We know that there are at least 75 planets outside our own solar system, orbiting their distant stars. The rate of planet discovery has sped up recently, and many more planets will likely be discovered in the weeks and years to come.
And yet, we have never seen any of these planets with our own eyes. Planets do not glow like a star - they only reflect light. That makes them a lot harder to see from far away. Any light reflecting off a planet also tends to be overwhelmed by the brightness of the host star.
So how do we know the planets are really there if we can’t see them? Several different techniques have been developed, and they all rely on one thing – how planets affect the stars they orbit.
The radial velocity technique has been the most successful detection method so far. This technique looks at how stars are affected by the gravity of an orbiting planet. Over the course of an orbit, the planet will pull at the star from different sides. Scientists measure the Doppler shift of the starlight to tell when the star is moving slightly away from us or toward us.
"As a star moves away from us, the starlight is Doppler-stretched to longer wavelengths, shifting the starlight toward the red end of the spectrum," explains Paul Butler, an astronomer with the Carnegie Institution of Washington and NAI member. "When the star moves toward us, the starlight is scrunched toward shorter wavelengths, shifting the starlight toward the blue. The Doppler shift that a planet imposes on a star is tiny. The ‘color’ change is imperceptible to the human eye."
Butler and his team have found many planets using the radial velocity, or "precision Doppler," technique. The planets detected by this technique have all been massive - the largest about 15 times more massive than Jupiter, the smallest about the same mass as Saturn. Although the planet’s mass affects the amount of tugging on a star, the radial velocity technique only indicates the minimum mass of the orbiting planet. For a more precise determination of a planet’s mass, Butler combines radial velocity observations with readings from another technique called transit photometry.
Transit photometry measures the apparent change in a star's brightness when a planet passes in front of it. The planet blocks some of the starlight reaching us, making the star seem slightly dimmer. This loss of light magnitude depends on the size of the planet.
In order for transit photometry to work, we must view the planetary system right at the orbital plane. If we’re watching from either too far above or too far below the planet’s orbit, the planet won’t pass in front of our view and we won’t witness any apparent dimming of the star.
Butler says transit photometry has provided his team with independent confirmation for one planet, HD 209458, and has allowed for the direct determination of the physical size and mass of the planet.
"We continue to work on finding transit planet candidates from our Doppler velocity measurements, primarily planets that orbit within 0.2 AU – 20 percent of the Earth to Sun distance -- of the host star," says Butler. "Transit measurements combined with Doppler velocity measurements also yield the orbital inclination and the true mass of the planet, as well as the physical size and bulk density of the planet."
Another limitation of the radial velocity technique is that the Doppler shift can’t be accurately measured for all stars, because many of them aren’t moving directly toward or away from us. Still, Butler says that this "orbital inclination" limitation doesn’t prevent the radial velocity method from detecting the movement of most stars.
"While the Doppler technique becomes less sensitive to planetary systems as the orbital plane becomes ‘face on’ relative to our line of sight, this is a minor effect," says Butler. "Very, very few planetary systems will be so ‘face on’ as to render them undetectable."
To overcome this limitation, Butler hopes to eventually combine his Doppler velocity measurements with observations from astrometry. Butler says that astrometric instruments are not currently capable of detecting extrasolar planets, but they should achieve sufficient sensitivity to detect planets within the next few years.
Like radial velocity, astrometry looks at how a planet’s gravity tugs on its star. But instead of measuring the Doppler shift of the starlight, astrometry measures the star’s position relative to distant background stars. As a planet completes an orbit around a star, the star appears to move back and forth in the sky. An astrometric instrument (such as an interferometer) can measure this change of position, which can then tell us something about the planet’s mass and orbital distance.
"Over the next 10 years, all we are going to ask of the inferometric astrometry technique is to solve for the orbital inclination of known planetary systems," says Butler.
Another type of extrasolar planet detection is gravitational microlensing. This technique uses foreground stars as a sort of magnifying glass to help detect distant stars and their planets. When a star that is closer to us passes in front of a more distant star, its gravity bends and amplifies the light from the distant star. This results in an apparent increase of light from the distant star.
