The small icy bodies that make up the Kuiper belt are the most distant objects known in the Solar System. A consistent picture is now emerging which suggests that these objects formed much closer to the Sun.
Since the first member of the Kuiper belt was discovered1 in 1992, many unexpected features of their orbits and physical properties have been uncovered. One surprise was the very low total mass observed in the belt — about one-tenth of the mass of the Earth, when it was predicted to be a hundred times larger than that. To explain the missing mass, it has been proposed that collisions between Kuiper-belt objects over the lifetime of the Solar System have gradually transformed most of their mass into dust; bombarded by solar radiation, these dust grains are eventually expelled from the Solar System. But such theories have never been able to satisfactorily explain the extent of the depletion of the original Kuiper-belt mass. On page 419 of this issue, Levison and Morbidelli2 propose an alternative. They show that some objects that now exist in the Kuiper belt might have been pushed there from original positions near Neptune's present orbit: the original Kuiper-belt region could, in fact, have been virtually empty, and only a small amount of mass was subsequently deposited there.
According to current theory, the planets of the Solar System formed from a primordial disk of gas and dust, as the dust accumulated into gradually larger objects. Of course, in going from dust to planets there must have been intermediate stages, such as a disk of fledgling planets, or planetesimals, of roughly asteroid size. In regions where the total mass in the disk was not large enough for the accumulation process to produce planets, this planetesimal disk would have been preserved to the present day. This would certainly be the case for the Kuiper belt: because of its less dense mass distribution and large distance from the Sun, the accretion process in the belt would have ended when object sizes were no bigger than Pluto. However, in contrast to the present estimate for the Kuiper-belt mass, this accretion theory would require there to be around 10 Earth masses in the Kuiper-belt region, to allow the formation of objects as big as those presently seen in the belt.
The paucity of mass in the Kuiper belt is not the only enigma. Lying at such great distances from the rest of the Solar System, and experiencing no great perturbations from other large bodies, Kuiper-belt objects were expected to have orbits that were nearly circular and located close to the average plane of the Solar System. In fact, the orbits are quite eccentric, and are inclined out of the Solar System plane (Fig. 1). Planetary scientists have wondered how these orbits could have been so dynamically excited. My own idea3 is that these objects formed much closer to the Sun and were then propelled outwards by a mechanism involving close gravitational encounters with the outward-migrating, primordial Neptune. Other work4 shows that, if there had originally been a large mass beyond Neptune's present position, this planet would have moved much further out than it is today (effectively, into the Kuiper-belt region).
Thus, the puzzle seemed to be nearly solved. On the one hand, the original planetesimal disk would be truncated near Neptune's present position (Fig. 1), and was massive enough to form the large bodies now observed in the Kuiper belt; on the other hand, some objects would have been transported to the belt from this dense inner planetesimal disk by a mechanism that induced high-inclination orbits through close encounters with Neptune3. However, there was one piece that didn't fit. In addition to the Kuiper-belt objects in high-inclination orbits, there is a roughly equal number of objects at low inclination. These could not have been pushed out by the same mechanism.
Levison and Morbidelli2 propose a solution. Their premise is based on another enigma — the fact that the Kuiper belt is considered to have an outer edge at about 7 × 109 km from the Sun (equivalent to 48 au, or astronomical units). This distance is significant: at this point, the orbital periods of Kuiper-belt objects are twice that of Neptune, a feature known as the '1:2 mean motion resonance'. As Neptune migrated outwards through the primordial Solar System, it pushed out some of the objects in this resonance trap5,6. This mechanism would naturally create orbit eccentricities, causing the resonant objects to approach close to Neptune and the other major planets, such that they would eventually be ejected from the Solar System after a final gravitational encounter with Jupiter.
Levison and Morbidelli argue, however, that the influence of other, so-called secular resonances, inside the 1:2 resonance, would have kept the eccentricities and inclinations of some resonant objects low. A secular resonance is also based on commensurability of periods — not of the periods of the objects' motion around the Sun, but instead of the motion of the orbits themselves around the Sun. Such resonances are powerful, and can induce large variations in orbital eccentricities and inclinations, either raising or lowering them. According to Levison and Morbidelli's simulations, a small fraction of the resonance-trapped objects, once released through the rather jumpy migration of Neptune, would eventually be left in fairly low-inclination orbits in the Kuiper belt, owing to the influence of secular resonances. A few other objects remaining in the 1:2 resonance would set up the outer edge of the Solar System as it is today.
Of course, this new set of ideas raises further questions. The main one is, how could the primordial Solar System be formed in a truncated disk? The few observations made of other planetary systems as they are forming indicate that they have radially expanded disks. Is the Solar System then a rare case? Regardless of the answer, what conditions could cause this truncation of a developing planetary system? This, I believe, will be a major topic in planetary science for years to come.
Luu, J. & Jewitt, D. Nature 362, 730–732 (1993).
Levison, H. F. & Morbidelli, A. Nature 426, 419–421 (2003).
Gomes, R. S. Icarus 161, 404–418 (2003).
Gomes, R. S., Morbidelli, A. & Levison, H. F. Icarus (in the press).
Malhotra, R. Nature 365, 819–821 (1993).
Malhotra, R. Astron. J. 110, 420–429 (1995).
About this article
Space Science Reviews (2015)
EAS Publications Series (2010)
Astronomy & Astrophysics (2005)
The Astrophysical Journal (2004)