While a collection of large rocks travelling around a star may be a good basis for planet formation, this stony ensemble cannot yet be compared to planets. What is needed is for these planetary bases to acquire a topping of atmosphere.
At the beginning of the assembly process, our planets were dust-sized specks at the mercy of the gas disc. Carried by the gas motion, these solid grains had to avoid a catastrophic meeting with the Sun until they were able to grow big enough to resist the gas drag.
As they now approach the size of planetary embryos, these roles reverse and gas can become trapped by the embryo’s gravitational pull. Once bound to the embryo’s vicinity, the gas forms an envelope around the rocky core to create the planet’s first primitive atmosphere.
Like the solid particles in the disc, the gas has a randomly directed motion in addition to its orbital path around the Sun. The speed of this movement in the gas is governed by the temperature. This is what makes a hot-air balloon inflate; warming the gas increases its speed and causes the molecules to bump harder against the fabric to expand the balloon. When the speed needed to escape the gravity of the planetary embryo exceeds the gas’s random motion, the gas becomes trapped to form an atmosphere.
The acquisition of an atmosphere helps the growing planetary embryo sweep up smaller planetesimals. Such rocks entering the atmosphere feel a drag like the air resistance on a skydiver. As it loses speed, the planetesimal is more easily pulled towards the embryo’s surface and collides to add to its core.
The braking of the incoming planetesimals also releases heat. In the same way that meteors become burning-hot shooting stars as they enter the Earth’s atmosphere, descending planetesimals heat up as they are forced to slow. This energy warms the atmosphere and increases the random motions of the gas. While this boost in speed is not usually enough to allow the atmosphere to escape the planetary embryo, it does stabilise the gas against the embryo’s gravity and stops it from being compressed. When gravity is balanced by the gas heat, the atmosphere neither expands nor contracts. This stable position is referred to as hydrostatic equilibrium.
As the planetary embryo grows through planetesimal accretion, its gravity increases. This temporarily breaks the hydrostatic equilibrium and pulls the atmosphere more tightly to the planet. As it is compressed, the gas heats up and balances the gravity again to make a new stable position. With the embryo’s stronger gravity extending its reach and the gas compressed to make more space, new gas can pour in to deepen the atmosphere.
For a planetary embryo with about a tenth of the Earth’s mass (as expected at the Earth’s current position at this time), the captured atmosphere is always much smaller than the planet’s solid mass. The gaseous topping therefore helps the embryo’s growth, but does not drastically change the planet’s evolution. However, as we step away from the Sun and cross the ice line, the story becomes very different.
The giants of gas
Further away from the Sun’s gravitational pull, the larger cores around Jupiter’s current location can trap much bigger atmospheres – so big, in fact, that the weight of the gas can never be supported by the heat from the incoming planetesimals.
At exactly what point this occurs is debated. A rough rule of thumb suggests everything can stay balanced until the atmosphere reaches the same mass as the embryo’s solid core. However, less mass is needed if the incoming planetesimals partially evaporate in the atmosphere as they descend towards the embryo’s surface, depositing their vaporised material into the gas. The heavier elements that make up the planetesimal ice and rock are coolants, quickly reducing the temperature of the gas, which slows its motion and weakens the support against gravity.
When this critical atmosphere size is reached, a balance point between the gas motion and gravity cannot be found. Instead, the combined mass of the embryo and atmosphere creates a gravitational force that overwhelms the gas motion. Hydrostatic equilibrium is shattered and the atmosphere is steadily compressed.
As the atmosphere packs down close to the planet, the embryo’s gravity is able to pull in fresh gas from the disc. This joins the atmosphere and also begins to compress. The new gas adds to the embryo’s combined mass, increasing its gravitational reach and allowing it to pull still more gas into the atmosphere. This produces another runaway process whereby the planetary embryo’s atmosphere is accrued at a faster and faster rate. The result is a massive atmosphere of thousands of kilometres; a gas giant planet has been born.
There are two ways in which this massive atmosphere accretion can be stopped. In one way the atmosphere can continue to grow until the gas disc disappears. Once the star begins to evaporate the disc, the pool of gas surrounding the planet steadily reduces. Before 10 million years are up, the disc vanishes to leave the planets with whatever atmosphere they have managed to gather.
