Yesterday’s article on supernovae ‘triggers’ for star and planet formation shed some light on how a shock wave moving through a cloud of gas and dust could not only cause the collapse and contraction of a proto-star but also impart angular momentum to an infant solar system. Today’s essay focuses on a somewhat later phase of system formation. Specifically, how is it that gas giants like Jupiter and Saturn can form in the first place, given core accretion models that have ‘trigger’ problems of their own?
Here’s the issue: To create a gas giant, you need plenty of hydrogen and helium, material in which a solar nebula would be rich. But we’re learning a lot about how planetary systems evolve, and the emerging reality is that the gas disks from which planets are made usually last a comparatively brief time, somewhere on the order of one to ten million years. That would imply that the gas giants had to accumulate their atmospheres within this timeframe.
But how? Jupiter’s atmosphere is massive enough that it requires a large solid core. Forming first, that ice and rock object, itself of planetary size, then causes the gravitational inflow of gas and dust. So we’re asking that a core perhaps ten times the size of Earth form in no more than than a few million years. Hal Levison (SwRI), lead author of a new study on this issue, calls this “the timescale problem,” as you can see in this news release from his parent institution. It’s a problem, says Levison, that “has been sticking in our throats for some time.”
Because if you look at rocky worlds like the Earth, the current thinking is that it needed at least 30 million years to form, and that’s a bare minimum — the number could reach 100 million years. During this period, small objects gradually interact, banging into each other to create larger rocks, and so on in a process that leads to planetesimals and ultimately to terrestrial-class worlds. So how do we get the gas giant’s core to form quickly enough to enable gas accumulation sufficient to produce the observed thick atmospheres of Jupiter and Saturn?
Image: This artist’s concept of a young star system shows gas giants forming first, while the gas nebula is present. Southwest Research Institute scientists used computer simulations to nail down how Jupiter and Saturn evolved in our own solar system. These new calculations show that the cores of gas giants likely formed by gradually accumulating a population of planetary pebbles – icy objects about 30 centimeters in diameter. Credit: NASA/JPL-Caltech.
The Levison paper, co-authored with Martin Duncan (Queen’s University, Ontario) and Katherine Kretke (SwRI) makes the case that a 10 million year timeframe for a gas giant’s core is sufficient if the infant planet accumulates small planetary ‘pebbles,’ here explained as objects of ice and dust about 30 centimeters in diameter. Objects in this size range quickly spiral onto a protoplanet when sufficient gas is present. The rapidly accumulating gas acts as the snare to gather the core materials, which are concentrated by drag and gravitationally collapse.
Levison and colleagues believe that this ‘aerodynamic drag and collapse’ model can produce cores of the needed size in timeframes as short as a few thousand years. Moreover, properly tuned, the method produces a Solar System not so different from what we see. That ‘tuning’ involves assuming pebble formation timed just right so that gravitational interactions among the growing planetesimals cause the larger of them to scatter the smaller out of the disk, slowing down their further growth. Ironically, we need pebbles that aren’t in too much of a hurry:
“If the pebbles form too quickly, pebble accretion would lead to the formation of hundreds of icy Earths,” said Kretke. “The growing cores need some time to fling their competitors away from the pebbles, effectively starving them. This is why only a couple of gas giants formed.”
We wind up with a system that forms between one and four gas giants some 5 to 15 AU from the Sun, which isn’t a bad match at all for our own Solar System, with its two gas giants and the ice giants Uranus and Neptune. This is a computer simulation that, unlike numerous earlier attempts at modeling the core accretion model, does produce gas giants in the timeframe and configuration needed. The formation of gas giants early in a system’s history, then, remains consistent with the basic model, with a short period of core formation no longer a deal breaker.
The paper is Levison, Kretke and Duncan, “Growing the gas-giant planets by the gradual accumulation of pebbles,” Nature 524 (20 August 2015) pp. 322-324 (abstract).
It is my understanding that to get migrating planets like “hot Jupiters” the migration has to be in the early system formation when sufficient materials were still in the disk to create the necessary drag. This model seems to help that explanation, if the gas giant forms while there is sufficient free material in the disk. OTOH, wouldn’t very quick core formation also potentially result in the core migrating to the inner system before the gas envelope can accrete?
Does this model also work for terrestrial worlds, or is the ice the critical component that makes the 2 formation models different?
The supernova paper was interesting, but mainly it reminded me of a “conversation” I had with a creationist (Dwayne Gish ?) back in the late 1970’s in Toronto. he had “explained” to a church meeting I had been dragged along how the complexity of the Bombadier beetle proved God’s design and disproved evolution. We ended up with me introducing the idea of the supernova triggering the formation of the solar system billions of years ago (certainly more than 10,000 years) to which he replied “But who created the supernova explosion?”. I felt he was a very slippery fellow who seemed more intent on bamboozling the believers than finding “truth”. But I was young and naive at the time.
