When I was a kid, I read scifi novels and watched TV shows depicting planets orbiting binary stars — two stars orbiting each other. When I got a bit older, and got to know the science a little better, I figured that was pretty unlikely.
Then when I got even older and actually learned how science works on a more subtle level, I realized the Universe is smarter than we are, and I began to wonder if such a system really could work. And then we started discovering hundreds and then thousands of planets orbiting other stars… with these worlds coming in such a dazzling and unexpected diversity. It became clear the Universe really was smarter than we are, or at least more clever, making planets in ways we didn't think possible.
And sure enough, mocking any form of scientific hubris, astronomers started finding planets orbiting binary stars. The Kepler Space Telescope looked at over 150,000 stars for years, searching for the dips in starlight that would occur if a planet passed directly in front of the face of the star in a sort of mini-eclipse — what astronomers call a transit. About 2,900 of those stars were what we call eclipsing binaries, where the orbit of the two stars is seen edge-on, so one star passes alternately in front of the other. We think planets would form in the same plane as the stars, so those are the kinds of binaries where Kepler could hope to spot transits.
Out of that horde, 9 eclipsing binary stars showed evidence of planets. They were all what are called circumbinary planets, meaning each planet orbited both stars (as opposed to orbiting just one of the two stars in the pair). Eight of these binary systems had only one planet, but in 2012, astronomers announced that the binary star Kepler-47 had two planets! The inner one was called Kepler-47b and the outer Kepler-47c.
It was looking like circumbinary planets didn't just happen in a fictional setting long ago in a galaxy far, far away. Kepler-47 is 3,300 light years away. Far, but in our own galaxy. And quite real.
But it gets better: The same astronomers have now found a third planet in the system!
Let me take a moment to describe the stars. One is very similar to the Sun in mass and size (and is called the primary). The other (the secondary) is about a third of the Sun's mass, and is much cooler and fainter. The primary is about 200 times brighter than the secondary; so much brighter that all the transits seen by the exoplanets are across the primary star. Transits across the secondary would be so faint they're essentially invisible. The two stars are about 12 million kilometers apart and take 7.5 days to orbit each other once. They're at least 3.5 billion years old, and possibly much older.
Now, for the planets. In the original discovery paper, the astronomers noted that besides the transits from the two planets, there was a single "orphan" transit that looked real but was not from either of the two planets they had found. After publishing that first paper, more observations by Kepler revealed two more transits that matched that orphan one. That allowed them to get a period for the planet. Going back over all the data again they found three more transits, very weak ones but there. That makes six total, all indicating the presence of the third planet: Kepler-47d* .
The new planet orbits the binary in between the other two. Weirdly, the depths of the transits of 47d changed with time; literally, the amount of light the planet blocked from the primary star appeared to increase. That depth relates to how big the planet is, and how it passes in front of the star. A big planet blocks more of the star, so the transit is deeper. But if a planet just nicks the edge of the star as seen from Earth the transit will be shallower.
It doesn't make sense that the planet is getting bigger, so the orbit of the planet must be changing. And sure enough, that's what's happening. The gravity of the system with two stars and (at least) three planets is a mess. The planets get tugged around by each other, and that makes their orbits wobble around. This is called precession, and it's the same thing you see when the axis of a spinning top makes a slow circle even as the top rotates rapidly.
As the orbit precesses, that changes how deeply the planet cuts across the face of the star as seen from Earth (for what it's worth, this is very similar to the reason we don't have a solar eclipse every month; the Moon's orbit is tilted and only lines up with the Sun at special times). So at first the transits of 47d cut a very shallow chord across the star's face, but as the orbit wobbles that chord got deeper… and is getting deeper still. Eventually, though, the transit will get shallower again and the whole thing repeats.
So what are these planets like? In order from the stars (so from inner, to middle (the new one) to the outer planet), they are 3.1, 7.0, and 4.7 times the diameter of the Earth. So, big. They take 49, 87, and 303 days to orbit the stars, too. That's interesting! It puts the outer planet towards the inner edge of the habitable zone, the distance from the stars where liquid water could exist on a solid body with an atmosphere. The inner planet is far too close in, but the middle planet is right on the inner edge of the "optimistic" habitable zone (that is, crossing your fingers and allowing for everything to go right so that the HZ is as wide as possible). So 47d is even warmer than 47c, but perhaps reasonably within human tolerances.
Unfortunately, transits don't tell you the masses of the planets, so the densities can't be determined. That's what tells you what the planets are made of; a very dense planet has lots of rock and metal, where a low-density planet is likely to be a gas giant. Still, given their large sizes, these are very likely to be more like Neptune and Saturn than Earth. Ah well.
But I have to wonder… maybe these planets have moons. And maybe some of these moons are pretty big. Maybe Earth-sized. That's not too far outside the realm of possibility, and if so, well, they're in the right place to maybe be nice and warm. Hmmmm.
And this all sounds familiar. Kepler-47c and d are hot, orbiting a binary, with one star yellow and the other red… heyyyyy. It's like if there's a bright center of the universe, this is the planetary system that it's farthest from.
Sadly, 47c and d are likely to be a gas giants of some sort and not desert outposts. Still. Planets orbiting binary stars may be pretty common; remember, Kepler could only see planets orbiting eclipsing binaries. This is a small fraction of all binaries, which by number make up a majority of stars in the sky. So who knows?
Every time we find another planet like this, it gives me a new hope.
*Planets are given the lower case letter after the star name in order of discovery. So even though this planet is in between the other two, it gets a d as its letter.