This method is certainly effective, since a planet cannot accumulate atmosphere if its gas reservoir has gone. It probably played the leading role for the outermost gas giants in our Solar System. Forming so far from the Sun, Uranus and Neptune will have had only a low density of rocks and gas to feed from, making the creation of their planetary embryos a slow process. It is therefore likely that they were still gathering their atmospheres when the Sun evaporated the remains of the gas.
As a small addendum, Uranus and Neptune are in fact so far out, that they probably did not form in exactly their present locations. Due to the length of time it would have taken to produce a planet of their size, the gas disc would have vanished before they could acquire a good atmosphere. It is more likely that they formed closer to Jupiter and Saturn and later moved outwards. Nevertheless, even at their estimated closer location, the evaporation of the gas disc probably terminated their atmosphere growth.
However, the above method is less probable for the inner two giants, Jupiter and Saturn. Our two most massive gas behemoths are thought to have a much larger ratio between their solid cores and huge atmospheres. It is therefore likely that they had plenty of time to acquire gas and that a different mechanism came into play to stop their feeding. This mechanism is suspected to be a gap that formed in the protoplanetary gas disc along each planet’s orbit.
When in orbit, the distance away from the star determines the length of time it takes to complete one circuit. As happens on lanes on a circular running track, material closest to the star has the shortest distance to travel to return to its original position. Gas orbiting between the planet and the star therefore draws ahead of the planet, while gas further out lags behind.
The gravity of the planet pulls on the gas as it moves through the disc. For gas closer to the star that is trying to move ahead, this tug pulls it back and slows it down. Conversely, the outer gas feels a pull to speed it up.
As the gas speed changes, it must alter its orbit so that its circular speed can once again balance the star’s gravity. Gas between the planet and star now has a lower circular speed and is forced to pull away from the planet and move closer towards the star. Meanwhile, the accelerated outer gas is able to move further away. This produces a gap around the planet where the gas density is much lower.
If the gravitational reach of the planet pierces the top and bottom of the protoplanetary disc, this gap can remain. The planet is so big that gas is unable to sneak into the hole without its speed changing and forcing it back out. The gap therefore persists and throttles the gas flow until the gas disc evaporates.
After its atmosphere stops growing, the planet will contract as its current atmosphere continues to cool and collapse. This causes the atmosphere to get denser, whereupon it gets harder to compress and begins to resist further shrinking. Deep inside a gas giant’s atmosphere, the compressed gas reaches high enough pressures to transform its hydrogen gas into an exotic liquid metal. These incredible forces reduce the contraction of Jupiter and Saturn to an extremely slow rate, with current estimates for Jupiter hovering at less than a millimetre a year. This minute shrinkage is still enough to heat the planet, which radiates more energy than it receives from the Sun.
This mechanism for forming a gas giant planet is known as the core accretion model, a term inspired by the image of the runaway gas accreting on to the solid core. One of the most compelling features about it is that the method is very similar to terrestrial planet formation, with the only difference being a runaway atmosphere gathering. However, its least compelling feature is the time it takes.
Initial estimates for how long it would take Jupiter, Saturn, Uranus and Neptune to form in their current orbits are in excess of 10 million years; a problem since the gas disc will be gone within that time. At one point, this was thought to make core accretion impossible, but since then some tweaks to the system have been discovered that can shorten the time needed.
The first adjustment is simply to improve the accuracy of the original model calculations. How fast the gas cools is partially controlled by whether dust grains in the atmosphere clump and sink, or if they remain suspended in the gas. In the latter case, the fog of grains prevents heat from escaping (technically referred to as increasing the atmosphere’s opacity), and slows cooling. Allowing the dust to fall towards the core gives cooling a boost and the atmosphere swiftly collapses into a runaway mode.
A more energetic solution is to move the planet. The planetary tow truck is the same mechanism that eventually causes the gap opening in the disc. While the planet pulls on the gas to attempt to open the gap, the gas in turn is tugging back on the planet. The inner gas tries to pull the planet forwards as the planet drags it back, while the outer gas drags on the planet as it is pulled forwards. If both the inner and outer gas pull equally, the planet will remain unaffected. However, the planet moves slightly faster than the local gas as it is not susceptible to pressure. This causes the dragging gas to be closer to the planet, dominating the inner accelerating forces. The planet therefore slows and gets an inward push.
