How to Find Extra-Solar Planets

Extra-solar planets can be found in any of three ways. The first, and by far the most common, is the Doppler Shift method; this precisely measures the color of the star, and the minute periodic shifts towards red and blue caused by the tug of the unseen planet. A second method that is more preferable but is much less likely to work is the Transit Method. This only works if the star and planet line up directly along the line-of-sight to Earth. This will then cause the planet to periodically pass in front of the star and for the star to pass in front of the planet, which results in dimming of the total light output of the system. The third way that had never worked until 2004 is direct imaging, where telescopes are used to directly resolve the planet around its companion star.

The chart on the right shows the exoplanets that have been found so far in terms of their size and how far they orbit from their parent star. It also shows Earth and Jupiter, which shows that we still have a long way to go before Earth-like planets are found. For more data on exoplanets, see this site's page on Data Trends.

Even if unseen, the effects of exoplanets on their host stars can be detected. This is how almost all of the exoplanets discovered so far have been found. When an object moves, the light reflecting off it or emanating from it is shifted - a phenomenon called the doppler shift. If an object is moving away from a detector, its light is shifted towards the longer, redder wavelengths; if an object is moving towards a detector, its light is shifted towards shorter, bluer wavelengths.

When a planet orbits a star, it does not orbit the exact center of the star. Rather, it orbits the "center of mass" of the system, which is where, if one were to place the two objects on opposite sides of a teeter-totter, the stand would be placed. So, while the planet does the bulk of the moving, the star wobbles a little bit. It wobbles directly towards us when the planet is just starting to swing behind the star, and it wobbles directly away from us when the planet is just starting to swing in front of the star.

Therefore, by very carefully observing a star and measuring the shift in the color of light, we can determine how quickly the star is moving. Based on this and the period of its movement (which corresponds to the planet's year), the minimum mass* of the planet can be determined.

The shifts are so small that it has only been in recent years that instruments have been sensitive enough to measure the shifts. And, the shifts will be larger with a larger planet, and they will require less observing time in order to detect them if the planet's period is short, such as only a few days. This is why the majority of exoplanets that have been discovered are "hot Jupiters" - planets larger than Jupiter that orbit very close to their host stars.

*Only the minimum mass can be found through this method. This is because we can only see motion along one direction, and so we have no way of knowing at what angle the system is tilted relative to Earth's orbit. If it is tilted so that the planet(s) cross almost directly in front of the star, then the mass that is calculated is the planet's true mass. However, if the system is tilted so that the system would look like a dartboard to us, then the mass calculated is much smaller than the actual mass.

Besides using extremely precise doppler measurements, a second way to detect an extrasolar planet is to observe it passing in front of its parent star. When it does this, the planet will block part of the light from the star. If you can observe this dimming and it repeats regularly then it might be a planet.

Other than a planet, the transiting object could be a binary star companion. This can be determined by measuring the object's mass. This is done with the doppler shift method: Once a star is observed to dim periodically by what is thought to be an unseen planet, then it can be targeted for doppler measurements.

A variation of this is to measure the light output of the system in infrared light. Planetary temperatures are such that they radiate energy mostly in the infrared part of the spectrum. Stars around which planets are being looked for radiate most of their energy in the visible part of the spectrum. Thus, stars are much dimmer and planets are much brighter when one uses infrared light (though the star still far outshines the planet). So, when the opposite eclipse happens - when the planet passes behind the star - the total amount of infrared light that is received from the system can dim by about one part in 500. With new instruments, this dimming is measurable.

Because the planet was seen transiting the star, the inclination of the system relative to Earth - how much the system is tilted - is known, so the precise mass of the planet can be determined. If the mass is under about 13 times Jupiter's, then it is considered a planet.

Imagine this picture: You have a friend who placed a grain of sand somewhere on the side of a mountain. This friend asked you to stand at a distance of five miles, and find the grain of sand from there. To make it even harder, your friend didn't tell you upon which mountain to look. Before you start looking for a new friend, keep in mind that this is how hard it is to spot a planet orbiting a distant star.

First Image of an Exoplanet - 2M1207 In 1997, based upon an image by the Hubble Space Telescope (at right), astronomers thought that they found one speeding away from its parent star 450 light-years away in the constellation Tarus, the Bull. The "planet" appeared to be leaving behind a gaseous trail 225 billion km (135 billion miles) long. Follow-up observations, however, showed that the object, known as TMR-1C, was too hot to be a planet. It is probably a background star whose light has been dimmed by interstellar dust, making it appear to be a planet in the vicinity of the star.

In April of 2004, an international team of astronomers discovered what appeared to be a planet orbiting a "brown dwarf" - a failed star that never accreted enough matter to begin fusion in its core. At the time, it was impossible to prove that the faint "planet" was not a background object.

In August of 2004, HST observations verified the previous ones, though they were taken too soon in order to conclusively demonstrate that it was a planet. However, observations taken in February and March of 2005 found the same thing, and so it has become the first exoplanet to be directly imaged. It orbits the brown dwarf 2M1207, and so it is called 2M1207b. It lies approximately 55 A.U. from the brown dwarf, and the system lies approximately 200 light-years from Earth. The mass is estimated at about 5 Jupiter masses.

The search is ongoing, and new planets are now being found relatively frequently. But, with the current methods, it has been estimated that the smallest planets that can be found are Neptune's or Uranus' size (3-5 times the size of Earth). Therefore, new techniques are being developed.

One such technique would work on the principle of interferometry -- if you combine two or more telescope's light, the resulting image would have the resolution of a telescope the size of the distance separating them; the idea is to build a set of telescopes (about five), place them on a satellite about a football field distance apart, and set them in orbit out beyond Jupiter. This would, in theory, allow astronomers to directly image the disks of the planets their searching for, and possibly to even see Earth-sized planets.

NASA's Kepler mission, which should be launched in 2008, is designed to do this, though with one telescope. It will take, in a single exposure, images of about 100,000 sun-like stars. It will be aimed at just one region of space, the star field in the Cygnus-Lyra region, and it has a nominal mission lifetime of 4 years. It will take images of this region for the entire mission, and it will search for the periodic dimming that would be indicative of a transiting planet. It is believed that the mission will find hundreds of exoplanets that are 1 to 2 times Earth's size, as well as hundreds larger.

Other ideas also stem from the interferometry approach, but are more grounded, literally. Since the military declassification of adaptive optics (the technology that allows, based on a laser beam aimed through the atmosphere, a computer to make very slight modifications to a telescope's main mirror, which will correct for atmospheric distortions), ground-based telescopes have greatly improved, and they have been able to greatly increase their seeing power. Therefore, there are many telescopes on the drawing board and already being built that will be able to increase the resolution of stars' wobbles, and also to possibly image planetary disks. It was with adaptive optics that the image of 2M1207b above was taken.