A boiling superEarth joins the exoplanet roster

By Phil Plait | November 30, 2011 12:19 pm

A collaboration between space- and ground-based telescopes has added a new world to the growing list of exoplanets: Kepler-21b, a planet bigger and more massive than Earth. It’s far smaller than Jupiter, though, putting it firmly in the "small, rocky planet" category. Not that it’s Earth-like: it orbits its star in just under 3 days, making it hot enough to have pools of molten iron on its surface!

Now, I don’t generally write about every new alien planet discovered — with over a thousand of them and counting, it would be all I ever do! — but this one interested me. For one thing, it’s not all that much bigger than Earth; it’s about 1.6 times our diameter. The size was able to be found because the planet transits its star: it passes directly between the star and us, blocking the star’s light a wee bit. The amount of light blocked depends on the size of the planet itself, so by carefully measuring that dip in brightness the planet’s size can be determined.

And did I say a wee bit? I mean a really wee bit! Here is a graph showing the planet’s effect on the starlight:

The vertical axis is the amount of light we see from the star, and the horizontal axis is time. You can see how the light drops a bit when the planet blocks the star. But look at the scale! The planet blocks a mere 0.005% of the star’s light! That’s an incredibly sensitive detection, and incredibly difficult to detect. Stars have all sorts of ways of varying their light output, from sunspots to intrinsic pulsing. All those effects had to be removed from the observations to find this weak leftover signal.

But that’s the power of multiple observatories. The star was observed by the orbiting Kepler observatory, designed to look for such planets transiting their stars. It was followed up by the ground-based Mayall and WIYN telescopes at Arizona’s Kitt Peak National Observatory for confirmation, and in total the planet was watched for over 15 months to determine its characteristics.

Even better, these combined observations tell us the mass of the planet itself. As it circles its star every 2.8 days, its gravity pulls on the star, subtly changing the spectrum of the star’s light. The more mass a planet has, the more gravity, and so the more it pulls on the star, and the bigger the effect on the spectrum.

In this case, the planet has a mass of no more than 10 times that of Earth, and is probably less. This implies it’s denser than our home world. Why? Because at 1.6 times our diameter it has 4 times our volume (volume increases with the cube of the diameter, and 1.6 x 1.6 x 1.6 is about 4). If it had the same density as Earth that would mean it would have 4 times our mass. Since the mass is likely higher than that, it is probably denser.

That makes sense to me. The planet is orbiting its star at a distance of only about 6 million km — far closer than Mercury orbits the Sun! That makes the surface of Kepler21b hot, probably about 1900 K (roughly 1600°C or 3000° F). That kind of heat tends to boil away lighter stuff like water, leaving denser material behind (unless the planet is big enough to hold onto those lighter materials; Jupiter could, for example, but this planet is far smaller). So the planet being denser than Earth isn’t surprising.

Playing with the numbers a bit, I find the surface gravity of the planet is 4 times Earth’s, too. If you weigh 120 pounds on Earth, you’d weigh nearly a quarter ton on Kepler-21b. Not a great place to lose weight!

… on the other hand, with a surface temperature literally high enough to boil lead, you’d lose weight fast. But then, you’d be a puff of vapor. Probably not the best diet plan.

Not that you’d be heading over there anytime soon anyway. Kepler-21b is 350 light years away, or 3.5 quadrillion kilometers (2000 trillion miles) away. Getting there would be tough. I suggest something easier, like doing 10,000 push ups a day.

Anyway, this is an amazing detection; the planet is pretty small, very far away, and its parent star very luminous. These all combine to make this a tough world to detect, but that goes to show you: we’re getting really good at this sort of thing.

How long before we find another Earth this way? I’m guessing not very long. A few years at most. If they’re out there, they can’t hide forever.

Credits: ESO/L. Calçada; Steve Howell and the Kepler team


Related posts:

Motherlode of potential planets found: more than 1200 alien worlds!
Kepler works!
Astronomers discover a wretched hive of scum and villainy
Two exoplanets discovered by “citizen scientists”

CATEGORIZED UNDER: Astronomy, Cool stuff, Science

Comments (54)

  1. Christopher Jablonski

    I tell you, the surface of Kepler-21b is really hot.

  2. andy

    I’d have to ask what assumptions are going into the claim that the mass is greater than 4 Earth masses. As far as I am aware, the RV signature of this planet has not yet been detected, and with only an upper limit we cannot tell whether we are looking at a dense, rocky planet or a more volatile-rich one. Low-mass, volatile-rich planets do seem to be able to survive in close-in orbits: witness 55 Cancri e for example. Certainly a lot of so-called “super-Earths” found so far have turned out to be in the latter category, perhaps they should better be described as “mini-Neptunes”.

  3. Chip

    Just as a footnote – is the Alpha Centauri system completely ruled out as far as planets go? Four light years is still pretty far away but it would still be exciting if some kind of terrestrial planet or a gas giant was discovered orbiting Alpha, Beta or Proxima. I recall a while back some Astronomers were studying the system via radial velocity and star transit.

