In the late 90s there was a race going on between two astronomy collaborations. Both were on the verge of making a discovery that would change the field of cosmology forever, though they may not have realised it at the time. The High-z Supernova Search Team and the Supernova Cosmology Project were both studying a peculiar sort of exploding star, known as a type 1a supernova, and trying to figure out the ultimate fate of the universe — would it expand forever, or eventually slow to a stop and reverse in on itself in a "big crunch"? The answer: neither. Both teams reported, in separate papers, one published in 1998 and the other in 1999, that the expansion of the universe was actually accelerating. That is, it is expanding faster today than it was yesterday, and tomorrow it will be moving apart even faster than it is today. This was a rather unexpected result.

Astronomers think that the type of supernova used to make this discovery, type 1a, occurs when a white dwarf — an old, dense star that was once similar to our own Sun — with a binary companion begins to drag material from its companion star on to itself, growing bigger and bigger until, eventually, it can no longer sustain itself. At this point, it implodes, then explodes in a bright supernova. The upper mass limit of a white dwarf, above which it will turn into a supernova, is called the Chandrasekhar mass, after the Indian astrophysicist who predicted its existence in 1930. It is caused when a phenomenon known as electron degeneracy breaks down. Electron degeneracy is a result of the Pauli exclusion principle, which states that two particles cannot occupy the same quantum state at the same time. In effect, the electrons in the white dwarf refuse to be pushed closer together, creating a repulsive force that can, for a while, balance the extra gravity caused by the accreting material. If enough material is accreted on to the white dwarf, however, the gravitational force will outweigh that from electron degeneracy and the white dwarf will implode. The Chandrasekhar limit is the point at which these two opposing forces exactly balance. Once a white dwarf passes the Chandrasekhar limit, and after it goes through the supernova stage, it becomes a neutron star or, sometimes, a black hole.

The European Southern Observatory (ESO) has a nice video showing how the white dwarf sucks material from its companion:

There are two different scenarios in which a type 1a supernova may be created. In the first, only one star in the binary system is a white dwarf, the other is a main sequence star like the Sun, or a more evolved star such as a red giant. In the second scenario, both stars are white dwarfs.

In a situation with only one white dwarf, there should be matter left over after the supernova. This material, which was not accreted on to the white dwarf before it imploded, is known as circumstellar matter because it surrounds the stars. Its detection near a type 1a supernova would indicate that the binary system from which the supernova was born contained only one white dwarf.

Knowledge of a supernova's predecessors could prove invaluable. Type 1a supernovae can be used to measure distances on a cosmic scale. Predictable patterns in the way type 1a supernova brighten and fade reveal their true luminosity, which can be compared with how bright they look to us to find their distance. Measuring distances using these supernova was how the two teams of astronomers made their discovery of the accelerated expansion of the universe. We say that this expansion is caused by dark energy, but the phrase "dark energy" could, at the moment, mean a number of different things. A better understanding of one of the main pieces of evidence for dark energy may help us to figure out what it actually is. Which brings us back to those type 1a supernovae...

Astronomers have looked for circumstellar material around certain type 1a supernovae before and found evidence to suggest that it is there. Dr Assaf Sternberg, of the Benoziyo Center for Astrophysics at the Weizmann Institute of Science in Israel, and his colleagues wanted to go one step further and find out whether all type 1a supernova have this material close to them.

By looking at spectra from type 1a supernovae, Sternberg and colleagues hoped to test their hypothesis. A spectrum contains the signature of everything that the light from a star had to pass through on its way to Earth. Spectra from supernovae with circumstellar material close by should provide a telltale sign of this in their spectra. The sign Sternberg and his colleagues looked for was absorption of light by sodium atoms, as this would indicate that there was cool, neutral gas present — the circumstellar material. They expected the lines in the spectra made by sodium absorption to be blueshifted, too. When something is "blueshifted" it is moving towards the observer — us — and away from its source, in this case the supernovae.

Sternberg and his colleagues looked at the spectra from 35 type 1a supernovae and 11 of another type, known as "core collapse" supernovae, that they got from the High Resolution Echelle Spectrometer (HIRES) at the Keck Observatory in Hawaii and Magellan Inamori Kyocera Echelle (MIKE) spectrograph on the Magellan telescopes which are part of the Carnegie Institution for Science. They also studied 6 previously published type 1as supernova spectra and 7 core collapse ones.

Sternberg and his colleagues first found out how far away each supernova was and what kind of galaxy it was from, by looking in astronomical databases and using existing images taken in the Sloan Digital Sky Survey. They then looked carefully at the spectrum from each supernova, keeping an eye out for features that would give away the existence of circumstellar matter.

What the team found was that, more often than not, the spectra contained blueshifted sodium absorption features. They checked to see whether these features could have been produced by clouds of gas in the host galaxy of the supernova or wind blown by the host galaxy, as well as checking whether interstellar matter was the cause, but found that the most likely explanation was an inherent feature of the supernovae themselves — the circumstellar matter they were looking for. Their results suggest that type 1a supernovae, or at least some type 1a supernovae in nearby spiral galaxies, are born of binary star systems with only one white dwarf.

Hopefully, this will be one of the first steps on a path that will lead to a better understanding of the origin of these supernovae, and we will no longer be left with the lingering question of whether they really are a reliable tape measure for cosmic distances.


Sternberg A, Gal-Yam A, Simon JD, Leonard DC, Quimby RM, Phillips MM, Morrell N, Thompson IB, Ivans I, Marshall JL, Filippenko AV, Marcy GW, Bloom JS, Patat F, Foley RJ, Yong D, Penprase BE, Beeler DJ, Allende Prieto C, & Stringfellow GS (2011). Circumstellar material in type Ia supernovae via sodium absorption features. Science (New York, N.Y.), 333 (6044), 856-9 PMID: 21836010

Video credit: ESO/M. Kornmesser