(Exo)Planet Nine strikes back!

There is evidence that an additional planet may be lurking in the outer reaches of our Solar System.  Astronomers call it Planet Nine (or sometimes ‘Planet X’), since the Sun’s previous ninth planet, Pluto, was demoted to the status of ‘minor planet’ by the International Astronomical Union back in 2006. If it exists, we should be able to detect Planet Nine within the next few years.  The trouble is that it’s hard to understand how a planet ever got out there.

The evidence for Planet Nine comes from the distribution of Extreme Trans-Neptunian Objects, or ETNOs.  These objects have orbits that are distant (out past Neptune’s) and stretched-out, and most importantly, their closest approaches to the Sun are so wide that they never cross the orbits of any of the planets. The elliptical orbits of ETNOs tend to align, making their closest approaches fall on the same side of the Sun.  This surprising alignment was noticed in 2014 by Chad Trujillo and Scott Sheppard after the discovery of just six of these objects, and they proposed the potential existence of a planet that was shepherding these objects and maintaining their orbital alignment. In 2016, Konstantin Batygin and Mike Brown did an in-depth analysis of the orbital dynamics to show how this shepherding works, and used it to predict the orbit of Planet Nine (but not its exact position along the orbit). 

Over the last decade, more ETNOs have been discovered.  Some astronomers have argued that the alignment that is seen is an artifact of how surveys are done, whereas others remain convinced that the alignment is real.  In a recent study, Batygin and colleagues showed that a separate population of outer Solar System objects – those with orbits very close to the plane of the planets’ but that cross Neptune’s orbit – are more reliably matched if Planet Nine exists.  Additional studies have further constrained the expected properties of Planet Nine; the current “best-fit” is a 4-5 Earth-mass planet on a modestly stretched-out orbit of about 300 astronomical units in size that is only slightly misaligned with the plane of the planets. That means it would take about 5,000 years to complete an orbit around the Sun, which would provide little heat and just look like a bright star in the sky.

The problem is that, even if the evidence for Planet Nine is relatively strong, no one can explain how a planet could have got there.  After Batygin and Brown’s bold prediction, there was a rush of studies attempting to explain Planet Nine’s origins.  (I myself jumped on this bandwagon and discussed three different origins scenarios on this very blog.) The main origins stories were as follows. Maybe Planet Nine formed locally from a massive disk around the young Sun. Maybe it was kicked out of the inner Solar System but captured by a dense ring of planetesimals or gas. Or maybe Planet Nine is an exoplanet that was captured by the Sun from another star.

None of these scenarios is plausible.  We have some (admittedly indirect) evidence that the Sun’s planet-forming disk was much smaller than Planet Nine’s orbit.  That means that Planet Nine could not have formed locally, nor could there have been a ring of planetesimals or gas at that location.  It’s not impossible that Planet Nine could have been captured from another star, but the odds are very low.  Pushing all of the factors in Planet Nine’s favor, the probability of a successful capture is at best about 2%.  Frowny face.

Enter my good friends and collaborators Andre Izidoro and Nate Kaib.  Together, we came up with a new idea, and ended up collaborating with scientific giants Alessandro Morbidelli and Andrea Isella. Our paper was just published in Nature Astronomy.

Our idea was to take a step back and look at times in the Solar System’s early history when planet-sized objects were being tossed around like juggling balls. There are two such phases: the growth of the ice giants, and the giant planet instability. 

The best idea for the growth of the ice giants is that they formed from a population of ice-rich planetary cores with masses of a few Earth-masses. These cores started off way out beyond Jupiter and Saturn, which grew faster and carved a gap in the Sun’s planet-forming disk.  The icy cores interacted gravitationally with the gas in the disk, which caused their orbits to shrink in a process called migration.  The icy cores migrated inward but ran into a barrier: the gas giants. These cores collided with each other and grew into the ice giants, but 4-5 cores got too close to Jupiter and Saturn and were gravitationally kicked outward and ejected from the Solar System.  In simulations of this process, there are often one or two additional, miniature ice giants at the end (but this is not a problem, as we shall see).

The giant planet instability was a seismic event in early Solar System history.  After the gas disk dissipated, the gas giants and ice giants were in a more compact orbital configuration, likely a chain of orbital resonances.  There was almost certainly one or two extra, mini-ice giants. There was also an outer disk of comet-like planetesimals containing a total of perhaps 20 Earth-masses. There is some debate about exactly how the instability was triggered – it may have been the dispersal of the Sun’s gaseous disk that did it, or another effect.  Regardless of the trigger, the instability involved the giant planets’ orbits spreading out, the ejection of one or two ice giants, and the clearing out of 99% of the outer disk of comets.  It looked something like this:

The missing ingredient in this story is the Sun’s birth cluster. The ice giants’ growth and the giant planet instability have been simulated before, but no one had accounted for the fact that the Sun was not an isolated star when these events took place.  The Sun formed in a cluster with a few thousand other stars that lasted for 10-100 million years before dissipating and releasing the stars into the wilds of the Galaxy. 