As the planet moves through the disc, it enters a new population of planetesimals. This fresh supply of food allows its accretion rate to increase again, reducing the time needed to begin runaway collapse by an impressive factor of 10. In this scenario, a planet like Jupiter would begin to form further out at around 8au, then begin a food trawl towards its current position at 5au. Upon the discovery of exoplanets, this concept of planet migration became a key component in formation theories: as both an asset to the formation process and one of its biggest thorns.
More recently, another mechanism has been suggested to accelerate the weight gain of the gas giants. Rather than consuming large planetesimals, pebble accretion theory suggests eating smaller rocks may be more fattening.
Planetary embryo growth slows once the incoming planetesimals can move fast enough to escape the embryo’s gravity. The difficulties begin at the oligarchic growth stage, and become worse for capturing larger planetesimals that are later scattered into the embryo’s neighbourhood.
However, even after larger planetesimals have formed in the disc, there is still a large population of smaller rocks. At around 10cm (4in) across, pebbles are an excellent-sized snack because they are still affected by gas drag. The drag slows the pebbles as they swing by the embryo, making it far easier to divert them into a collision course. Embryos can therefore accrete this size of rock extremely efficiently, gaining mass a hundred times faster at Jupiter’s current position.
In practice, all three of these mechanisms are likely to occur to shorten the time needed to pile on a huge atmosphere. This makes core accretion the preferred option for the formation of the majority of gas giant planets. Yet, some worlds would still challenge its limits.
Building distant planets
Despite core accretion being a serious player for giant planet formation, not everyone was happy. In particular, the further out in the disc you went, the harder it became to form a planet. For small rocky outer worlds such as Pluto, their far-flung positions can be blamed on interactions with the massive planets. As the gas giants balloon in size, their gravitational reach pulls first large planetesimals, then smaller rocky embryos towards their core. Too large to be affected by gas drag, the majority of these objects are moving too fast by the time they reach the gas giant to be captured. Instead, they accelerate past the planet to shoot all over the Solar System.
Pluto was shot outwards with a large collection of dwarf planets and planetesimals to sit beyond Neptune. Other planetesimals were scattered inwards or out of the Solar System entirely. So large was the gravitational tug from the mammoth Jupiter, that the embryos located in the inner Solar System were knocked about and collided to form the terrestrial planets.
This could just about explain our Solar System: planetary embryos formed via successful collisions of dust grains to create planetesimals. The gas giants acquired their huge atmospheres in a runaway process during core accretion, and their massive bulk triggered a game of gravitational ping-pong to complete the inner planets’ growth and push out a ring of dwarf planets and rocks. Then, we discovered the exoplanets.
Fomalhaut b is thought to be a giant planet orbiting its star at a staggering distance of 119au. In comparison with our own Solar System, our outermost gas planet, Neptune, is as a measly 30au from the Sun. At hundreds of au, it is not possible to build up a large enough core to pull in a massive atmosphere, yet Fomalhaut b weighs in with an upper mass estimate three times that of Jupiter. This hit the core accretion model hard, both due to Fomalhaut b’s location in the sparse outer disc, and by tripling the target mass that needed to be reached.
Fomalhaut b was not discovered by either the radial velocity or transit technique. Instead, it was the first exoplanet to be seen directly. Direct imaging of exoplanets is extremely difficult, since their dim radiation (from reflected starlight and heat) is usually overwhelmed by the star. The further away a planet orbits, the better the chance of spotting its faint signature.
The planetary nature of Fomalhaut b has been questioned, since it is surrounded by a large dust cloud. Is a planet buried in the fog, or is this the rubbly ruins from planet-making collisions? Either way, Fomalhaut b was not the last object to be found far out in the disc.
