Over the last couple of years, our knowledge of alien worlds has skyrocketed. In this week's "Ask a Physicist," we'll talk about how we find exoplanets and why we might miss a second earth, even if it's right next door.

If you've been paying even the slightest attention, you'll probably have noticed that in the last year or so, the number of known or suspected extraterrestrial worlds out there has more than tripled. This being io9, your first thoughts are no doubt about starting a super-race with green-skinned alien women.


But before we set course for parts unknown, we should probably figure out the destination. It turns out that finding a second earth is fairly tough, even if there are a lot of them hiding out there. A good launching point was suggested by question from Aseel Tungekar who asks:

I have a question regarding the discovery of exoplanets. One method, as far as I understand, uses the change in a star's brightness. How do we take into account all of the planets that don't actually go in front of a planet?

The sad fact is that the earth is very small, and would be very tough to detect from another star system. It doesn't produce much in the way of gravity and even less in the way of light. In the last few years, there have been a few amazing discoveries where we've actually taken (incredibly unresolved) pictures of alien worlds. But these worlds have been several times more massive than Jupiter, and so insanely far from their stars that you're probably not interested in relocating there.


You (hypothetical io9 reader) are looking for planets with atmospheres, planets in the Goldilocks zone, close enough to their sun to have liquid water, and planets with a rocky surface that you could stand on with something approaching earth-normal gravity. Let's face it. You're not in this for advancement of the scientific frontier, you're in it for the unobtainium.

The planets you find strongly depend on how you look for them, and how lucky you get. If it turns out that a particularly awesome planet is just next door but just happens to be orbiting its star on the wrong axis, we're just SOL, since the odds are heavily stacked against us discovering it.

So how do we find planets in the first place?


The classic approach, and the one that has still produced the majority of planets to date, is through the wobble method. Gravity works both ways, you see. The earth doesn't really orbit the sun. The sun and earth both orbit around a point about 450 kilometers from the center of the sun every year. Stars and planets in other systems do the exact same thing. Just like in a speed trap, we can use the Doppler shift to measure the periodic back and forth motion of a distant star. No wobble, no detectable planet.


Some of these motions, especially in systems with more than one planet, can get pretty complicated, but the basic idea is simple enough. The period of the star's wobble tells us how long the year is on the planet, and ever since since the days of the original Kepler, if we know the length of a year, we know how far a planet is from its star.

In order to detect it at all, the planet needs to wobble enough to measure it. The closer the planet is to the star, the faster the wobble will be. This is just the reverse of the effect that the planets closest to the sun orbit it the fastest. This does mean that the majority of the planets we're going to find are going to be very close to their host stars — we're talking closer than mercury is to the sun, and thus, insanely hot. It is my most sincere hope that you don't need to turn to the internet to tell you that it's hot near stars. Also, don't run with scissors. The other big factor, the biggest one, really, is mass. Jupiter causes about 140 times the wobble in the sun than earth does, even though Jupiter is much further away.


This combination of closeness and mass gave us a ton of Hot Jupiters, which, while awesome and incredibly instructive about how planets form and the sort of variety that's out there, aren't exactly conducive to human habitability.

Every now and again, though, we get to be both smart and lucky, as we did last year when the Lick-Carnegie Exoplanet Survey announced Gliese 581g, or planet "Zarmina" as people have started calling it. Gliese 581g is a 3-ish earth mass planet around 20 light-years away, and it's tantalizingly close to its star's Golidlocks zone. What made the wobble noticeable in this case is that the star it was orbiting was a red dwarf, only about a third the mass of our own sun, and with a much more concentrated habitable zone.



The reason that there has been so much excitement since the Kepler mission was launched (and virtually every time there has been either an official data release or even rumors about the data), is that Kepler could detect something just like earth.

The concept of Kepler (and a few other searches like it) is insanely simple. Suppose we get lucky and a planetary system just happens to line up perfectly with our line of sight. Once a year (every one of the planet's years, that is) the planet will transit in front of the star, eclipsing it ever so slightly. To give you an idea of the effect, aliens with their own version of Kepler would see the sun dim by a little less than 0.01% when the earth passes in front of the sun. In reality, Kepler can find planets even dimmer than that. From this relatively small bit of information, we can get the physical size of the planet (which gives us a pretty good guess about the surface gravity), the length of the year, and, because we know a thing or two about how light works, we can make a decent stab at the temperature.


