Centauri Dreams

The most interesting thing about the worlds known as KOI 55.01 and KOI 55.02 is not just the fact that they are — if current thinking holds — the smallest planets yet detected around an active star other than our Sun, but that they are evidently survivors of the most extreme kind of experience. KOI 55, their host star, is of subdwarf B class, the exposed core of a red giant that has lost most of its gaseous envelope. The two planets that circle it are in such tight orbits that they would have been engulfed when the central star went through its red giant expansion.

What a scenario, one we’ve often contemplated in these pages as we look toward the future of our own Sun. We tend to think in terms of planets that survive the red giant phase by orbiting far enough from the primary not to be swallowed up in it — smaller worlds like Mercury, Venus and the Earth would not survive the experience. But KOI 55.01 and KOI 55.02 evidently were swallowed, and probably represent the remains of gas giants that underwent the plunge. At present, they are thought to have radii 0.76 and 0.87 times the Earth’s radius.

Elizabeth Green (University of Arizona) explains:

“When our sun swells up to become a red giant, it will engulf the Earth. If a tiny planet like the Earth spends 1 billion years in an environment like that, it will just evaporate. Only planets with masses very much larger than the Earth, like Jupiter or Saturn, could possibly survive.”

Green participated in this work as a member of a team led by Stephane Charpinet (Institut de Recherche en Astrophysique et Planétologie, Université de Toulouse-CNRS). The team’s findings make an intriguing case for the proposition that planets may directly influence the development of their host stars in this late phase of stellar evolution. Here they seem to have done it by contributing to the mass loss that is necessary for a subdwarf B star to form.

Here’s Green again on how this would work:

“As the star puffs up and engulfs the planet, the planet has to plow through the star’s hot atmosphere and that causes friction, sending it spiraling toward the star. As it’s doing that, it helps strip atmosphere off the star. At the same time, the friction with the star’s envelope also strips the gaseous and liquid layers off the planet, leaving behind only some part of the solid core, scorched but still there.”

Finding remnant planets around a stellar core, planets that have passed through the maelstrom of red dwarf expansion, is surprising enough, but the work also came from an unexpected direction. The team’s objective had been to use Kepler data to study astroseismology, examining the rhythmic expansions and contractions that pressure and gravitational forces produce inside a star as it burns. It’s useful stuff because with enough data, astronomers can estimate the star’s mass, temperature and size, as well as learning something about its internal structure.

Astroseismology takes time as researchers accumulate information about the star’s variations in brightness and compare these to theoretical models of stellar interiors. Green had already been studying hot subdwarf stars in the galactic plane and had accumulated spectra of KOI 55 with instruments on Kitt Peak before the Kepler mission was launched. Kepler was able to show the star’s pulsational modes with great clarity and it was in the midst of examining the Kepler data that periodic modulations occurring every 5.76 and 8.23 hours began to turn up. The researchers were able to show that the modulations could not have been caused by internal pulsations.

Two planets were the best explanation, each orbiting closer to their star than Mercury is to the Sun. Conditions on these worlds today must be hellish, as KOI 55 is much hotter than the Sun, but their tight orbits tell us that things were once worse still, when the planets would have been engulfed during the star’s expansion.

“I find it incredibly fascinating that after hundreds of years of being able to only look at the outsides of stars, now we can finally investigate the interiors of a few stars – even if only in these special types of pulsators – and compare that with how we thought stars evolved,” Green said. “We thought we had a pretty good understanding of what solar systems were like as long as we only knew one – ours. Now we are discovering a huge variety of solar systems that are nothing like ours, including, for the first time, remnant planets around a stellar core like this one.”

The paper is Charpinet, “A compact system of small planets around a former red-giant star,” Nature 480, 496-499 (22 December 2011). Abstract available. This University of Arizona news release offers further details.

I’m delighted that we keep finding solar systems so different from our own. The discovery of two new planets that are roughly the size of the Earth just confirms the feeling — in a galaxy of dazzling fecundity, every system we look at has its own peculiarities to instruct and delight us. The system around the star called Kepler-20 (from its designation by the space observatory studying planetary transits) is a case in point. Yes, it has small, rocky worlds, but it also has three larger planets, and all five orbit closer than the orbit of Mercury in our own system. Kepler-20 is a G-class star somewhat cooler than the Sun located some 950 light years from Earth in the constellation Lyra.

Moreover, while we once assumed that smaller planets orbited close to stars while larger gas giants orbited further out in the system (again based on our own system and our assumptions about it), our new discoveries point to different scenarios. In Kepler-20 we have a system where the larger planets (all smaller than Neptune) orbit in alternating fashion with the rocky planets. We get a big planet and then a little one, then another large one, another rocky planet, followed by a third large world. Kepler 20-b, 20c and 20d are the large planets, with diameters of 24,100, 39,600, and 35,400 kilometers respectively, and orbital periods of 3.7, 10.9, and 77.6 days.

