≡ Menu

‘Snowball Growth’ and the Centauri Stars

With three groups now looking hard at Alpha Centauri for planets, let’s hope our nearest stars don’t do for us what Gliese 581 has. First we had a habitable planet in Gl 581c, then we didn’t. Then Gl 581d looked a bit promising, and may skirt the outer edges of the habitable zone, although the jury is still out. Gl 581g looked to be the winner, the fabled ‘Goldilocks’ planet, but now the evidence for it seems weak and its existence is called very much into doubt. Gl 581 keeps dealing out winners and then calling them back, a frustrating period for all concerned.

What we’d like to find at Alpha Centauri, then, is something unambiguous. But while we wait for answers, the issue of how planets form in close binary systems like Alpha Centauri is under the microscope. Centauri A and B have a mean separation of 23 AU, closing to within 11.2 AU (think of another star as close to ours as Saturn) and receding up to 35.6 AU (roughly Pluto’s distance). Proxima is much further out at 13,000 AU and is the nearest individual star to the Sun. We know there are feasible planetary orbits around Centauri A and B. The question is how small, rocky worlds might form in this environment in the first place.

Image: Apparent and true orbits of Alpha Centauri. Motion is shown from the A component against the relative orbital motion of the B component. The apparent orbit (thin ellipse) is the shape of the orbit as seen by the observer on Earth. The true orbit is the shape of the orbit viewed perpendicular to the plane of the orbital motion. Credit: Wikimedia Commons.

Growing Planetesimals Out of the Dust

We saw yesterday that the initial phases of planet formation are at issue. In the core accretion model, planetesimals form from a dust disk and, through a series of collisions, eventually form into planetary embryos. Much remains to be learned about the transition from dust to planetesimal, as well as the planetesimal collision stage. A new paper by Ji-Wei Xie (Nanjing University) et al. looks at the transition between planetesimal formation and the beginning of the collisions between them, a period the paper refers to as the ‘snowball phase.’

Where this departs from older work is that many earlier scenarios for growing planetesimals start with the planetesimals in place and go from there. The new paper questions this approach, positing the snowball phase as a transition period when isolated planetesimals exist in a disk where most of the mass is still in the form of dust grains:

We outline a new growth mode, which dominates during this early phase, during which planetesimals experience negligible gravitational or collisional interactions with one another, and grow mainly (or solely) via the accretion of dust or ice that they sweep up−in the manner of a rolling snowball. In this paper, we refer to this dust-fueled planetesimal growth phase as the “snowball” growth phase.

What the paper finds is that in close binaries, snowball growth overcomes a key problem, that high impact velocities prevent kilometer-sized planetesimals from growing into planetary embryos:

Indeed, this dynamically excited environment, which is hostile to mutual planetesimal accretion, is on the contrary favorable to growth by dust sweeping. If efficient enough, snowball growth allows planetesimals to grow large enough, 50-100 km, to be protected from mutually destructive impacts.

Viable Planetary Growth Around Close Binaries

Going from planetesimals to planetary cores, then, becomes a viable process once the objects are large enough (the 50 to 100 kilometers mentioned above) to survive high speed collisions. The ‘snowball’ model allows the planetesimals time to grow through the accretion of ice and dust that is swept up in the manner of a rolling snowball. Snowball growth as defined here begins with the formation of the first planetesimal and ends when planetesimal collisions become the dominant growth mode. The paper derives the snowball growth rate, shows how long it will last, and considers how the planetesimals can grow around single stars and binaries.

All of this winds up looking promising for environments like Alpha Centauri, and again, its scenario is possible because the authors are defining a growth phase where the planetesimals are emerging from the dust disk but are still sparse — not yet in an era of mutual planetesimal annihilation — growing by sweeping up dust in the protoplanetary disk. Snowball growth is a mechanism that, the authors point out, needs more detailed studies to improve its modeling. I suspect that before we get too far along in that process, we’ll have news from our planet hunting teams about the presence (or lack of same) of planets around Centauri A and B.

The paper is Ji-Wei Xie et al., “From Dust To Planetesimal: The Snowball Phase?” accepted by The Astrophysical Journal and available as a preprint.


Comments on this entry are closed.

  • Tarmen October 26, 2010, 11:25

    Good to see a Chinese University , Nanking at the forefront. China is back in the game after 200 yrs. They’re learning fast. Humanity needs them. Maybe it’ll be a Chinese probe that gets to Centauri system first. If surveys show it’s worth a probe.

  • Carl Keller October 26, 2010, 12:38

    Though it is a single-star system, Epsilon Eridani may shed more light on planetary and belt formation. Having a young neighborhood system is bound to be opportunities for many theories, and the formation of new ones.

