The Kepler announcements yesterday were greatly cheering to those of us fascinated with the sheer process of doing exoplanetology. The ‘verification by multiplicity’ technique propelled the statistical analysis that resulted in 715 newly verified worlds, and we have yet to turn it loose on two more years of Kepler data (check Hugh Osborn’s excellent Lost in Transits site for more on the method). For those who focus primarily on habitable worlds, the results seemed a bit more sparse, with just four planets found in the habitable zone. And even where we find such, there are reasons to wonder whether a ‘super-Earth’ could actually sustain life.
Apropos of this question, a team of researchers led by Helmut Lammer (Austrian Academy of Sciences) has just published the results of its modeling of planetary cores, looking at the rate of hydrogen capture and removal for cores between 0.1 and 5 times the mass of the Earth found in the habitable zone of a G-class star. Cores like these inevitably attract hydrogen from the surrounding protoplanetary disk, although some of it will be stripped away by ultraviolet light from the hot young star they orbit. The question is, can enough of this primordial hydrogen envelope be blown off to allow the formation of a more benign secondary atmosphere?
The results are not promising for even relatively small super-Earths. Planetary cores with a mass similar to the Earth capture a hydrogen envelope and can also lose it — the paper suggests that terrestrial planets like Mercury, Venus, Earth and Mars lost their proto-atmospheres because of ultraviolet light. But high-mass cores similar to many of the super-Earth discoveries like Kepler-62e and -62f, wind up with atmospheres much thicker than ours. The habitable zone isn’t enough to guarantee a habitable world, as Lammer notes:
“Our results suggest that worlds like these two super-Earths may have captured the equivalent of between 100 and 1000 times the hydrogen in the Earth’s oceans, but may only lose a few percent of it over their lifetime. With such thick atmospheres, the pressure on the surfaces will be huge, making it almost impossible for life to exist.”
Image: The mass of the initial rocky core determines whether the final planet is potentially habitable. On the top row of the diagram, the core has a mass of more than 1.5 times that of the Earth. The result is that it holds on to a thick atmosphere of hydrogen (H), deuterium (H2) and helium (He). The lower row shows the evolution of a smaller mass core, between 0.5 and 1.5 times the mass of the Earth. It holds on to far less of the lighter gases, making it much more likely to develop an atmosphere suitable for life. Credit: NASA / H. Lammer.
The constraints here get to be pretty tight. Let me quote the paper on this:
Therefore we suggest that ‘rocky’ habitable terrestrial planets, which can lose their nebula-captured hydrogen envelopes and can keep their outgassed or impact delivered secondary atmospheres in habitable zones of G-type stars, have most likely core masses with 1±0.5M⊕ and corresponding radii between 0.8–1.15R⊕. Depending on nebula conditions, the formation scenarios, and the nebula life time, there may be some planets with masses that are larger than 1.5M⊕ and lost their protoatmospheres, but these objects may represent a minority compared to planets in the Earth-mass domain.
The super-Earths we’re likely to find in habitable zones, according to this work, are going to be uninhabitable places with hydrogen dominated atmospheres. The paper goes on to note that, based on its results, there may well be Earth-size and mass planets that have not been able to shed their captured, nebula-based hydrogen envelope. The upshot: Extreme caution is advised when speculating about how habitable planets evolve, particular when mass and density are still problematic. The variety of outcomes is wide even for small worlds inside the habitable zone.
The paper is Lammer et al., “Origin and Loss of nebula-captured hydrogen envelopes from “sub”- to “super-Earths” in the habitable zone of Sun-like stars,” accepted at Monthly Notices of the Royal Astronomical Society (preprint).
Comments on this entry are closed.
The upshot: Extreme caution is advised when speculating about how habitable planets evolve, in particular when mass and density are still problematic. The variety of outcomes is wide even for small worlds inside the habitable zone….(from excellent Centauri Dreams post)
Skeptical Magazine understands…..3.24 civilizations in the entire galaxy is their best guess….I guess we’re the .24…..Maybe the Chinese will slingshot us forward again…..JDS
Does “hydrogen dominated” really imply “uninhabitable”? For humans, to be sure, but for life in general?
