≡ Menu

Are ‘Waterworlds’ Planets in Transition?

Ponder how our planet got its water. The current view is that objects beyond the ‘snow line,’ where water ice is available in the protoplanetary disk, were eventually pushed into highly eccentric orbits by their encounters with massive young planets like Jupiter. Eventually some of these water-bearing objects would have impacted the Earth. The same analysis works for exoplanetary systems, but the amount of water delivered to a potentially habitable planet depends, in this scenario, on the presence of giant planets and their orbits.

Dorian Abbot (University of Chicago) and colleagues Nicolas Cowan and Fred Ciesla (both at Northwestern University) note the consequences of this theory of water delivery. One is that because low mass stars are thought to have low mass disks, they would have fewer gas giants and would produce less gravitational scattering. In other words, we may find that small planets around M-dwarfs are dry. On the other hand, solar-mass stars and above could easily have habitable planets with amounts of water similar to the Earth.

Waterworlds and their Future

‘Waterworlds’ are planets that may have formed outside the snow line and then migrated to a position in the habitable zone. A planet like this could be completely covered in ocean. In any case, we can expect habitable zone planets could have a wide range of water mass fractions; i.e., the amount of water vs. the amount of land. The Abbot paper studies how variable land surfaces could influence planetary habitability, and the authors attack the question using a computer model for weathering and global climate, assuming an Earth-like planet with silicate rocks, a large reservoir of carbon in carbonate rocks, and at least some surface ocean.

Image: A waterworld may be a planet in transition, moving from all ocean to a mixture of land and sea. Credit: ESA – AOES Medialab.

Interestingly, the researchers found that partially ocean-covered planets like the Earth are not dependent upon a particular fraction of land coverage as long as the land fraction is greater than about 0.01:

We will find that the weathering behavior is fairly insensitive to land fraction when there is partial ocean coverage. For example, we will find that weathering feedbacks function similarly, yielding a habitable zone of similar width, if a planet has a land fraction of 0.3 (like modern Earth) or 0.01 (equivalent to the combined size of Greenland and Mexico). In contrast, we will find that the weathering behavior of a waterworld is drastically different from a planet with partial ocean coverage.

What that means is that planets with some continent and some ocean should have habitable zones of about the same width, no matter what the percentage of land to water. The conclusion is based upon the fact that silicate weathering feedback helped to maintain habitable conditions through Earth’s own history. The weathering of surface silicate rocks is the main removal process for carbon dioxide from the atmosphere, and it is temperature dependent, thus helping to buffer climate changes and expanding the size of the habitable zone around a star.

Seafloor weathering also occurs, but the authors point out that it is thought to be weaker than continental weathering and to depend on ocean chemistry and seawater circulation more than surface climate. That would mean carbon dioxide would be removed less efficiently from the atmosphere of a waterworld, which would produce higher CO2 levels and a warmer climate. A planet like this would be less able to buffer any changes in received solar radiation (insolation) and would thus have a smaller habitable zone.

Planetary Evolution at Work

All this is leading up to an absorbing conclusion about waterworlds. Assuming that seafloor weathering does not depend on surface temperature, planets that are completely covered by water can have no climate-weathering feedback. Thus the conclusion that a water world has a smaller habitable zone than a planet with even a few small continents. But a waterworld may be, depending on its position in its solar system, a planet in a state of transition. Abbot and company posit a mechanism that would put a waterworld through a ‘moist greenhouse’ stage which would turn it into a planet with only partial ocean coverage, much like the Earth. Here what would have been complete loss of water is stopped by the exposure of even a small amount of land:

We find… that weathering could operate quickly enough that a waterworld could “self-arrest” while undergoing a moist greenhouse and the planet would be left with partial ocean coverage and a clement climate. If this result holds up to more detailed kinetic weathering modeling, it would be profound, because it implies that waterworlds that form in the habitable zone have a pathway to evolve into a planet with partial ocean coverage that is more resistant to changes in stellar luminosity.

A waterworld thus becomes an Earth-like planet after going through a ‘moist greenhouse’ phase — this occurs when a planet gets hot enough that large amounts of water are lost by photolysis in the atmosphere and hydrogen escapes into space. As water is lost and land begins to be exposed, the moist greenhouse phase can then be stopped by reducing the carbon dioxide through silicate weathering. This is the process the authors call ‘waterworld self-arrest.’