  4. Jeff

    told you so, yeah, like I’m the only one! everybody has been optimisitic about Kepler’s success, and this is the beginning.

    good one, a very close earth-like planet except the 2000 Kelvin temp.

    Getting closer.

    I predict you’ll be doing a blog in 2012 saying something like “KEPLER HITS THE BULLSEYE” and it’ll be earthlike and 300 kelvin and we’ll be speculating about oceans on it. In fact, we’ll be a little frustrated that we won’t have an actual closeup image of that one.

  5. andy

    @Chip: heard a rumour from a conference that the studies have so far managed to rule out any planets above 4 Earth-masses in orbital periods less than 300 days. Presumably this is for Alpha Cen B, which is for various reasons a better RV target than Alpha Cen A and is the main focus of the various campaigns on this system.

    Gas giants Avatar-style have been ruled out since quite a long time back…

    (Also Alpha Cen B is NOT Beta Centauri. Beta Centauri is a blue-white B-type giant star that is utterly unrelated to the Alpha Centauri system.)

  6. lepton

    Chip, Alpha Centauri system A and B still possibly host habitable earth like planet. Each habitable zone is well inside stable orbit zone. We just don’t have equipment big enough to reach a conclusion.

    Also, Phil, stellar variation, like spot, etc, is much less than transient event of an Earth like planet around a Sun like stars ON THE TRANSIENT TIME scale. So, make no mistake, Kepler is an INCREDIBLE machine and one of my personal favorites, but, no need to over dramatize the capability.

    Jeff, Kepler will tell us how common Earth like planet will be, so, armed with this data, we can search in our vicinity with high confident on result. I am sure I would be literally in tears if it turns out that Alpha Centauri A or B has habitable earth like planets.

  7. I’ll bet that planet’s surface is flatter then Kansas. Especially if it has an atmosphere. Imagine molten iron rain, at 4 Gs! That’ll wear down any rocks and mountains pretty darn fast.

    Interesting that they keep finding terrestrial planets that are many times the mass of Earth, when Earth is the most massive terrestrial planet in the Solar system. I wonder if that makes us a bit of an odd bird, with Terra herself actually being somewhat on the low side of the curve for terrestrial planets?

  8. #@6 lepton: Also, Phil, stellar variation, like spot, etc, is much less than transient event of an Earth like planet around a Sun like stars ON THE TRANSIENT TIME scale. So, make no mistake, Kepler is an INCREDIBLE machine and one of my personal favorites, but, no need to over dramatize the capability.
    I’m confused. Are you saying that the light curve dip of a starspot is shorter in duration then the transit of a planet? Because unless that star is spinning really darned fast, that sounds impossible to me.

  9. lepton

    @Joseph,

    I meant stellar variations are bigger than transient signal for terrestrial planet, but, they are pretty much long time scale event, so, within a few hours (typical transient time), star is very stable.

    Also, nice scene of molten iron rain at 4G! :-)

    Earth is actually indeed at the low end of habitable planet due to it’s barely big enough to support tectonic activity to generate magnetic field to protect life as we know. Of course, we don’t know whether life can exist in a completely different form, I am personally very open minded on this.

  10. Jess Tauber

    So, how about life on such a world. Something involving refractory chemical compounds that can stand the molten metal and still form coherent, information and energy- rich combinations. At least we wouldn’t have to worry about them invading earth- a horrendously frigid planet like ours would have no attraction (except, of course, gravitationally!).

  11. andy

    Earth is actually indeed at the low end of habitable planet due to it’s barely big enough to support tectonic activity to generate magnetic field to protect life as we know.

    Two separate issues here: the geomagnetic field and the plate tectonic system. The geomagnetic field is generated by convection in the liquid outer core, plate tectonics are a process that occurs in the lithosphere (which comprises the crust and the outer region of the mantle). The two processes operate in different regions of the planet, and there is no particular reason to expect that plate tectonics are necessary for generating a geomagnetic field: the obvious counterexample here is Mercury which has a geomagnetic field but no plate tectonics.

    As for Earth being a low-mass planet for plate tectonics, the usual example that is given is Venus, but this is not a particularly good comparison. Conditions at the Venusian surface are somewhat different to those at the surface of the Earth, and these conditions make it less likely for Venus to host plate tectonics regardless of mass. One of the key differences is the lack of liquid water, which appears to play a significant role in lubricating the Earth’s subduction zones. As far as I am aware, there is no evidence for subduction zones on any other planet, unlike rift features which appear to be fairly common. Furthermore the higher Venusian surface temperature implies a reduced thermal gradient through its crust which means less energy available to drive plate tectonics. Unfortunately we do not have access to any other terrestrial planets with liquid water present at the surface to do a fair comparison: probably the only other ones in our solar system are Europa and Enceladus (most of the other icy moons with oceans probably have the liquid water separated from the rocky interior by a layer of high-pressure ices) – I would certainly love to know what kind of geology is going on down there, but getting to the ocean floor of these worlds is going to be challenging to say the least…

    Theoretical models of super-Earths seem to be inconsistent about whether these worlds should support plate tectonics: some have suggested that the higher energy available from the interior should be able to drive tectonics even in the absence of water, others have suggested that the higher gravity would tend to lock up the tectonics. Jury is still out here.

