Wednesday, January 27, 2010

Kepler's First Planets

This month the Kepler mission announced its first batch of extrasolar planet discoveries. Many people were surprised that only five new planets were announced, since the mission is monitoring more than 150,000 stars. But there are very sensible reasons why the initial catch was limited.

The first thing to understand about the newly discovered planets is that all of them were found in just the first 43 days of Kepler observations. The satellite was launched in early March, but the first two months were spent largely on engineering to ensure that the instrument was operating as expected. This was followed by a 10-day "commissioning" run in early May, and then a 33-day initial science run before the spacecraft made its first quarterly roll in mid-June to keep the solar panels facing the Sun. Because the search method requires a minimum of three transits, only planets with orbital periods shorter than about 14 days were detectable from the initial 43 days of observations. In fact, all of the planets that were announced had orbital periods between 3.2 and 4.9 days.

Fair enough, but shouldn't Kepler have discovered hundreds of such planets during the first 43 days? It probably did, but the mission requires detailed follow-up observations from ground-based telescopes to confirm each planet discovery. This is necessary to weed out "false positives" -- stars that look like they may host planets, but are actually something else. A simple example is an eclipsing binary star, where one star periodically passes in front of the other just like a transit. If there is a bright star close to the binary, or just along the line of sight, it dilutes the eclipses so they look like they are caused by a smaller object like a planet. The follow-up observations involve measuring the wobble of the host star caused by the orbiting object -- the traditional method of detecting planets around other stars. If the orbiting object is a star, the wobble is enormous; If it's a planet, the wobble is tiny.

Kepler is searching for planets in a large patch of the sky in the constellations Cygnus and Lyra. This area of the sky is well placed for ground-based telescopes during the summer months, but by the time the early data from Kepler was processed it was already August. The season for these follow-up observations was very short, so the team could only confirm a limited number of the planet candidates. In the end, they decided to announce only five of the more than 70 candidates that emerged from the initial analysis. By next summer, the mission will have many more candidates in longer orbits -- but they will also have a full summer season to confirm the discoveries.

By next January, expect to hear about hundreds of new planets. It will still be too early for the Earth-like planets, since that requires three 1-year orbits. But planets as tiny as the Earth may be found in faster orbits, and if they are hosted by smaller stars they might even be habitable. Stay tuned.