How we discovered Planet Nine

In 1820,  Alexis Bouvard, director of the Observatoire de Paris, noticed something funny about Uranus. Although the planet had been discovered only 40 years earlier, Bouvard tracked down 130 years of serendipitous observations of Uranus — times when some astronomer charted a star that turned out to have been the planet instead — and used them to prove that Uranus traveled in a nearly circular orbit around the Sun. Uranus was not, then, one of those pesky comets that kept appearing in the sky. Rather, it was a true distant planet, like the six others known before it. 

But there was a problem. The old recordings of the position of Uranus were slightly off. Bouvard concluded that either the old observations were in error, or yet another unseen planet was tugging Uranus — just slightly — from where it should have been. In the decades that followed, Uranus continued to stray from its path. No more excuses could be found. Something was out there.

Then, using Bouvard’s data, French mathematician Urbain Le Verrier predicted the position of Neptune. And German astronomer Johann Galle sighted the new planet on only a single night of searching. This was arguably the most spectacular success of classical celestial mechanics. With the acclaim Le Verrier gained for the discovery, it’s not surprising that astronomers quickly jumped on the idea that the new eighth planet might point the way to a ninth. 

By 1910, many astronomers had made conflicting predictions for a ninth planet, and Edward C. Pickering and Percival Lowell separately led unsuccessful searches for what was variously called Planet O, Planet P, or Planet X. The predictions and searches for a giant planet beyond Neptune continued all the way until 1993, when Jet Propulsion Laboratory scientist Myles Standish published an update on the expected positions of everything in the solar system using the latest data obtained from careful measurements of the position of the Voyager spacecraft. Standish showed conclusively that all the planets were exactly where they were supposed to be. More than a century of searching for Planet X had been in vain. Nothing was out there.

But that wasn’t really the case. At nearly the exact moment that Planet X was being put to rest, astronomers found the first new object beyond Neptune since the accidental discovery of Pluto (during a search for Planet X). As discoveries mounted, planetary scientists quickly realized that this population of objects in what we now call the Kuiper Belt is vast. To many of us who had begun to study this newest known collection of objects in the solar system, another thing became obvious: There was no chance that Pluto was going to be the only large object in the Kuiper Belt. 

By 1998, I began a large-scale survey of the sky from Palomar Observatory with the clear goal of detecting these largest objects, but also with the hope that perhaps — just perhaps — there was something new and big still to be found in whatever the next region was beyond the Kuiper Belt.

Another five years later, astronomers Chad Trujillo and David Rabinowitz joined me in the search, and we quickly discovered an incredibly distant object. It was so far away, in fact, that we were hopeful it was the elusive new planet I had begun seeking five years earlier. It wasn’t Planet X, of course, because we knew that nothing out there was big enough and close enough to perturb the orbit of Neptune, but this new object — designated 2003 VB12 — was the most distant object ever seen by humans in orbit around the Sun. We thought that perhaps it was a tiny, twice-Pluto-size ice ball in a distant circular orbit.

A new planet! Or at least what we would have called a planet back then. Only it wasn’t. Continued observations showed that the new object — now called Sedna — is only 40 percent the size of Pluto and is not on a circular orbit at all, but on an extremely elongated orbit that takes it on a 10,000-year journey around the Sun. Most importantly, the closest approach of Sedna’s elongated orbit to the Sun — the orbital perihelion — is not within the Kuiper Belt at all, like it is for every single other Kuiper Belt object with an elongated orbit. Instead, it appeared to have been pulled out to 76 astronomical units (AU) — nearly twice the average distance of Pluto. (One AU is the average Earth-Sun distance, about 92.5 million miles.) Something massive was, or had been, tugging on the orbit of Sedna. 

In the scientific paper announcing the existence of Sedna, we laid out the imagined possibilities: Maybe Sedna was pushed around early in the history of the solar system by stars that were nearby; maybe a single unusually close stellar encounter knocked Sedna into place more recently; or perhaps a yet unseen Earth-sized planet in a circular orbit near Sedna’s perihelion distance was flinging objects onto distant orbits. None of these explanations turned out to be correct. Instead, Sedna was our first good clue of something else entirely: a much more massive, much more distant planet shaping the orbits of all of the most distant elongated Kuiper Belt objects. 