    As for geomagnetic fields, studies seem to consistently find that super-Earths larger than about 2-3 Earth-masses or so ought to have entirely solid cores due to the increased pressure in the centre of these planets. This is bad news for magnetic fields: Earth’s magnetic field is generated by convection currents in the liquid layer of the outer core – without a liquid layer in the core this mechanism will not work, and a solid core would be well above its Curie temperature and thus be non-magnetic. Therefore it may well be the case that terrestrial planets much more massive than Earth are a bad bet for habitability.

  12. Infinite123Lifer

    Loved the article.

    Sorry if someone already asked the question but

    “When is Kepler-21b going to finally be consumed by its star?”

    If planets can rip apart moons via gravitational forces I wonder if a planet such as Kepler-21b who is working its way slowly inward (speculation) could become malleable to some degree to either overcome the destructive gravitation forces and shape itself more like putty or remain dense & malleable enough to just fall into its star mostly whole?Maybe -21b would not explode apart into a doomed ring system falling ever inwards. But most importantly

    “Will we witness whatever fate this Kepler -21b will take on?”
    I heart astronomers
    I heart graphs

  13. MadScientist

    Wow – that’s some photometry job getting <0.001% precision with so few photons coming in.

  14. Sam H

    @6 Lepton: There’s been a simulation study done that suggests that terrestrial planets should form in stable orbits for at least 200 million years – one of which may be Earth-size, and lie in the HZ (these prospects being very good for Centauri B). Link to article about study is here:

    http://www.sciencedaily.com/releases/2008/03/080307121613.htm

    Even though Pandora & Polyphemus are impossible in this system, the possibility of an Earth-like planet, and thus a potential habitat for life, and thus a huge impetus to any plans for interstellar probes (and just maybe, a potential site for a future interstellar colony) – all these things are still within the realm of physical possibility. I’m not easily moved to the point of tears, but if this is real, then when we find it I might just be the most excited man alive. I just hope they name it Gaia Centauri (which coincidentally is the name of a significant planet in the epic hard-SF space opera I will eventually write ;))

    @Andy: so super-Earths won’t have magnetic fields. Well, considering that it’s a combination of geomagnetism and the atmosphere that protects us from cosmic radiation, carbonic life should still be possible on such a world, given a thick-enough atmosphere. As well – might life be able to evolve in a high-radiaton environment anyway? Perhaps radiation might be a potential catalyst for some forms of biochemistry (though this applies more to life as we don’t know it, obviously). And evolution can work wonders…:)

  15. @#9 Lepton: I meant stellar variations are bigger than transient signal for terrestrial planet, but, they are pretty much long time scale event, so, within a few hours (typical transient time), star is very stable.

    Ohhh, ok, I see what you mean.

    Earth is actually indeed at the low end of habitable planet due to it’s barely big enough to support tectonic activity to generate magnetic field to protect life as we know. Of course, we don’t know whether life can exist in a completely different form, I am personally very open minded on this.

    Makes you wonder about all those skinny, big-headed aliens that people claim to encounter. Sounds much more likely that if we find intelligent life, it might be physically adapted to high-gravity environments. Reminds me of “Footfall,” a Niven/Pournelle sci-fi novel about an invasion by aliens that look a lot like baby elephants.

    @#10 Jess Tauber: At least we wouldn’t have to worry about them invading earth- a horrendously frigid planet like ours would have no attraction (except, of course, gravitationally!).

    On the other hand, the specter of enormously strong, high-jumping lava monsters is pretty scary, you hafta admit :)

  16. Relativity

    @Chip

    Kepler, I think, will only be looking over a limited (in astronomical terms) span of 3000 LY over the plane of the Milky Way near the neighborhood of the Orion Spur (approximately covers the area around the constellation Cygnus and Lyra). Kepler uses fixed distant light-points (quasars) for its telemetry system to fix its gaze at this area for most, if not all, of its lifespan.

    a Centauri system is way down in the southern hemisphere in the constellation of Centaurus. They are as far south as the Southern Cross in declination! Too bad Kepler isn’t designed to be more flexible but its a start. There will be another telescope in the future that will refine what Kepler has found – perhaps that multi-sensor-array thing being proposed. We’ll just see after the Webb telescope is launched in 2018? [crossing fingers]

    I am with you though that the day someone will announce a planetary system at either or both alpha or beta components (they are almost a clone of our Sun to boot!) will be the day I will start packing my luggages for I will volunteer – if I still happen to be around this planet, that is. :)

  17. Robert

    @Joeseph: We are finding these massive planets close to their stars because they are the ones that are ‘easy’ to find. Big planets close to their stars make the star wobble more, and are more likely to transit, and all-in-all declare their presence much more loudly.

