the neutrinos from SN 1987A arrive before its light was seen?

Question 2 of 5

A supernova spectrum shows strong hydrogen lines. What type is it?

Question 3 of 5

Cobalt-56 has a half-life of about 77 days. What fraction of the cobalt-56 remains 231 days after it formed?

Question 4 of 5

How did Tycho Brahe show that the new star of 1572 was not inside Earth's atmosphere or near the Moon?

Question 5 of 5

Supernova debris moves at 20,000 km/s. Using 1 AU = 1.496 × 10⁸ km and 1 year = 3.156 × 10⁷ s, about how far does it travel in a year?

← Previous topicNext topic →