Although we have not performed a full analysis of the kinetic (non-equilibrium) effects, the order-of-magnitude analysis we have done indicates that a habitable zone waterworld could stop a moist greenhouse through weathering and become a habitable partially ocean-covered planet. We note that this process would not occur if the initial water complement of the planet is so large that continent is not exposed even after billions of years in the moist greenhouse state…

It’s also true that waterworlds at the outer edge of the habitable zone would not be in a moist greenhouse state in the first place. We’re likely to find waterworlds, then, but some of them may be in the process of transformation, becoming planets of continents and oceans. And any Earth-sized planet discovered near the habitable zone would be a good candidate to have a wide habitable zone and a stable climate if it has at least a small area of exposed land. That makes discovering the land fraction of any Earth-class planet we observe through future planet-finder missions a priority. The authors believe that missions of the Terrestrial Planet Finder class should be able to determine the land fraction by measuring reflected visible light.

The paper is Abbot et al., “Indication of insensitivity of planetary weathering behavior and habitable zone to surface land fraction,” accepted at The Astrophysical Journal (preprint). Thanks to Andrew Tribick for the pointer.


Comments on this entry are closed.

  • Phil August 17, 2012, 11:22

    I have always thought that Venus was a fascinating place, and much understudied. I would love to see a mission put a balloon into its’ atmosphere. I found this article very interesting, as it reminded me of the story of Venus’ past. It is thought Venus had a lot of water at one point, but it all vanished.

    From wiki, “The weak magnetosphere around Venus means the solar wind is interacting directly with the outer atmosphere of the planet. Here, ions of hydrogen and oxygen are being created by the dissociation of neutral molecules from ultraviolet radiation. The solar wind then supplies energy that gives some of these ions sufficient velocity to escape the planet’s gravity field. This erosion process results in a steady loss of low-mass hydrogen, helium, and oxygen ions, while higher-mass molecules, such as carbon dioxide, are more likely to be retained. Atmospheric erosion by the solar wind most probably led to the loss of most of the planet’s water during the first billion years after it formed.”

    This sounds like the story in your article Paul, except the planet did not have a magnetic field strong enough to hold on to the water while it was part of the atmosphere.

    From the article above, “We note that this process would not occur if the initial water complement of the planet is so large that continent is not exposed even after billions of years in the moist greenhouse state…”

    Maybe we can learn something about Venus’ past as well. Did Venus have so much water that the green house effect completely ran away?

    Again from wiki, “…[Venus’] atmospheric mass is 93 times that of Earth’s atmosphere…” and “~96.5% carbon dioxide”.

    So, maybe the answer is much more simple. Venus has too much CO2 in its atmosphere for silicate weathering to remove no matter the size of ocean the planet may have once had.

    Or maybe something else is going on. I am no expert so I will leave this banter at that. These are just the thoughts that came to my head while reading this article. :)

  • Tim Whitworth August 17, 2012, 11:23

    Does anyone know how wave action would relate to this? What fraction of the continental mass is lost to wave action? Would waves with an unlimited fetch (fetch means “the distance over which the wave builds up, in miles”) have very large waves that would overtop continental shelves?

  • Alex Tolley August 17, 2012, 15:00

    The weathering of surface silicate rocks is the main removal process for carbon dioxide from the atmosphere, and it is temperature dependent, thus helping to buffer climate changes and expanding the size of the habitable zone around a star.

    I don’t agree that only pre-biotic geochemistry is important. Once life formed, and it was very early in earth’s history, life started to affect carbon cycles. Once photosynthesis started, life has become the dominant mechanism for carbon cycling. The low C13 levels of early carbonate rocks indicates that life was present on earth and actively involved in the carbon cycle.

    Now I appreciate the authors may be arguing for their hz calculations assuming abiotic (conservative) conditions, but until we can rule out life originating in extreme environments (or surviving after transfer) , this might be rather premature. It is currently estimated that earth will exit the hz for current life in less than 1 bn years, as atmospheric CO2 levels are so reduced to prevent warming that terrestrial plants will be unable to survive, although this shouldn’t effect marine organisms, AFAICS.

  • James M Essig August 17, 2012, 15:49

    Hi Paul;

    Just thought I’d stop by TZ-CD to make a brief comment. I think the prospect of water worlds has profound ramifications especially with regard to in situ deuterium sequestration as fusion fuel for star ships. Also, the idea of folks eventually gazing out the window of a water world lander can provide much additional muzings.

  • Spugpow August 17, 2012, 19:19

    Phil: a balloon was already put on Venus. http://en.wikipedia.org/wiki/Vega_program#Balloon

  • Interstellar Bill August 18, 2012, 0:07

    How is a water world supposed to get a high-oxygen atmosphere?

    Ours came about when massive organic burial prevented oxygen from recombining with the dead organisms that produced it, allowing O2 to accumulate, and said burial was a consequence of the extraordinary geological episode known as the Great Unconformity, which couldn’t have happened with only 1% land cover, vs the 25% back then.

