The planetary pair—discovered using NASA's Kepler space telescope and announced in the journal Nature last December—are just under Earth's radius. Both orbit a so-called subdwarf B star dubbed KIC 05807616, which sits about 4,000 light-years away.
When sunlike stars run out of hydrogen fuel, they enter a red giant phase, in which their gas envelopes can swell to several hundred times their original size.
Eventually a red giant's gas envelope will slough off entirely, leaving behind a dense stellar corpse known as a white dwarf. Sometimes, however, a red giant will lose its gas envelope prematurely to form a subdwarf B star, like KIC 05807616.
The scientists who discovered the roughly Earth-size planets in Kepler's data had proposed that both worlds were once gas giants, like Jupiter or Saturn, that had been pulled nearer to their star when it ballooned during the red giant phase.
Plowing through the dying star's swollen atmosphere burned away the planets' liquids and gases, that team suggested, leaving behind the two rocky pits that Kepler sees as Earth-size worlds.
But a new study, by astrophysicists Ealeal Bear and Noam Soker of the Israel Institute of Technology, offers an alternate explanation.
It's possible that both worlds actually come from a single gas giant planet at least five times more massive than Jupiter that was stripped naked by the dying star, the researchers say.
The lone planet's rocky core was then ripped apart by the star's gravity into several Earth-size chunks.
Planets' Resonance a Problem?
Bear and Soker developed the new theory because of their concerns over the Earth-size planets' orbital resonance, a gravitational interaction that involves two objects orbiting a third body in a predictable pattern.
The Kepler planets have an orbital resonance that's almost exactly 3:2—that is, one of the planets completes three orbits around the star in the time it takes the other planet to complete two orbits.
The planets' discoverers had suggested at the time that the worlds had already been in a 3:2 resonance before they were engulfed by their star's envelope of hot gas.
But Bear and Soker argue that such a scenario is improbable, because the act of being swallowed by the bloated star would likely have destroyed any existing resonance between the planets.
The engulfment process "is a violent one that proceeds rapidly," Soker said.
(Also see "How Planets Can Survive a Supernova.")
According to the new study, the lone giant planet would have also played a major role in the evolution of its parent star.
As it was consumed, the planet deposited energy into the stellar envelope, which helped strip away the star's gas layers, leaving behind a naked stellar core.
In this scenario, at least two of the pieces of the gas giant's core survived and continued to orbit the star, while the others may have fallen into the star or been ejected out of the system altogether.
More Worlds May Circle Dying Star
Valerie Van Grootel is an astronomer at the University of Liège in Belgium and one of the original discoverers of the Earth-size planets. She said the new idea was an "interesting alternative interpretation" of her team's results.
"Their objection to our explanation is relevant," Van Grootel said. "It's probably difficult to explain that the two planets kept their 3:2 resonance along the whole evolution process."
She added, however, that it's also possible the 3:2 resonance was acquired after the engulfment phase.
Study co-author Soker countered by saying that the presence of a possible third planet in the system—hinted at by the Kepler data—is more easily explained by his theory than by the two-planet model.
Our model "naturally accounts for three—and even more—planets in the system," since those worlds could be more pieces of the original gas giant, he said.
Rory Barnes, an astronomer at the University of Washington in Seattle, called the new hypothesis a "credible alternative" to the two-planet model, but he said that more detailed computer modeling will be needed to determine which scenario is more likely.
"A few more objects like KIC 05807616 would also be welcome" to add more data to the models, Barnes said.
Bear and Soker presented their research at the Planets Around Stellar Remnants conference in Puerto Rico in January and have submitted their study to a scientific journal for review.