Artist's impression of Haumea, a dwarf planet in the Kuiper belt, which is flattened like a football that has been "a bit deflated and stepped on".
Credit: Harvard-Smithsonian CFA
BEYOND THE ORBIT of Neptune lies a shadowy zone of frozen 'worldlets' orbiting so slowly that few of them, if any, have made complete circuits of the Sun since Captain James Cook sailed into Botany Bay.
It's called the Kuiper Belt, named for astronomer Gerard Kuiper who, 50 years ago, proposed it as the source of the icy moons of Jupiter, Saturn, Uranus, and Neptune.
Not that he believed anything was still there. In the billions of years since the Solar System formed, he presumed gravitational encounters had caused all of these bodies either to fall inward and be captured as moons or be completely ejected from the distant Sun's feeble hold.
But the Kuiper Belt still exists. In an explosion of discoveries in the past decade, astronomers have identified about 1,000 Kuiper Belt objects, and the number is mounting.
The most famous is Pluto, reclassified as a Kuiper Belt object after being dethroned as a planet.
But there are several others large enough that any might once have been dubbed Planet X, had astronomers stumbled across them decades ago, in a then vain search for a tenth planet.
The largest is Eris (formerly known as Xena), named for the Greek goddess of strife and discord. Eris orbits nearly three times farther out than Pluto and, at around 2,300 km in diameter, is close enough to Pluto's size that there has been dispute over which is bigger.
Behind Eris and Pluto in size, the list includes Haumea (oval-shaped like a rugby ball, 2,000 km long), Sedna (another football, 1,600 km long), Makemake (a 1,500 km sphere), Pluto's moon Charon (1,200 km), Orcus (950 km), Quaoar (890 km), Ixion (around 800 km), and many others.
These planetoids intrigue astronomers because, like asteroids, they're 'fossil remnants' of the early Solar System - leftovers from the planet-forming process.
Finding more has become a priority for many astronomers. One is Matthew Holman, an astrophysicist at the Smithsonian Astrophysical Observatory at Harvard University in Boston, whose team is using a new telescope called Pan-STARRS-1, atop a 3,000 m volcano in Hawaii.
With an aperture of 1.8 m, Pan-STARRS-1 isn't huge. What makes it unique is that it has an extremely wide field of view, capable of examining a swath of sky almost 40 times larger than the full Moon in a single image. It also has four 1.4 billion-pixel digital cameras, the largest ever built.
All of this technology was designed to scan for potentially dangerous Earth-impacting asteroids. But it's also ideal for spotting Kuiper Belt objects down to 300-500 km in size.
In the first few months of operation in 2010, Pan-STARRS-1 found 10 new Kuiper Belt objects above its minimum size, Holman reported in January 2010 at a meeting of the American Astronomical Society in Seattle, Washington. "[And] this represents just the tip of the iceberg," he said. "I think we'll find several hundred to 1,000."
Another telescope, called the Large Synoptic Survey Telescope, under development in Chile, should be able to push the size limit down to 50-100 km once it becomes operational in 2020.
"We would expect about 30,000 objects," says one of the project's leaders, astronomer Lynne Jones, of the University of Washington in Seattle.
It's not just a matter of counting objects. Astronomers want to plot their orbits. Not surprisingly, many Kuiper Belt objects lie in the same plane as the rest of the Solar System, about five to seven billion kilometres from the Sun. These comprise what is called the Cold Classical Kuiper Belt - 'cold' not because it's more frigid than the rest, but because the worldlets' ordered orbits are dynamically calm.
But others zip high out of the main plane, like satellites orbiting above the Earth's poles. That's important information, says Jones, because it allows scientists to test models of planet formation and the evolution of the Solar System.
"There's currently a lot of debate about what happened in the early history of the Solar System," she says. "We used to think things were relatively quiet, but we can see that this is not the case. They are dynamically much more excited than we would expect."
Figuring out what kicked so many Kuiper Belt objects out of the plane of the Solar System isn't the only problem that's facing Solar System theorists. Kuiper's original prediction notwithstanding, modern scientists have struggled to explain why there's anything in the Kuiper Belt at all.
"Classical planet-formation models have a hard time growing big objects at this distance from the Sun," says Alex Parker, a graduate student in planetary science at the University of Victoria in British Columbia, Canada, who presented his thesis work in October 2010 in Pasadena, California, at a meeting of the American Astronomical Society's Division for Planetary Science.
"Unless you start with a lot more mass in the primordial disc and then get rid of it later, these models can take longer than the age of the Solar System to grow the kinds of objects we see."
