Indiana University Research & Creative Activity


Volume XXVII Number 1
Fall 2004

<< Table of Contents

Richard Durisen
Richard Durisen
Photo © Tyagan Miller

simulated image of protoplanetary disk
Planet formation: This image illustrates a protoplanetary disk undergoing spiral wave instabilities as it would appear using a millimeter wave telescope.
Image created by John Rosheck

Rings Around the Stars

by Hal Kibbey

Peering into the depths of space, an astrophysicist witnesses the birth of a planet. In the middle of an enormous disk of gas swirling around a star, a huge planet looms into view, many times larger than Jupiter . . . . Wait a minute, the scientist thinks. That can't be right. The gas is supposedly all at the same temperature, and that's not realistic. Let's try again, this time with temperature differences in the gas. Peering into the depths of space, an astrophysicist witnesses clumps of dense gas appearing and disappearing in a disk. None of the clumps lasts even one orbit around a star, and none becomes anywhere near the size of a planet . . . . Something's missing.


For theorists trying to simulate how planets form around young stars, this kind of experience is familiar. The turbulent gases and swirling disks are mathematical constructions using hydrodynamics and computer graphics, and the "telescopes" through which they peer are computer monitors displaying the results of their calculations as colorful animations.

Richard Durisen, professor of astronomy and chair of that department at Indiana University Bloomington, is a leader in the use of computers to model planet formation. Seated at his computer, he demonstrates one simulation after another.

A green disk of gas swirls around a central star. Eventually, spiral arms of yellow begin to appear within the disk, indicating regions where the gas is becoming denser. Then a few blobs of red appear, at first just hints but then gradually more stable. These red regions are even denser, showing where masses of gas are accumulating that might later become planets.

The models are mesmerizing. It's easy to imagine looking down from a vantage point in interstellar space and watching the process actually happen. But there is still a lot of work to be done.

"For many years I've been doing three-dimensional hydrodynamics of astrophysical objects, supported by NASA's Origins of Solar Systems program," Durisen says. "Since the early 1990s, computer power has been increasing rapidly, and our techniques have become good enough that we can make numerical computer models of severely distorted, rapidly rotating fluid objects--and that's what protoplanetary disks are."

These are the disks on Durisen's monitor. "They're the disks of gas and dust that astronomers see around most young stars, from which planets form," Durisen explains. "They're mostly gas, with dust sprinkled in, and they're like a giant whirlpool swirling around the star in orbit. Our own solar system formed out of such a disk."

Scientists now know of more than 100 planets around other stars, and almost all of them are at least as massive as Jupiter. "Gas giant planets are more common than we could have guessed even 10 years ago," Durisen says. "Nature is pretty good at making these planets."

The key to understanding how planets are made is a phenomenon called gravitational instabilities, according to Durisen. Scientists have long thought that if gas disks are massive enough and cold enough, these instabilities happen, allowing the disk's gravity to overwhelm gas pressure and cause parts of the disk to pull together and form dense structures, which could become planets. Large-scale computer simulations are helping Durisen and others make significant advances in testing this idea.

One of the challenges they face is a recent discovery by astronomers that giant gas planets such as Jupiter form fairly quickly by astronomical standards. They have to--otherwise the gas they need will be gone.

"Astronomers know that disks around young stars tend to be massive, and those disks tend to go away over a period of a few million years," Durisen says. "So that's the chance to make gas-rich planets. Jupiter and Saturn and the planets that are common around other stars are all gas giants, and those planets have to be made during this few-million-year window when there is still a substantial amount of disk around."

This need for speed causes problems for any theory with a leisurely approach to forming planets, such as the core accretion theory that was the standard model until recently.

"In the core accretion theory, the formation of gas giant planets gets started by a process similar to the way planets such as Earth accumulate," Durisen explains. "Solid objects hit each other and stick together and grow in size. If a solid object grows to be about 10 times the mass of Earth, and there's also gas around, it becomes massive enough to grab onto a lot of the gas by gravity. Once that happens, you get rapid growth of a gas giant planet."

The trouble is, it takes a long time to form a solid core that way--anywhere from about 10 million to 100 million years. The theory may work for Jupiter and Saturn, but not for dozens of planets around other stars. Many of these other planets have several times the mass of Jupiter, and it's very hard to make such enormous planets by core accretion. And observations have shown that many protoplanetary disks don't last--in some cases they're gone after barely a million years. Finally, new data from spacecraft and new models of Jupiter's interior are making scientists wonder whether Jupiter ever had a solid core.

The theory that gravitational instabilities by themselves can form gas giant planets was first proposed more than 50 years ago. It's recently been revived because of problems with the core accretion theory. The idea that vast masses of gas suddenly collapse by gravity to form a dense object, perhaps in just a few orbits, certainly fits the time limitation, but it has some problems of its own.

According to the gravitational instability theory, spiral arms form in a protoplanetary disk and then break up into clumps that are in different orbits. These clumps survive and grow larger until planets form around them. Durisen sees these clumps in his simulations--but they don't last long.

"Under some extreme conditions, we see a spiral arm break up into small pieces, but they fly around and shear out and re-form and are destroyed over and over again," he says. "If the gravitational instabilities are strong enough, a spiral arm will break into clumps. The question is, what happens to them?"

A gravitationally unstable disk is a violent environment. Interactions with other disk material and other clumps can throw a potential planet into the central star or tear it apart completely. If planets are to form in an unstable disk, they need a more protected environment, and Durisen thinks he may have found one.

As his simulations run, rings of matter form in the disk at an edge of the unstable region and grow more dense. If solid particles accumulating in a ring quickly migrate to the middle of the ring, it could speed up the core accretion process and form the core of a planet much sooner.

"There's an awful lot to be worked out, but it's another way to get a planet going," Durisen says. "It's a secondary consequence of the gravitational instabilities, so it marries the two main theories of planet formation. Nobody else has seen rings grow steadily in one place in their simulations."

One possibility is that the rings act as "particle collectors"--stable places where rock and ice can settle to form the core of a planet. As it turns out, gravitational instabilities are a very effective way to move mass around.

"In our solar system, planets that orbit closer to the sun move faster than the planets farther out. In protoplanetary disks, the same thing happens," Durisen explains. "The disk is not rotating the way a DVD rotates in a player, as a solid object. It rotates as a liquid, where the inner part is swirling faster than the outer part. That means that when gravity is strong enough to pull together parts of the disk into dense structures, the structures tend to be trailing spirals.

"Because of that," he continues, "the fluid elements in the inner part of a spiral are subject to a backward gravitational force from the outer part of the spiral, which reduces the inner fluid's angular momentum. That makes the material in the inner part of the spiral move inward and the material in the outer part of the spiral move outward."

Oddly enough, what controls gravitational instability in a protoplanetary disk is not the disk's gravity, but how easily the disk heats or cools. The spiral arms that lead to the formation of planets appear when the disk is cooling. The faster it cools, the faster the spirals form and break up into dense fragments. But if the disk is heated significantly by radiation from the central star, for example, then any dense structures that might begin to form could be disrupted by gas pressure.

In other words, any process that either heats or cools a disk may affect its stability. Durisen and his research group (see must consider as many of those processes as possible as they search for the answer to how giant planets are born.

Hal Kibbey is a media relations specialist in the IU Office of Media Relations in Bloomington.