But Aseel was right to be concerned. An alien looking for earths would only have about a 1 in 200 chance that the solar system would be tilted just right (from its perspective) to actually detect us. These are good odds, I suppose, from the perspective of not wanting to be invaded. The approach, then, means that we take the number of earths that we find, and essentially multiply by about 200 to get the true number out there. The exact details depend on the size and orbital radius, but the point is that there's an awful lot out there that we're just not going to detect directly. This is why to get a good sample, Kepler needs to look at 145,000 stars more or less continuously.

Even so, the number of planets discovered by transits is skyrocketing. The Extrasolar Planets Encyclopedia currently pegs the number of confirmed planets discovered in this way as only 124, but so far, Kepler has announced nearly ten times that many candidates, the majority of which will presumably be confirmed through followup observation.

One of the cool things about all of these measurements is that for lots of systems, the orbits aren't exactly periodic. There are little hiccups of a few minutes here and there. These hiccups provide clues that there might be other planets in the same system, perhaps in resonances.


Though it's producing the most candidates, Kepler is not the only game in town. One of my personal favorites is the MEarth project. Remember that with Gliese 581g, one of the big advantages was that it was circling a dwarf star? The MEarth project exploits this to the fullest and only looks for transit events around M Dwarfs. These would be a great place to live, by the way. Unlike our sun, which will live for a paltry 10 billion years in total, M Dwarfs can live up to a trillion years or so. No need to relocate. The other advantage is that because the stars are (relatively) so dim, these searches can look one not only at the transit, but also can see the effect of the reflected starlight off the planetary surface, allowing them to make ridiculously good models of the atmosphere, giving us a very good insight as to whether they are habitable or not.

Image courtesy the microFUN Planet Collaboration.

We get lucky

Sometimes you find planets when you aren't even looking for them. For instance, the Optical Gravitational Lensing Experiment
(OGLE) was looking for massive, dim objects in the Milky Way using a technique known as microlensing. I've talked a bit about gravitational lensing before. When a star passes behind a massive object (whether it's another star, a black hole, or whatever), the gravity focuses the light, and for a few days, the star appears to get brighter and then dimmer again.


In 2005, OGLE observed a star with an extra little bump in the light curve. This corresponded to what was then one of the least massive planets yet discovered, OGLE 2005 BLG 290Lb, only about 5 1/2 times the mass of the earth, though since it is something like 20,000 light-years from here and only 50 degrees above absolute zero, probably not the best prospect for human habitation. Since then, a few others have been found through this technique, and the takeaway of it all is that planets, and perhaps even earth-like ones, are all over the place.

This is just the first step.

It seems as though the Galaxy seems to abound in earth-like (or at least, earth-mass) planets. But this is just the first step in the Drake equation. I'm sure plenty of you could take me to school in the finer points of the equation (and hands up, those of you who wrote to Robert Zemeckis to complain about his butchering of it in "Contact"), but for those who need a gentle reminder, we can do a back-of-the envelope calculation of how many intelligent species are out there by multiplying things together like the fraction of stars with habitable planets, the probability of life forming and the like. The point is that just because a planet is small enough for a rocky surface and cool enough for liquid water doesn't mean that it has its own life. As I've written elsewhere, there are some other complications (including the odds of creating life and the future longevity of an intelligent species).


But even so, let me finish with a sobering, and somewhat dickish reminder. The closest stars are a bit over 4 light-years away. Our current record-holder for most distance manned trip is to the moon which is a whopping 1.3 light-seconds away. Even if we find a perfect world, getting there is going to be a a chore, perhaps an insurmountable one.


Dave Goldberg is the author, with Jeff Blomquist, of "A User's Guide to the Universe: Surviving the Perils of Black Holes, Time Paradoxes, and Quantum Uncertainty." (follow us on twitter, facebook, twitter or our blog.) He is an Associate Professor of Physics at Drexel University. Feel free to send email to askaphysicist@io9.com with any questions about the universe.

Top image: Antonova Elena/Shutterstock; planets by Darren Whitt/Shutterstock