    I doubt that we have the technology to detect an identical solar system to ours elsewhere in the galaxy. Maybe if a Venus or Earth twin transited in front of Kepler, but, even then, we’d need many years of readings to make that decision – three transits would be the minimum to even suggest it, and that’s three years of watching.

  18. Joseph G (#7):

    Interesting that they keep finding terrestrial planets that are many times the mass of Earth, when Earth is the most massive terrestrial planet in the Solar system. I wonder if that makes us a bit of an odd bird, with Terra herself actually being somewhat on the low side of the curve for terrestrial planets?

    Well, the Earth was created by the Racnoss, so it wasn’t a “natural” event. :-)

    Imagine molten iron rain, at 4 Gs!

    What a difference one letter can make: “Singing in the Rain” vs “Singeing in the Rain”. :-)

  19. OtherRob

    The planet blocks a mere 0.005% of the star’s light!

    For comparison sake, how much of the sun’s light would the Earth block?

  20. amphiox

    Interesting that they keep finding terrestrial planets that are many times the mass of Earth, when Earth is the most massive terrestrial planet in the Solar system. I wonder if that makes us a bit of an odd bird, with Terra herself actually being somewhat on the low side of the curve for terrestrial planets?

    It don’t think this makes us any more of a “odd bird” then the fact that early on we kept finding gas giants many times the mass of Jupiter.

    Our methods still remain biased towards bigger planets. A true earth analog is still only barely within our detection capabilities.

  21. -jeffB

    Infinite123Lifer asked:

    “If planets can rip apart moons via gravitational forces I wonder if a planet such as Kepler-21b who is working its way slowly inward (speculation) could become malleable to some degree to either overcome the destructive gravitation forces and shape itself more like putty or remain dense & malleable enough to just fall into its star mostly whole?”

    At planetary scale, material strength isn’t really an issue. When you’re stressing a planet with tidal forces, EVERYTHING is soft like putty, or more accurately, like liquid.

  22. Chris A.

    @OtherRob (#18):

    “The planet blocks a mere 0.005% of the star’s light!”

    “For comparison sake, how much of the sun’s light would the Earth block?”

    Earth is 1/109th of the sun’s diameter, so 100 x (1/109)^2 = 0.008% –just a tad more.

  23. Chris A.

    The “molten iron temperature” thing should be taken with a grain of salt, considering that a planet orbiting that close would almost certainly be tidally locked (1:1 spin-orbit ratio). So the side permanently facing its sun would be molten iron hot, but the dark side would be icy cold (unless there’s an atmosphere to distribute the heat around).

  24. Torbjorn Larsson, OM

    Getting closer.

    @ lepton:

    Earth is actually indeed at the low end of habitable planet due to it’s barely big enough to support tectonic activity to generate magnetic field to protect life as we know.

    As andy notes, field and plate tectonics are different issues.

    – What plate tectonics contribute is weathering and CHNOPS (the major cell compounds of C, H et cetera elements) recycling. One specific point is that plate subduction zones contribute energetic polyphosphate compounds that could have jump started cellular metabolism. These doesn’t seem to be sourced elsewhere, and they show up in our cells. (As ATP/ADP metabolism today, even if inorganic polyphosphates can substitute and _do_ substitute in some prokaryotes).

    To add to the plate tectonic observations, this week results of finding putative hot spots on Venus came out. These hot spots would likely release core heat without the catastrophic remodeling earlier speculations had. Less like plate tectonic recycling but more like plate tectonics thermal behavior. The Hawaiian island chain is a result of a mantle hot spot.

    Curiously the martian Tharsis is generally also thought to be a mantle super plume hot spot result. Birds of a feather flock together.

    – Magnetic fields doesn’t protect life. What protects life from UV (today)* and cosmic rays (CR) is the atmosphere. (With the difference that the heliosphere magnetic field is responsible for ~ 90 % of CR shielding.)

    Magnetic fields protects the atmosphere, somewhat. IIRC they protect mostly against Coronal Mass Ejections that stands for ~ 2/3 of atmospheric loss.

    One way to compensate for losses is hence simply to have a thicker atmosphere. Venus demonstrates that. It also shows that in that case water loss may be heavy due to loss of hydrogen, so presumably you would want to have the whole package of denser initial atmosphere and a more massive terrestrial that keeps a better gravitational grip on the atmosphere.

    Incidentally super-Earths are also good for magnetic fields. There is controversy here how much liquid core remains, but one model that nicely predicts differences between Venus and Earth (IIRC) has it that convection results in large enough fields even for, say, tidal locked planets.

    ———————–
    * Early life had to live beneath water or crust due to lack of shielding oxygen atmosphere. Luckily this is where most biomass circulate anyway.