What would the effect of the Sun’s birth cluster be on the evolution of ejected planets?  That is what we set out to test, using a code designed to simulate the growth of the ice giants and the giant planet instability, and also include the effects of a star cluster.  These simulations (run by Andre Izidoro) are pretty amazing – here is one nice example of the giant planet instability, simultaneously showing the evolution of the cluster, a zoomed-in view of the Sun (top right), and a mega-zoomed in view of the planets close the Sun:

That wide-orbit planet at the end of the simulation is a very good analog to Planet Nine!  How did the planet actually get there?  During the giant planet instability, the planet was scattered out by Jupiter onto a wide, eccentric orbit.  As Jupiter repeatedly scattered the planet, its orbit became wider and more stretched-out, but it still crossed Jupiter’s path.  Then, a passing star gave the planet a gravitational kick, which changed the shape of its trajectory so that it no longer crossed Jupiter’s orbit.  This stabilized the scattered planet’s orbit (although it did remain vulnerable to perturbations from other passing stars).

We found a 5-10% probability that any given scattered planet will be trapped on a wide orbit close to that of Planet Nine.  (The exact number depends on how many stars were in the Sun’s birth cluster, how dense it was, and how long it lived). We think that 3-5 icy cores were ejected during the ice giants’ formation, and another one or two during the giant planet instability.  That amounts to a ~40% chance of trapping a planet on a Planet Nine-like orbit!  This is how Planet Nine got where it is today – IF it really exists, of course.

If it exists, Planet Nine should be found in the next few years.  The brand-new Vera C. Rubin telescope, in Chile, is something astronomy has never seen before.  Its mirror is 8 meters across, putting it on par with the biggest telescopes on the ground.  But what makes it special is that it plans to snap an image of the whole visible sky every couple of days.  To accomplish this, it has a super-wide field of view, the biggest camera ever built, and incredible machinery that allows it to move across the sky faster than any other big telescope.  The Vera Rubin telescope will dramatically improve our understanding of how things change in space – including variable stars, supernovae, and objects that move in the sky such as asteroids, comets, and interstellar objects.

The Vera Rubin telescope will start operations later this year (2025). It’s been estimated that the telescope has a higher than 60 percent chance of detecting Planet Nine if it exists.  It’s disappointing that the odds are not 100%, but if Planet Nine’s orbit is more stretched-out than currently estimated, and it happens to be on the more distant part of its orbit at the moment, or if it is a little smaller or colder than expected, then it would be extremely challenging to detect.  Nonetheless, Vera Rubin should find a plethora of new Extreme Trans-Neptunian objects – enough to confidently claim that Planet Nine must exist, or to rule it out.

Planet Nine is not the only wide-orbit planet in town.  A couple dozen exoplanets have been detected on wide orbits around their stars.  Two stand out as having orbits that are similar to Planet Nine’s – that is, hundreds of astronomical units in size and stretched-out: HD 106906 b and BD+60 1417.  As you can see from the image below, the origins story that was imagined for HD 106906 b’s orbit is pretty similar to our own scenario.

We ran suites of simulations to test the efficiency of trapping exoplanets on wide orbits.  Giant exoplanets have much more violent histories than our own Jupiter and Saturn.  We know this because giant (Jupiter-mass) exoplanets tend to have much more stretched-out orbits than our own giant planets.  These stretched-out orbits are essentially scars from their violent pasts (see here), dynamical instabilities in which one or more gas giants were typically launched out of their home planetary systems (to become free-floating planets).  These instabilities go like this:

In addition to systems containing gas giant exoplanets, we also tested other configurations, such as planets orbiting close binary stars – like Tatooine from Star Wars.  We also tested systems containing only ice giants rather than gas giants. 

It seems that the Solar System is the “sweet spot” for ending up with a wide-orbit planet.  Simulations of giant exoplanets and planets in Tatooine binaries had much lower capture rates of scattered planets.  This is because the ejection in these systems was so fast that there were rarely any stars passing by at the right time, to provide a gravitational kick from the outside.  Likewise, systems with only ice giant-mass planets also had a very low rate of capture on wide orbits.  The reason is because the ejection process in these systems was too slow: the star cluster itself had dissipated before planets were on wide enough orbits to be captured.

Taking the demographics of stars and exoplanets into account, we predict that at least one in every thousand stars has captured a wide-orbit planet like Planet Nine. More and more should be found in the coming years, using a range of different techniques (in addition to exoplanets in their stars’ Oort clouds).

With the Vera Rubin telescope about to start its survey, Planet Nine is at the edge – its discovery would be of huge scientific importance, but it also may never have existed, and simply been the shadow of incomplete data among Trans-Neptunian objects.  Either way, it’s an exciting time for planet-hunting, both in the Solar System and around other stars! 

---

Additional Information

• Our new paper: Very wide-orbit planets from dynamical instabilities during the stellar birth cluster phase
• The Solar System’s story — a series of blog posts explaining current thinking about how the Solar System formed and where it’s going.
• The Vera Rubin Telescope