Planetary science

Archaeology of the asteroid belt

The size of asteroids in the Solar System's main asteroid belt may help constrain one of the least-understood aspects of planet formation — the transition from pebble-sized dust balls to mountain-sized planetesimals.

As residents of our own planet Earth, we can all take an interest in how planets are born. The discovery of planets orbiting stars other than the Sun, and the rich variety of worlds that have come to light over the past two decades, make this question even more fascinating. Unfortunately, our understanding of how planets form is frustratingly incomplete, not least because it is extremely difficult to observe this process in action. Writing in Icarus, Morbidelli et al.1 shed light on the matter by showing that the early stages of planet formation in the Solar System have left a permanent imprint on the size distribution of objects in the main asteroid belt, which lies between the orbits of Mars and Jupiter.

Planet formation is thought to go through a series of growth stages, beginning with micrometre-sized dust grains that are seen in large numbers in protoplanetary (planet-forming) disks of gas and dust orbiting young stars. Laboratory experiments show that gentle collisions between dust grains cause them to stick together through electrostatic forces, forming millimetre- to centimetre-sized aggregates2. Objects that grow larger than about one kilometre — traditionally called planetesimals — have appreciable gravitational fields that hold them together and help them acquire additional material. Low-velocity collisions between planetesimals lead to the formation of Moon-to-Mars-sized 'planetary embryos', and ultimately planets, in a process that is reasonably well understood. However, the transition from pebble-sized dust aggregates to mountain-sized planetesimals is problematic and remains an unresolved issue. This is unfortunate, because all subsequent stages of planet formation depend on it.

Protoplanetary disks are predominantly composed of gas that can frustrate the growth of centimetre- to metre-sized objects in two ways. Gas pressure typically decreases with distance from the central star, and this pressure gradient means that gas orbits the star more slowly than does a solid body. As a result, solid objects experience a headwind, which causes them to drift towards the star. Drift rates increase with size and are greatest for metre-sized bodies, which take as little as 100 years to end up close to the star3, where they quickly evaporate. This short lifetime of metre-sized bodies means there is a very limited window of opportunity for them to grow into planetesimals, for which radial drift is no longer a factor. Worse still, gas in a protoplanetary disk is probably turbulent. Centimetre-to-metre-sized bodies are strongly stirred by turbulent eddies, so they collide with one another at high speeds, causing them to break apart rather than stick together4. For these reasons, it seems unlikely that objects will grow larger than about one metre as a result of the gradual accumulation of dust grains.

Recently, two models have been developed5,6 that may overcome this 'metre-size barrier'. Each enlists turbulence as an aid to growth rather than a hindrance. The details of the two models differ, but in both cases turbulence acts to concentrate sub-metre-sized particles into diffuse, gravitationally bound clumps 104–106 km in diameter. These clumps subsequently shrink owing to their own gravity, forming solid objects 100–1,000 km in size, thereby leaping across the metre-size barrier in a single bound. Significantly, both of the models predict that small planetesimals, in the 1–10-km size range, should be rare or even absent.

The new models5,6 for planetesimal formation are still in their infancy. Ideally, one would like observational data to test their viability. This is where the asteroids come in, because many of these objects are thought to be left over from the time planetesimals and planetary embryos formed. The main asteroid belt has probably gone through four phases of evolution in its history (Fig. 1, overleaf). In stages 1 and 2, planetesimals, and then planetary embryos, formed in the belt as they did elsewhere in the Solar System. In stage 3, the asteroid belt lost more than 99.9% of its mass. Morbidelli et al.1 argue that this material must have been removed dynamically rather than collisionally, probably owing to gravitational perturbations from Jupiter. As a result, the shape of the planetesimal size distribution was preserved during stage 3. The same dynamic process increased the velocities of the planetesimals relative to one another to the high values seen in the asteroid belt today7. Stage 4, encompassing most of the history of the Solar System, involved occasional high-speed collisions between planetesimals, causing catastrophic disruptions, and leading to the distribution of the objects we see in the modern asteroid belt.

Figure 1: Evolutionary stages of the main asteroid belt.