Sedna was the first good clue, but it was not the first actual clue. Three years earlier, one of the first large Kuiper Belt surveys detected an object called 2000 CR105. This object’s orbit, too, is distant and elongated and appears to have been pulled — or pushed — away from Neptune, though only by a small amount. It was such a small amount, in fact, that some astronomers argued that the mildly unusual orbit was caused simply by long-term interactions with Neptune. Nonetheless, Brett Gladman, now at the University of British Columbia, suggested that perhaps 2000 CR105 had been pulled away from Neptune by an approximately Mars-sized planet on the fringes of the Kuiper Belt. Interestingly, in what was thought to be a simple coincidence until just this year, 2000 CR105 and Sedna were both being pulled in the same direction. 

In the 15 years since these early discoveries, thousands of new objects have been discovered in the Kuiper Belt, including two with elongated orbits that appear to have had their perihelia, their closest approach to the Sun, pulled slightly away from the Kuiper Belt. Coin­cidentally — or not — both of these were tugged in the same direction as 2000 CR105 and Sedna.

In 2012, Trujillo, now at Gemini Observatory, and Scott Sheppard, at the Carnegie Institution, found the first new truly Sedna-like object, 2012 VP113. The fact that it, too, was pulled in the same direction as all of the rest was again overlooked as mere coincidence. Trujillo and Sheppard did notice what they thought were some unusual orbital alignments, though, and, as was now common, suggested that a yet unseen planet — perhaps twice the mass of Earth and lying at a distance of 200 AU — was the natural explanation. 

At this point, Konstantin Batygin, a newly hired assistant professor at Caltech, and I became interested. We were pretty sure that the Trujillo and Sheppard suggestion wouldn’t work and quickly convinced ourselves that we were right (results that were also reached by Meg Schwamb, working in Taiwan). No planet could cause the alignments that Trujillo and Sheppard thought they were seeing.

Looking carefully at the data, however, we became intrigued by these distant objects whose perihelia had been tugged away from the Kuiper Belt and by the fact that they were all coincidentally aligned. We quickly calculated that the probability of such an alignment occurring just due to chance would be only about 1 percent, a small, but not overwhelmingly small, chance. I recall saying to Batygin, “This is interesting, but we really need one more object to be aligned to make it statistically convincing.”

In a seemingly unrelated analysis, Rodney Gomes in Brazil noticed the existence of an unusually large number of objects with distant orbits whose closest approach to the Sun had been pushed inward even closer than the orbit of Saturn and whose orbits were twisted such that they were nearly perpendicular to the disk of the solar system.

No one really had any viable suggestion for the origin of these peculiar objects. But Gomes had an interesting hypothesis: Perhaps a distant, massive planet was twisting these orbits perpendicularly and pushing their closest approaches inward.

For anyone paying close attention to the outer solar system, it was 1820 all over again. Bouvard simply had Uranus to guide his way. This time there were multiple lines of evidence suggesting that perhaps an undiscovered planet was out there. Just as in 1820, the size, the orbit, and the mass were all completely unknown, but the hints that something was out there seemed to be coming from all directions. 

Meanwhile, Batygin and I saw the new object we had been waiting for when, in 2014, astronomers reported the discovery of 2013 RF98, whose orbit is distant, elongated, and aligned nearly precisely like the rest. All of the most distant orbits were aligned. All of the distant orbits whose perihelia had been pulled out of the Kuiper Belt were aligned in the same direction. This time we calculated that the probability that this alignment was just due to coincidence was down to 0.007 percent. The signs in the sky were clear: Something was out there.

Batygin and I got to work. With months of pen-and-paper calculations, and then more months of detailed computer simulations, we realized that everything we were seeing could be explained by a planet a little less massive than Neptune on an eccentric orbit that takes it from around 200 AU at its perihelion out to 1,200 AU at its aphelion — its farthest point from the Sun — over an approximately 20,000-year orbital period. Such a planet would capture Kuiper Belt objects with distant elongated orbits into stable orbits elongated in the opposite direction from the planet. 