If the paper turns out to be validated by follow up studies, then this will turn out to be a seminal paper on the statistical probability on twin earths.
Many scientists have wondered if there are filters on the development of life including, Nick Bostrom I am sure most are familiar with his paper (http://www.nickbostrom.com/extraterrestrial.pdf ) it includes many
topics that have been kicked around here on occasion. Verify this filter
and a lot questions are answered.
If you really need worlds between .8 and 1.05 ME to create tolerable and sustained atmosphere’s so that a hydrosphere emerges and Tide locking turns out to be no-starter. Then such a world on average was just kicked a substantial distance from us by this paper. I was open to the idea that there could be hundreds of near-twin earths in the Kepler Spacecraft field of view, now that number looks closer to upper double digits to me.
Perhaps life can exist on super-Earths. After all, animal life exists in Earth’s oceans at depths commensurate with pressures on the order of 10,000 PSI. The atmospheric temperatures near the bottom would need to be manageable though.
“With such thick atmospheres, the pressure on the surfaces will be huge, making it almost impossible for life to exist.”
We should tell the organisms in the deep ocean abysses that their existence is “almost impossible”!
A super-Earth without a moon. How will that affect it’s rotation, axis precession, tides and yearly weather. Also our Earth has a larger core because of a collision with a Mars like object. A super-Earth without a moon might have a less dense core so it can be larger but the gravity will still be the same as our Earth. All things to consider how life might have more trouble evolving there than on Earth.
I am a little sceptical of H/He atmospheres on near in planets of around earth gravities, I mean planet formation is an energetic series of events which would leave the planet very hot~1000 degrees plus. Would the light gases hang around after each impact I think not.
As for H/He been a hindrance for life I am not so sure, life would have an insulating atmosphere and certain bacteria can metabolise hydrogen. Higher life forms that use oxygen would have a problem as it would combine quickly with the hydrogen to form water. Pressure as well can be tolerated by barophilic organisms such as Halomonas salaria, protein deformation been a major issue though.
Surprise result, but hardly the end of the world.
Just to clarify, the challenge of high hydrogen pressures, for life, is not the pressure, but the greenhouse effect and the attenuation of starlight. Rayleigh scattering will mean light doesn’t effectively reach the surface, and geophysical heat will escape very slowly. Such planets will remain very hot at their rocky surface, possibly never cooling sufficiently for water vapour to condense. At least not for several aeons.
Yes, like Adam I’m wondering if a planet with a thick hydrogen atmosphere could have liquid water on its surface (making the assumption that a rocky planet with liquid water is a key prerequisite for life). Also wondering what other differences more hydrogen and higher pressures would make. As always trying to extrapolate from conditions on Earth (past and present) is shaky but it’s all we really have to work with…
RobFlores: “If you really need worlds between .8 and 1.05 ME”.
Well, it isn’t quite that bad, read carefully. The conclusion of the paper mentions between 0.8 and 1.15 Re (i.e. radius, not mass), which corresponds to about 0.5 to 1.5 Me. The main body of the text is a bit more tolerant, going up to about 1.2 Re, or about 2 Me.
That’s why I mentioned that only the smallest Kepler category, < 1.25 Re, should be used for statistics on earthlike planets.
I think the problem with these kinds of theoretical models is that they can steer researchers on the wrong path and damage the course of good science. We will know what the atmospheres of super earth type planets are when we gather enough data on them so that is where research energy should be placed.
“Higher life forms that use oxygen would have a problem as it would combine quickly with the hydrogen to form water.”
Which simply means that such life forms would not evolve, as they didn’t evolve on Earth while we had a reducing atmosphere.
“Pressure as well can be tolerated by barophilic organisms such as Halomonas salaria, protein deformation been a major issue though.”
I expect that if life evolves under such pressure, it gets automatically taken into account.
As has been pointed out by several, this statement does not hold up.
