≡ Menu

Kepler-78b: Rocky World in Unusual Orbit

When I was first learning about astronomy by reading every book I could find on the subject — I spent a lot of time at the library in my youth — I was fascinated to hear stories of a planet closer to the Sun than Mercury. The French scientist Urbain Le Verrier pondered the existence of such a world in the 19th Century, wondering if it wouldn’t explain peculiarities in Mercury’s orbit. When an amateur astronomer named Edmond Modeste Lescarbault claimed he had observed a transit of such a planet in 1859, Le Verrier’s investigation satisfied him that the detection was legitimate. He went on to announce the discovery of the planet he dubbed Vulcan in Paris in 1860.

Despite a string of other observing claims later in the century, the existence of Vulcan is now discounted, although the possibility of asteroids in tight solar orbits hasn’t been ruled out. As for Mercury, it fell to Einstein to demonstrate that apparent anomalies in its orbit could be explained by his theory of General Relativity. Vulcan was an intriguing story in those days of my early astronomy reading, as I asked myself how anything could survive closer to the Sun than Mercury, and conjured up hellish landscapes dominated by a sky-swallowing star above.

But if we don’t have Vulcan in our own Solar System, it’s clear that many other systems do have planets in extraordinarily close orbits. Yesterday I listened in on the news conference describing Kepler-78b, one of the new class of ‘ultrashort period’ planets that Kepler has put into our databases. To qualify as an ultrashort period planet, the world has to orbit with a period of less than twelve hours. Defying the imagination, Kepler-78b bests even that figure, orbiting a star somewhat smaller and less massive than the Sun in a mere 8.5 hours.

78b

Image: Kepler-78b is a planet that shouldn’t exist. This scorching lava world, shown here in an artist’s conception, circles its star every eight and a half hours at a distance of less than 1.6 million kilometers. According to current theories of planet formation, it couldn’t have formed so close to its star, nor could it have moved there. Credit: David A. Aguilar (CfA).

Dimitar Sasselov (Harvard-Smithsonian Center for Astrophysics) told the audience that we had now found about a dozen objects orbiting various stars with periods of between three and ten hours, creating serious questions about how such planets form and survive. It’s even conceivable, said Sasselov, that Kepler-78b is the core of a former gas giant. Amidst the questions, what we do know is that its radius is about 1.2 times that of Earth and its mass 1.7 times Earth’s. With both size and mass measured, it’s possible to calculate the density, and that figure works out to 5 grams per cubic centimeter, which is a density much like our planet’s.

Kepler-78b, then, is most likely made of rock and iron. The work to demonstrate this points to the synergy between two kinds of observation and also relies on a dual investigation that produced confirming results. The planet was first observed by Kepler using the transit method, in which a planet moves in front of its star as seen from Earth and thus creates a dip in its lightcurve. Andrew Howard and team at the University of Hawaii at Manoa then used the Keck ten-meter instrument with the HIRES spectrograph to measure the star’s radial velocity, thus allowing the determination of the planet’s mass. The density reading follows from the size and mass.

Howard told the news conference that thirty hours of observations over eight nights went into the Keck work, which was a difficult measurement because the star is young and has many starspots. With an orbital period this short, the team was able to observe an entire orbit in a single night, however, and the starspot ‘noise’ could be filtered out. At the same time, Francesco Pepe (University of Geneva), using data from the HARPS-North spectrograph in the Canary Islands, was leading a companion study that verified the Keck team’s radial velocity observations. “The gold standard in science is having your findings reproduced by other researchers,” said Howard. “In this case, we did not have to wait for this to happen.”

78b_2

Image: This illustration compares our Earth with the newly confirmed lava planet Kepler-78b. Kepler-78b is about 20 percent larger than Earth, with a diameter of 9,200 miles, and weighs roughly 1.8 times as much as Earth. Credit: David A. Aguilar (CfA).

So we have an Earth-sized planet with Earth-like density in an orbit that remains a mystery. The star Kepler-78b orbits was larger than it is now when the planetary system was forming. “[The planet] couldn’t have formed in place because you can’t form a planet inside a star,” said Sasselov. “It couldn’t have formed further out and migrated inward, because it would have migrated all the way into the star. This planet is an enigma.” It is also a world that will eventually be torn apart by gravitational forces as it is drawn inexorably closer to the star, although according to this CfA news release, that won’t happen for another three billion years.