    @ Jess Tauber:

    So, how about life on such a world. Something involving refractory chemical compounds that can stand the molten metal and still form coherent, information and energy- rich combinations.

    Not enough variety. Only organics has that.

  25. OtherRob

    @Chris A., #22:

    Earth is 1/109th of the sun’s diameter, so 100 x (1/109)^2 = 0.008% –just a tad more.

    Thanks. :-)

  26. Superluminous news and a great exoplanetary find. :-)

    Although it does seem to be the latest and one of the more massive of what I think of as “Mustafar” class worlds joining the likes of Corot-7b, 55 Cancris e and Kepler-10 b. (Click on my name here for a great youtube clip on the latter incl. an artists animation of what it may look like.)

    With a surface that most likely consists of molten lava at least for the star-facing hemisphere and cold cratered rock on the night side (assuming tidally locked rotation which seems most probable) the term “SuperEarth” really seems a misnomer with “Lava planet” being a far more accurate label. ;-)

    Could also be all that’s left of a Hot Jupiter that’s quite literally had its atmosphere blown away. Assuming this I suspect we’d be down to a metallic hydrogen / molten rock / carbon / even perhaps plasma mass like nothing we’re familiar with here.

    Remarkable find and world although neither nice to visit nor live there! 8)

  27. Dr. Strangelobe

    Jess Tauber @ 10 ;

    You should read “Iceworld” by Hal Clement, an alien from a high-temperature planet has to go to Earth and do things.

  28. Bryan Yager

    Wikipedia states that Mercury has a density slightly lower than that of Earth, namely 5.427 g/cm3 versus 5.515 g/cm3 for Earth, so your claim that a planet could double in density as a result of all volatiles boiling off into space sounds a bit iffy to me. Perhaps other mechanisms are in play that could explain these results?

  29. Could also be all that’s left of a Hot Jupiter or Hot Neptune that’s quite literally had its atmosphere blown away. Assuming this I suspect we’d be down to a metallic hydrogen / molten rock / carbon / even perhaps plasma mass like nothing we’re familiar with here.

    Imagines a comet tail atmosphere spiralling around into space as it bubbles off a constantly boiling ever shallowing ocean of metallic hydrogen overlying a core of hyper-pressurised flamed from both sides fluid through to a central plasma core at inconcievably high pressures and temperatures.

    A gas giant that has wandered inwards from the Cold Zone where it first formed scattering or engulfing any might-have-been-habitable worlds into the Black. Plunging into its star and stopping just short of being consumed in a orbit leaving it nearly brushing its slightly larger and hotter than our Sun, star. (HD 179070*) Blow-torched ever hotter, its cloudtops fizz away into the void forming a comet like tail stretching ever further from the star before disappating into the endless vacuum of space. Shrinking over aeons the heat from the planets core and the heat from the sun on its doorstep combine to vapourise Kepler 21b with agonising slowness by unfathomable extremes. Losing mass and ripped down to a searing core the doomed world ‘s remnant, too massive and pressurised to blow apart, too hot to keep itself intact shrinks ever more sizzling away until its star expands into gianthood and ends its hellish torment in a final conflageration.

    Poor ole Kepler 21 b. I actually feel sorry for it! ;-)

    Wonder if we can ever tell whether Kepler 21 b’s history and nature is something like I’ve imagined above & how much we can learn and compute about its history and evolution and future.

    PS. Anyone know the primary stars – Kepler 21 / HD 179070’s exact spectral type?

    * Source linked to my name – on Universe Today article ‘New Planet Kepler-21b Confirmed From Both Space And Ground’ by Tammy Plotner on the 1st of December, 2011.

  30. @3. Chip :

    Just as a footnote – is the Alpha Centauri system completely ruled out as far as planets go? Four light years is still pretty far away but it would still be exciting if some kind of terrestrial planet or a gas giant was discovered orbiting Alpha, Beta or Proxima. I recall a while back some Astronomers were studying the system via radial velocity and star transit.

    Click on my name or cut’n’paste :

    Theoretically, Alpha Centauri should have planets

    into this blogs search box for an article the BA posted here on the 7th of March , 2008 at 2:13 PM discussing a study that concluded Alpha Centauri B should – at least in theory – have rocky planets quite likely in the HZ.

    I’m sure if they do find something Alpha Centaurian exoplanet~wise we’ll know about it fairly quickly & I’m somewhat surprised there’s been nothing found yet although Earth and smaller mass planets are exceedingly tricky to find even with the technology we have today.

    @18. Ken B : What a difference one letter can make: “Singing in the Rain” vs “Singeing in the Rain”.

    Good one. :-)

    @24. Torbjorn Larsson, OM, #14. Sam H & 22. #Chris A. : Thanks for your informative comments there. This blog needs a ‘like’ button! :-)

  31. Chris

    As an additional note, Alpha Centauri A and B are not eclipsing binaries. More than likely any planets around these stars would be orbiting in the plane with the stars so we couldn’t detect them by the transit method.

