Not that long ago, it seemed the glory days of NASA’s New Horizons mission were in the rearview mirror, left behind with its historic Pluto encounter in 2015. Then, early last year, the spacecraft streaked by Arrokoth, a bit of flotsam drifting through the Kuiper Belt—the diffuse ring of primitive icy bodies beyond Neptune, of which Pluto is the largest member. What New Horizons found at Arrokoth—initially reported last year and now reinforced with 10 times more data in three studies published last week in Science—is a critical clue to the greatest cold case in the solar system: the mystery of how planets are born.

“I never expected that our encounter with Arrokoth would be shoulder to shoulder with the Pluto flyby in terms of its importance,” says New Horizons principal investigator and study co-author Alan Stern, a planetary scientist at the Southwest Research Institute. “I didn’t expect to make an earth-shattering discovery about planet formation in the Kuiper Belt, and yet we have. At Arrokoth, we stumbled onto maybe the biggest prize of the entire New Horizons mission.”

Through careful studies of Arrokoth’s shape, geology, color and composition—as well as sophisticated computer simulations—researchers have developed a clearer picture of how this relic from the early solar system must have formed. And with that knowledge, they have also gained a better understanding of how the building blocks of worlds took shape around the sun more than four billion years ago.

How to Make a Planet

The recipe for making planets is deceptively simple: Jostle a massive cloud of gas and dust so that it collapses in on itself like a spherical avalanche, compressing most of its material into a central newborn star. Next, stand back and watch as the cloud’s remnant angular momentum spins and flattens the leftovers into a whirling disk around the star. Within a few million years, it is thought, worlds coalesce within the disk via a process called hierarchical accretion. Dust particles collide and stick, gradually glomming together into pebbles and, eventually, planets. Easy, right?

Except there seems to be a crucial bottleneck in this planetary assembly line: the jump from pebbles to kilometer-scale building blocks called planetesimals. This step is where many theorists expect hierarchical accretion to temporarily break down, because meter-scale boulders knocking together at orbital speeds are more likely to shatter back to gravel than to get larger. Planetesimals, by contrast, should be bulky enough that their intrinsic gravity corrals the fragments produced by collisions, pulling them back into the fold and allowing growth to continue all the way to planethood.

“Gravity is a universal force and acts like a glue to grow planetesimals bigger and bigger once they form,” says David Nesvorný of the Southwest Research Institute, who was a co-author of one of the new studies. “But that's not true about the initial stage, when you just have dust particles in a disk sticking together through molecular forces to make pebbles. Gravity isn’t very important there. So what’s the ‘glue’ that lets things grow to produce 10- or 100-kilometer objects?”

Top-Down or Bottom-Up?

The leading alternative to the “bottom-up” assembly process of hierarchical accretion is a “local cloud collapse” mechanism that would build planetesimals from the “top-down.” In this approach, pebbles in a protoplanetary disk bypass the collisional bottleneck by settling into self-gravitating clouds, rapidly compressing under their own weight to directly collapse into planetesimals. Originating in the 1950s and refined with pioneering theoretical work in the 1970s, the idea initially struggled to explain how the pebbles could clump up in the first place. But 15 years ago, more sophisticated models emerged showing how gas drag within a disk—a phenomenon called the streaming instability—can concentrate pebbles into dense groups, much like flocks of birds or a peloton of cyclists moving together against a headwind.

From there, a pebble cloud will collapse, popping out planetesimals—plural, because the conservation of angular momentum spins out two or more dense, kilometer-scale bodies from the infalling material. Thus, if planetesimals form via collapse, most of them should begin as binary systems—some of which will then either slowly merge together or lose their companions through gravitational interactions. And according to state-of-the-art numerical simulations recently performed by Nesvorný and his colleagues, if their progenitor pebble clouds formed via the streaming instability, these binaries should tend to orbit each other in a prograde direction—that is, in the same direction as they orbit the sun. (Models of binary formation from other mechanisms predict the opposite: a tendency for retrograde orbits.) Remarkably, an analysis of data from the Hubble Space Telescope and other sources has shown that the Kuiper Belt’s oldest binaries exhibit exactly this effect, with the vast majority displaying prograde orbits. When first revealed last year, this overlapping evidence from high-performance supercomputers and telescopic studies of Kuiper Belt objects was hailed by some experts as the best evidence yet for the reality of the streaming instability and local cloud collapse models of planetesimal formation.

