By far the most common way that extra solar planets are observed today is when they pass directly in front of their host star and block a small fraction of the star’s light. Detecting a planetary transit (also known as an eclipse) isn’t an easy task. It requires both a very precise alignment and a very accurate detector. Using the small angle equation, we can deduce that we would only see the reduction in light when the planet and star line up within an angle of 206,265 × (d / D), where d is the diameter of the star, and D is the star-planet distance. Using the Sun and Jupiter as an example, this gives 206,265 × (1.4 × 106) / (7.8 × 108) = 370 seconds of arc. Such a small angle is only 370 / 206,265 = 0.0018 of a full circle. If planetary systems are flung at random orientations in space, only 1 in 500 will have a favorable alignment for us to see the eclipse. Also, the eclipse would only occupy a small fraction of the planet's orbital period. The circumference of a Jupiter orbit is 2πD = 4.9 × 109 kilometers, so the planet would cross the star for (1.4 × 106) / (4.9 × 109) = 0.0003, or 0.03% of the orbit time. This is a single day out of a 12-year orbit. The easy part is detecting the dimming of the star’s light. Jupiter is ten times smaller than the Sun so it projects to 1% of the Sun’s area. With modern CCD detectors seeing a star dim by 1% is straightforward.
Despite the challenges, this technique first proved successful in 1999, when a planet two-thirds the mass of Jupiter momentarily dimmed its parent star. Initially, the majority of the planets observed with this technique were already discovered via the Doppler shifts they created in their host star. The organization TransitSearch.org was established to facilitate amateur astronomers and academics working to observe or rule out transits of stars with known planets. What’s perhaps most remarkable is the fact that these detections can be done with small, consumer telescopes just 4 inches in diameter. This is because it is the precision of the measurement that matters most, not the sensitivity. If the star is bright enough to be seen by the 4-inch telescope, the telescope only needs to be able to measure the slight changes in brightness that a planet would cause. This is actually remarkably hard to accomplish. Small fluctuations in the atmosphere, vibration of the ground (perhaps caused by footfalls or a passing truck), and errors in data reduction can all cause false detections or noise that makes the transit impossible to discern above the noise. But for a careful observer, transits can be detected.
Transit searches are now the most utilized technique for finding planets. The Kepler (NASA) and COROT (ESA) missions are dedicated to trying to find new planets strictly via their transits. Both missions are observing dense fields of stars. By looking at many many thousands of stars at once, the 1 in 500 statistic becomes less of a concern. Also, the space environment is very stable so tiny variations in brightness can be detected. Earth is ten times smaller than Jupiter, so an Earth transit of a Sun-like star would dim the star by 0.01% rather than 1%. Kepler and COROT were both designed to detect transits by Earth-like planets. Based in part on results of these missions, and other observations, it is now believed that nearly every star in the sky has at least one planet.
We’ve also learned from observations of distant globular clusters made by the Hubble Space Telescope and other observatories that globular clusters — systems of perhaps hundreds of thousands of stars without many heavy elements (called metals by astronomers) appear to be devoid of planets. This one characteristic — heavy element abudance — appears to be the most significant limiting factor on planet formation. Since planets are built from dust grains and rocky material, if there are not enough atoms of heavy elements in the nebula, planets may not form.
Planets that transit offer astronomers a fabulous opportunity to study alien atmospheres. As the planet passes in front of its host star, light will pass through the planet’s outer atmosphere, and the atmosphere will absorb light at specific colors that correspond to the chemical ingredients in the atmosphere. This technique has allowed the detection of water and carbon dioxide in a couple of gas giants, but it is still too difficult to attempt on terrestrial planets. However, it is believed by many astronomers that the first detection of life will be made through this type of observation. The detection of molecular oxygen in an alien atmosphere would be strong evidence of a microbial metabolism at work.
A superficially related technique is gravitational micro lensing, which enables astronomers to detect planets at much farther distances. As the name indicates, the method works on the same principle as gravitational lensing, but on a smaller scale. Einstein’s theory of general relativity predicts that the gravity of a massive object will bend light, like the lens of a telescope. The effect is visible when a star or planet passes in front of a distant background star. The light from the distant star is bent and focused, so the observer sees its brightness temporarily increase. The amount by which the background star’s intensity increases indicates the mass of the invisible object between the observer and the background star.
The first planet detected with this method had 1.5 times the mass of Jupiter, and it orbits about 3 A.U. from its parent star. This star is 17,000 light years from Earth, more than three times the distance to any extra solar planet previously discovered. The power of the gravitational lensing method lies in its ability to detect not only very distant planets like this one, but also very small planets, eventually even Moon-sized planets. The drawback is the necessity for a planetary system to be in a specific position with respect to a background star; the technique only works when they are perfectly aligned. Because this happens so infrequently, telescopes must monitor millions of stars every night, in the hopes of catching the rare event when a planet passes in front of one of them. The other limitation of micro lensing is that the event is transient. The foreground planet and background stars are "ships that pass in the night" so the characteristic brightening will only occur once. By contract, the Doppler effect and transits repeat every time the planet completes an orbit.