I doubt the gas giants migrate inwards following the dust but more likely the hydrogen and helium still forming the sun, dust tends to resist inward migration as do the planets as they are denser. I suspect that once the stars heat and light become strong enough it blows the gas back out which then causes a migration pattern back out as the planets follow the mass of material. However if there is not enough mass of gas left to draw the planets outwards then they can be stuck in tight orbits. I feel that it is just down to the luck of the dice role or mass distribution to begin with, I don’t believe we will find a set pattern of planet distributions, after all the gas clouds are quite knotty structures. It is good to at last have a more descriptive theory of the gas giant formations as there was a few bits missing to the puzzle. Juno may answer the question of their internal structures more definitively as well with the gravity science module and will reach Jupiter in July 2016.
http://www.nasa.gov/mission_pages/juno/spacecraft/index.html
I wonder how this would “impact” the formation of gas giants around the components of widely spread binaries.
Is the “fine-tuning” of the pebble formation based on anything other than the values required for the simulation to produce solar systems that resemble our own, or is it a variable we could reasonably expect to vary between differing proto-systems? In other words, is it likely that there are there systems out there that did form hundreds of icy Earth-sized worlds (most of which would have beeb subsequently ejected out of the system as rogue planets I take it) rather than gas giants?
Which raises another speculative origin for 10+ Earth mass bodies appearing in the proto-systems during the short gas giant building phase… Could proto-systems possibly capture rogue planets, failed gas giants ejected from other systems, which then have a head start gravitationally speaking when it comes to bulking up and vacuuming up the necessary hydrogen and helium to get a second chance at the Big Leagues? If so, what would we observe examining such a system? Weird retrograde or hugely eccentric orbits compared to other “homegrown” planets? Or would the disc’s spin and gravitational interactions between planets over millions or billions of years smooth that sort of thing out?
As soon as the cores start to form, they will induce turbulence in the gas, right? The turbulence will tend to move small particles in random directions, partially defeating the collapse of particles co-moving with a static disk of gas. Was turbulence properly taken into account?
@Dan H August 20, 2015 at 18:57
‘In other words, is it likely that there are there systems out there that did form hundreds of icy Earth-sized worlds (most of which would have beeb subsequently ejected out of the system as rogue planets I take it) rather than gas giants?’
I find the stars that don’t have gas giant most intriguing as they could harbour more earth massed worlds but they are just very hard to see. Kepler is giving us a statically evaluation of planets but skewed towards larger worlds as it is not that good at finding smaller worlds.
‘Could proto-systems possibly capture rogue planets, failed gas giants ejected from other systems, which then have a head start gravitationally speaking when it comes to bulking up and vacuuming up the necessary hydrogen and helium to get a second chance at the Big Leagues?’
I doubt probability wise there will be an exchange of planets between systems as the velocities and angles must be right, but a planet, I like to call them ‘flanets’, a flung out planet, thrown into deep space of a star forming nebulae perhaps could feed on gas and dust forming Brown dwarfs or Red dwarfs.
‘If so, what would we observe examining such a system? Weird retrograde or hugely eccentric orbits compared to other “homegrown” planets? Or would the disc’s spin and gravitational interactions between planets over millions or billions of years smooth that sort of thing out?’
I would think if they was a capture event it would be very eccentric and it could be equally retrograde.
If Kepler were looking at our system, even at the best of edge on orientation, it would see nothing. The fact, then, that Kepler sees so many planets in other systems appears to indicate that our system may be relatively planet-poor: After all, the presence of hot planets near a star does not mean that there aren’t other, unseen planets more like ours, quite possibly in greater numbers.
Like others here, I am intrigued by the idea of “hundreds of icy Earths”. “Icy Earths” in the HZ will presumably have lots of water, which will escape over time at wildly different rates. We could then have many planets with all sorts of different amounts of water on them.
I think many exciting finds lie ahead once we have an instrument that can actually reliably detect Earth-sized planets.
Two things intrigue me here. Firstly, the mention of ‘hundreds of smaller planets would be formed’ seems to fit with earlier thoughts on our systems early formation; thoughts that may have changed however and that I’m unaware of. Interactions between these smaller bodies led to ejection of some, injection of others into the early Sun and the remainder coalescing into the planets we have today. Is this idea (oversimplified by me) now on shaky ground?
Secondly, yesterdays mention of ‘R-T fingers’ got me thinking when the article said…
“Angular momentum imparted by the shock indentations, they argue, would allow the disk of gas and dust to form around the Sun rather than being pushed directly into it. In their models, without the necessary spin produced by the shock front, the disk materials would simply disappear into the Sun.”
At the risk of sounding foolish, wouldn’t this be ideal for gas giant production? My line-of-thought goes like this… if the R-T fingers prevent material collapsing in short order onto the early proto-Sun we have a temporary disc. If two areas in the disc that are to become Jupiter and Saturn are seeded by instabilities and start to grow, they will each be much smaller proto-discs in their own right. But, without smaller R-T fingers associated with them, wouldn’t those discs just collapse quickly… the exact opposite of what we need to form the solar-disc as mentioned yesterday? Would that be a mechanism for gas-giant growth spurts within the alloted time? If these smaller fingers are present then we would have a non collapsing disc which would lead to starvation of the seed… no gas giant forms before the disc material is dispersed.
Given the range of variables at work here I’m also of a mind that a ‘cookbook’ for planetary formation won’t ever be able to be written… the apocryphal tale of Heisenberg’s deathbed quote rings in my ears, namely, “Why quantum mechanics and why turbulence?”.