First steps in terrestrial planet formation observed?
Posted: January 13, 2002

Astronomers say they have discovered what may be "planetesimals" -- precursors of Earth-like planets -- forming at Earth-like distances and temperatures around a Sun-like star 430 light years away.

Michael Meyer of the University of Arizona announced the group's results at the 199th national meeting of the American Astronomical Society in Washington, D.C.

The group includes Eric Mamajek, Philip Hinz, and William Hoffman of the University of Arizona (Tucson, AZ), Dana Backman and Victor Herrera of Franklin and Marshall College (Lancaster, PA), John Carpenter and Sebastian Wolf of Caltech (Pasadena, CA), and Joseph Hora of the Harvard/Smithsonian Center for Astrophysics (Cambridge, MA).

The star, HD 113766A, is located 430 light years away in the direction of Centaurus. The star is number 113766 in Henry Draper's late 19th-century catalog of spectral classifications for bright stars. The letter 'A' indicates that it is one member of a binary star pair. Both stars are similar to, but somewhat hotter, more massive, and more luminous, than our Sun.

Infrared image from the MIRAC/BLINC camera obtained at the Magellan 1 telescope and a schematic representation of the system. Credit: University of Arizona
Although bright by astronomers' standards, the two stars' combined light is not visible to the naked eye. The pair are outlying members of the Scorpio--Centarus association, a nearby and well-studied group of young stars with ages of 5 - 20 million years visible from the southern hemisphere.

In 1998 astronomers Vincent Mannings and Michael Barlow included HD 113766 in their published catalog of "Vega-like" stars, otherwise normal stars showing signs of excess heat emission from planetary debris material. This class of objects was first identified by Fred Gillet and collaborators using data from the Infrared Astronomy Satellite (IRAS), an international collaboration led by NASA, in 1984.

Members of the present team Backman and Herrera independently identified the HD 113766 system as being especially interesting because much of the circumstellar material has comfortably Earth-like temperatures around 300 degrees Kelvin (80 Fahrenheit).

Meyer and colleagues made their observations with the 6.5-meter Magellan I telescope at Las Campanas in Chile during August 2001 using an advanced infrared camera system, MIRAC/BLINC, built by Hoffman, Hinz, Hora, and collaborators.

They found, to their surprise, that heat emission from dust debris appears around the slightly brighter of the two stars, component A, but not around B

The two stars are just 1.3 arcseconds apart, so close that they were separately resolved in infrared images only by combined virtues of the large aperture of the Magellan telescope, excellent observing conditions in Chile, and the fine new camera. The astronomers also found significant amounts of radiation from material hotter than Earth temperature that had been missed in previous observations.

The HD 113766 system, estimated to be about 10 million years old, does not show evidence of a massive hot gas+dust disk like the one around the prototypical very young Sun-like star T Tauri. Rather, it appears to be in a subsequent developmental stage in which the gas has mostly dispersed and solid particles are supposed to be accumulating into asteroid- and planet-sized objects. Planetary formation models indicate that at an age comparable to this system, Jupiter and Saturn were mostly finished, but the Earth and its neighbors in the inner solar were only partly constructed.

HD 113766 is similar to another famous inner debris disk system surrounding one component in a multiple star system, HD 98800. Why one component in each system appears to harbor a circumstellar disk system and the other does not is still a mystery. The only other known star system with a similar distributionof inner planetary debris material, apart from our solar system, is zeta Lep, a more massive and somewhat older star, announced earlier this year by Christine Chen and Michael Jura of UCLA, Meyer said.

Meyer presented a schematic (posted on the Internet at the address below) based on models of the location and density of the planetary debris around HD 113766A computed by members of the team using the new infrared measurements from Magellan, millimeter-wave observations from the SEST telescope, plus archived data from IRAS and the NASA-sponsored 2-Micron All-Sky Survey.

HD 113766 is much too far away to map the actual structure of the dust belt even using the best telescopes; instead, the structure must be inferred from calculations constrained by the observations. The team finds that the solid material around the star HD 113766A has temperatures ranging between 805 and 195 K (+990 and -110 F), which suggests it is distributed between distances of 0.35 and 5.8 Astronomical Units (AU) from its parent star (the hottest material is closest to the star). For comparison, Mercury, Earth, and Jupiter respectively orbit 0.4, 1.0, and 5.2 AU from our Sun. The companion star, component 'B', is located about 170 AU from 'A'.

Models of the debris distribution indicate it has approximately constant density (mass per area) between its inner and outer edges. This is exactly the distribution that would be produced by an effect known as "Poynting-Robertson" (P-R) radiation drag which can influence dust particles only if there is no gas to impede motions.

Most importantly, P-R drag sets a timescale for destruction of small particles. The observed grains should all spiral in toward the central star in a few hundred thousand years at most. This means that they can't be primordial grains persisting since the birth of the system 10 million years ago unless they are still embedded in a gas-rich disk, which seems unlikely.

"The fact that we observe copious amounts of dust means either that we are seeing the system in a brief moment after the sudden creation of huge amounts of it, or, more likely, that dust is produced and replaced continuously, for example by destructive collisions of larger parent bodies," Backman said.

"This line of reasoning applies as well to our solar system's zodiacal dust cloud, where particles produced by asteroid collisions continuously replace previous generations of dust that drift to destruction in the Sun in times much shorter than the age of the solar system," he added.

The astronomers estimate that the density of the HD 113766A debris belt is 250,000 times that of our solar system's zodiacal dust cloud. Roughly 200 times the total mass of asteroids in our present asteroid belt is required around HD 113766A to collide and replenish the observed small particles at the required rate. This means the amount of loose raw material in the form of asteroids and smaller objects located in the terrestrial-planet zone around HD 113766A equals about one-tenth the Earth's mass.

"The structure of the belt of material is consistent with current theories about planet formation in our solar system as well as the dynamical interactions of dust with large planetesimals," Meyer said. "Although there is no direct evidence for planets surrounding HD 113766, the observations suggest the emergence of a planetary system not unlike our own."

Future observations with NASA's Space Infrared Telescope Facility will be required to rule out the presence of remnant gas and refine models of the dust distribution, he added.