    So in your scenario those travelers will have to make their own O2.
    Worse yet, without O2 in the air there could never be anything for them to eat, there being no multicellular life without appreciable O2.

    Earth had a lot more bottlenecks than most are willing to admit.

  • torque_xtr August 18, 2012, 8:25

    (thinks of super-Arrakis with no axial tilt, 99% of water photodissociated, thin abiogenic oxygen atmosphere and temperate poles at Mercury’s distance, Venus twin at the distance of asteroid belt, and ocean Neptune with essentially no sunlight, but with water ocean kept liquid by 1 kbar of insulating H2/He, smaller radius but otherwise visually indistinguishable from our Neptune.)
    Water-based habitability clearly depends on surface composition and atmospheric conditions more than on stellar flux…

  • Alex Tolley August 18, 2012, 10:21

    @interstellar bill
    If I understand your point correctly, one implication is that geo-engineering to remove CO2 with oceanic algal growth is futile.

    I’m not so sure about your last point. One can eat anaerobic microorganisms with suitable processing. We don’t know that multi-cellularity requires aerobic metabolism. Multi-cellular macro-organisms appear to have evolved after O2 had built up in the atmosphere. But had anaerobic multi-cellular organisms evolved, they may well have been out-competed.

  • Michael August 18, 2012, 15:49

    Water worlds should be quite stable provided they have sufficient gravity and a strong enough magnetic field. If water is broken down into hydrogen and oxygen the hydrogen may escape leaving oxygen which could form an ozone shield to prevent further losses or at least reduce it.

  • Adam Crowl August 18, 2012, 19:35

    Hi All

    Some responses…
    Interstellar Bill, carbon burial didn’t spike during the Unconformity AFAIK. Carbon burial occurred before there was much land-life, so clearly oceanic biospheres can do it. Photolysis will also produce significant oxygen – a wet-greenhouse planet undergoing hydrogen loss will build-up potentially ~tens of bars.
    torque_xtr, nicely put. “Habitable” has a quite different meaning to “human tolerable” these days – looking at its usage it actually means “biocompatible” without any restriction on the kind of “bio” it’s compatible with. Thus a broad range of habitats are out there in the cosmos, but human-friendly conditions are a small fraction of those.
    Michael, even with magnetic fields there will be significant water-loss via photolysis and solar-wind pick-up around the magnetic poles. Earth loses gases via the “polar wind” which flows out from the magnetic poles. Gas/water loss happened much more rapidly in the past because the Sun’s solar-wind was ~1000 times stronger than it is today. Even from Mars or Venus the current loss-rate via the solar-wind is negligible.

  • Brett Bellmore August 19, 2012, 10:29

    No, as I understand it, in the upper atmosphere gasses start to segregate themselves according to molecular weight, because lower molecular weight gasses are moving faster at a given temperature. This happens above the ozone layer, at the “turbopause”, and as a result the highest portions of the atmosphere, subject to being swept away, consist mostly of hydrogen and helium. This is going to be the case regardless of how much oxygen is present in the lower atmosphere. Though a strong magnetic field would slow down the stripping rate by keeping the solar wind well away from the upper reaches of the atmosphere.

  • Rob Henry August 19, 2012, 16:13

    @Adam Crowl, from what little I know of stellar dynamics, I am struggling to put your claim that the solar wind was once about 1000 times stronger than today in anywhere but its very brief T-Tauri phase. If so, wasn’t this well before completion of planet formation?

    @Brett Bellmore, my understanding of Earth’s case was that hydrogen (but not helium) loss was only limited by its diffusion rate from the lower atmosphere to exobase. Under those circumstances, the ozone layer and Earth’s trap can work in conjunction to retain water against loss via photodisassociation.

  • amphiox August 19, 2012, 18:21

    So in your scenario those travelers will have to make their own O2.

    For travelers with the technology to get there, making O2 on a planet covered in water is easy.

    Earth had a lot more bottlenecks than most are willing to admit.

    Until we actually discover another alien biosphere to compare earth to, we will never know whether or not those events represent true universal bottlenecks or simply earth-specific ones.

    For example, respiration using H2 as the basis for electron transfer redox reactions is 1/3 as energy efficient as respiration using O2, and probably the second most energy efficient potential energy source for life (with O2 being number one) using relatively common elements and molecules that lifeforms would be expected to have access to. In the absence of O2 and turbo-charged O2 respirators to compete with, H2 respiration may well be sufficiently energetic to support multicellular life.

  • amphiox August 19, 2012, 18:23

    Organic carbon sequestration is actually only one of two known mechanisms for producing an O2 rich atmosphere. The other of course being UV photolysis followed by subsequent loss of H2 to space, and this one happens during moist-greenhouse events.