  32. andy

    Although it does seem to be the latest and one of the more massive of what I think of as “Mustafar” class worlds joining the likes of Corot-7b, 55 Cancris e and Kepler-10 b

    Thing is, those planets do not naturally group together. Kepler-10b is consistent with being a rocky planet, i.e. a super-Earth, it probably fits the lava-ocean planet model. 55 Cancri e on the other hand is not: it requires at the very least a deep layer of supercritical water to explain the observations, i.e. a mini-Neptune (initially it was announced to be a high-density rocky planet, but this was swiftly contradicted by Spitzer measurements and subsequently an error was found in the data-analysis pipeline that led to the initial underestimate of the planet’s radius). CoRoT-7b has somewhat uncertain mass, lately the estimates have led towards the high end of the scale which suggests it groups with Kepler-10b rather than 55 Cancri e.

    The remaining low-mass planets for which we have both mass and radius measurements have all turned out to be in the mini-Neptune class: GJ 1214b, the Kepler-11 system, HD 97658b, Kepler-18b.

  33. Lightndattic

    Quick question about estimating size based on transits-

    Is the size estimate based upon the full outline of the planet blocking the star’s light and the assumption that we lie in the planet’s orbital plane providing that full outline?

    I would imagine the planet’s size estimate could be low if we’re only seeing a portion of the planet eclipsing the star. Chris A calculated the Earth blocks 0.008% of the sun’s light providing it’s the full outline of the Earth. In truth, the Earth could block anywhere from 0% up to 0.008% based on the observation point given intergalactic distance (not counting extremely close perspectives where the Earth could totally block the sun’s face).

  34. So all we need to do is find the Mass Relay outside of Pluto and we’re good to go :D

    … wait… I’ve been playing too many video games.

  35. andy

    Wikipedia states that Mercury has a density slightly lower than that of Earth, namely 5.427 g/cm3 versus 5.515 g/cm3 for Earth, so your claim that a planet could double in density as a result of all volatiles boiling off into space sounds a bit iffy to me. Perhaps other mechanisms are in play that could explain these results?

    The other factor that comes into play is compression. Earth is a more massive planet than Mercury and the interior is under greater pressure. This acts to compress the Earth, increasing its density. The real comparison between the two is uncompressed density, and here Mercury is clearly made of denser material (which would be expected given that a large fraction of Mercury’s mass is the iron core). The uncompressed density of Mercury is around 5.3 g/cm³ while that of the Earth is 4.4 g/cm³.

  36. Regner Trampedach

    Lightndattic @ 35: The shape of the transit curve will change with latitude so that is fitted for and accounted for. Oh, – and we are not talking intergalactic distances here. All Kepler targets are well within the Milkyway and only hundreds of light years away. Just a few blocks down the Orion arm from us.
    lepton @ 6 and Joseph G @ 15: The noise that you can see in the light curve is from mainly two sources: sound waves that are excited by the roaring convection in the surface layers of the star, and the granulation at the surface resulting from that same surface convection. The granulation signal is a true noise source affecting all time scales. The sound waves, on the other hand, gets filtered by the size of the acoustic cavity (the star) and only standing waves remain. Their periods are of the order of some minutes for main sequence stars. As the data for Kepler-21b are long-cadence data (sampled every half an hour) you don’t see the nice sinusoidal sound waves in the signal and it just looks like noise. And due to the cadence it looks like noise on the time-scale of the transient. With the very long 15 month time-series the higher frequency sound waves can be recovered despite the low sampling rate. And this stability of Kepler is phenomenal and absolutely crucial to its mission.
    It is exactly a testimony to the fantastic instrument that Kepler is, that the detections of planets are limited by the properties of the stars itself (convection and sound waves) and not by the Kepler telescope. Phil is not over dramatizing anything, but the details didn’t come out quite right. Understandably so – they didn’t in the press release either…
    And since I have mentioned sound waves: Those are actually the most exciting results from Kepler. We can very accurately measure the standing sound waves of stars, which means we can do astero-seismology, just like we do seismology for Earth and have done it for the Sun for the last 30 years (helio-seismology, which has taught us a lot of atomic physics and led to the discovery that neutrinos oscillate between flavors). From that we can find very accurate mases and radii for stars that are not in the rather rare eclipsing binaries. That was impossible before! We can even determine their ages and we can learn about their internal structure – WE CAN SEE INSIDE THE STARS! And that I find exciting.
    Cheers, Regner

  37. @18 Ken B: Well, the Earth was created by the Racnoss, so it wasn’t a “natural” event.
    Sadly, I had to Google that. ‘Fraid I never got on board with the Doctor; Star Trek: TNG was more my thing growing up. And now Dr. Who has been on for so many decades, it’s just darned intimidating trying to catch on to what’s going on :-P

