Why NASA Hasn't Been Able to Find Earthlike Planets

Since 1995, astronomers have been building a rapidly growing catalogue of known planets orbiting stars -- suns, really -- other than our own. And, quite naturally, once we were able to prove that other stars had planets, the next obvious question that came up was: how many are like ours? The search for other planets is intricately tied up with the search for other life.

Which is why it is particularly disappointing, as a lay member of the public, to watch the years of my life creep by with no new discoveries of planets that actually look like Earth. Most of the planets discovered are huge, much larger than anything even in our solar system. Most are also far too close to their sun: we wouldn't expect to see any life on them because all the basic building blocks of life will have been fried for billions of years. So does NASA's lack of discovery of planets around other stars mean that there probably aren't any?

Not necessarily. The reason why we're having trouble finding Earth-like planets is because we're only gradually coming to grips with the sort of technology needed to find those Earths in the first place. For most of the last fifteen years, we've been searching for new planets with technology that's the astronomical equivalent of being legally blind. Even now, we generally can't actually see the planets directly with telescopes; we have to adopt various tricks to find them. As our technology improves, we've been finding planets more and more like our own. Which suggests that there are actually a huge number of planets out there that, at the very least, are Earth-sized and have Earth-like orbits. We just have to find them.

Before the 1980s: (Un)informed Speculation

Understanding the search for planets means understanding how scientists look for them, and for this, it's best to start at the beginning. The idea that other stars might have other planets is pretty much as old as our scientific study of our own planets, following Galileo: in fact it was put forward by Italian mathematician Giordan Bruno, who, along with Tycho Brahe, was instrumental in beginning to question the geocentric model of the universe (one in which Earth sat at the centre and our sun, all the planets, and all other stars gradually circled around it). Isaac Newton accepted it as perfectly logical, too.

But that was at a time when we barely had the technological capacity to observe all of our own planets. (Pluto, which is no longer a planet anyways, was only discovered in 1915.) By that time, there had actually been a few theories about real extrasolar planets (as they're technically called, or alternatively as "exoplanets"). British and American astronomers were particularly interested in a star named 70 Ophiuchi, because of apparent orbital anomalies it displayed, but the theory was eventually proved mathematically impossible. At this point all of the early speculative planet "discoveries" are considered to have been false on the evidence presented. This doesn't mean there aren't any planets around them -- it just means that if there are, we don't have any evidence for them yet.

The basic problem is that planets, unlike stars, are far too small and far too dark to be able to see, even with telescopes. To illustrate the problem, it's worth observing that even the most common types of stars, the red dwarfs that are smaller and perhaps one-tenth the brightness of our own sun, are relatively difficult for astronomers to find. (And none of them can be seen with the naked eye.) And then, somewhere between red dwarfs and large planets, there lies another category of objects called brown dwarfs, which literally could be pretty much anywhere, for all we know. The smallest known brown dwarf is only eight times as massive as Jupiter, and NASA speculates that there are at least a few of them between us and the closest known star, Proxima Centauri. We just haven't been able to see them yet.

Looking for Planets Yesterday: They Have to be Big, and They're Often Hot

By the 1980s, human technology had advanced far enough that it was possible to make some fairly detailed observations of reasonably close stars, which made the first of the effective modern tricks available. All of these, however, come with serious side effects for people who are eager to see evidence of planets like our own.

Another thing most of the techniques share is the observation that a star with a planet around it will "wobble." The reason that planets move in regular circles -- orbits -- around stars is because of the extremely high gravity of those stars. But planets are heavy too, so they should constantly be pulling at their star just a little. The result is that the star will be spinning in a very tiny little circle of its own as it's tugged around by its planets. From the perspective of our telescopes, a star should look as though it wobbles, ever so slightly.

The trouble with this sort of technique is that, if you're looking for Earth-like planets, it's going to be hard to find them. The most obvious wobbles are going to be caused by huge planets (in other words, this is a method that's only going to find you a lot of Jupiters). You can demonstrate the problem by swinging an object in a circle on the end of a string. You'll find yourself pulling against the weight as you spin it around. But the heavier object, the more you'll find yourself pulled. Another thing you'll discover is that spinning a heavier object on a short rope means you have to spin it much more quickly, as well.

And the farther away a planet is, the less its effect on the star will be. So this sort of technique is likely to find you planets that are very close, and very big. A large category of such planets now exists, known as Hot Jupiters -- so named because they tend to be so close to the sun that they will be heated beyond belief. Hot Jupiters have three things that make them easy to detect by the wobble method: they're big, they're close, and they orbit very quickly, which makes the wobble especially obvious and easy to spot

Bellerophon, or 51 Pegasi b, is a classic Hot Jupiter: it is at least half the size of Jupiter (150 times as heavy as Earth), but it is closer to its star than Mercury is to ours, and its year lasts only four of our days here on Earth. These planets are too big to have formed that close to the star on their own, and the current theory is that they in old star systems; they probably formed a long way out and then gradually crept inward toward their stars in a process known as planetary migration.

Looking for Planets Today: Crossing the Sun

In the past five years or so, the research frontier has shifted from the wobble methods above to what are called "transit" detections, which you could think of as "crossing" methods. With new and particularly sensitive telescopes, it is possible to take very detailed measurements of the amount of light a star is giving off -- and that light is normally pretty constant, at least for the stars we expect to find planets around. But every so often, a planet will move between us and its star. As soon as it does that, the amount of light that is reaching us will decrease slightly. We can track those reductions and guess at the presence and size of a planet.

