Finding another Earth

Here's another topic that's been in and out of the news over the past several years (and will likely come up again in the near future). Exoplanets (otherwise known as an "extra-solar planet" or a planet orbiting a star other than our own sun) have been all the rage over the past decade or so. That's mainly because it's incredible that we can even detect them. Thirty years ago, the thought of being able to detect a planet outside of our solar system was pure science fiction. And there's a reason for that: planets are small. Really, really small compared to the distances between us and them. So small that directly photographing them is near impossible. But now, finding new planets is child's play. Since the first extra-solar planet discovery in 1988, we've had almost 3,500 confirmed exoplanets discovered. So... how is it done? There are two main methods.

Doin' the Do-si-do

Until fairly recently, the primary way that extra-solar planets were discovered was by their influence on their companion star. We normally think of the planets in our solar system orbiting the stationary sun at the center, but that's not actually true. All gravitationally bound systems orbit the system's center of mass, called the "barycenter." Now a star is so much more massive than any planets it might have orbiting it, that the system's barycenter is typically very close to the star (or even inside the star), but the point is that the star itself also orbits the barycenter in a kind of cosmic do-si-do. The larger the planet in the system (or the smaller the star), the farther that barycenter will be from the star and the larger its orbit. So being able to determine the orbit of a star around its system's barycenter, can allow astronomers to infer the size and position of exoplanets.

In a two-body system (just a star and a single planet), the star will move in a simple elliptical path, or a tight little circle. But the more complex the system is, the stranger the orbit of the star will be around its barycenter. Take our system, for instance. We have a lot of planets, and a few really big ones too. The insert is an image showing the weird and irregular path that the barycenter of our solar system has taken over the past few decades relative to the sun due to the constant motion and differential pulling of all of the planets in our solar system.

But still, this wobble is pretty small. For a two-body system, we're talking a couple hundred thousand miles at the most--which is the distance between the Earth and the moon--a drop in the bucket on an astronomical scale. So how do we see the wobble? Indirectly through the Doppler effect. Just like the siren of a passing police car is higher pitch as it approaches and then goes lower pitch as it passes, light does the same thing. The light waves of a light-emitting object approaching you squeeze together causing the frequency of the light to go up (just like the siren of the police car). This is called "blue shift" because the light shifts towards the blue end of the spectrum. The opposite also happens. As a light-emitting object recedes from you, the frequency of its light decreases. This is called "red shift" since the light shifts in frequency towards the red end of the spectrum. As a star wobbles around its barycenter, it will alternately move closer and farther away from us, causing its light to noticeably red shift and blue shift on a regular basis. We can use this periodic red shifting and blue shifting to calculate the stars orbital velocity as it approaches and recedes from us, which tells us its orbit.

So the red shifting and blue shifting tell us the star's orbit, and the star's orbit tells us about the planet. Got it? The below is the radial velocity graph of the star 51 Pegasus, which is only about 50 light years away. Its periodic red shifting and blue shifting led to the discovery of Bellerophon--the first exoplanet found orbiting a sun-like star.

Total Eclipse of the Heart

Since the Kepler spacecraft was launched in 2009, looking at stars' radial velocities to discover exoplanets has become less common. Now the most common way to find exoplanets is by looking at what's called a transit. 

A transit is what it's called when a planet crosses directly between us and a star. The most famous example is the transit of Venus. Since Venus is closer to the sun than we are, it periodically passes between us and the sun, partially eclipsing it. Now we're so close to our own sun, that when Venus transits (the next time Venus transits will be December 11th, 2117, so don't hold your breath), you can actually see it with a regular telescope and a filter. But when an exoplanet passes in front of its companion star, we don't directly see it. Again, we have to infer its presence by its effects.

As the exoplanet transits, it partially eclipses the star, causing its brightness (or flux) to decrease. Seeing the period of when it transits tells you the orbital period of the planet, which tells you its distance from its companion star. Seeing how long and how deeply the flux of the star changes tells you the planet's size. And lastly, the orbital velocity of the star in question (the red shifting and blue shifting I described above) tells you the amount of pull the exoplanet has on the star, which tells you the exoplanet's mass. The insert is an illustration of a star's light curve as an exoplanet passes in front of it.


These methods, while very fruitful, are far from perfect. There is one huge disadvantage to both of them: perspective. It would be nice if all solar planes were lined up perfectly with our line of sight, but they're not. They're all randomly distributed. If an exoplanet's orbital plane is even slightly inclined to our line of sight then we won't actually see it transit and we will miss that discovery. For a planet our size at a distance similar to our distance from the sun, it has been estimated that for a random distribution of orbits, we'd only see about 0.5% of them transit. That means for every near-Earth candidate we discover due to a transit, we're missing 199!

Ditto for orbital velocity measurements. While we see the light of the star no matter how its wobbling, we only see red shifting and blue shifting of the light as the star moves towards and away from us. So if the star is wobbling out of the plane compared to us, we'll only see a partial effect (and none if it's wobbling perpendicular to us). The orbital period that we see will be the same, but the magnitude of red shifting and blue shifting will be less, which would underestimate the true value of the mass of the exoplanet depending on angle of orbital plane to our line of site. But here's the problem: unless we directly visualize the star wobbling (and with our current technology we can't), then we have no idea what that angle is. And so all of our discoveries by this method are just estimates.

But that is partially solved by using the transit method. If we're discovering an exoplanet by a transit, we know that the orbit of the star is almost exactly in line with our line of sight, so our radial velocity measurements of that star hold up nicely. This is why using the radial velocity method by itself (in the absence of a transit) has mostly fallen out of favor. So while we're getting a more accurate picture of the planets we're discovering now, we're missing almost all of them.

Future directions

But don't be disheartened! If history tells us anything, it's that people are clever. Eventually we'll solve these problems. Even now, people are coming up with ever more intriguing uses for our current technology. As a planet transits in front of its star, the light from the star passes through the planet's atmosphere (if it has one) and some of that light gets absorbed at predictable wavelengths depending on the composition of that atmosphere. This wouldn't only change the amount of light we see from the star as the exoplanet transits, but subtly change its spectrum as well. As we further refine our techniques, it may be possible very soon to not only accurately determine a planet's existence, mass and orbit, but whether that planet may have a habitable atmosphere as well.

As always, science is amazing.