    >>Imagine molten iron rain, at 4 Gs!
    What a difference one letter can make: “Singing in the Rain” vs “Singeing in the Rain”.
    Hah! Yep, better pack your titanium umbrella and asbestos waders ;)

    @29 MTU: Imagines a comet tail atmosphere spiralling around into space as it bubbles off a constantly boiling ever shallowing ocean of metallic hydrogen overlying a core of hyper-pressurised flamed from both sides fluid through to a central plasma core at inconcievably high pressures and temperatures.
    A gas giant that has wandered inwards from the Cold Zone where it first formed scattering or engulfing any might-have-been-habitable worlds into the Black. Plunging into its star and stopping just short of being consumed…

    You certainly do paint a picture!!
    Can you refresh my memory – what exactly is it that can slow something that big enough for it to fall into its primary??

    @24 Torbjorn Larsson: Cool, thanks for the elucidation.

    The magnetic field/atmospheric erosion issue reminds me of this great series I read: Kim Stanley Robinson’s Mars trilogy. In it, comets are redirected to bring water and other volatiles to Mars. It’s acknowledged that eventually, the water will dissociate into hydrogen and oxygen in the upper atmosphere, and the hydrogen will escape. Apparently it’s only a concern on timescales of millions of years, though. I wonder if a planet with huge amounts of water (say a super-earth with several Earths’ worth of seas) could simply race the process and develop life before the oceans dissociated? Or could it do so if it were hit by a steady stream of comets?
    Sheesh. Astronomy always seems to bring up more questions than answers!!

    @36 Katherine Lorraine: So all we need to do is find the Mass Relay outside of Pluto and we’re good to go :D

    IIRC, the first mass relay discovered was first taken to be a KBO. Not only that, but asteroids and KBOs are sometimes photographed without the objects being seen for what they are in the images, until later.
    We may have pictures of said mass relay, in some astronomical archive somewhere, and not even know it! ;)
    /Yes, before my Xbox 360 died, I sat there with Mass Effect 2 and read ALL the historical codex entries :-P

  38. andy

    While Kepler-21b is fairly interesting as these things go, the really interesting super-Earth discovery that was announced recently was Gliese 667 Cc. I expect we will be hearing quite a lot about this one in the near future…

  39. @ Katherine Lorraine: Nice blog, btw. I shot ya a comment. I’m just getting into blogging for the first time, myself, so I’m reading all of ‘em I can get my eyeballs on.

  40. @ Regner Trampedach: …The sound waves, on the other hand, gets filtered by the size of the acoustic cavity (the star) and only standing waves remain. …

    …With the very long 15 month time-series the higher frequency sound waves can be recovered despite the low sampling rate. And this stability of Kepler is phenomenal and absolutely crucial to its mission…

    …And since I have mentioned sound waves: Those are actually the most exciting results from Kepler. We can very accurately measure the standing sound waves of stars, which means we can do astero-seismology…

    I really wish we could post pictures, so I could post a picture of my gobsmacked face. This is the coolest thing I’ve read all week! For example, I’ve heard of low frequency solar waves, but it never occurred to me that each star would have a distinct frequency. As a musician, this really excites me. Each star has a different pitch, like strings on an galactic harp (even if they’re a few dozen octaves below anything audible). I love it!

    Oh yeah, and being able to see inside distant stars. That’s kinda neat too :D

  41. Spaceman Spiff

    @37 — speaking of compression….

    It is due to compression (and the resultant changes in the matter equation of state) that the exponent, alpha, that appears in the radius — mass relation:
    R ~ M^{alpha}
    becomes smaller for ever more massive planets, depending on their composition. alpha is 1/3 for gravitationally uncompressed objects (like a rock). This is the exponent the BA took, when he estimated 4 Earth masses (he simply cubed the radius ratio). For a (cold) object whose mass is in the range of several Earth’s, alpha is ~ 0.27 (i.e., R ~ M^{0.27}), meaning then that the mass might easily be 6 Earth masses, even if its density were exactly that as Earth. Of course, then there’s the possible effect of bloating via sufficient heating by the parent star, but this may not be very important for terrestrial-like exo-planets.

    btw: Planets of roughly a Jovian mass and greater have an exponent alpha ~ 0 — that is, they are all about the size of Jupiter. This includes brown dwarf stars with masses up to 80 Jovian masses.

  42. Regner Trampedach

    Joseph G @ 42: Very glad to have made your day – with science, no less!
    You say “…each star would have a distinct frequency.”
    Well, we can actually distinguish 20-30 modes (tones) for each star, and some tens of thousands for the Sun :-) For the Sun we can see modes that have node-lines across the surface, just like on a drum-skin when an over-tone is struck. We can see modes with hundreds of node-lines both in latitude and longitude on the Sun, resulting in the
    tens of thousands of modes. For other stars we can only see the total light of the star and everything averages out, except the modes with only a few or no nodelines across the surface.
    Google ‘asteroseismology +sound’ and you can also find soundfiles with (slightly :-) transposed Kepler observations.
    Cheers, Regner

  43. @44 Regner: Well, we can actually distinguish 20-30 modes (tones) for each star,
    Ahh that’s right. It is a three dimensional object (well, as opposed to, say, a string, which normally just vibrates along one or two axes (I think).