These problems led theorists to propose that today's Kuiper Belt objects must have formed closer to the Sun, possibly in the same zone as Jupiter, Saturn, Uranus, and Neptune. Once formed, they were flung outward by dramatic shifts in the orbits of the outer planets, most importantly Neptune.
The theory, known as the Nice model (because it was developed, in part, by scientists from Nice, France) became a leading explanation for the Kuiper Belt. "In some of these simulations, they even suggest that Uranus and Neptune switched places," Parker said.
There's just one problem. At the October 2010 meeting, Parker announced that he had found at least one class of Kuiper Belt objects for which this couldn't have occurred. These objects are binary pairs, such as Pluto and its moon Charon.
Such pairs are common in the Kuiper Belt - estimates are that at least 30% of objects in the Cold Classical Kuiper Belt exist in pairs or larger groupings. But Parker found seven such pairs that are so widely separated they circle each other in a slow waltz, with orbital periods from four to 17 years. "They're delicate," Parker said.
Any gravitational encounter with a giant planet would have ripped them apart and sent them hurtling in different directions. "They would not be there today if [they] were ever hassled by Neptune," Parker said.
Nice model theorist Alessandro Morbidelli, of the Laboratoire Cassiopee in Nice, France, agrees. "The Parker result calls for at least some tweak in the Nice model," he says. "Either the cold Kuiper Belt (or part of it) formed in situ, or it was pushed to its current location by mechanisms less violent than envisioned."
That doesn't overturn the Nice model, as the details of how Kuiper Belt objects might have been pushed outward depends on unknowns in the evolution of Neptune's orbit. "Many variants are possible," he says.
Still, it's a problem. "We have to come up with a way to tweak the Nice model," says Stephen Tegler, a planetary scientist at Northern Arizona University.
Enter Mike Brown, a cheery astronomer from the California Institute of Technology who looks a bit like a young Bill Gates. Brown's Twitter handle is 'plutokiller' because of his role in kicking Pluto off the list of planets and into the Kuiper Belt. But he's also co-discoverer of Eris, Sedna, Makemake, Haumea, and nearly a dozen others.
Like seemingly everyone studying the Kuiper Belt, Brown (pictured below) too has come up with surprises to upset the theoreticians' apple carts. In his case, the problems are these worldlets' densities. Traditional theories of planet formation say that planets condensed bit-by-bit out of a primordial solar disc, or nebula, of dust and gas that circled the Sun like the brim of a sombrero.
By the time you went as far out as the Kuiper Belt, that disc should have been rather tenuous, meaning that 1,000 km objects couldn't have been popping up all over the place.
"To get something that large, you would have had to accrete from a very large swath of the outer Solar System," Brown says. This should have averaged out any random variations in the nebula's composition. "You would think they would be some of the most uniformly comprised objects in the Solar System."
Pluto fits this model perfectly. Its density is about 2.0 grams per cubic centimetre (g/cm3), indicating that it's composed of a fifty-fifty mix of rock and ice, just about what theoreticians would expect.
But it turns out that not everything is like Pluto. The first clue came from Haumea. It's not merely elongated, like Sedna. It's also flattened, Brown says, not just like a football, but like one "that's a bit deflated and stepped on". It's also spinning end-over-end, so fast that it completes a full revolution every four hours. Even more interestingly, it turns out to have a density of about 3.0 g/cm3. "That's 100% rock," Brown says.
As far back as 2006, Brown suggested that Haumea (then known as 2003 EL61) might be the core of a larger planetoid that got clobbered, sometime in its early history. A glancing blow, he suggested, not only set it spinning, but knocked off its outer icy mantle, creating two tiny moons that are almost pure ice.
In a 15 March 2007 study in Nature, he reported additional evidence of this collision: the discovery of at least five other fragments from it - a group of Kuiper Belt objects whose orbits indicate they were blown away from Haumea near the dawn of the Solar System, 4.6 billion years ago.
The fragments (the largest of which might be 400 km across) appear to be made of nearly pure ice - just right to be more of Haumea's missing mantle.
One such anomalously dense world could be the result of a fluke collision.
But as more and more numbers come in, the densities of other large Kuiper Belt objects appear to be all over the map, ranging from well below 1 g/cm3 (the density of frothy ice) to nearly as dense as Haumea. In particular, Eris and Quaoar are unexpectedly dense.
