Measuring distances is easy enough on Earth, but rather more tricky for astrophysicists and cosmologists attempting to map the Universe. For them, type Ia supernovae (SNIa) are akin to yardsticks. These supernovae burn with a known luminosity — not unlike the wattage of a light bulb — so as an observer on Earth looks at SNIa, their apparent brightness varies according to how far they are away. The locations of many far-flung SNIa can therefore be used to establish a cosmological distance scale with which to track the Universe's ever-expanding growth.

Although the usefulness of SNIa is undisputed, how they come about has been a subject of debate for more than two decades. These supernovae are known to result from a thermonuclear explosion that occurs when a white dwarf star somehow reaches a critical mass, roughly 1.4 times that of the Sun. Two ideas have so far been proposed for how dwarf stars might gain mass: first, that they do so by 'accretion', slowly draining material from a nearby companion star in a close binary system; second, that, in cases in which a binary star system is made up of two white dwarfs, the two could merge.

In 2008, Marat Gilfanov was scribbling down some numbers relating to an X-ray emission glow found in the Andromeda galaxy when he realized he might have the answer to how SNIa arise. It struck him that in the accretion model, a white dwarf would give off a distinctive X-ray emission signature for about 10 million years before the SNIa explosion. By contrast, the merger of two white dwarfs would be expected to give off an X-ray emission for just a short time before the explosion. “Since we were already studying the X-ray emission from the Andromeda galaxy, I did a simple back-of-the-envelope calculation,” explains Gilfanov, an astrophysicist at the Max Planck Institute for Astrophysics in Garching, Germany. He calculated the predicted X-ray emissions if all SNIa in the galaxy came from white dwarfs accreting mass from another star. “My number was almost three orders of magnitude higher than what was actually observed from that galaxy,” he says.

To be sure, Gilfanov and his graduate student Ákos Bogdán surveyed the X-ray emission data from six nearby galaxies, all of which are known as 'early-type' galaxies and contain very little neutral gas and dust, which can obscure X-ray emission from accreting white dwarfs. They found that, across the six, the measured X-ray flux was 30–50 times lower than would be predicted in the accretion scenario. Thus, they conclude that the vast majority — at least 95% — of supernovae in these early-type galaxies must result from the mergers of binary white dwarfs (see page 924).

Gilfanov and Bogdán spent almost a year combing through the theory and data to account for different modes of accretion and different types of galaxy. “Many times you go to bed thinking, 'I made a great discovery!' and then, in the morning, when you double-check, it disappears,” says Gilfanov, adding that this one did hold up to scrutiny.

The story is not closed, however, because the scenario could be different for late-type galaxies and there might be other ways in which SNIa can be generated. “In computer models, scientists can make white dwarfs explode at below the critical mass,” says Gilfanov. “But, so far, these supernovae look very different from the ones we actually see out there in the Universe.”