With temperatures as high as 3100 K on the surface, Kepler-78b is some 40 times closer to its star than Mercury is to the Sun, making my musings about Vulcan’s sky seem relatively tame. The papers are Howard et al., “A rocky composition for an Earth-sized exoplanet,” published online in Nature 30 October 2013 (abstract), and Pepe et al., “An Earth-sized planet with an Earth-like density,” published online in Nature 30 October 2013 (abstract).

tzf_img_post

Comments on this entry are closed.

  • Alexander McLin October 31, 2013, 10:00

    Is it possible, assuming that the world is the core of a former gas giant, that the outgassing of the planet’s atmosphere envelope caused the planet to spiral closer to the star?

  • David Cummings October 31, 2013, 10:23

    This planet is so close to its sun that it must be tidally locked and always presenting the same face to the sun. I wonder if the dark side is cool, or if the heat from the star bakes the whole planet through and through so that even the dark side is extremely hot.

  • David Cummings October 31, 2013, 10:42

    “[The planet] couldn’t have formed in place because you can’t form a planet inside a star,” said Sasselov. “It couldn’t have formed further out and migrated inward, because it would have migrated all the way into the star. This planet is an enigma.”

    Maybe that’s what Kardashev Type II civilizations do… move rocky planets into close solar orbits to set up energy factories of some kind.

  • Rob Flores October 31, 2013, 10:52

    Yes the fly in the ointment when trying to find Twin earth’s, a plethra of
    hot terretrials (which no one expected.) They mentioned that it was possible that it’s the core of a gas giant. I guess we can estimate as to how large the gas giant was. Based on their measurements it sounds like the original planet something like 2-3 times the size of Neptune, based on the
    core that’s left. Which makes sense since Neptune class planets seem to be a common feature in the Kepler findings.
    What about Mercury in our own solar system? Could it have been micro jovian, say 1/3 the mass of neptune. Is 4.5 billion years enough time to blow away a gas envelope. What features would such a remaining core have ? and does Mercury have them.
    As hostile as this world is to life as we know it, I’ll bet it’s more likely that surface life dwelling life can arise there than the surface of Venus, albeit life as we don’t know it.

  • Alex Tolley October 31, 2013, 11:33

    “So we have an Earth-sized planet with Earth-like density in an orbit that remains a mystery”

    Purely speculative/SF. What if this is an example of malevolent aliens destroying a world, and we not seeing it as artificial? We keep saying that we can recognize the artificial, so that the “ants living near a highway” model does not apply to us, but is that true?

  • Eric Hughes October 31, 2013, 11:52

    This planet seems a good illustration that we would do well, when thinking about habitable zones, to avoid being parochial in favor of liquid water as the sole basis of life. I anticipate we’ll discover that some planets such as this one have seas of molten glass, hosting complex chemistry around Al-Si-P-S instead of our own C-N-O. With temperatures at 3000 K, above the boiling point of glass, we might even expect atmospheric transport in some situations, that is, clouds of glass vapor and rains of molten glass.

  • andy October 31, 2013, 13:55

    Regarding the suggestion that this is the exposed core of a gas giant planet – how much and how fast are giant planet cores expected to decompress when the gaseous envelope is removed? That is, should we expect that a giant planet core would have a similar density to a terrestrial planet, or could it remain in a high density state for a long time?

  • Steve Bowers October 31, 2013, 14:42

    Alex Tolley; “What if this is an example of malevolent aliens destroying a world, and we not seeing it as artificial? ”

    To move a planet of this mass from an Earth-type orbit to an epistellar orbit would require a vast amount of energy, more than enough to sterilise the planet to a depth of several kilometers. So as a method of planet-killing it would be very inefficient.

  • Enzo October 31, 2013, 15:28

    There’s the second Kepler Science Conference in a few days :
    http://kepler.nasa.gov/Science/ForScientists/keplerconference/

    More results should come.

  • Geoffrey Hillend October 31, 2013, 17:13

    If we got rid of the .8 greater than Earth mass how big would this planet be? It would have a diameter less than our Earth but be still be equal to it in mass which makes me wounder if there is a extra solar planet smaller than our Earth in diameter that is rich in Iron and heavy elements with heavy gases in it’s atmosphere and a bone dry planet.