    Thanks for the tip! The Kepler sit itself has some of these: http://kepler.nasa.gov/multimedia/Audio/sonifications/
    Even without the different overtones that we can detect in the sun, those are some surprisingly complex sounds!

    (For the record, I think links from nasa.gov should just automatically bypass moderation :-P)

  44. Sam H

    I’d type a longer comment if I had time but:

    SOMEONE NEEDS TO TAKE THESE STELLAR SOUNDWAVES, MAKE AN EPIC, SPACEY AMBIENT TRACK FROM IT, AND NAME IT MUSIC OF THE SPHERES. IT MIGHT EVEN BE ME. WHAT IS THIS I DON’T EVEN….DAMN….THE UNIVERSE ITSELF MIGHT BE MUSICAL…. :o :o :o

  45. andy

    Spaceman Spiff @43: just checked against the paper “Mass-Radius Relationships for Solid Exoplanets” by Seager et al. (2007) – 4 Earth masses would result in a density lower than a pure silicate planet (which is itself an unlikely composition for a real planet): it would require at least some volatile content. Inverting the mass-radius relationship given there suggests that for a 1.636 Earth-radius planet, the mass would need to be:

    ~20 Earth-masses for a pure iron “theorists’ planet”
    ~10 Earth-masses for a Mercury-like compositon (67.5% iron, 32.5% silicate)
    ~6.4 Earth-masses for an Earth-like composition (30% iron, 70% silicate)
    ~4.8 Earth masses for a pure silicate “theorists’ planet”

    Not much point doing the calculation for icy compositons because these would not be solid at the temperatures of Kepler-21b, violating the assumptions of the mass-radius relationship. Nevertheless any mass below 4.8 Earth masses essentially must include volatiles. Above that there is a degeneracy between volatile content and iron core fraction.

  46. Spaceman Spiff

    andy @47: Yes, in my little example, I was assuming a fixed density/composition roughly that of Earth, just to make the point that the mass would be substantially (~50% or so) higher than than estimated by the BA due to the effects of gravitational compression. As you noted, there is the additional effect of the change in density with large changes in composition, in the sense that the exponent in R-M relation about a particular radius is somewhat smaller (R-M relation is flatter) for denser compositions.

  47. Messier Tidy Upper

    @ 34. andy : December 1st, 2011 at 12:23 am

    “Although it does seem to be the latest and one of the more massive of what I think of as “Mustafar” class worlds joining the likes of Corot-7b, 55 Cancris e and Kepler-10 b.”
    Thing is, those planets do not naturally group together. Kepler-10b is consistent with being a rocky planet, i.e. a super-Earth, it probably fits the lava-ocean planet model. 55 Cancri e on the other hand is not: it requires at the very least a deep layer of supercritical water to explain the observations, i.e. a mini-Neptune (initially it was announced to be a high-density rocky planet, but this was swiftly contradicted by Spitzer measurements and subsequently an error was found in the data-analysis pipeline that led to the initial underestimate of the planet’s radius). CoRoT-7b has somewhat uncertain mass, lately the estimates have led towards the high end of the scale which suggests it groups with Kepler-10b rather than 55 Cancri e. The remaining low-mass planets for which we have both mass and radius measurements have all turned out to be in the mini-Neptune class: GJ 1214b, the Kepler-11 system, HD 97658b, Kepler-18b.

    Thanks for that informative comment. Much appreciated. :-)

    I like exoplanet hunter Sara Seager’s term for such mini-Neptunes – “Gas Dwarfs” & guess that may make many of the rocky or lava worlds “Rock Giants” by extension. :-)

    @47. andy : Cheers for your greatly informative comment too. :-)

  48. Albert J. Hoch Jr.

    I’m resending this as my first attempt seems to have failed.

    Dear Mr. Plait,
    My recent issue of “Sky and Telescope” shows an astonishing image of a “binary asteroid” as shadows cast by the light of a single star. Can you comment on this interesting application of stellar occultation? How accurate is the resulting image? Since the light rays from a distant star are nearly parallel, the size of the shadow should just about match that of the asteroid! It would seem to me that the precision is almost entirely dependent on the astronomers clocks. (and, of course the number of observation locations) So we should be able to “see” any occulting object regardless of distance. Perhaps groups of stars could be used as a kind of “telescope”? Note: This last would require special camera software to assign timing to each star imaged on the camera chip! (A cute trick. Scanning won’t do it, well . . . maybe super fast scanning.)

    Here’s a great opportunity for another organization of amateur and pro astronomers.

    Sincerely
    Albert J. Hoch Jr.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

ADVERTISEMENT

See More

ADVERTISEMENT
Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »