Mirror Earth (25 page)

“I always thought the debate over Pluto was a stupid argument,” said Marcy, “and most of my colleagues thought it was a stupid argument as well. We would e-mail privately to each
other that all this hullabaloo about Pluto was much ado about nothing.” He was always wryly amused, however, by the fact that while everyone was arguing about whether Pluto was a planet, the eight planets everyone did agree on were, as Marcy put it, “two qualitatively different types of beasts. And no one seemed to be bothered by the fact that they are all so different and yet we call them all planets.” To realize that the Earths are so qualitatively different is really important, he said, because it bears so strongly on the question of how easily and how often they form. “It's easy to pretend we have the answer to that, theoretically, but we don't.”

What we do know, thanks to Marcy's Eta-Sub-Earth Survey, and Dave Charbonneau's MEarth Project, and the radial-velocity work of Mayor, and the results that were spilling from Kepler's light detectors, is that the dividing line between Neptunes and Earths might fall a lot closer to Earth than to Neptune. “We have GJ 1214 b,” Marcy pointed out, “which is Dave Charbonneau's water world. And Hat-P-11 b, and Gliese 436 b, which my group found, and they all have radii smaller than four Earths, and densities of about two, which means they formed in an environment that had water and almost certainly gas as well—not at all the way the Earth formed.”

If Earth-size worlds can be vastly different from the original Earth, in short, the search for life could get very complicated. This unsettling idea was circulating in the exoplanetology community well before Dave Charbonneau found the water world GJ 1214 b. In 2003, for example, a Harvard postdoc
named Marc Kuchner wrote a paper for the
Astrophysical Journal
. “Discussions of extrasolar planets,” it said, “often quietly assume that any object with [one Earth mass] orbiting in a star's habitable zone will be terrestrial, i.e., composed mostly of silicates and iron-peak elements like the Earth. However, we suggest that the habitable zones of nearby stars could harbor other similar-looking beasts.” The beasts he had in mind in this paper were water worlds, covered with oceans many hundreds of miles deep and surrounded by atmospheres rich with steam. The French planetary scientist Alain Léger proposed a similar idea in a paper published in
Icarus
in 2004. “A new family of planets is considered,” he and his colleagues wrote, “which is in between rocky terrestrial planets and gaseous giant ones: ‘Ocean-Planets.'”

A few years later, Kuchner, by now a postdoc at Princeton, co-authored a paper with Sara Seager. Titled “Extrasolar Carbon Planets,” it pointed out that the giant interstellar clouds of gas and dust that collapse to form solar systems aren't identical. They do have the same general mix of elements, but the proportions can surely vary. One of the likely variables is the ratio of carbon to oxygen. Our Sun has about half as much carbon as oxygen, which reflects what our original cloud was made of and what the solar nebula was made of as well. But it's easy to imagine an interstellar cloud with a different ratio. With a higher percentage of carbon, the planets that would be born of that extrasolar nebula could in principle be mostly carbon, not rock, with a core of pure diamond. Kuchner and Seager went on to ask:

What other possible kinds of planets are there? The planet zoo now contains silicate planets (e.g., Earth and Mars), hydrogen and helium planets (e.g., Jupiter and Saturn), water planets [Neptune and Uranus] (which perhaps we might think of as oxygen planets), iron planets, and carbon planets. A glance at a table of solar abundances suggests that next we might consider helium, neon, and nitrogen planets.

At the time they were writing, of course, the zoo “contained” carbon planets and iron planets in the sense that they were theoretically possible. Seager and Kuchner had described water planets; the iron planets they were referring to had been described by David Stevenson, a Caltech planetary scientist. “Yes,” Stevenson said in a recent interview, “you could have a planet that was essentially an Earth-size cannonball.” Stevenson was familiar with Kuchner and Seager's work, and agreed with the general proposition that Earth-size planets could come in all sorts of flavors. “There's a tendency to look at our solar system and think, ‘We've got one of these and one of those' and decide that's the whole range of possible planet types.”

But our solar system lacks certain obvious types of planets. Never mind giant cannonballs or giant lumps of carbon with quadrillion-carat diamonds inside. “What about a super-Ganymede?” Stevenson asked, referring to Jupiter's, and the solar system's, largest moon. It's easy to imagine something like that forming out of the early solar nebula; it just happened not to. But given the enormous number of exoplanets that have been found so far with the limited searches we've been
able to do, he said, “if you have a good physical reason to think something is possible and haven't found it yet, you probably will.”

Even after Dave Charbonneau found GJ 1214 b, however, it wasn't absolutely certain that he'd found a water world. Based on its overall density, it could be half rock, half water, surrounded by an atmosphere thick with steam (since the planet is so close to its star). But it could also have a smaller rocky core with a huge thick atmosphere of hydrogen gas. The way to distinguish between the two would obviously be to take a look at the atmosphere, and fortunately, the star and planet are only forty light-years away. This doesn't make such an observation easy, but it at least makes it possible: You look at the star's light as it passes through the planet's atmosphere, taking a so-called transmission spectrum, just as Charbonneau did when he first detected sodium in the atmosphere of HD 209458 b, and see what elements or compounds are there.

If GJ 1214 b has a hydrogen-rich atmosphere, you should expect to see water. If the atmosphere is rich in water, you shouldn't. This naturally sounds a bit crazy, but Zach Berta, the MEarth team member whose “nice shooting” had spotted the planet in the first place, explained why it's not. “A hydrogen-rich atmosphere,” he told me, “will still have some water in it, and because the atmosphere is physically big [hydrogen is much lighter than water vapor, so it expands to fill a bigger volume] you can see the water easily.” A water-rich atmosphere, by contrast, would show very weak water features. That's not, said Berta, because there isn't much there, but
because the atmosphere is dense and squashed down, it's relatively small. In early 2012, Berta finally did observations that confirmed the squashed-down, water-rich version.

There's also some evidence for carbon planets. In 2010, Princeton postdoc Nikku Madhusudhan and University of Central Florida faculty member Joe Harrington used several telescopes to look at the transmission spectrum of WASP-12b, a hot Jupiter found a year earlier by the UK-based Wide Angle Search for Planets. They found more than twice as much carbon and one hundred times as much methane (which is made of carbon and hydrogen) as you'd expect to see in a planet like this. If there's an Earth-size planet in this system, it could plausibly have a core of pure diamond, with diamond continents sloping down to seas of tar.

The very fact that Wasp-12b exists, however, might mean that its solar system has no Earths at all—and the same could be true of any solar system with a hot Jupiter. The reason it took so long to find the first exoplanet was that nobody imagined a planet as big as Jupiter hugging tightly to its star. It certainly couldn't have formed there, and the first explanation theorists could come up with was that it had somehow spiraled in from its original location, much farther out. This wasn't necessarily good news for finding Earth-like planets, since an inspiraling Jupiter would have disrupted the orbits of anything in its way, even flinging other planets out into deep space.

More recently, though, evidence has turned up to suggest a different scenario. It's been uncovered in part by Josh Winn at MIT, the astronomer-turned-science-journalist-turned-astronomer who is working on the TESS mission. In our own
solar system, the planets orbit in the same direction as the Sun rotates, and in pretty much the same plane as the Sun's equator. That makes sense: The Sun and the planets all condensed out of the same original spinning pancake of gas and dust, so everything should move in the same direction.

But by using something known as the Rossiter-McLaughlin effect, which was first posited all the way back in the 1890s, Winn, Marcy, and others realized that between a quarter and half of all the hot Jupiters are way out of line. They orbit at sharp angles to their stars' equators, and in some cases even revolve in a direction opposite to the stars' rotations. “This burst on the scene about a year ago,” Marcy said. “The Europeans found a few, we found a few. That was disturbing, but as a good scientist, you sweep it under the rug. Maybe it's just a fluke.” But then, he said, five consecutive WASP planets were all found to be misaligned. “Now you could no longer sweep. We all smacked ourselves on the forehead.”

