Update on the search for planets in the Alpha Centauri system

Thanks in part to Planetary Society members’ support, Yale’s Debra Fischer and her team were able to secure many nights of observations of our nearest stellar neighbors – the main Alpha Centauri stars – to look for planets. They also tried to confirm the discovery announced by another team of an exoplanet around one of the Alpha Centauri stars. Below, Debra gives us an excellent update on the intriguing results of their observations, as well as future plans, but I don’t want to spoil it for you, so here is the update from Debra Fischer:

The proximity of the alpha Centauri star system – just 4 light years away – makes this an exciting first destination for robotic space craft venturing outside the solar system. The “α Cen” system is comprised of three stars that are gravitationally bound. All are just a bit older than the Sun with a rich chemical composition that is known to be favorable for the formation of planets. Like all multiple star systems, the α Cen star system is hierarchical. There are two central stars, α Cen A and B, that orbit each other once every 80 years. These two stars are similar in mass to the Sun and very bright. The third star, “Proxima” is about half of the mass of the Sun and is so far from α Cen A and B that the estimated orbital period of Proxima around A and B is hundreds of thousands of years.

In 2007, we started planning the search for planets around our nearest neighbors. We focused our attention on the two bright stars, A and B, where we knew that we could obtain planet-detecting precision in our velocity measurements. We understood that we were racing against time: the orbit of α Cen A and B is oriented in such a way that the two stars appear to be getting closer to each other. This is a problem because we needed to clearly resolve the stars in order to obtain good velocity precision. In 2007, the stars were a respectable 10” apart and easy to separate with a decent telescope.

In 2008 we refurbished an old spectrograph reaching a modest 5 meters-per-second precision at the 1.5-m telescope at Cerro Tololo Inter-American Observatory (CTIO) in Chile. In parallel, we began designing CHIRON, the CTIO High Resolution spectrometer. With funding from the NSF in 2009, my team designed, built and commissioned the instrument by March 2011. The α Cen stars were getting closer, but still 7” apart. Because our precision was “only” 1 meter-per-second, we decided to try to further stabilize the instrument. In 2012, we added a new octagonal fiber scrambler, developed with help from the Planetary Society and our precision improved. The stars were getting closer - now a worrisome 6” apart, but the quality of our data was excellent and we obtained hundreds of measurements of A and B.

In December 2012, the Geneva team surprised us by announcing the detection of a 1-Earthmass planet in a scorching 3.24-day orbit around α Cen B using the HARPS spectrometer. We examined our data from 2012 and did not see the 3.24-day signal. However, the weak signal was a challenging detection and the Geneva team had five times more data than we had collected. We decided to redouble our efforts.

As soon as α Cen A and B emerged from behind the Sun in 2013, we set out to confirm the detection. We modified our search strategy to specifically look for the 3.24-day signal by observing α Cen B “all night, every night.” We again had help from the Planetary Society to purchase extra nights that allowed for this intense observing strategy. We coordinated our observing runs with the Geneva team; Yale grad student Matt Giguere observed from Cerro Tololo at the same time that Xavier Dumusque was observing with HARPS on a neighboring peak in the foothills of the Chilean Andes. In May 2013, the stars were only 5” apart. On the clearest nights, we were able to obtain good quality data; on marginal nights, we began seeing significant contamination from the other star.

So what did we find? We have examined all of our data from 2012 and 2013 and we still do not see evidence for the planet that was detected by the Geneva team. I am often asked if this null result is consistent with an analysis of the Geneva HARPS data by Hatzes (2013) that did not find a statistically significant 3.24-day signal. Our simulations suggest that we should have seen this signal in our data – but only at a marginal level. One thing we did learn is that our observing technique, which uses an iodine cell for wavelength calibration, is not adequate to reach the precision that we need. In order to make a definitive statement, we need a higher precision instrument with new wavelength calibrators.

Now, in February 2014, the α Cen stars are “rising” again in the southern hemisphere skies. However, α Cen A and B are now at their closest projected separation – less than 4”. It is no longer possible to use either CHIRON or HARPS to observe one star without significant contamination from the other star. For the past 5 years, we raced against time. Now we must wait. In 2015, the two stars will begin pulling away from each other so that in a few years, we will be able to search again.

Both the Geneva team and my team will use the next few years to develop innovative instruments with the goal of reaching 10 centimeter-per-second precision – a factor of ten gain over current precision. The Geneva team is designing a high-resolution instrument, ESPRESSO, for the 8-meter telescopes at Paranal in Chile. My team is designing EXPRES for the Discovery Channel Telescope. As the acronyms imply, we are both aiming for the extreme precision needed to robustly detect Earth-mass planets orbiting at habitable-zone distances.

Help support projects like this that hunt for exoplanets:

Donate to The Planetary Society Exoplanets Program!