Tuesday, 5 January 2010

There are several ways to spot an exoplanet.  The two most common are the radial velocity and transit techniques.  Radial velocity measurements detect changes in the star’s light due to its back-and-forth motion in response to its planet’s gravitational tug.  RV observations and astrometry—similar idea, but you can actually see the star move, not just detect the motion in its light spectrum—have brought us the bulk of exoplanet discoveries (nearly 400) and provide a minimum mass estimate for the planet. 

Transits, on the other hand, occur when a planet passes in front of its parent star, causing a dip in the amount of starlight we see.  This method has unearthed (pun intended) 69 exoplanets, as of exoplanet.eu’s last update on 11 January, including the five Borucki announced at the AAS.  The Kepler mission, as well as the European space telescope COROT and some ground-based programs like the MEarth project, use the transit method.

One method that I learned about back in astrobiology class but never gave much thought to is microlensing.  Microlensing occurs when a planet passes in front of a background source—a star, say, or a galaxy—and causes a brief spike in brightness by acting as a magnifying glass.  This magnifying effect results from “gravitational lensing,” in which the light of the background object is bent by passing through the well the planet makes in spacetime as the planet moseys along.  If the planet’s parent star also passes in front of the background object, the lensing caused by the planet will appear as a spike on top of the peak caused by the star’s lensing effect. 

The major downside for those who are repeatability-obsessed is that microlensing events are once-in-a-lifetime occurrences.  That means if you miss it, tough luck.  You can imagine the choking that happens when scientists try to swallow that.

B. Scott Gaudi of Ohio State University made me rethink my dismissal of microlensing, however.  In his presentation Tuesday afternoon he gave a compelling—and chortle-worthy—consideration of why microlensing might help us balance out the considerable biases in the types of exoplanets we are observing. 

“Due to small sample sizes, extreme selection biases in some cases, and in the face of limited observational resources, it’s going to be very difficult in practice to determine the demographics of planets,” Gaudi said. “. . . And without doing something fundamentally different than what we’ve been doing, we’re not going to answer some basic questions,” like whether migration of gas giants closer to their stars is common and what fraction of planetary systems look like ours.

For example, radial velocity measurements “do not constitute a homogeneous [census] of planets” because they were taken with a variety of accuracies and observation parameters.  In addition, only about 50 planets have been used to create the largest homogeneous assessment, he claimed.  You need to look at a good 10,000 stars, not go after specific stars to answer detailed science questions or after a handful of stars with “exquisite” accuracy (both of which radial velocity studies do the most, he said), in order to have a decent sample size, he continued.

Transits are also biased in that, until we have the slew of data for which Kepler keeps us hoping, verifiable observations are dominated by certain kinds of planets, like those in tight orbits around their stars.

Microlensing does not have the same limitations, Gaudi argued.  Because microlensing events are once-in-a-lifetime, they are “completely unbiased” because astronomers “have no idea what’s going to happen” beforehand, he joked.  Ground-based microlensing measurements are also sensitive to Earth-sized exoplanets and in particular to planets beyond the “snow line,” the distance from the parent star where ices form.  For a Sun-like star, the snow line distance is about 2.5 times the Earth-Sun separation.  This sensitivity arises because the magnification depends on time, and a planet further out will move more slowly (Kepler’s third law of planetary motion), therefore it will move more slowly across the object being lensed.  And, in fact, the majority of the 10 exoplanets discovered by microlensing lie beyond their parent stars’ snow lines, Gaudi said.

While ground-based microlensing observations are not particularly sensitive to Earth-sized exoplanets, space-based observations could detect planets as small as one-tenth the mass of Mars, he claimed.  Such a range could fill in a lot of gaps, he concluded.

I agree . . . if we can raise the number of sightings enough to make a dent.