In 2009, the 8.2m (27ft) Japanese telescope Subaru began combing the skies for far-out planets. The survey was dubbed the ‘Strategic Explorations of Exoplanets and Disks with Subaru’, or SEEDS for short. The plan was to directly image discs around stars and any visible gas giant planets. By 2016, SEEDS had found four planets significantly larger than Jupiter and orbiting between 29au and 55au away from their star. These outer planets might not be numerous, but they were here to stay.
With the current model for core accretion pushed back beyond its limits, scientists hunted for an alternative gas giant factory. The proposal was to more closely mimic the way stars are formed.
Images of disc galaxies similar to our own Milky Way display a dazzling collection of spiral arms. The spiral is usually a compression wave, which is of the same type as sound. These spiral arms can appear when the gas’s own gravity is strong enough to break apart the smooth structure of the disc.
Such an effect is known as a disc instability: a rather fancy term for saying that gravity breaks the disc apart. In the resultant spiral arm, natal gas clouds are gathered together to create more of the dense pockets where stars are born.
The idea behind the second form of gas giant formation is that the protoplanetary disc could behave in the same way. Spiral arms would develop in the gas disc encircling the star, and these would compress the gas, which would then collapse directly into a giant planet. Unlike the much lower densities commonly found in the natal star-forming cloud, densities in the protoplanetary disc could potentially rise high enough to form a small, planet-sized object.
It is hard not to be enticed by such a prospect, since it neatly avoids every other problem we have been trying to deal with. With no need to first build up a solid core, sticking mechanisms and the gas drag on planetesimals can be negated. The formation time for the gas giant can now be as little as a thousand years, a fraction of that needed for the core accretion model and well within the lifetime of the gaseous protoplanetary disc. Moreover, it should be quite possible to create planets of 1–10 Jupiter masses, encompassing Fomalhaut b and other exoplanet mega-worlds.
The problem (and there is always a problem) is that it is questionable whether a protoplanetary disc can form instabilities in the same way as a Galaxy. There are two main factors that decide if a disc can become unstable: mass and temperature. If the disc is too light, its gravity is not sufficient to upset the smooth distribution and form a spiral. Conversely, if the temperature is too hot, the random motion of the gas can swiftly smooth out the compression wave before it can form a planet. It is also not certain if planets formed in this manner can survive. Closely forming planets may merge into bigger objects or they may shred each other apart.
Models of a protoplanetary disc encircling a star like our own Sun suggest that anything inside 40au is very unlikely to become unstable. However, in the disc’s younger and heavier days, it could have fragmented beyond 100au: a figure that fits well with Fomalhaut b’s location. For the planets found by SEEDS, their location puts them on the cusp between where core accretion and disc instability seem to be possible. Their formation mechanism is therefore perhaps understood, but the jury is out as to which method we should apply.
The gaseous planet formed via a disc instability initially has no solid core. It can acquire one by capturing planetesimals that slowly sink to the planet’s centre. While our own gas giants are too close to the Sun to have formed via disc instabilities, the core for a Jupiter-sized planet that was formed via this mechanism is around 6 Earth masses; within the guess range for our own Jupiter’s solid centre.
So between the two theories of core accretion and disc instability, which is correct or could both be at work? The only real reason to reject the idea that both mechanisms occur is aesthetics: it is just ugly to have two different gas giant formation methods. Yet, neither model alone can explain the gas giants in our own Solar System and match those around other stars. A compromise might be that the two methods complement and help one another: disc instabilities compress gas in spiral waves, which in the right circumstances can collapse into a planet. Where collapse does not happen, any disc instabilities may still assist core accretion, as the gathered gas boosts the rate at which planets can gather atmosphere around a solid core.
At this point, our planets are becoming recognisable as the worlds of our Solar System. The four outer worlds formed the fastest, gathering in planetesimals and smaller embryos from an ever-increasing gravitational reach. With this extra bulk, huge atmospheres poured down to drench the rock and ice cores in gas. In the inner Solar System, the stronger pull from the Sun restricted the gravitational influence of the planets, causing formation to proceed more sedately. The gas giants’ gravity then ruffled the orbits of the inner planetary embryos to begin a final set of collisions. Out of these impacts, four terrestrial planets would emerge, coated with thin atmospheres. However, not a single one of these planets is yet capable of supporting life.