“I’m under no illusions that there will be a universal, instantaneous agreement about this,” says Andrew Youdin of the University of Arizona, a co-originator of the streaming instability hypothesis, who helped perform this breakthrough work. “You don’t want everyone to just jump on the bandwagon, anyway. It’s a more gradual thing. That’s the way science should work.”

In light of the data from New Horizons’s Arrokoth flyby, however, the bandwagon may soon be standing room only. “These two things fit together,” says Will Grundy of the Lowell Observatory in Flagstaff, Ariz., a co-author of the three new Arrokoth studies and leader of the Kuiper Belt binary analysis. “The evidence of prograde binary-orbit orientations is perfectly consistent with the streaming instability as the formative mechanism. And all the evidence that Arrokoth gives is that it formed through cloud collapse—although it doesn’t tell us how that cloud formed.”

The Case for Cloud Collapse

Formerly known as 2014 MU69 (or its informal designation Ultima Thule) before its official naming, Arrokoth is a 36-kilometer-long “contact binary,” composed of two icy, flattened, lightly cratered and gently touching lobes. The arrangement gives Arrokoth the appearance of a squashed snowman. Its surface is extremely and uniformly red—probably because of organic molecules that formed over eons of steady pummeling by cosmic radiation. And perhaps most importantly, the contact binary is a member of the “cold classical” family of bodies in the Kuiper Belt—objects in sedate, circular orbits that have scarcely interacted with anything else since their formation more than four billion years ago, at the solar system’s dawn.

“The debate over how planetesimals form has mostly been based on computer models—because every small object in the solar system we’ve gone to for ‘ground truth’ has been heavily heated and eroded by sunlight and impacts,” Stern says. “Then we go to Arrokoth, and it’s clear this thing has been cold as long as it has existed and is in a very rarefied part of the solar system where there has never been an intensive collisional environment. It’s a time capsule from more than four billion years ago, and it cannot be explained, in aggregate, by hierarchical accretion models.”

In every detail, Stern and his colleagues say, Arrokoth fulfills expectations set by cloud collapse models. Its smooth lobes, so delicately perched atop each other, show no signs of the violent high-speed smashups predicted by hierarchical accretion—they must have collided very placidly, drawn together with a closing speed as low as a meter per second as they spiraled through the gas in the embryonic solar system’s natal disk. And the lobes are each flattened in the same way—precisely as if they both spun out from the same collapsing cloud. In color and composition, they appear, everywhere, the same—whereas they should be more varied if formed from smaller objects colliding from across remote parts of the solar system. “This is like a CSI episode,” Stern says. “There are too many lines of evidence all pointing to one perpetrator here, not the other. Everything lines up for cloud collapse.”

That conclusion itself is somewhat surprising. “We knew we’d probably be able to learn something about planetesimal formation from Arrokoth,” says John Spencer of the Southwest Research Institute, a co-author of the three recent Science papers. “But we didn’t expect it to be so blindingly obvious when we got there. None of us imagined, I don’t think, that Arrokoth would be so pristine and that the story it told would be so clear.”

There is, of course, a potential catch: Arrokoth is the only object of its kind ever seen close-up, and making enormous extrapolations from a sample size of one is inherently risky. “I’m confident in this being a major advance in our understanding of planetesimal formation, but someone will probably ask, ‘Well, this is just one object. How can you know it’s typical?’” says William McKinnon of Washington University in St. Louis, who also co-authored the three new studies. “Well, we didn’t pick [Arrokoth] because we knew what it would look like. We picked it because we could reach it with New Horizons. If it had turned out to be a space potato covered with craters, we’d be telling a different story now—but it didn’t.”

More certainty could come from New Horizons as it journeys deeper into the Kuiper Belt. With heat and power for its instruments provided by the gradual decay of long-lasting nuclear isotopes, the mission could continue its explorations well into the 2030s (provided NASA keeps funding its operations). The spacecraft’s 10 kilograms or so of remaining propellant are unlikely to suffice for another post-Pluto flyby of a Kuiper Belt object, but the team is still ardently seeking other possible targets using some of the largest ground-based telescopes on Earth. Meanwhile they are employing New Horizons’s far more modest 21-centimeter telescope to remotely study Kuiper Belt objects passing by in the distance. Such studies will not return gorgeous images. But they could still surpass any observations from Earth’s vicinity, providing measurements of shapes, spins and surface properties for perhaps 50 or 100 additional objects—enough to form a statistically significant sample and, just maybe, to settle the planetesimal debate for good.