Moreover, it would pull the perihelia of these Kuiper Belt objects away from the Kuiper Belt. And, in a result we did not expect, it would alter these elongated orbits, twisting them to be perpendicular to the plane of the planets and driving their perihelia inside of Saturn’s orbit. We didn’t know off the tops of our heads if any such strange orbits existed because, embarrassingly, we had overlooked Gomes’ paper, which had come out just as we were in the intensive last phases of our analysis. But when we saw the objects pointed out by Gomes, we grew excited. Our planet theory not only predicts objects like that, but it also predicts exactly how those orbits should be aligned. We quickly plotted the locations of these distant, twisted orbits, and, to our astonishment, the orbits were precisely where we predicted them to be. 

For us, this moment marked moving from working on an interesting hypothesis about some unusual orbital alignments to instantly realizing that we were talking about something that was really out there. This was something waiting to be found, something that both explained the old observations we were working on and also crystallized correct predictions about things we were completely unaware of. We like to think of this as the day that Planet Nine was born.

With the many effects that Planet Nine is having on the outer solar system, we can infer many things about its properties. In practice, because the solar system is a complicated place, understanding these properties has involved massive amounts of computer simulation. We simulate a slightly larger planet, a slightly closer planet, a slightly more inclined planet, and each time we compare the results of our simulations with observations of the solar system that we know. 

From these constraints we have determined that Planet Nine is about 10 times the mass of Earth, that its orbit is inclined by approximately 30 degrees to the plane of the planets, that it has an average distance of something like 600 AU from the Sun, and that when it is at its most distant point from the Sun, it lies toward the outstretched arm of the constellation Orion.

All of this relatively detailed knowledge might make it seem like we could, like Le Verrier, simply say to the world, “Go look; it will be THERE!” But we can’t. Le Verrier had the advantage of being able to analyze the full orbit of Uranus around the Sun to see its deviations. If we waited 10,000 years to fully track Sedna around its orbit, we, too, would be able to pinpoint Planet Nine. 

Instead, though, we have only a snapshot of the orbits of a variety of different objects, and we must infer what should have happened in the past. In prac­tical terms, that means that although we know the orbital path of Planet Nine through the sky, we don’t know where it is in its orbit. We no longer have to search the entire sky to find Planet Nine, but there’s still a lot of work to do. 

The search will not be as hard as it might have been, however, as many sky surveys over the past few years have covered large swaths of the sky and might have detected Planet Nine had it been in their region. We know, for example, that when Planet Nine is at its perihelion, it is as bright as 18th magnitude, lying in the southern sky near the constellation Ophi­uchus. Such an object would have been detected years earlier. Most likely, Planet Nine is now closer to its aphelion, where it would glow dimly, likely close to 25th magnitude. 

While that is very faint, detecting such an object is well within the capabilities of the 8-meter Subaru Tele­scope on Mauna Kea and its impressive Hyper Suprime-Cam, a mosaic of 112 CCD cameras covering nearly two square degrees of sky with every exposure. We have already begun our search using this telescope. Other astronomers are likely to follow. 

Is Planet Nine really out there? It’s always wise to be skeptical, but still, we are quite convinced that the answer is yes. Something must be responsible for all the unusual orbits that we now see in the outer solar system. Planet Nine is by far the most likely explanation. 

So if it is really out there, when will we find it? The world has been alerted, and multiple teams are on the hunt. Perhaps during the next five years, someone, at some telescope somewhere, will spot a faint blip in the sky that moves to a slightly different spot the next night. When they first see it, they will gasp. Then they’ll recheck all the data and gasp again. They’ll scramble to beg and borrow a few hours on big telescopes here and there to confirm the blip’s slow march across the sky. Finally, after checking and double checking and checking 10 more times, they’ll make a dramatic announcement to a now-anticipating world: Planet Nine is found; Planet Nine is real!