Brown suggests that these objects didn't form by gradual accretion. Instead, he thinks, they originated from large chunks, several hundred kilometres in diameter - large enough to have already differentiated into rocky cores and icy mantles. These collided stepwise in big collisions, creating what he calls 'pyramidal growth' that in only a few steps produced the largest bodies we see.
Each collision would have blasted some debris into space - the source of today's smaller Kuiper Belt objects (975 m across being the smallest found in visible light so far), such as those associated with Haumea, plus many smaller fragments yet to be found. Sometimes, two ice-depleted objects would have collided, producing a larger ice-depleted object.
Other times, icy chunks of one-time mantle would merge into larger, rock-depleted worlds. Still other collisions would produce in-between mixes of rock and ice. But this only works if the main growth process came from collisions between large objects.
That way, a large world could be grown in a small number of steps - small enough that the law of averages wouldn't turn them all back into equal mixes of rock and ice, like Pluto. It's an interesting idea, but it begs the question of where the first round of objects came from.
Brown's theory only works if the dominant collisions were between large objects, not dust bunnies or tennis balls, or even things the size of Earth's dino-killing asteroid (about 10 km in diameter).
More support for this comes from Parker's loosely bound binaries, which would long ago have been knocked out of each other's orbits if they had been pelted early on in the accretion stage by kilometre-sized objects.
This means the Kuiper Belt had to have somehow produced a lot of large objects early on, before the collision cascade began. The mid-sized ones we see today are the original survivors. Small ones are chips from big collisions. And big ones are the result of Brown's collision cascade.
But how could that have happened?
The answer, Brown and Parker believe, lies in emerging models of planet formation. Like their predecessors, these models start with a dusty solar disc. But instead of focussing solely on stepwise accretion, these models include the effect of eddies, vortexes, and other turbulent flows in the early disc. Brown compares these turbulences to brooms sweeping dust particles together.
That makes for a short-cut process of planet formation, he says, by creating dense zones in which 'big hunks' fall together without need for intermediates. "The first things [to be formed] are 100, 200, and 300 km [in size], instead of millimetres," he says.
These new models, Brown adds, also solve another long-standing problem, in which the traditional accretion models haven't actually been able to show how dust bunnies merged into objects large enough to form kilometre-wide bodies.
It's easy for the models to make dust bunnies, he says. And it's easy to model what happens in an infant solar system once protoplanets reach kilometre size.
But there's a no-man's-land in between, he says, in which objects tend to lose orbital speed and quickly spiral into the Sun due to drag from the solar disc's gas cloud. That problem has been the elephant-in-the-living-room of planet-formation models.
"It's been a mystery nobody wanted to think too hard about," Brown says. But it's not a problem with the new turbulence-based models needed to produce the impactors that kickstart Brown's pyramidal-growth cascade.
There are still problems, though. The Kuiper Belt is a gigantic zone, covering far more volume than all of the interior portions of the Solar System combined. That raises the question of how these giant collisions even happened, if the objects formed in situ.
The Haumea collision clearly occurred after Haumea reached its present location - the ability to trace fragments from it clearly shows that.
But if big collisions formed other large Kuiper Belt objects, nobody has yet found additional fragments, Morbidelli says.
And that means the Nice model remains intact, suggesting that these collisions occurred before Neptune flung the resulting bodies into their present orbits, in the process scattering the fragments.
Thus, Morbidelli believes that despite the tweaks that are required to accommodate Parker's findings, Brown's collision cascade actually lends the Nice model "substantial support".
Testing all of this will take a lot of work. But in July 2015 we'll also get a trove of data when the New Horizons spacecraft gives us our first close look at two Kuiper Belt objects: the no-longer-planet Pluto and its moon Charon.
Brown thinks this will fill yet another gap in our understanding by allowing astronomers to tally the number and size of craters, answering questions about the relative frequency of large and small objects in the Kuiper Belt.
Pluto itself probably won't be all that useful for this because it has a tenuous nitrogen/methane atmosphere that produces frost deposits probably thick enough to have wiped out evidence of old craters. But Charon is too small for even a tenuous atmosphere.
"Charon presumably has an ancient surface that's not really been much perturbed for the past 4.6 billion years," Brown says.
"So you can see the cratering record. I think that when you combine that with some of the more intensive studies we'll do between now and then, we'll know the basic outlines [of Kuiper Belt history]."
And that, he says, could have broad implications, including shedding more light on our own planet's origins. "We could well be completely rewriting how planets form," he says.
Rick Lovett is a science journalist based in Portland, Oregon, who frequently writes for COSMOS.