Two exciting space missions are currently undergoing these sorts of studies from orbit: the French COROT, and the American Kepler telescopes. Kepler was launched last year, and it is sensitive enough that (we hope) it will be able to measure the detection in light that happens even when an Earth-sized planet passes in front of its star. As an indication of how powerful these methods were, similar studies done using the older Spitzer telescope have estimated the surface temperature of an extrasolar planet, based on the intensity of the light change.

This method gives us new opportunities, but it still suffers from major disadvantages. One is a new one: the planet has to pass between us and the star. This means we have to be lucky enough for two things to happen: we have to line up perfectly so that that actually happens (if the planets orbit at the wrong angle, it just will never happen). Roughly 10% of planets that are very close in (like the hot Jupiters) will show up on these scans. Unfortunately, less than one in a hundred planets orbiting at the distance of Earth will.

Another problem which then crops up is that planets orbiting at Earth-like distances will have very long years. (Long relative to the Hot Jupiters, that is: Earth's own orbit seems relatively short when you put it up against those of planets like Saturn and Neptune, which circle the Sun every 30 and 165 years, respectively.) So there is only going to be a relatively short period of time where a given planet will be at the right point in its orbit for us to be able to see it.

Looking for Planets Tomorrow: Reasons for Hope

The best thing we have going for us, fortunately, is the odds: we have essentially a countless number of candidate stars to inspect, so we're bound to luck out eventually even if we're looking for something that is relatively uncommon. Several relatively recent discoveries bear out that theory, because they're getting to be more and more Earth-like in their appearance. There's even a category that houses most of these discoveries: the Super-Earths, which are up to ten times our size. (Any bigger than that and you get into gas giant territory, which are where Jupiter, Saturn, Uranus, and Neptune are.)

To be Earth-like, however, we begin with two basic criteria: it has to fit into this category of roughly Earth-like size, but it also has to fit within an orbit roughly like Earth's, where it's neither too hot nor too cold to plausibly believe that life could exist. Within our own solar system, Mars is hovering close to the outside edge of this Earth-like envelope, while Venus is just on the inside of it. Remember that both of these criteria make it difficult to fulfill the conditions of finding a planet: they're not big, and they're not close to their sun. (And with an orbit that big, they're likely to have a year more similar to our own as well, which also complicates things.)

One of our best bets yet is named COROT-7b, so named because its parent star (its sun) was the seventh in the list of stars surveyed by the French COROT mission since 2006. Right now, COROT-7b is one of the smallest planets we've ever found, a little bit less than twice the size of Earth. Our estimates of its mass have also proved that it is heavy, but not heavy enough to be all iron (the most common metal), implying that it might have something light on it, like water. (Searching for necessary life ingredients like water is another necessary step in finding Earth-like planets, but one that it's not worth worrying about until you've settled the astronomical criteria first.) Unfortunately, COROT-7b is way too close to its sun: it's 1/23 the distance of Mercury from our Sun, and it's year is shorter than our day.

Which leaves the Gliese 581 system, which a Chilean observatory has been studying. Gliese 581 is a red dwarf, meaning normally we'd consider it too cold and dark to be capable of supporting life on an orbiting planet. The advantage of red dwarfs, however, is that because they're cooler, the Earth-like planets will be closer in -- and easier to find. Around Gliese, we've found several record-breakers: 581e is the smallest planet we've ever found (less than two times as heavy as Earth itself), and both 581c and 581d within the habitable "warm" zone of its star.

Unfortunately, that puts 581c so close to Gliese 581 that it's probably tidally locked, a process that is identical to what's happened to our moon: it spins around itself at precisely the same rate that it spins around the sun (in our moon's case, the Earth). As a result, one side always faces the sun, and one side always faces outwards. In our case, that just means we always see the same part of the moon -- we always see the "man in the moon." But in a planet's case, that means that wherever you are, it will either be day forever, or night forever.

Which leaves us 581d, so far. 581d is on the other side of the envelope -- far enough away that there's some dispute about whether it's slightly too cold to maintain liquid water reliably. Assuming it's just warm enough, which to be fair is not likely, it's our best candidate so far. Which is why, last year, Bebo sent a digital radio signal which will reach Gliese 581c in just 20 years. On the off chance that there's anyone there advanced enough both to receive it and understand it, we could be hearing from little green men some time around 2050.

What If We're Not Being Creative Enough?

Of course, this obsession with finding "Earth-like" planets isn't necessarily doing us any favours. Perhaps it just results from our lack of imagination about where we're likely to find life. For example, after Earth (and the distant past on Mars), the two most likely places to find the liquid environments necessary for life to evolve are water on Europa and methane on Titan. Those are moons, orbiting Jupiter and Saturn respectively, so they're nowhere near the "Earth" zone we've been searching around other stars. And in a billion years or so, when our own sun expands to the point that it chargrills the Earth, the temperature on Titan might heat up enough to have even more Earth-like conditions.

Any little bacteria on Titan will have to grow fast, though. A few hundred million more years, and our Sun will be out of fuel. Kaput. By then, certainly our descendants will either be all dead, or they'll have moved on to somewhere else.