University of Pittsburgh
October 25, 2000

When is a planet not a planet? researchers see stars instead

Contact:  412-624-4147

PASADENA, Oct. 26 -- More than half of the recently detected extrasolar planets appear not to be planetary objects at all, according to a preliminary astrometric study conducted by researchers at the University of Pittsburgh's Allegheny Observatory, the Lunar and Planetary Institute, and the Korea Astronomy Observatory. The study is being presented at the Division for Planetary Sciences of the American Astronomical Society meeting,

October 23-27, in Pasadena, California.

The study suggests that the companions are not planets, but brown dwarfs (objects more massive than planets but smaller than stars), double stars, or low-mass stars. The confusion arises because of the inability of the radial velocity techniques, the basis of the original announcements, to determine the masses of the companions. Radial velocity observations cannot distinguish between a planet in an orbit that is viewed nearly edge on from a brown dwarf or stellar companion in an orbit that is nearly in the plane of the sky.

George Gatewood of the University of Pittsburgh's Allegheny Observatory, Inwoo Han of the Korea Astronomy Observatory, and David C. Black of the Lunar and Planetary Institute used astrometric data from the European Space Administration (ESA) Hipparcos spacecraft, as well as radial velocity data, to determine the inclination of the companion orbital planes to the line-of-sight of the observer and thus to determine masses for the proposed extrasolar companions.

"A striking result is the finding that orbital inclinations of the systems studied are extremely low -- that is, the orbital planes of these companions appear to be oriented nearly face on to the observer," said Black. "This contradicts the assumption that the line-of-sight angle is random in the radial velocity studies. It suggests that radial velocity studies that have reported detection of low-mass companions may be biased toward small inclinations in their selection of target systems."

"Part of the problem arises because masses smaller than that of Jupiter are very difficult to detect, thus most of the objects that have been detected and are suspected to be planets have masses greater than that of Jupiter, the most massive planet in our solar system," said Gatewood. "Until very recently, there were few known objects between one Jupiter in mass and 80 times as much, and those are small stars."

Now that some objects in this mass range have actually been discovered, astronomers are as confused as everyone else about what to call them, Gatewood said. "The answer has to do with how they form, but that is difficult to determine. Unfortunately, some astronomers have started calling objects less than 10 Jupiters in mass 'planets', a sort of mystical dividing line. To be scientifically correct, we should just call these objects 'substellar.'"

In reducing the data to determine the companion orbits, the authors arrived at four groupings of objects ordered by increasing value of the ratio of the estimated orbit's semi-major axis to its standard error.

Nine stars are estimated to have companions with true masses of 10-15 Jupiters or less. This is slightly higher than the minimum masses given by the radial velocity studies for these stars, but not excessively so.

A second group of 11 stars appear to have companions in the range of 15-80 Jupiters, masses substantially higher than the radial velocity results, which are thus likely to be brown dwarfs rather than planets.

A third group of four stars yields companion masses that are clearly stellar, probably M dwarf stars, although statistical artifacts have not been completely ruled out.

A final group of six will require additional astrometric observations before companion masses and inclinations can be determined with certainty.

An independent study by Mayor and colleagues (Halbwachs et al. 2000, also using Hipparcos data) of systems discovered by radial velocity searches with minimum masses in the brown dwarf range, shows a similar strong tendency (7 of 11 objects) toward systems with very low inclination angles.

This result obtained from combining astrometric data with the radial velocity findings is consistent with earlier studies that have noted that the orbital periods and eccentricities of so-called "extrasolar planets" are distributed in a way that is statistically indistinguishable from binary stars. This suggests that many, perhaps most, of the systems that have been identified as having planetary companions are indeed stellar or brown dwarf binaries.

Han, Black, and Gatewood stress that this is a preliminary result. Its intent is to guide future observers in their attempt to provide better data for the study of the origins of our own planetary system. Their decision to publish the initial results is based upon the surprising trend discovered by the study, not by the result for any individual system. They have already begun follow up studies with astrometric telescopic equipment at the University of Pittsburgh's Allegheny Observatory.