Skip Navigation

Astronomers Report Observational Evidence that Terrestrial Planets Form Around Nearby Stars

News Services
University of Arizona

Frank Low, 520-622-7074,
Dean Hines, 520-621-3445,
Glenn Schneider, 520-621-5865,

Lori Stiles, News Services
Tel: 520-621-1877   FAX: 520-626-4121

Astronomers believe they have found the makings – or the dust and debris – from a failed terrestrial planet circling a nearby system of two young stars similar to our early sun. The belt of material around the binary star HD 98800B, which is only 150 light-years away, is strikingly similar to the zodiacal dust bands in our solar system’s asteroid belt. These bands of dust and the asteroids that produce them resulted from a failed planet between Mars and Jupiter, say astronomers from The University of Arizona in Tucson, and they argue that a similar explanation is needed to understand their new observations.

Frank Low, Dean Hines and Glenn Schnieder of the UA Steward Observatory made the discovery as part of the Hubble Space Telescope NICMOS Instrument Definition Team’s Environments of Nearby Stars program. They are reporting the results this week at the 194th American Astronomical Society meeting in Chicago. The team also is publishing a paper on this work in the Astrophysical Journal Letters, “NICMOS Observations of the Pre-Main Sequence Planetary Debris System HD 98800.”

“Because we find so many similarities between HD98800B and our own solar system we conclude that the material responsible for the peculiar infrared emission around these two very young stars must have been produced in a way similar to the way asteroids formed around the sun,” Low said. “This is strong observational evidence that planets like Earth form around nearby stars.”

The zodiacal dust bands in our solar system, discovered 16 years ago in Infrared Astronomy Satellite (IRAS) observations by Low and colleagues, formed and have been replenished for four billion years by collisions of asteroids between Mars and Jupiter. Scientists say the belt of asteroids would have been another planet in our solar system, except that Jupiter’s tremendous gravitational forces prevent planet formation so close by.

During its flight in 1983 IRAS also found the huge infrared emission from HD98800. However, only very recently, by making observations with NICMOS on the Hubble Space Telescope and closely related ground based observations, have astronomers fit the pieces of the puzzle into a model that is remarkably similar in temperature, size, structure and mass to the asteroid families and the zodiacal dust bands. Low emphasized that, “It is only because HD98800 is so much younger than the Sun, about 5 million years instead of 4 billion years old, that it has enough dust and fine particles that we can see it clearly from space and from the ground.”

The astronomers used the high-resolution NICMOS camera in observations spanning 306 days to study luminosity, or light intensity, for each component of the HD 98800 system. They successfully measured how much energy each of the two stars individually generates in the system, which was no easy feat.

The binary stars are “K dwarf” stars, not too different from our sun. When K dwarf stars age, they typically end up slightly cooler and slightly less massive than the sun. These companion stars are the same age, about 5 to 10 million years old, according to earlier research.

But the astronomers didn’t see the planetary debris disk in their NICMOS images at all. And that, Schneider said, “is where the story got really interesting.” Like evidence in the Sherlock Holmes case from the dog that didn’t bark in the night, the evidence for the planetary debris disk that NICMOS didn’t get is very telling.

The long light wavelengths radiating from the HD 98800 planetary debris disk are in the far infrared, beyond the wavelength range detectable with NICMOS, the Near Infrared Camera and Multi-Object Spectrometer. The data from IRAS gave the energy output and temperature of the planetary debris disk. Almost all of the material is at a single temperature – 165 degrees Kelvin – just slightly cooler than our solar system zodiacal dust bands at 200 degrees Kelvin.

“That all this debris is at a single temperature is an amazing result,” Schneider said. “It behaves as a perfect ‘blackbody’ – absorbing light that falls on it from the stars and then re-radiating energy at particular wavelengths characteristic of the nature of this material.”

That the debris particles are all at one temperature shows that they are the same distance from the binary stars, their heat source. The team found that the HD 98800 planetary debris disk is 4.5 AU from star B, the cooler and slightly larger star in the binary system – just as our solar system’s planetary debris at the asteroid belt is at the correct distance, about 4.5 AU, for its temperature. At this temperature, the debris particles, which previously have been shown to be silicates, the stuff Earth is made of, must be bigger at least than 200 microns, which is about the diameter of a human hair, Hines said. For all that’s known at this point, Schneider added, the particles could be millimeters in size, and many must be much larger to remain in their orbits.

The total mass of material in the belt is between one-half and one Earth mass, the researchers say. The innermost and outermost particles in the disk are probably no more than one AU apart. If you were to stand at the center of the HD 98800B star, Hines said, you would see the band extend 12 degrees above and 12 degrees below the vertical.

If the HD 98800 binary system is around 5 million years old and current astronomical understanding is correct, Hines said, “Planets should have formed in this system by now. Given evidence that this planetary debris disk is so nearly identical to the zodiacal dust bands of our asteroid belt, we can logically infer the next step: Our asteroid belt was formed by a failed terrestrial planet. Now we see something that’s essentially analogous, or almost identical in properties, to our asteroid belt. A terrestrial planet is trying to form here – or it did form and was disrupted, probably by perturbing forces of the binary system of stars.”

“In an evolutionary sense, what we are really seeing is the results of a terrestrial planetary formation process or attempted formation process,” Schneider added. “There’s been a kind of gap between finding planets by radial velocity measurements and observing actual mechanisms of planet formation.”

EDITORS NOTE: The American Astronomical Society meeting poster on this story is viewable and downloadable via Netscape or other browers at:

The poster itself points to a journal preprint server from which the paper on this research can be displayed or downloaded. Hines’ artist rendering of the planetary debris disk around HD 98800b can be viewed at

(Contact Hines for a publishable print.)

Low this week is in Tucson; Hines and Schneider return to Tucson Friday, June 4.