Stellar explosions known as type Ia supernovae have proved invaluable to astrophysicists as markers of cosmic distance. Their brightness and consistency in observed properties allow astronomers to use them as "standard candles" to determine distances to objects in the sky. Just a decade ago type Ia supernovae took center stage when researchers harnessed them as evidence that the universe is accelerating in its expansion under the sway of a previously undiscovered influence known as dark energy.

But the mechanisms for forming type Ia supernovae are still not well understood, and a new simulation suggests that a variety of scenarios could lead to these events, which explode with an estimated force of 10 octillion (1028) megatons—more than a billion times as bright as the sun. The standard model holds that a white dwarf—a dense, shrunken remnant of a star that has exhausted its fuel—grows beyond its sustainable mass (approximately 1.4 times the mass of the sun) by accreting matter from a stellar neighbor or by merging with a co-orbiting white dwarf. The result is a runaway nuclear reaction that ignites the previously dim dwarf in a flash of radiation that can be seen billions of light-years across the universe. The use of supernovae as standard candles relies in part on the supposition that a roughly equal amount of fuel is being consumed in each event.

In a paper published in the November 10 Astrophysical Journal Letters, a group of researchers proposes an addition to the family of type Ia progenitors. Rather than a mass transfer or outright merger between a pair of stars locked in an orbital dance, two unbound white dwarfs residing in a dense stellar grouping could collide head-on. The shock of the impact, according to the team's supercomputer simulations, could trigger a thermonuclear reaction that mimics more traditional type Ia supernovae but may burn more fuel than is consumed in the standard 1.4–solar mass event. Some binary-merger scenarios have also been implicated in supersize type Ia explosions, and such variations in supernova starting conditions could complicate their use in future high-precision distance measurements.

Study co-author Enrico Ramirez-Ruiz, an astrophysicist at the University of California, Santa Cruz, points out that in some regions of space, such as in globular clusters, the stellar population density is roughly a million times that of our solar system's relatively sparse environs. "The density of white dwarfs in the cores of globular clusters could be very high," he says, "so the probability of them interacting in a collision is non-negligible."

Ramirez-Ruiz and his co-authors are cautious in their assessment of the contribution of their proposed mechanism to the sum total of observed type Ia supernovae, estimating that colliding white dwarfs likely account for one in a hundred such explosions or fewer. "Nonetheless, it's intriguing," says astronomer Douglas Leonard of San Diego State University, who did not contribute to the study.

"Type Ia supernovae are very exciting objects observationally for cosmology and distance measurements, and yet the embarrassing little secret is that we still don't really know what is exploding," Leonard says. It is widely believed that white dwarfs are at the heart of the events, but the mechanism by which the stellar remnants are ignited remains an open question. "It's a great time to be a theorist," he adds. "If you can figure out some clever way to get them to explode, all bets are still on the table, and you may just turn out to be right."

The more the stellar explosions are studied in detail, he notes, the more diversity they reveal and the more it appears that multiple paths may lead to type Ia supernovae. "There may be many different ways that you can lead a white dwarf to explode, and this one certainly seems reasonable," Leonard says.

Ramirez-Ruiz points out that despite the low estimated abundance rate of collision-induced supernovae, large-scale bulk surveys in the coming years may net hundreds of thousands of type Ia events, including hundreds of the new variety annually. "Once you have such a large number of type Ia's that you're assembling to do cosmology, how standard your model is depends on whether other channels will pollute your sample," he says.