As for light or oxygen, neither of these were used by the earliest life, so their absence could do no more than limit its advancement (or maybe not). It should have no effect on the presence or absence of life.
Hydrogen is a possible candidate for certain energetic reaction in alternative, speculative life forms on Titan:
Hydrogen at tens of bars is actually used for deep diving, mixed with a tiny % of oxygen (partial pressure of 0.2 bars or so), it’s called hydrox :
You would need other traces gases to balance the effect of the high pressure on the working on the brain but it’s breathable by humans.
The main problem, like Bob mentions, is the greenhouse effect of the thick hydrogen atmosphere, although that might help somewhat to move the HZ outwards :
Higher temperatures also mean that the hydrogen will combine with any oxygen faster cleaning it up from the atmosphere.
With a super-earth in a orbit around a K or M dwarf more of the H/He envelope will be blown off.
The problem with thick atmospheres is not (only) the pressure, but temperature. Very roughly, suppose there’s a layer where radiation absorption, re-radiation, and reflection happens in a given atmosphere – it will have approximately the temperature of airless body at similar distance from the central star, since temperature on both of those wil be set by radiation balance with space. Anything below that layer will be at higher temperature: air that moves downwards from that layer will be compressed, and it’s temperature rises, air that moves upwards towards that layer, will decompress, and be lower in temperature (try googling with “dry adiabatic lapse rate”).
At earth, most radiation is actually absorbed at the surface, so it’s temperature survives this logic. Deeper atmosphere, like Venus, and you are out of luck…
I would imagine that Oxygen is the key to this debate . Most of Earth’s original hydrogen atmosphere might have reacted with oxygen to become water ,when cooling made this possible . This process prevented the hydrogen loss which probably happened at Mars and Venus , and it does not seem obvious that a surpluss of hydrogen (relative to the quantity oxygenating material ) is so easyly maintained over a long period . Lammer’s study does not deal with the dynamics of planetary chemisty , but seems to focus entirely on the fysics of the planetary formation . Therefore the maximum size of a waterworld might be bigger than his result sugests .
In which case, how do you explain the discovery of KOI-314c? Incidentally in regards to Henry Rouhivouri’s point – the host star KOI-314 is an M-dwarf.
It may also be worth considering Pierrehumbert and Gaidos (2011) “Hydrogen Greenhouse Planets Beyond the Habitable Zone” which outlines a scenario where life can cause the hydrogen greenhouse environment to collapse as methanogens convert hydrogen and carbon dioxide outgassed from the planet’s silicate core into methane. If the planet’s habitability depends on the hydrogen greenhouse, then biospheres on such planets may be self-limiting.
Deeper atmosphere simply implies that you need to be further from the star, and possibly wait longer before biology shows up. Eventually even stars cool, planets with thick atmospheres will be quicker to do so.
Perhaps you could get stable convection zones going, driven by radioactive decay, which would cool parts of the surface earlier. Or, how about a heavy atmosphere planet that’s tidally locked to it’s primary? Photochemistry on the hot side, active species condensing on the cold side, could drive life.
Which, in effect, means it is not a problem at all. Just an adjustment to the definition of the HZ for these planets.
After pressure, light, oxygen, and temperature have been shown to be red herrings, I am not sure that anything remains to support the notion that super-Earths are unsuitable for life. Did I miss anything?
Eniac: “After pressure, light, oxygen, and temperature have been shown to be red herrings, I am not sure that anything remains to support the notion that super-Earths are unsuitable for life. Did I miss anything?”
I wonder, would such super-earths be suitable for any higher life? This in particular with regard to the H2 atmosphere instead of O2. More specifically metabolism: would a strongly reducing atmosphere of H2 be suitable for releasing sufficient energy for higher (Eukaryote) cells and larger, active organisms?
I agree that life can arise in the most unexpected places. But environmental conditions set limits on energy (both chemical and solar)
availability. These limits will keep life at the low end of the development
scale. So these super earth’s may harbor life, but It will be of a class
of life that results in nothing more interesting than a paramecium, as their highest form of life.
