Almost all of them have been discovered using what can be colloquially called the "wobble method." This is an indirect method, which means the presence of a planet is inferred -- in this case, by a planet's effect on the star it orbits.
Not unlike a teeter-totter with a heavy person nearer the center fulcrum for balance, the star is nevertheless slightly offset around its balance point (called the "barycenter") by the comparatively lightweight planet orbiting it.
It is this slight stellar wobble that leads astronomers to conclude there must be an unseen something on the other end of the astronomical teeter-totter to produce the periodic effect -- and the period in question is the orbital "year" of the planet.
But what, exactly, is detected? To answer that question, we need to look into the discovery of an 18th-century gentleman named Christian Doppler.
Doppler found that waves traveling toward you would be compressed, and widened if moving away.
He started with sound waves. Picture a train where the engineer is "throwing" sound waves at you. He times them so that he sends you one every second.
When the train is standing still, you can look at your watch and see that you are getting a sound wave every second; since this is how frequently you are catching them, let’s call it the sound wave’s "frequency." (Frequency comes in units per second; a "per second" is also called a "Hertz" after an early pioneer of wave physics.)
Now the engineer decides to start the train moving toward you, without altering the rate or frequency at which he is sending sound waves. Since the train is moving toward you, each subsequent sound wave being emitted has less and less distance to travel to you.
This means that, according to your watch, the frequency of the waves’ arrival times increases (according to the engineer’s watch, the frequency remains the same). You perceive this as a higher pitch.
As the train passes you and starts moving away, the distance each wave has to travel toward you increases and the opposite occurs. The waves arrive less frequently, and the pitch of the sound waves goes down.
This change in frequency caused by the movement of the wave-emitting object is known as the "Doppler shift."
It also works with light waves, with higher frequency light appearing bluer while lower frequency light appears redder.
Now back to that planetary wobble. Scientists break light waves up into their component colors, and if you examine the rainbow colors of the Sun (called the "solar spectrum") with a prism, you'll notice dark lines within the spectrum where color is almost completely missing.
These missing frequencies are largely light that has been absorbed by the Sun's cooler outer layers.
Every element -- hydrogen, calcium, magnesium and so on -- absorbs precise frequencies of light, removing them from the Suns rainbow. And if you were to move the light source away, these spectral lines would (Doppler) shift to lower frequencies, moving toward the red. The faster the light is receding, the redder the spectra appear.
So what would you conclude if you were looking at the spectral lines from a nearby star that shift first toward the red, then toward the blue, then red, then blue and so on?
You might correctly conclude that Earth was moving with respect to the star (this time you are moving rather than the light source -- the effect is the same).
However, after you correct for this and other movements and still see a tiny (less than a billionth of the width of a human hair) but unmistakable oscillation, you would have to conclude that the star itself was moving. What could be moving the star back and forth by this tiny amount? A planet.
The wobble can't tell us how massive the orbiting planet is because the Doppler shift only works if the object is moving toward us or away.
Picture a car merging onto the freeway from an on-ramp. You are already on the freeway and see the car approaching. The car may be moving at 60 miles per hour, but it is only closing the gap between you at a speed of perhaps 5 miles per hour.
If you were to make a Doppler shift measurement of the car, you would detect only the 5 miles per hour, not the rest of the "sideways" motion of the car’s velocity.
Similarly, the wobble method of finding extrasolar planets detects only the star’s motion toward or away from us, not the full extent of the planet's pull. We know that the planet has offset the star at least as much as has been measured.
The actual mass of the planet is thus a minimum, not an precise figure.
This method of detecting extrasolar planets has been extremely successful to date, providing astronomers with numerous surprises. Large Jupiter-like planets have been detected in orbits with "years" as short as 3 days and some star systems have been found to have at least three of these giants.
Giant planets have also been found orbiting one component of double star systems whose stars are far enough apart to allow the planet to orbit somewhat stably.
Are there planets out there in orbits like our own Jupiter? Within the next couple of years, astronomers using the wobble method should be able to tell. By then, they will have looked long enough to see a motion toward, then away from us; a task that takes Jupiter about 11.8 years.
However, it is possible that an extraterrestrial "wobble-method" astronomer looking at our solar system might not believe that Jupiter is here, for it seems that the planet has a disguise.
It just happens that in our own solar system, the orbital period of Jupiter, our most massive planet, nearly matches the cycle of sunspots our Sun produces.
The Sun, like many other stars, has a cycle of sunspots -- caused by the twisting of magnetic field lines into "knots" of darker regions -- that has a period of about 11.2 years.
These knots change the Sun's spectral lines, making them difficult to distinguish from the spectral line changes due to Jupiter’s orbit.
As mentioned in an earlier article, a "Jupiter" correctly placed within a planetary system comes in very handy to first deflect comets formed in the outer regions around the star toward the warmer, but more metal-rich planets.
Later, such a planet "cleans up" its planetary system by deflecting rogue comets out of the inner regions, making the tiny inner planets safer for critters to develop. (If it were not for Jupiter, there might be mass extinctions caused by comet impacts -- similar to the one that likely wiped out the dinosaurs -- as often as every few thousand years.) Thus, an extraterrestrial civilization might not be able to tell very easily, using the wobble method, if our solar system was a good place to live or not. So might they not then send a radio signal our way?
Fortunately, there are other ways to tell we are here, including a way for an extraterrestrial civilization to detect Earth, which causes a wobble too small to detect with the wobble method. Some of these other extrasolar planet detection methods will be the subject of the next series.