The Rossiter-McLaughlin effect works like this: When Geoff Marcy or Michel Mayor measures a star's redshift or blueshift—the shift in light that happens when a star is moving away from you or approaching—they're actually looking at an average. The star itself may be moving away or coming toward you or standing still. But even if the star is standing still, its rotation means that one edge is moving toward you and the other is moving away.

Now, imagine a planet transiting across the face of the star. If it's orbiting in the same direction as the star, it will block the approaching, blueshifted half first, then the receding, red-shifted half. So if you look carefully at the star's spectrum, you
should see a subtle change in the mixture of redshifted and blueshifted light as the planet moves across. If the planet is traveling backward, though, you should see a dimming of the redshifted half first, then the blue: The change will happen in reverse. And if it's orbiting at some sharp angle, the pattern will be somewhere in between. The Rossiter-McLaughlin effect can even tell you whether a planet is traveling across the star's meaty middle or merely grazing its edge. Every possible combination of route and direction across the star's face has a unique signature.

The significance of planets that go backward and are otherwise out of kilter means that inward migration might not be the explanation for the hot Jupiters after all. “Back in 1995,” said Marcy, “Doug Lin [of Santa Cruz] jumped up and said, ‘I have this migration model.' It had problems, but we all kind of accepted it. It doesn't explain retrograde motion, though, which shows that our best idea for the past fifteen years was mostly wrong—or at least, wrong half the time.” The best alternative explanation is that planets, as Marcy puts it, “slingshot themselves.” That is, when they approach one another too closely in the early days of a solar system's existence, they tend to fling one another around with their gravity. “I find it lovely and slightly embarrassing,” said Marcy, referring to the fact that exoplaneteers may have been barking up the wrong tree for a decade and a half, “but this is what makes science wonderful. Sometimes you have to throw the baby out with the bathwater.”

It probably won't be the last time, either. The problem with creating a convincing theory of planet formation and motion
is no longer that we have too few planets to work with. With hundreds in hand and hundreds more coming from Kepler, there are now plenty. The problem is that astronomers are still working with a very biased sample: They can see only the planets that are easiest to see. At first, these were the hot Jupiters. Now, thanks to better radial-velocity measurements and to transits, they can find smaller planets, and planets farther out from their stars.

But even with the progress we've made in planet detection, we'd still fail to find seven of the eight planets in our own solar system with all existing techniques if it were just a few tens of light-years away. “I wouldn't say it's foolish to be making theoretical predictions and working on the theory of planet formation,” said Scott Tremaine, a theorist at the Institute for Advanced Study and Eric Ford's thesis adviser—the one who went off hiking while Ford was doing his research. “But I think there has been a culture in which a lot of theorists think of it as a little bit like playing the lottery. You make a prediction and if it turns out to be right, then you're a hero, and if it turns out to be wrong, it's going to kind of be forgotten about and so you don't really suffer from it.”

With that in mind, here's a really intriguing theory from David Stevenson, the same Caltech planetary theorist who talks about iron cannonballs the size of Earth. Back in the late 1990s, Stevenson wrote a paper based on the slingshot idea. As Jupiter was forming—according, at least, to conventional planetary theory—smaller, Earth-size objects should have formed as well, with atmospheres made of the same hydrogen that now shrouds Jupiter in a layer of gas thousands of miles deep.
Most of these objects would have been absorbed into the giant planet, but some could have been slingshotted entirely out of the solar system. “This part is not in dispute,” he said. “Jupiter and other giant planets were perfectly capable of ejecting stuff. So that's rather straightforward.”

The part that is not so straightforward is Stevenson's idea that these Earth-size bodies, hurtling through interstellar space no longer tethered to any star, would retain their dense, hydrogen-rich atmospheres. Hydrogen, it turns out, is a greenhouse gas, just like carbon dioxide or water vapor: It keeps heat from escaping out into space. On our own planet, that heat comes from the Sun. On these lonely worlds, it would come instead from within, through the decay of radioactive elements in their cores—the same heat source that keeps Earth's core partially molten. “The physics is not in doubt,” he said. “It's just a question of whether the planet would retain its atmosphere when ejected. I have no doubt it happens; I just don't know how often. I suspect it's fairly common.”