“The Kepler data are showing us some planetary systems have arrangements of planets very different from that seen in our solar system,” said Jack Lissauer, planetary scientist and Kepler science team member at NASA’s Ames Research Center in Moffett Field, Calif. “The analysis of Kepler data continues to reveal new insights about the diversity of planets and planetary systems within our galaxy.”

As you would expect, it’s the small planets that command the headlines this morning, their presence seen as another step in the goal of finding an Earth-sized planet in the habitable zone. These worlds at least fit the bill in terms of size, although water on the surface is out of the question, as temperatures are thought to reach 760 degrees Celsius on the inner world and 426 degrees Celsius on the outer. With an orbit of 6.1 days, Kepler 20e is equivalent to 0.87 times the size of Earth, with a diameter of 11,100 kilometers. Kepler-20f, orbiting the host star every 19.6 days, has a diameter of 13,200 kilometers, quite close to the size of the Earth. Both planets are expected to be rocky, with masses less than 1.7 and 3 times that of Earth.

Image: This chart compares the first Earth-size planets found around a sun-like star to planets in our own solar system, Earth and Venus. NASA’s Kepler mission discovered the new found planets, called Kepler-20e and Kepler-20f. Kepler-20e is slightly smaller than Venus with a radius .87 times that of Earth. Kepler-20f is a bit larger than Earth at 1.03 times the radius of Earth. Venus is very similar in size to Earth, with a radius of .95 times that our planet. Prior to this discovery, the smallest known planet orbiting a sun-like star was Kepler-10b with a radius of 1.42 that of Earth, which translates to 2.9 times the volume. Credit: NASA/Ames/JPL-Caltech.

So there we are, the smallest planets yet confirmed around a Sun-like star, a likely case of planetary migration in which planets form much farther from the host star and migrate inward because of interactions with the protoplanetary disk. Although Kepler’s transit methodology is finely tuned, the radial velocity signature of planets like the smaller two around Kepler-20 is out of the detection range of current technology. Kepler’s ‘long stare’ at the starry patch in Cygnus, Lyra and Draco was able to turn up the planetary candidates here, but intense computer simulations by way of follow-up were necessary to demonstrate the likelihood of the smaller detections being planets.

The paper is Fressin et al., “Two Earth-sized planets orbiting Kepler-20,” published online in Nature (20 December 2011). Abstract available.

So-called ‘super-Earths,’ planets larger than the Earth but smaller than Neptune, pose problems to our theories of planet formation. The most recent illustration of this came in the announcement that the candidate planets found by Kepler had now reached 2,326. Remember, many of these will not be confirmed — they’re candidates — but taken in the aggregate, what is interesting here is that one-third to one-half of these candidates fit the super-Earth category. And just as we had a problem with ‘hot Jupiters’ in trying to figure out their orbital position, many of these new planets are likewise in orbits close to their parent star, where the models say they shouldn’t be.

Things seemed so much simpler when we just had a single solar system to worry about, our own. Then, the idea of core accretion could readily account for everything we saw. The dust in the protoplanetary disk was thought to have aggregated into small planetesimals which, in the course of time and numerous collisions, bulked up into planets. It made sense that planets in the inner system would be smaller because the inner part of the disk was thought to have less material for growth, whereas an outer planet would become larger, growing into a gas giant massive enough to pull in a thick atmosphere from the surrounding disk.

As this article by Eric Hand in Nature News points out, the basic model was challenged by finding gas giants in tight orbits, forcing the development of a migration model in which these huge worlds formed out beyond the ‘snow line’ and later made their way into the inner system. Now we have to account for the latest super-Earth findings, as the article points out:

…these models predicted that anything reaching super-Earth size should either become a gas giant or be swallowed by its star, creating a ‘planetary desert’ in this size range. Kepler’s discoveries wreck those predictions. “It’s a tropical rainforest, not a desert,” says Andrew Howard, an astronomer at the University of California, Berkeley. “We hope the theory is going to catch up.”

Hand goes on to talk to Jack Lissauer (NASA Ames), who suggests that some super-Earths could have begun as smaller cores in the outer system that simply never reached the kind of runaway growth that would lead to a Jupiter-class world. Lissauer thinks a planet like this could grow to super-Earth size, and his ideas could explain some of the super-Earths we’ve found with low densities, implying a small rocky core surrounded by a large gas envelope. Even so, we’re still not able to explain denser super-Earths of the kind that are now beginning to be detected.

Kepler’s gradual revelation of smaller and smaller worlds keeps pointing us in the direction of new planet formation models like Norm Murray’s. The astrophysicist (University of Toronto) has tweaked the migration model to have the process of planet formation occur after planetesimals have migrated inward on their own, with the subsequent accretion occurring near the host star. Who knows whether this theory will stand up to the next wave of Kepler results, but if there is one thing we’ve learned from the exoplanet hunt so far, it’s that surprises are abundant, and too doctrinaire a view at this stage is simply asking for revision down the road.