But the paper implies that there is narrow corridor to arrive at environmental conditions similar to Earth’s when it hosted the first living
organisms. That is an atmosphere dominated by Nitrogen, Plus a slew of
Organic compounds and water vapor. Once again, that sweet spot of
.8 RE – 1.2 RE with similar density of Earth, is needed.
@Ole Burde March 2, 2014 at 16:21
‘I would imagine that Oxygen is the key to this debate . Most of Earth’s original hydrogen atmosphere might have reacted with oxygen…’
There was little oxygen in the early atmosphere after formation, it is most likely that oxygen was liberated from the hydrogen in water which then escaped or was converted to hydrocarbons by life.
@andy March 2, 2014 at 17:19
‘I am a little sceptical of H/He atmospheres on near in planets of around earth gravities, I mean planet formation is an energetic series of events which would leave the planet very hot~1000 degrees plus. Would the light gases hang around after each impact I think not.’
‘In which case, how do you explain the discovery of KOI-314c? Incidentally in regards to Henry Rouhivouri’s point – the host star KOI-314 is an M-dwarf.’
It is quite possible it is a water world that migrated in from the frost line which I discussed in an earlier A .C thread. My argument is not that hydrogen can shroud a ‘super-earth’ to great thickness Uranus and Neptune for example testify to that but the creation process in the first place. There is a lot of heat released during an impact with temperatures that are in the 10 to 100 000 K range which would expel a significant amount of H/He away on each impact. So unless the accretion process took longer so that the heat could dissipate after each impact the net effect would be less H/He accretion.
@Brett Bellmore March 1, 2014 at 14:26
“Higher life forms that use oxygen would have a problem as it would combine quickly with the hydrogen to form water.”
‘Which simply means that such life forms would not evolve, as they didn’t evolve on Earth while we had a reducing atmosphere.’
The first organisms did not use oxygen to any great extent.
Michael ”There was little oxygen in the early atmosphere after formation, it is most likely that oxygen was liberated from the hydrogen in water which then escaped or was converted to hydrocarbons by life.” …. True enough , but if we look at the earth system BEFORE it coolede down below 2000 degree celsius , there must have been a rocky core surounded by an ekstreemely thick atmosphere of mostly two parts hydrogen and one part oxygen . The pressure might have been ten times that of the Venus ‘s atmosphere , and in such conditions much hydrogen would be blown away by the solar wind . One relevant question is therefore how long time a planet remains warm enough for this to happen on a big scale . If there is time enough , it might be that super-earth size planets (up to a certain maximum size ) in the habitable zone automaticly will loose any EXCESS hydrogen ,for which there are no oxygenating agent aveiable to form heawier molecules later down the cooling process.
“…making it almost impossible for life to exist”.
That should be “life as we know it”. Earth shows that life can exist in a vast range of habitats (even undersea volcanic vents!), so it’s impossible for life to exist on these planets. It will just be something very werid…
Further to Tom West:
I often read people’s argumentation that we cannot say anything about the possibilities for (higher) life on other planets.
Though the physical circumstances on other planets and the appearance of lifeforms may be quite different, and life almost limitless in its diversity, the fundamental laws of nature are universal, and hence biochemistry will have to follow the same rules.
Moreover, life on our own planet indeed shows that ´life can exist in a vast range of habitats (even undersea volcanic vents!)´, however, it also show that this is not unlimited.
In fact our own planet Earth might be indicative of those limits. Life, at least higher life, may exist in hottish environments, but it does not exist under the very hottest conditions (e.g. active volcanoes), or the very coldest (interior of Antarctica), or the very driest (Atacama desert). Even bacteria and viruses can hardly grow there and would probably not have originated in such extreme environments.
Finally, surviving and adapting in an extreme environment is not the same as originating there.
For those interested in exoplanet hability ,I would like to draw your attention on a recent paper:
Possible climates on terrestrial exoplanets http://arxiv.org/abs/1311.3101 where authors deal with a broad range of possibilities.
And don’t forget to check the update of the habitable exoplanet catalog at
http://phl.upr.edu/projects/habitable-exoplanets-catalog , from 12 the count is now at 20 with 4 new planets from the last Kepler release and 4 new planets covered here:http://phl.upr.edu/press-releases/multiple-HZ
PHL is extremely interesting, however, I object against their very (VERY) optimistic definitions of habitability, particularly with regard to planet size:
even in the so-called ‘Conservative Definition’, habitable planet size is defined as from 0.1 to 10 Me, whereas the study mentioned in this post and other recent research would limit this to about 0.3 – 3 Me, and even that optimistically.
The ‘conservative HZ’, from 0.99 – 1.69 AU, I would call realistic (actually still optimistic on the outside, maybe a bit too conservative on the inside, that could be 0.97 or 0.95 AU).
Even their own conservative definition leaves only 10 (out of 20) planets.
If we apply the stricter criteria of 0.3 – 3 Me, then only 3 or 4 planets are left:
– GJ 667C f
– GJ 667C e
– Kepler-62 f
– Gl 581 g
Of which only Kepler-62 is remotely solar type, the others are orbiting M dwarfs and will certainly be tidally locked.
The truly habitable planets are still to be discovered.
Thanks for posting the links, waijai.
None of the planets in the habitable exoplanet catalog look like promising candidates. They’re all bordering on too big (by mass or radius), too hot, or both.
And check out the scatterplot of Kepler planet candidates (Figure 1 on p. 3) in the Forget, Leconte article. There’s a conspicuous absence of planets in the lower right hand corner, where Earth-like planets would be found.
It looks like the planet forming process favors a range of planets from mini-Neptunes to Jupiter sized gas giants or larger, but Earth sized terrestrial worlds are rare.
Robert Feyerharm wrote
‘It looks like the planet forming process favors a range of planets from mini-Neptunes to Jupiter sized gas giants or larger, but Earth sized terrestrial worlds are rare.’
Truth is Mr F, because of the Kepler (in the field vs designed) sensitivity we cannot definitively confirm your statement, as much I myself would like to
But as they say the circumstantial evidence is mounting
1) We haven’t found very many planets, Neptune sized and smaller in the
Habitable Zone of G or K, type stars.
2) Kepler completed almost 4 years of monitoring, that should be enough shake out most of real candidates, with a very reduced number of false positives. If HZ’s of most suns in the Kepler field of view are replete with planets we would have found them by now, we have not.
3) Unfortunately we cannot tease out a solar system similar to ours
(w/0 a Mercury though) from the data. So the solar systems with NO detections are the GREAT UNKNOWN.
Fortunately those Stars Kepler found to contain planets can be looked at by ground based telescopes. They are a good starting point for hunting for lower mass planets due to the likelihood of co-planarity of their orbits with those planets we already know about.
Some interesting points were raised at the Randi Forum where I asked about this very thing:
Take a look at what ben m said. Basically, the surface of such a planet is going to be a super-high-pressure blast furnace. No hope for any kind of life as there’d be no complex chemistry.
(to clarify: I mean a hydrogen-laden “Super-Earth”)
I can’t see in the article that they have modelled that the accreting matter would heat the envelope more as it moves through to the cores surface therefore making it puffier than a discrete atmosphere surface interface.
I am also not so sure about the amount but at those temperatures quoted wouldn’t the oxides in silicate material react with the hydrogen to form water and separate the ‘metals’ which move deeper into the core subsequently releasing thermal potential energy. Now if the oxides are broken apart and the hydrogen captured to form water, then water worlds would also be likely after all if we took all the oxides of the crust and mantle and combined it with hydrogen it would form a massive water world.
Thinking about oxide separation wouldn’t Late heavy bombardment events also have silicates rain through the hydrogen atmospheres possibly reacting to form water, again forming water worlds.
Michael – Interesting … Without being an expert on planet formation, my first check on your hypothesis is look at the gas giants in our own solar system, all of which are presumed to have solid iron-silicate cores. None of our gas giants are water worlds as far as I know. As I recall at the intense pressures within the cores of these worlds, hydrogen may exist as a liquid. I don’t know in these pressure/temp regimes if H2 will react with silicate material according to geochemistry observed on Earth.
