The asteroid belt surrounds the inner solar system like a rocky, ring-shaped moat, extending out from the orbit of Mars to that of Jupiter. But there are voids in that moat, most notably where the orbital influence of Jupiter is especially potent; any asteroid unlucky enough to venture into one of those so-called Kirkwood gaps (named for mathematician Daniel Kirkwood) will be perturbed and ejected from the cozy confines of the belt, often winding up on a collision course with one of the inner, rocky planets (such as Earth) or the moon.

But Jupiter's pull cannot account for the extent of the belt's depletion today or for the spotty distribution of asteroids across the belt—unless there was a migration of planets early in the history of the solar system, according to new research.

Study co-authors David Minton and Prof. Renu Malhotra, planetary scientists at the University of Arizona's Lunar and Planetary Laboratory in Tucson, report in Nature today that an orbital migration of Jupiter and Saturn four billion years ago may explain the observed distribution of asteroids.

The researchers designed a computer model of the asteroid belt under the influence of the outer "gas giant" planets, allowing them to test the distribution that would result from changes in the planets' orbits over time. A simulation wherein the orbits remained static, Minton says, did not agree with observational evidence. "There were places," he says, "where there should have been a lot more asteroids than we saw."

On the other hand, a simulation with an early migration of Jupiter inward and Saturn outward, the result of interactions with lingering planetesimals (small bodies) from the creation of the solar system, fit the observed layout of the belt much better. The uneven spacing of asteroids "is readily explained by this planet-migration process that other people have worked on," says Minton, a graduate student. In particular, "if Jupiter had started somewhat farther from the sun and then migrated inward toward its current location," the gaps it carved into the belt would also have inched inward, leaving the belt looking much like it does now.

Joseph Hahn, a specialist in planetary dynamics at the Space Science Institute in Boulder, Colo., says that the new research bolsters the case for early planetary migration. "The good agreement between the simulated and observed asteroid distributions," he says, "is actually quite remarkable." Jack Wisdom, a planetary scientist at the Massachusetts Institute of Technology, says that most in the field buy into the planetary-migration theory in general. "The really interesting question, not addressed in this paper, is the pattern of migration," he says—whether the asteroid belt can be used to rule out one of the competing theories of migratory patterns.

One issue raised by the new study, Hahn says, is the speed at which the planets' orbits changed. Minton and Malhotra's simulation presumes a rather rapid migration of a million or two million years, but "other models of Neptune's early orbital evolution tend to show that migration proceeds much more slowly," over tens of millions of years, Hahn says. "I suspect that follow-up studies of the solar system's early history will also have to reconcile these two very different timescales, which will hopefully lead to greater understanding of the solar system's early evolution."