Earlier this year, I wrote about how an apparent change in the orbital characteristics of a planet around TrES-2b may be indicative of a new planet, much in the same way perturbations of Uranus revealed the presence of Neptune. A follow up study was conducted by astronomers at the University of Arizona and another study on planet WASP-3b also enters the fray.
The new study by the University of Arizona team, observed the TrES-2b planet on June 15, 2009, just seven orbits after the observations reported by Mislis et al. that reported the change in orbit. The findings of Mislis et al. were that, not only was the onset of the transit offset, but the angle of inclination was slowly changing. Yet the Arizona team found their results matched the previous data sets and found no indication of either of these effects (within error) when compared to the timing predictions from other, previous studies.
Additionally, an unrelated study led by Ronald Gilliland of the Space Telescope Science Institute discussing various sampling modes of the Kepler telescope used the TrES-2b system as an example and had coincidentally preceded and overlapped on of the observations made by Mislis et al. This study too found no variation in orbital characteristics of the planet.
Another test they applied to determine if the orbit was changing was the depth of the eclipse. Mislis’ team predicted that the trend would slowly cause the plane of the orbit to change such that, eventually, the planet would no longer eclipse the star. But before that happened, there should be a period of time where the area blocked by the planet was covering less and less of the star. If that were to happen, the amount of light blocked would decrease as well until it vanished all together. The Arizona team compared the depth of the eclipses they observed with the earlier observations and found that they observed no change here either.
So what went wrong with the data from Mislis et al.? One possibility is that they did not properly account for differences in their filter when compared with that of the original observations by which the transit timing was determined. Stars have a feature known as limb darkening in which the edges appear darker due to the angle at which light is being released. Some light is scattered in the atmosphere of the star and since the scattering is wavelength dependent, so too is the effects of the limb darkening. If a photometric filter is observing in a slightly different part of the spectrum, it would read the effects differently.
While these findings have discredited the notion that there are perturbations in the TrES-2b system, the notion that we can find exoplanets by their effects on known ones is still an attractive one that other astronomers are considering. One team, lead by G. Maciejewski has launched an international observing campaign to discover new planets by just this method. The campaign uses a series of telescopes ranging from 0.6 – 2.2 meters located around the world to frequently monitor stars with known transiting planets. And this study may have just had its first success.
In a paper recently uploaded to arXiv, the team announced that variations in the timing of transits for planet WASP-3b indicate the presence of a 15 Earth mass planet in a 2:1 orbital resonance with the known one. Currently, the team is working to make followup observations of their own including radial velocity measurements with the Hobby-Eberly Telescope owned by the University of Texas, Austin. With any luck, this new method will begin to discover new planets.
UPDATE: It looks like Maciejewski’s team has announced another potential planet through timing variations. This time around WASP-10.