After decades of frustration and false alarms, astronomers may finally have a new method in their toolkit for finding planets around other stars: astrometry. Using the mighty 200-inch Hale Telescope at the Palomar Observatory in California, a team led by Steven Pravdo and Stuart Shaklan (Jet Propulsion Laboratory) claims that it has pegged a 6-Jupiter-mass planet orbiting the diminutive red-dwarf star VB 10.

Artist's concept of planet VB 10b (left) orbiting its host M-dwarf star VB 10 (right), 20 light-years away in the constellation Aquila. VB 10 is the lowest-mass star known to host a planet.

Intern A (Johannes Hirn)

The astrometry method shares one important attribute with the method that astronomers have used to find most of the nearly 350 known exoplanets: they look for stellar wobbles. The radial-velocity technique, which has revealed about 300 exoplanets, finds planets by seeing how their gravity tugs their host stars toward and away from Earth, a movement that shows up as a Doppler shift in the star’s spectrum. In contrast, the astrometry method finds planets by how they induce wobbles in their host stars’ motions on the sky. In other words, radial velocity reveals a star’s toward-and-away motion, whereas astrometry identifies stars that are wobbling “left and right” or “up and down” as they move across the sky over many years.

“This method is optimal for finding solar-system configurations like ours that might harbor other Earths,” says Pravdo.

View of the Palomar Observatory in San Diego county, California, with the dome of the 5.1-meter Hale telescope in the foreground.

In fact, the newfound planet orbits VB 10 every 9 months at about the same distance that Mercury orbits the Sun. Given the host star’s extremely low luminosity and cool temperature, the planet receives only a small fraction of the energy Earth intercepts from the Sun. However, a massive planet such as VB 10b generates a copious amount of internal heat, meaning its cloudtops may experience temperatures that would be comfortable to humans.

To be honest, the planet itself is not the significant aspect of this discovery. Given VB 10b’s high mass and the host star’s low mass (only about 8% that of the Sun, barely qualifying it for “stardom”), the system more closely resembles a very-low-mass binary star than a star-planet system. The planet probably formed from a collapsing gas cloud, like a star, rather than from the accretion of material inside a disk, like a planet.

Artist's concept of the VB 10 system and the solar system, with all bodies drawn to scale. Planet VB 10b (top left) has a size similar to that of Jupiter, but with 6 times the mass. VB 10b's orbit is comparable in size to that of Mercury (bottom left), but much slower (nine months instead of three) due to the low mass of the host star VB 10. VB 10 (top right) is an M-dwarf 10 times smaller and 12 times lighter than the Sun (bottom right).

Intern A (Johannes Hirn)

What’s really important about this finding is that after more than 50 years of efforts, it shows that the astrometry method can actually bear fruit. In past years, astronomers have announced planet discoveries using astrometry, the most famous being Peter Van de Kamp’s claims of a planet orbiting Barnard’s Star, one of the closest stars to Earth. In 1996, a claim of a planet orbiting the nearby star Lalande 21185 appeared on the front page of the New York Times. Unfortunately, none of these results survived critical scrutiny by other astronomers.

If VB 10b is confirmed, it means astrometry becomes the sixth technique to find exoplanets, joining radial velocity, pulsar timing, transits, gravitational microlensing, and direct imaging.

Finding planets via astrometry requires patience, because it takes many years of painstaking, ultraprecise observations to notice these tiny wobbles as the planet orbits its star. Pravdo and his team observed VB 10 for 12 years before making its announcement, and they measured motions of just a few milliarcseconds, akin to measuring the width of a human hair from 2 miles away. And even though their paper has been accepted for publication in the prestigious Astrophysical Journal, it has not yet been made public, so the astronomical community has not yet had a chance to evaluate the claim and see if Pravdo and his colleagues have properly accounted for possible observational errors.

Artist's concept of the SIM spacecraft.

NASA / JPL-Caltech

As Pravdo points out, astrometry holds great potential for finding Earth-size planets in the habitable zones of their host stars. Veteran exoplanet hunter Geoff Marcy (University of California, Berkeley) says, “The only known method for detecting Earth-like planets around nearby Sun-like stars is with astrometry. The proposed Space Interferometry Mission (SIM-Lite) proposed to NASA will measure positions of stars to within 1 microarcsecond. It is technically ready to be launched. Thus, astrometry holds the future of detecting other Earths in its hands.”

Exoplanet researcher David Kipping (University College, London, England) also notes that the other methods would not have been able to detect VB 10b. “This really points out how well astrometry can complement the other methods,” he says.

The planet and its host star lie about 20 light-years from Earth in the constellation Aquila. The star is now the lowest-mass star known to host a planet.


Image of Scott C. Schumacher

Scott C. Schumacher

May 29, 2009 at 1:19 pm

A 6 Jupiter-mass planet orbiting a low-mass star ought to be a prime candidate for radial velocity confirmation, I think. The only difficulty I see is that it will take a couple or few years for confirmation due to the planet's orbital period.

You must be logged in to post a comment.

Image of Robert Naeye

Robert Naeye

May 29, 2009 at 3:06 pm


Thanks for writing with this very insightful comment. Unfortunately, the radial-velocity method probably won’t be able to confirm this planet. The main problem is the fact that the host star is a feeble red dwarf of spectral type M8. It puts out less than one-millionth the light of the Sun, so even though it is only 20 light-years from Earth, it is very faint, with a visual magnitude of 17. This means astronomers have very few photons for collecting the high-quality spectrum they need to reveal the subtle motions induced by the planet’s gravity. So my guess is that this planet will never be confirmed by radial velocity. But I certainly hope someone tries and proves me wrong!

Robert Naeye

You must be logged in to post a comment.

Image of Kent Karlson

Kent Karlson

May 29, 2009 at 6:28 pm

Please check the author's name of the new study. You spelled it wrong.

You must be logged in to post a comment.

Image of Rod


May 31, 2009 at 7:31 pm

FYI readers. This URL maintains a list of the reported exoplanets When I reviewed the data the average mass reported for 348 entries is 3.1024 Jupiters and average eccentricity = 0.2215. The average host star mass = 1.0932.

You must be logged in to post a comment.

Image of Marc Dubbeldam

Marc Dubbeldam

June 12, 2009 at 10:05 am

Whether or not the planet can be confirmed by means of radial velocity measurements strongly depends on the geometry of the planet’s orbit, i.e. if the planet’s orbital plane is normal to the line of sight (from the observer to the star), the planet’s motion will not induce any radial velocity components in the star’s motion. However, this geometry would produce the largest possible motion (wobble) on the sky. Ergo, the two methods are complementary.

Anyway, in the light of previous false alarms, I remain sceptical until the detection has been independently verified.

You must be logged in to post a comment.

You must be logged in to post a comment.