  • kurt9 October 26, 2010, 12:39

    I think that the distance between A and B varies between 11AU and 23AU create tidal effects that would probably prevent the formation of planets in either system. Most likely A and B have lots of asteroids and comets, but no planets.

  • Bounty October 26, 2010, 14:17

    What would a great amount of comets and asteroids look like in a system like that? Would they ‘bulge’ or resonate in sync with the orbiting star? If so, what would that look like from here?

  • Ron S October 26, 2010, 15:28

    “Where this departs from older work is that many earlier scenarios for growing planetesimals start with the planetesimals in place and go from there. The new paper questions this approach, positing the snowball phase as a transition period when isolated planetesimals exist in a disk where most of the mass is still in the form of dust grains:”

    It is true that in the paper’s abstract the authors say that these earlier phases of planetesimal formation were “either ignored or underestimated in previous studies”, this is really just an unfortunate choice of language. It isn’t that formation of small planetesimals was preemptively excised, but rather that the mechanisms were too poorly understood to be formally modeled. The authors are proposing a new model of planetesimal formation, and making initial attempts to run a simulation employing that model.

    It is also true that extending any model down to very small planetesimals rapidly eats up processing capacity. Earlier studies will of course have had greater constraints on number of planetsimals (and therefore their minimum size) since processing capacity declines as one looks further into the past. Go back far enough and all the researchers had was pencil and paper. This is not the fault of those earlier researchers!

  • Adam October 26, 2010, 16:22

    A problem yet to be convincingly solved is the problem of gas drag in such disks. Larger than dust and the particles feel a head-wind from the gas which spirals them into the Sun on less than 1000 year time scales. This is a major issue being actively researched with a variety of solutions proposed, but none with much consensus. It might be that planets need to form via gravitational instability promoted by magnetic field interactions or forming further out from the Sun and then captured into lower orbits. There’s a lot of theory and not a lot of hard data to compare against.

  • andy October 26, 2010, 16:29

    Supposing terrestrial planets can form in such systems, I wonder if we should expect to find an increased frequency of Mercury-type planets and large Moon-type satellites as a legacy of a more violent giant impacts stage?

  • spaceman October 26, 2010, 17:18

    What is lacking in the lively debate regarding whether or not planet formation can occur in close binaries is a single comprehensive study that looks at this question from the dust aggregation stage all of the way through to the final growth to planetary-sized objects. So, in essence, we have a number of studies each with their own assumptions and input-parameters looking at different stages of the process–which is good, but there is still no thorough study that brings together the latest observational constraints plus theory and modern high-power computing to help answer the question of whether or not small planets can form, via core accretion, in systems like Alpha Centauri A/B. Is there such a thorough study that takes on this question from the dust stage all of the way through to the planet stage?

    Also, Alpha Centauri A/B based on our measurements of their metallicity, should have had plenty of material to form a few large terrestrial planets. How about coming at the question from this angle: if all of those metals and material did not end up as planets, then what else happened to them?

    1) Were the materials kicked out of the system?
    2). Did the materials spiral into the stellar surfaces?
    3). Did some combination of 1 and 2 occur?
    4). Did they coagulate into large boulders and planetesimals?
    5). If the answer to 4 is yes, then surely a dust disk should be detectable around the nearest star due to colliding planetesimals replenishing the dust, no? (I thought the Spitzer telescope was going to look for dust in the Alpha Centauri system, but I have yet to hear any news regarding this alleged observation.)

  • Dave Moore October 26, 2010, 18:24

    When considering the formation history of Alpha Centauri, you may want to consider this recent Sky and Telescope article, which covers the latest theories of our own solar system’s formation history.


    If the hypotheses in the article art true, then our solar system was relatively deprived of inner system material, compared to other systems, and if we don’t find planets around Alpha Centauri, then spaceman’s comments seem especially pertinent.

  • andy October 27, 2010, 13:36

    @spaceman: I’m not even sure there is any simulation of planetary formation that follows the entire process all the way through from the dust stage to the full-formed planetary system, simulating both the inner system and the region beyond the ice-line at the same time. Still very computationally challenging to consistently model all the effects at sufficient resolution.

    @Dave Moore: that does actually bring up an interesting point, that the only system we currently know that looks anything like the inner solar system is the pulsar planetary system at PSR B1257+12. While we’re not yet at the stage that we can detect inner solar system analogues, the low-mass planetary systems that are being found (e.g. HD 10180, HD 40307, CoRoT-7) have super-Earths in very short period orbits, far shorter than those of Mercury.

    As for the issue about the small size of Mars, perhaps this bodes well for the habitability of other planetary systems: if Mars were a more massive planet it could still be habitable at the current time. Maybe that suggests that a significant fraction of planetary systems that formed without gas giants host multiple habitable planets for long periods of time.