  • Alex Tolley October 31, 2013, 17:24

    @Steve B0wers. I agree a brute force method would indeed require much more energy than sterilization. I am assuming that by sterilization you are talking about severe bombardment by targeted asteroids, something we would be able to do this century. If the aliens have a mini black hole of asteroid sized mass, and it could be targeted to the world’s core and keep open a vent to the surface, it might well be able to move a planet to a close solar orbit after consuming perhaps 10^-4 of the planet’s mass and spewing out energy. Improbable certainly. Inefficient, hugely. But one uses a sledgehammer to crack a nut when you want everyone else to toe the line. ;-)

  • Rob Flores October 31, 2013, 17:50

    Now that the observational period of the KST is over, I have to assume the Kepler program staff is redoubling it’s effort to find ANY sign of a near twin of Earth in a Sun like stars’ HZ. Or did they throw everthing they had at sifting the data for twin Earth’s while the KST was in operation?

    Kepler was in operation for 4 years plus change. Would another 2 years really made a difference in finding Twin Earth’s as opposed to more Mars
    like terrestrials or even Jovians.

    Baring any surprises tomorow then Kepler 62f, at 1.4 RE is the defacto best Earth Like in all KSC discoveries. Since its sun appears to be K dwarf that is considerably older than SOL and that it’s orbit sits where Venus is in our solar system. What is the likelyhood that K62f is tide locked or nearly so.

  • Kamal Ali November 1, 2013, 4:10

    I wonder what angular extent the disc of the star would occupy
    if one were standing on the planet and the star was directly overhead?
    [assuming no clouds etc]

    I also wonder whether the surface would be cloudy or hazy or whether it would be cloudless ?

  • Anthony Mugan November 2, 2013, 5:58

    In terms of the suggestion that he inconsistency of this planet’s location with theory may indicate some artificial process…I am not clear as to how such an idea could be tested in this context, given that anomalies like this usually prompt improvements in our theoretical models which account for the anomaly. There would need to be a definitive test such as something very odd spectrographically I assume?

  • torque_xtr November 2, 2013, 6:50

    The cores of evaporated gas giants and neptunes would be very different from as-formed terrestrials. Possibly it’s too early to speculate, but the first of the likely things are huge salt pans or the seas of molten salt, which was extracted from cores or brought directly with ices from nebula. The saltiness of neptunes couldn’t be less than saltiness of comets, and multiplied by some Earth’s masses, that would suffice for meters, and more likely, hundreds of meters of salt.

    And molten salt is a perfect solvent for all that Al-Si-P-S-O complex chemistry :-)

    All the less soluble phases would crystallize out of the increasingly supersaline and cooling ocean on the later stages of evaporation and later, out of molten salt ocean on the terminator of a tidally locked core, creating gorgeous deposits of crystals scattered all over the surface. Much is dependent on the amount of heavy gases, such as N2, O2 (from photolysis) and Ar on the later stages of evaporation, and on the acidity/basicity of resulting salts – if they are acidic, a very noxious atmosphere of halogens and nitrogen oxides is likely, if they are basic, than all gases are chemically absorbed, even N2, and only noble gases remain…

  • Holger November 2, 2013, 17:20

    Rob Flores:
    I’m also beginning to fear that Kepler will not find any true Earth analogs. On the other hand, Borucki was still optimistic after Kepler’s demise to not only find one, but even find a good estimate for the frequency of Earth analogs, which would require several such discoveries.

    “Kepler was in operation for 4 years plus change. Would another 2 years really made a difference in finding Twin Earth’s as opposed to more Mars
    like terrestrials or even Jovians.”
    If any transiting Earths exist in Kepler’s field of view, than that would have greatly increased the chances. It took ~5 transits to confirm Kepler-62f; Earth-sized planets would take even longer.
    Since Venus isn’t tidally locked and Kepler-62 is less massive than the Sun, I think Kepler-62f is very unlikely to be tidally locked.

  • Andrew Palfreyman November 3, 2013, 4:04

    There’s a huge number of exoplanets – estimatedly about 400 billion in this galaxy alone. Less than 1% are visible via the transit method, on purely geometric grounds. There’s a ton of further discovery work to do. I suppose it’s all eyes on TESS next, after the Kepler data analysis is complete.

