mollwollfumble exoplanet announcement. This post busts four modern myths about planet formation. The myths are:
1. Jupiter formed further out and migrated inwards until it reached resonance with Saturn, and then Jupiter and Saturn both moved out.
2. “Hot Jupiters” all formed much further out and migrated inwards.
3. The Solar System is a very atypical planetary system.
4. Neptune could only have reached its current position if a large planet was ejected from the Solar System.
You will appreciate that the arguments given here are only brief summaries.
Myth 1. Jupiter formed further out and migrated inwards until it reached resonance with Saturn, and then Jupiter and Saturn both moved out.
Significant outward movement of Jupiter and Saturn violates the law of conservation of energy, and so this myth need be given no more attention than a perpetual motion machine. Another factor pointing the same way is that the asteroid belt is composed of a large number of very distinct asteroid types, and the asteroid belt has been proved to have formed very early in the Solar System. If Jupiter had been closer to the Sun it would have destroyed the substructure of the asteroid belt, as evidenced by the fact that even the gravity of tiny Ceres has a measurable dissipative effect on this substructure.
Myth 2. “Hot Jupiters” all formed much further out and migrated inwards.
A hundred years ago there were many hypotheses about how the Solar System formed. For example, one that was discarded by the 1950s was that the Solar System formed as a close encounter between the Sun and a passing star pulled a streamer of material from the two stars that coalesced into the planets – that was discarded because such close encounters are extremely rare. The final two hypotheses were cold accretion and hot accretion. The Solar System began as a mixture of dust and gas. In the cold accretion system the dust coalesced first into progressively larger grains until gravity became sufficient for them to pull together into planets. In the hot accretion system the gas coalesced first into hot balls that attracted dust grains into them. A study of meteorite material from the asteroid belt showed conclusively that the cold accretion system was correct, and so hot accretion was totally rejected.
But then the Kepler spacecraft and other telescopes discovered massive numbers of “hot Jupiters”, which are planets of about the size of Jupiter that orbit exceptionally close to their parent star. This is in conflict with the theory of cold accretion because in the cold accretion theory, Jupiter-size planets can only form near, and on the cold side of, what is called the “ice line”. The ice line is the orbit at which water freezes into ice. As a patch to the theory, it was proposed that these hot Jupiters all did form much further out beyond the ice line but then migrated inwards due to drag from the gas that then permeated the planetary system. The immediate problem with that patch is that energy loss increases as the orbit gets smaller, so any planet that formed a long way from the star would tend to either move not far enough (small gas drag at large orbital radius) or plunge into the star and be lost (large gas drag at small orbital radius).
This leads into the first of my two graphs.

This graph gives a histogram of raw results from the Kepler spacecraft. The heights of the histograms are set equal for easy comparison. “All candidates” includes both planets confirmed by other astronomical facilities and planets found by Kepler that have not yet been independently confirmed. First note the orbital period. that of Earth is 365 days whereas the peak in the observed distribution of planets is a much shorter 10 days. These planets are indeed very hot.
But also note the orbital period for binary stars. This period is even shorter than that for the planets. At this point in the argument it is important to keep in mind that there is a continuum of celestial objects between stars and planets. Kepler counts brown dwarfs as planets because they have a size practically indistinguishable from that of Jupiter. Obviously, no-one has suggested that binary stars form further out from their parent star than the ice line before migrating inwards. Instead, it’s accepted that binary stars form by hot accretion, by the gravitational self-compression of gas.
The histogram shown above is totally consistent with the hypothesis that most Kepler planets formed by hot accretion rather than cold accretion. The loss of many of these planets with orbital periods below 10 days is due to them being dragged into their parent star by the large gas drag at small orbital radius.
So let’s accept that most Kepler’s planets, unlike most of the planets in the Solar System, formed by hot accretion rather than cold accretion.
There is one further conclusion to be drawn from that. About 50% of binary stars are in highly eccentric orbits. This suggests that perhaps that proportion of the planets that formed by hot accretion also have highly eccentric orbits.
3. The Solar System is a very atypical planetary system.
None of the thousands of planetary systems found by Kepler and other telescopes even remotely resemble our Solar System. This has led directly to the statement that our Solar System is extremely peculiar.
To test this, let’s look closer at how Kepler works. It detects transits, dips in the light curve of a primary star, as a planet or orbiting star passes across its face. The further a planet is from the primary star, the lower the likelihood that it will pass directly between us and that star. Using the simplifying assumption of ignoring partial eclipses, this observational bias is easy to correct for. The result is shown below.

The histograms for both “binary stars” and “all candidate planets” show a sharp cut-off at a period of about 2 years. That’s because Kepler was observing for 6 years, and an object was only accepted as valid if there were three transits observed within that 6 years.
What I want to draw your attention to in the above graph is that the histograms for both “binary stars” and “all candidate planets” are still increasing at the cut-off. this strongly suggests that there are more planets and stars, perhaps far more, with orbital periods greater than 2 years than there are with orbital periods less than 2 years. Which suggests that our own solar system is not at all atypical.
The adjusted histogram for “confirmed planets” tells a completely different story. Why? As I don’t know how the confirmation process works I can’t give a definitive answer, but one thing is immediately clear. The confirmation process has been going on for a much shorter time than the original Kepler observations, so the people doing the confirmation haven’t had time to confirm the planets with longer orbital periods.
4. Neptune could only have reached its current position if a large planet was ejected from the Solar System.
From the very beginning, the theory of cold accretion has had problems. One that I first heard of many decades ago is the following. Cold accretion works well for explaining the origins of Mercury, Venus, Earth, Mars, the asteroid belt and Jupiter. But Saturn is larger than it should be. And the situation for Uranus and particularly Neptune is worse. Simply put, the Solar System hasn’t been around for long enough for Neptune to have grown to its current size at its current position. The obvious patch was that Neptune formed closer to the Sun but was later thrown out into a further orbit by gravitational slingshot off Jupiter. But detailed calculations showed that that was possible if and only if another large planet was completely ejected from the Solar System.
I say that the chance of Jupiter, Uranus, Neptune and lost planet being in exactly the right initial positions for Uranus and Neptune to end up in their current positions is vanishingly small. More likely is that hot accretion aided the formation process, and no planet was ejected. The observation from spacecraft Stardust that material coming in from the Oort Cloud had been subjected to high temperatures during formation supports the contention that hot accretion had some role to play even that far out, but it does not completely rule out cold accretion there.