New study ups the predicted number of exoplanets discovered by Gaia

In a study to appear in The Astrophysical Journal, and published today on arXiv, it is estimated that Gaia will have discovered some 20,000 Jupiter mass exoplanets by the time it completes its survey in 2019. Former Gaia project scientist Michael Perryman teamed up with Lennart Lindegren from Lund Observatory and Princeton University astronomers Joel Hartman and Gáspár Bakos to make a revised estimate of the expected numbers.

In just 20 years since the first planets were discovered orbiting other stars, the known numbers have climbed to almost 2000, as a result of the pioneering radial velocity surveys and more recently the impressive haul of transiting planets discovered by the Kepler and CoRoT satellites, and other photometric transit surveys carried out from the ground. The burgeoning field of exoplanet science is set to make another remarkable advance with Gaia, using the astrometric measurement technique, where planets around another star show up as a tiny wobbling motion of the star as the planet orbits around it.

Drawing on the latest models of the Galaxy's stellar content, simulating the occurrence of planets around them according to today's best estimates, and replicating the way that the Gaia satellite makes its observations, the new study predicts that many thousands of new planetary systems will be discovered. In the paper, titled "Astrometric exoplanet detection with Gaia", the authors estimate that 21,000 ± 6,000 high-mass (1-15 Jupiter mass) long-period planets should be discovered out to distances of about 500 pc from the Sun.

Perryman and Lindegren made the first estimates of the numbers of exoplanets expected from Gaia at the time of the mission's acceptance by ESA in 2000, and various other estimates have been made since. But this is the first time that a complete Galaxy model, in-depth statistics of the exoplanet occurrence rates, and full details of the satellite's measurement and data analysis methods, have been brought together for a comprehensive assessment. And the results indicate that the numbers may be significantly higher than previously thought.

"It is not just the number of expected exoplanet discoveries that is impressive", says lead author Michael Perryman. "This particular measurement method will give us planet masses, a complete exoplanet survey around all types of stars in our Galaxy, and will advance our knowledge of the existence of massive planets orbiting far out from their host stars".

The expected yield is very sensitive to the total duration of the Gaia observations. "If the Gaia measurement duration is extended from its planned 5 years", says Lennart Lindegren, "our simulations show that every additional year makes it even more sensitive to the motion of the orbiting planets, and to planets orbiting even further out from their host stars." For a hypothetical 10 year mission, the predicted numbers that should be discovered by Gaia would rise to an astonishing 70,000.

Bakos and Hartman lead the ground-based HAT exoplanet transit detection network, and are particularly excited about the 50 or so Gaia discoveries that are expected to transit. "They are Jupiter mass planets with 1-2 year orbital periods, typically orbiting low-mass M dwarfs, and some are therefore likely to have spectacularly deep transits," says Hartman.
Although almost a side-show to Gaia's main task of surveying the stellar content of our Galaxy, the results show that it will be centre stage in the future development of exoplanet science too.

For further information, please contact: Michael Perryman


Schematic of the path on the sky of a solar mass star at a distance of 50 pc, with a proper motion of 50 milli-arcsec per year, and orbited by a planet of 15 Jupiter masses, eccentricity 0.2, and semi-major axis 0.6 AU. The straight dashed line shows the system's barycentric motion viewed from the solar system barycenter. The dotted line shows the effect of parallax. The solid line shows the motion of the star as a result of the orbiting planet, the effect magnified 30 times for visibility. Labels indicate (arbitrary) times in years.


With two or more planets, the star's motion with respect to the system barycenter can be more complex; in this case the projected stellar reflex motion of the 3-planet system HD 37124 is shown over the 30-year interval 2000-2030 (from Perryman & Schulze-Hartung 2011).


Histograms of the predicted planet detections as a function of (a) G magnitude; (b) distance; and (c) planet mass, where the rising distribution up to the adopted occurrence threshold of 15 Jupiter masses contrasts with the mass distribution for the simulated planets (d, which is a 1 in 10,000 re-sampling of the simulated planet distribution, showing the input power-law behaviour as the dashed line); (e) semi-major axis; (f) eccentricity.


Orbit fits to 100 simulations of one of the astrometric planet detections inferred to be transiting, showing scatter plots of the astrometric signature, period P, eccentricity e, and cos i, along with (top right) the transit time displacement from the reference time. In all diagrams the long dashed lines show the true values. Noteworthy is the high S/N and well-behaved solution despite the relatively faint magnitude (G = 15 mag), the relatively long orbit period, and the small number of field crossings.