  • Eniac November 3, 2013, 12:07

    David Cummings:

    This planet is so close to its sun that it must be tidally locked and always presenting the same face to the sun. I wonder if the dark side is cool, or if the heat from the star bakes the whole planet through and through so that even the dark side is extremely hot.

    I am wondering this, too. Does anyone know?

    Given that the Earth’s mantle and core are every bit as hot as this planet, I would say no, “baked through” cannot be a factor. In fact, extending this argument, it seems that the far side must actually be much colder than Earth, since it does not receive heat from the surface at all, and there must exist a temperate zone somewhere in between. A land of permanent sunset, where just a sliver of the star shows over the horizon to provide light and heat.

    Or not?

  • torque_xtr November 4, 2013, 4:25

    “A land of permanent sunset, where just a sliver of the star shows over the horizon to provide light and heat.”
    Yes, just so, but it also depends on on atmospheric circulation and, possibly, geothermal heat. If the atmosphere is Venus-like, the nightside is almost as hot as dayside, if it’s thinner, the nightside is possibly cold enough to freeze all gasses. But if there’s another star circling the primary/super-earth system at the right distance – which is not uncommon – like Alpha Centauri system, but closer – so the system will be in it’s habitable zone, the dark side of super-earth would see just the common day/night cycle and have temperate climate.

  • kzb November 4, 2013, 9:10

    There is a new arxiv paper on the minimum mass of a gas giant core. According to this, the planet under discussion is too small at 1.7 Earth masses.

    “On the Minimum Core Mass for Giant Planet Formation at Wide Separations”

    http://arxiv.org/abs/1311.0011

  • Eniac November 4, 2013, 22:26

    Yes, just so, but it also depends on on atmospheric circulation and, possibly, geothermal heat.

    I think we have accounted for geothermal by simply noting that our own planet is just as hot in the interior as this one, which has not kept our surface from “freezing”.

    Atmospheric circulation is a valid issue. I have a hard time imagining what kind of atmosphere would be able to circle a planet with such extremes in temperature without either condensing or evaporating away. It seems like anybody’s guess, or have people seriously studied this?

  • kzb November 6, 2013, 8:41

    Eniac, somewhere out there on the web there is a planetary atmosphere calculator. You plug in the planet size, mass and other factors, and it gives you the temperature range that a gas could survive in. I’ve no time to find it again right now, I’ll try later.

  • Eniac November 6, 2013, 23:06

    kzb: I have heard of this calculator, but I do not think it deals with asymmetric illumination of this magnitude. My guess would be that anything that is not permanently condensed on the cold back side will quickly be evaporated from the hot side as soon as it makes its way there…. But I could be wrong. The problem is that we have no example in our own solar system, and anything we can say is really wild speculation, at best disguised as scientific reasoning.

  • kzb November 7, 2013, 9:03

    Planetary atmosphere retention plotter:

    http://astro.unl.edu/naap/atmosphere/animations/gasRetentionPlot.html

    Using this and mentally extrapolating the temperature, it does not look unlikely to me that this planet could have retained some heavy atmospheric gases ?

  • stephen November 7, 2013, 9:27

    Here’s an Alien Planet Designer. Is this the one you’re thinking of?

    http://www.johnbray.org.uk/planetdesigner/

    The article states that the planet couldn’t have migrated inward because it would be swallowed up. Well, maybe it hasn’t finished migrating inward. How old is the star? How long should it take to go from, let’s say 10 AU to zero?

    Its migration inward might have been slowed down by close encounters with other planets. And a close approach to another star in the past might have influenced it.

    What wavelengths of light would be able to penetrate a molten salt ocean? Or would everything in the oceans be blind?

  • andy November 12, 2013, 16:55

    @kzb: looks like it is based on Jeans escape which is the slowest of the various atmospheric escape mechanisms. I.e. if Jeans escape predicts you will lose the gas, you will definitely lose it, but even if Jeans escape predicts it will be retained then other mechanisms may still act to remove it.

    Most of these calculators throw in an arbitrary fudge factor to get the results to look vaguely like what we see in the solar system, but this can hardly be regarded as a rigorous analysis of atmospheric loss.