1, Dust grains coalesced into planetesimals, objects of 1–1,000 km in diameter, through an unknown process; 2, planetesimals merged to form Moon-to-Mars-sized planetary embryos; 3, the vast majority of planetesimals and all embryos were dynamically removed from the asteroid belt; 4, the surviving planetesimals underwent occasional high-speed collisions with one another, leading to catastrophic break-up and the formation of many small objects. By studying stages 2–4, and comparing the outcome with the size distribution of present-day asteroids in the main asteroid belt, Morbidelli et al.1 have determined the sizes of planetesimals at the end of stage 1, placing strong constraints on models of how planetesimals form.

Previous work8 has shown that most asteroids larger than 100 km in diameter have survived more or less intact since the end of stage 3, whereas smaller asteroids are likely to be fragments formed in catastrophic collisions between larger bodies. Thus, it seems that, before stage 4, most of the mass in the asteroid belt existed in bodies larger than 100 km in diameter, and that smaller bodies were relatively rare. Because the dynamic depletion of stage 3 was independent of size, bodies larger than 100 km in diameter also dominated the asteroid belt at the end of stage 2.

Morbidelli and colleagues' contribution1 is to provide a detailed model of stage 2 (planetesimals to planetary embryos). They consider a wide variety of possible planetesimal size distributions, on the basis that we don't know enough to say how big planetesimals were when they formed. Combining the results of this model with the size-independent depletion of stage 3, the authors find that the current size distribution of the asteroid belt can be reproduced only if most planetesimals were at least 100 km in size at the end of stage 1, with a size distribution extending up to 1,000 km (the size of the belt's largest object, Ceres), and possibly beyond. This is precisely the size range predicted by the new models5,6 for planetesimal formation. Furthermore, the new study1 suggests that planetesimals formed with a size distribution that is essentially identical to that seen in the asteroid belt today for diameters larger than 100 km.

The rarity of planetesimals smaller than 100 km in diameter at the end of stage 1 seems to rule out the possibility that dust aggregates somehow made it across the metre-size barrier by gradually sweeping up material from their surroundings. Instead, objects must have grown very rapidly from sub-metre-sized pebbles into 100-km-sized bodies, possibly in a single leap.

Morbidelli and colleagues' model is not the last word on the subject of planet formation. In particular, the authors considered the evolution of the asteroid belt in isolation, neglecting processes that may have exchanged material between the asteroid belt and other parts of the Solar System. However, by showing that the size distribution of the asteroids preserves information about the earliest history of the Solar System, the authors have opened a new window on the least-understood aspect of planet formation. This gives us hope that we may yet understand how Earth and its cousins came to be.


  1. 1

    Morbidelli, A., Bottke, W. F., Nesvorný, D. & Levison, H. F. Icarus doi:10.1016/j.icarus.2009.07.011 (2009). | Article |

  2. 2

    Blum, J. & Wurm, G. Annu. Rev. Astron. Astrophys. 46, 21–56 (2008).

    CAS  ADS  Article  Google Scholar 

  3. 3

    Weidenschilling, S. J. Mon. Not. R. Astron. Soc. 180, 57–70 (1977).

    ADS  Article  Google Scholar 

  4. 4

    Brauer, F., Dullemond, C. P. & Henning, T. Astron. Astrophys. 480, 859–877 (2008).

    ADS  Article  Google Scholar 

  5. 5

    Johansen, A. et al. Nature 448, 1022–1025 (2007).

    CAS  ADS  Article  Google Scholar 

  6. 6

    Cuzzi, J. N. et al. Astrophys. J. 687, 1432–1447 (2008).

    ADS  Article  Google Scholar 

  7. 7

    Petit, J.-M., Morbidelli, A. & Chambers, J. Icarus 153, 338–347 (2001).

    ADS  Article  Google Scholar 

  8. 8

    Bottke, W. F. Jr et al. Icarus 175, 111–140 (2005).

    ADS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Chambers, J. Archaeology of the asteroid belt. Nature 460, 963–964 (2009).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Sign up for the Nature Briefing newsletter for a daily update on COVID-19 science.
Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing