A supernova that lit up the skies in the year 1006 lived and died fast, leaving no companion star behind, astronomers have found. The result is the latest clue in a puzzle that has been troubling astronomers for some time – how does this type of stellar explosion happen?

Supernova 1006 exploded, as seen from Earth, in the year 1006 (hence the name). It was probably the brightest star ever visible to humans and could have been bright enough to read by at night. Astronomers know that it was a type Ia supernova – a particular type of stellar sticky end that involves taking another star along for the ride.

Type Ia supernovae are one rung on the cosmic ladder we use to measure distances in the universe. They were instrumental in the discovery of the accelerating expansion of the universe – caused by what we call dark energy – in 1998.

But the origins of this type of cosmic explosion have long been a bit of a mystery. Astronomers think that there are two ways type Ia supernovae are created. Both ways involve a white dwarf, a dead star similar to what the Sun will one day turn into, gathering up mass from an unsuspecting companion before reaching its limit and exploding in a dazzling supernova. One way is fast, stealing material from another white dwarf in orbit around the soon-to-explode one, and breaking it up in the process. The other way is slow, with the white dwarf dragging material from a companion that is more like the Sun as it is now, or a giant star, over time.

Crucially, the second, slow, method would leave behind a telltale sign. The companion star should survive the explosion and be visible alongside the supernova remnant. So that is what astronomers have been searching for.

Jonay González Hernández at the Instituto de Astrofísica de Canarias and colleagues searched for a companion star to supernova 1006. They came up empty handed.

SN 1006 is 7,100 light years away from us. González Hernández and colleagues started their search at the centre of the X-ray emission from the supernova remnant left behind after SN 1006 exploded, out to a radius of just over eight light years.

They were looking for a star at the same distance as the supernova. “In the 1000 years since the explosion occurred, there has not been much time to move far, at least in astronomical standards,” says González Hernández. It should also have shown some heavy elements such as nickel in its atmosphere, and a large ‘kick’ velocity.

“We found only four stars at a distance compatible with that of the supernova remnant, but all of them were ‘normal’ giant stars,” says González Hernández, and did not show the peculiarities predicted for stars that have survived a supernova.

Taking their study with that of four other supernovae, González Hernández and his colleagues conclude that only 20% of type Ia supernovae happen through the slow channel that leaves a companion behind. They arrived at this figure because only one of those five supernovae appears to have a companion star.

But five is a very small sample, points out Ryan Foley, an astronomer at the Harvard-Smithsonian Centre for Astrophysics in Cambridge, Massachusetts, who was not involved in the study. “Getting a rate from that has a lot of intrinsic uncertainty,” he says. The true rate could vary from that substantially. 1

González Hernández admits that the figure of 20% is “indicative” and that astronomers need to study more cases to see if it stands up.

There have also been questions raised over the five supernova that have already been studied.

The one companion star that has been found – that alongside supernova 1572 – has been called into question. The jury is still out on whether or not it really was involved in the supernova or just an innocent bystander. And there is another caveat: if any of these supernovae had companion stars that were smaller than our Sun, they would have been too faint to see. “If [a companion to supernova 1006] is somehow still hidden,” says González Hernández, “it has to be a star probably smaller than our Sun.”

While the 20% figure may have shaky foundations, this study does provide one more data point in the puzzle of whether type Ia supernovae form through the fast or slow channel – or both. And the solution to that puzzle matters.

If all type Ia supernovae are not created equal, then the path a particular supernova took will need to be taken into account when calculating cosmic distances. The answers we get are unlikely to change dramatically, but if we want accurate results we need to learn exactly how our tools work.

  1. In a sample of 5 systems, 1 potentially has a companion star. 1 in 5 is 20%. But there’s some chance in that. For instance, if the true rate were 50% (instead of 20%), about 17% of the time you would actually observe what we see: 1 in 5. If you include this statistical uncertainty, the proper claim is probably something like 9-66%.


Jonay I. González Hernández, Pilar Ruiz-Lapuente, Hugo M. Tabernero, David Montes, Ramon Canal, Javier Méndez, & Luigi R. Bedin (2012). No surviving evolved companions of the progenitor of SN 1006 Nature DOI: 10.1038/nature11447


Composite image of the SN 1006 supernova remnant. Credit: X-ray: NASA/CXC/Rutgers/G.Cassam-Chenaï, J.Hughes et al.; Radio: NRAO/AUI/NSF/GBT/VLA/Dyer, Maddalena & Cornwell; Optical: Middlebury College/F.Winkler, NOAO/AURA/NSF/CTIO Schmidt & DSS