
This post could also be titled: How blogging and social media help make science happen.
Here is the story. Planet Nine was conjectured in late January 2016. In February 2016 I started to think about where Planet Nine may have come from. This led to 3 blog posts, each exploring a different origins story for Planet Nine (see here, here and here).
Two of the origins stories were pretty standard, proposing that Planet Nine formed closer to the Sun and was kicked out. But one was a little different: it proposed that Planet Nine could be an extra-solar planet that formed around another star and was captured by the Sun. Here is an cartoon of the idea:

After the blog post was published I discovered that I wasn’t the only one with the idea. The blog post sparked a discussion that kick-started a collaboration between myself and two colleagues in Lund, Sweden: Alex Mustill and Melvyn Davies.
Fast-forward 6 weeks and we have written a scientific paper exploring the idea. Our paper is entitled Is there an exoplanet in the Solar System? We show that Planet Nine may indeed be an exoplanet captured by the young Sun. Here is the chain of events that makes this idea work:
- Within its birth cluster of a few thousand stars, the Sun passed relatively close to another star. The encounter in question must have been at a distance of a couple hundred Astronomical Units (1 Astronomical Unit = the Earth-Sun distance). A closer encounter would disturb the orbits of comets beyond Neptune (the “cold classical” part of the Kuiper belt).
- The other star in the encounter must have had a Neptune-like planet on an orbit wider than any of the planets in the Solar System. The planet’s orbit was wider than about 100 Astronomical Units (for scale, the real Neptune’s orbital distance is 30 AU).
- The Sun captured the wide-orbit planet onto an orbit like Planet Nine’s.
Let’s go through these events one at a time to see how likely they are to have happened.
First, within their stellar nurseries, almost all Sun-like stars have an encounter with another star at the right distance. Looking good so far.
Second, what are the odds that the encounter star hosts a planet like Neptune on a very wide orbit? Naively this would seem unlikely. However, we know that, around other stars, Neptunes are very common, at least on orbits similar to Jupiter and Saturn. In fact, about half of all stars are thought to have (at least) one Neptune, making Neptunes 5-10 times more common than Jupiters.
We also know that most (at least 75% of) systems with Jupiter-like planets go unstable (to explain their elliptical orbits; see this blog post). So, we can guess that most systems of Neptunes probably also become unstable. When a system of planets becomes unstable, the orbits of planets cross and the planets gravitationally scatter each other until one planet is completely thrown out into the Galaxy (see this animation). This happens quickly when Jupiters scatter each other, but is slow for Neptunes (as Neptune is 20 times less massive than Jupiter). For Neptunes the instability phase lasts for ten to hundreds of millions of years. For most of that time at least one planet is in the process of being kicked out and is hundreds of Astronomical Units away from its star. That is right where it needs to be for the Sun to capture it.
Third, how often does the Sun actually capture a planet during a close encounter? This depends on details like how massive the other star is, how fast the stars are moving, how close they come to each other, and the exact orbit of the planet. Here is what a capture looks like:

In our computer simulations the Sun captured a planet from the encounter star about 5 to 50% of the time, depending on the setup. But most of the time the captured planet’s orbit was not compatible with Planet Nine’s. The captured planet ended up an orbit like Planet Nine’s less than 10% of the the time.
When we put all of these probabilities together, we find that the likelihood that Planet Nine was captured from another star is at most about two percent. The probability that the Sun captured a planet at all is up to 15%, but getting just the right orbit for Planet Nine doesn’t happen every time.
Let’s get philosophical: what does it mean for there to be a two percent probability of capturing Planet Nine? First of all, remember that we’re not even sure that Planet Nine is there. However, with the discovery of another distant object in the Solar System its existence is looking more and more likely (see here). Now, what does it mean for our computer simulations to match reality 2% of the time? Most importantly, we need to remember that the Solar System exists! Something really did happen, and whatever that was may not have been a very probable event. For instance: the odds of any one team winning the World Series in a given year are only about 3%, but one team wins every year. And the odds of a given player hitting a home run in a particular situation are very small, often less than 1% (depending on the player and the situation), but it happens in most games. And the odds of a given person being elected president are almost zero, but someone is elected every time. Low probability events happen, and evaluating probabilities after the fact does not tell us all that much. What our study showed is that is it possible that Planet Nine was captured from another star. All of the physics checks out. Planet Nine could really be an extra-solar planet. (Planet Nine from Outer Space!)
So how do we tell if this is really what happened? We are now working to figure out how to tell apart different origins stories for Planet Nine (see here, and here for some other ideas). We think that there are clues in the orbits of outer Solar System bodies whose trajectories are affected by Planet Nine. This is the same class of body that hinted at Planet Nine’s existence, and as more of these objects are found we hope that we will be able to see the fingerprints not only of Planet Nine’s existence but also of its origins.
You can download our scientific paper here.
Questions? Comments? Words of wisdom?
If small objects were also captured from the other star could they be distinguished from the natives?
Good question! The short answer is: probably not. It’s not impossible to do this but it depends on knowing the orbital history of the planets, whose gravity shaped the orbits of the small bodies. And although there are some pretty good ideas out there, we still come up short (for now).