@Robert Feyerharm March 5, 2014 at 18:03
Uranus like Neptune is much larger than a super Earth and looks to be made mostly of “ices/water” and rock, with about 15 % hydrogen and a little helium. I find the ratio surprising because there should have been a lot of hydrogen out there as it is a lot colder than closer in.
Now hydrogen under high temperatures and pressures should react with silicates to form water and hydrated minerals to an equilibrium state. I am not a mineralogist or an expert on planet formation as well but it does strike me a plausible process. Does anyone know a geologist/planetologist that could perhaps enlighten us on this potential process?
Paul and others: I just completed my own analysis of *all* Kepler planet candidates found up to now (3845), and looked in particular at earthlike planets (R between 0.8 and 1.25 Re) and planets in the HZ of their star. I also related the different Kepler size categories to orbital (AU) distance, which we normally don’t see in the press releases. The results are rather interesting and I could share them here or somewhere else (it is quite a bit), if people are interested.
With respect to your Kepler research:
1. Does it seem that Kepler is not sensitive enough to find planets of 1.25 Earth R or less except when such planet is within .25 AU of the parent star?
2. Can you draw any conclusions about the frequency of planets .80-1.25 Earth R in the habitable zone of sun-like stars?
3. In what is apparently the most common type of solar system (Tau Ceti, 82 Eridani) where there are no Jupiter size bodies but there are super-earths and/or mini-Neptunes within 1 AU of the star, could there still be an earth-sized body in the habitable zone?
Thank you for your statistics and educated speculations.
R Kelley, also to Ashley Baldwin (see previous post about new Kepler planets) and Paul;
I have done my own analysis using the Kepler data (from the Caltech University site of Kepler candidates) and the Kepler criteria for different size categories, being painfully aware of the fact that I am not a professional astronomer, but an “informed amateur”.
A recent publication, cited under a previous post, concluded that by far most candidates are genuine planets. This has encouraged me to indeed use all Kepler candidates (plus confirmed) up to now (i.e. through Q16, including the recent new ‘discoveries’), 3845 in total.
The outcome, I think, is rather interesting and in certain aspects somewhat different from what is often cited about Kepler the results. I am very willing, even eager to humbly submit the results, plus my approach, to Paul and others for comment, criticism, correction, etc.
I have just received a mail from Paul, so I will do some last checks this coming weekend and then submit things to Paul, for his perusal and decision what to do with it.
A few tips of the veil and in response to R Kelley’s excellent questions:
1. It is well-known that the transit method is strongly biased against planets in wider orbits, the chance of detection decreasing with increasing orbit (apart from those planets being discovered and confirmed later, because of minimal requirements with regard to nr of orbits). However, in my limited knowledge, and I would love to be corrected here, the transit method is not so strongly biased against smaller planets *beyond a minimum threshold size*, contrary to the RV method which is clearly and strongly biased against both smaller planets and wider orbits.
Besides, my analysis shows that the drop-off in abundance of smaller planets with increasing orbit is very steep indeed, probably even correcting for orbit, more about this later. This could only be explained if, as R Kelley is suggesting, Kepler is particularly unsuitable for detecting small planets even in slightly wider ( > 0.3 AU) orbits, which seems unlikely (?).
2. Yes, some interesting findings, in fact I found 2 very earthsized candidates smack in the HZ of solar type stars, plus some 7 ‘borderline’ cases, but they are all different from the published and confirmed Kepler planets. The other way around: *none* of the published confirmed Kepler ‘earthlike’ planets qualifies, according to the Kepler team’s own criteria for size and HZ.
3. The 6 million dollar question! And yes, my own big question and fascination: can there be earthlike planets in the so very common ‘compact’ systems of super-earths and mini-Neptunes, particularly around the common (later type, low metallicity) solar type stars? I have looked at the afore-mentioned 9 earthlike candidates around solar type stars. More about this later, but still not very conclusive I am afraid.