Any planets that orbit the more distant star will perturb the gravitational lens, creating a brief variation in the amplified starlight. The duration of this change depends on the mass of the planet and the distance between the planet and its star, as well as the star velocity perpendicular to our line of sight.
However, Butler doesn’t think the gravitational microlensing technique is a very practical means for locating extrasolar planets.
"Gravitational microlensing is frankly not of much value," says Butler. "Microlensing events are notoriously complicated, involving complex theoretical calculations and interpretations of the data. The data can seldom be cleanly interpreted as the signature of a planet. Typically the host star can not even be seen in a microlensing effect, so we don't even know anything about the star, let alone the orbiting planet."
Gravitational microlensing events occur relatively quickly and do not reoccur, so it is almost impossible to confirm the data. In addition, says Butler, the other planetary detection techniques cannot confirm any of the planets discovered by gravitational microlensing.
"Such detections can not be followed up by any technique that we can imagine over the next hundred years," says Butler. "This is because microlensing can only detect planets that orbit stars many thousands of light-years away, while astrometry and direct imaging can only work on the nearest stars out to about 100 light years."
Still, according to William Borucki, a research scientist in the Planetary Studies Branch of NASA Ames and NAI member, having many so many different methods of extrasolar planet detection is a good thing. Where one method has a drawback, another method can provide information to fill in the gap.
"You wouldn’t want to have just one way of doing things," says Borucki. "For instance, you wouldn’t want to draw with just chalk. If someone told you that you couldn’t have a pencil or pen or typewriter or anything except chalk, you wouldn’t be very happy. Each detection method has its own strengths and weaknesses. They compliment each other very nicely."
Borucki is working on developing Kepler, a space telescope dedicated to transit photometry. While Kepler has not yet been approved by NASA, the telescope is designed to search for Earth-like extrasolar planets in the habitable zone of their stars. Some data suggest that such terrestrial planets may be out there, but currently these worlds are just below the limits of our detection.
"Kepler’s goal is to find out if planets with Earth-sized masses are rare or common," says Borucki. "There could be oodles of life out there, it could be that most stars have such planets. But then we have to ask ourselves, why haven’t they come to talk to us?"
"Personally, I think it is probable that there are a lot of Earths out there," Borucki continues. "But just imagine what a tremendous discovery it would be to find that there are no other Earth-like planets out there – no other planets capable of sustaining life. It would change the way we thought of ourselves, and the Universe, forever."
Butler says he is working to answer two questions: first, what fraction of nearby stars have planets? Also, what fraction of planetary systems are similar to the Solar System?
"The discovery of Solar System-like planets such as Jupiter and Saturn will require 10 to 30 years, roughly the orbital periods of Jupiter and Saturn," says Butler. "We thus expect to have the first statistically relevant answers to these questions around the end of this decade."
Detecting extrasolar planets takes time. In order to prove it is a planet that is affecting the star, information must be recorded for at least one full planetary orbit. Orbital times vary – the Earth takes one year to complete an orbit, while Jupiter takes about 12 years - so it can take many years to collect the necessary data.
Butler says that in the short term, the most important thing we can do is improve the precision of our Doppler technique to find smaller and more distant planets.
The next major breakthrough, says Butler, will be interferometric astrometry detections, which he says will begin in earnest in about 10 to 15 years. In addition, space based photometric transit telescopes like Kepler may also find Earth-like planets by the end of this decade.
Furthest away in the future are projects where we will actually get to see the planets themselves, says Butler.
"Direct imaging techniques like the Terrestrial Planet Finder are probably about 20 years or more away," he states. "Such techniques ultimately offer the opportunity to take direct spectra of extrasolar planets and thus directly search for signs of life."
SEND THIS STORY TO A FRIEND
History of planet detection
Explanation of Doppler technique with animations
Space interferometer Mission (SIM)