  • Eniac August 19, 2012, 20:12

    Phil: The next best thing to seeing a balloon mission on Venus (which you have missed by a few decades, as has been pointed out), is to read about it (and other equally exciting ones) here:


  • Eniac August 19, 2012, 20:21


    The weathering of surface silicate rocks is the main removal process for carbon dioxide from the atmosphere, and it is temperature dependent, thus helping to buffer climate changes and expanding the size of the habitable zone around a star.

    According to what I have read, and as Alex has also alluded to, the removal of CO2 through surface weathering is negligible compared with removal through biological mechanisms followed by burial. It is also clear, I think, that neither Venus nor Mars have had much CO2 removed from their atmospheres. I have a hard time understanding how surface weathering can nevertheless play such a large role in climate control. Am I missing something?

  • Paul Gilster August 19, 2012, 20:40

    Alex makes a good point, though I don’t know that CO2 removal is negligible through surface weathering. The authors point to this paper:

    Walker, J., Hays, P., & Kasting, J. 1981, J. Geophys. Res., 86,

    as their source, and also to:

    Feulner, G. 2012, Rev. Geophys., 50

    when they describe surface weathering as the main removal process. I haven’t read either of those papers and so can’t comment. They do talk about ‘some debate about the details,’ so maybe this is much more open than I realized. Alex’s point about life affecting the carbon cycle doesn’t come into play in the current paper.

  • amphiox August 20, 2012, 2:47

    The question of oxygen is actually an interesting one in terms of habitability of these evolving water worlds.

    Since the mechanism by which they change from fully water worlds to more earth-like worlds with some exposed land surface requires moist-greenhouse loss of water to space, and this process involves UV photolysis of water with loss of hydrogen to space, but retention of O2, immediately after the transition to a terran-like condition, these worlds would be expected to have a high-O2 atmosphere (which may not necessarily last long, as the O2 will react with the exposed land surface, and perhaps with dissolved substances like Fe as well, though I anticipate that these substances would have in fact been saturated out by reacting with O2 earlier, during the middle of the moist-greenhouse period, when O2 would have been constantly produced in the atmosphere)

    If the water-world had already undergone abiogenesis and produced prokaryotic lifeforms, it is possible that atmospheric oxygenation would produce an oxygen cataclysm and a major mass extinction, though what survivors, if any, there are would have the opportunity to take advantage of the oxygenated atmosphere, assuming it persists long enough.

    But if the water-world had not undergone abiogenesis prior to its transformation, the oxygenated, oxidizing atmosphere would likely strongly inhibit abiogenetic processes that might otherwise occur in the now-terran-like world.

  • Zanstel August 20, 2012, 6:10

    O2 is a wonderfull system to prevent Hydrogen loss, because at lower levels in the atmosphere, Hydrogen reacts naturally with oxygen and tends to make water, and in this layers, minimal relative water vapor tends to freeze and drops, so in a oxygen rich atmosphere, exists a water trap making free hydrogen at high altitudes far less concentration that in a not O2 rich atmosphere.

    Probably, a Mars or Venus terraformed with high O2 concentration will have much less loss of Hydrogen.

  • Rob Henry August 20, 2012, 20:26

    Eniac asks “According to what I have read, and as Alex has also alluded to, the removal of CO2 through surface weathering is negligible compared with removal through biological mechanisms followed by burial. It is also clear, I think, that neither Venus nor Mars have had much CO2 removed from their atmospheres. I have a hard time understanding how surface weathering can nevertheless play such a large role in climate control. Am I missing something?”

    And from what I have read you have. Notice how wonderfully stable the climate of Venus is, and how stable that of Snowball Earth should be. Between these two stable configurations for Earth climate, exists a series of metastable configurations for Earth – that seem to be made truly stable by some powerful and not fully understood negative feedback mechanism(s). So far the strongest of these that has been identified looks as if it is rock weathering, so you can understand a certain level of anguish from some corners if you wish to remove it from you model of what an alien biosphere might look like.

    If it turns out that alternative climate feed back mechanisms are not sufficiently powerful, you might want to put your alien biosphere in one of Earths two other stable configurations. This opens a can of worms since you then also have to decide if the increasing insolation over a scale of billions of years presents a problem or not.

  • Rob Henry August 20, 2012, 21:59

    Amphiox writes “For example, respiration using H2 as the basis for electron transfer redox reactions is 1/3 as energy efficient as respiration using O2, and probably the second most energy efficient potential energy source for life”. But I feel that that underrates H2 as an ideal basis for higher life. I feel that we should also be mindful of the following.

    1) hydrogen, unlike oxygen, has such a low degree of natural toxicity, that even creatures that have evolved in its absence, such as humans, can tolerate a very high level of it in their breathing mix.
    2) biology texts repeatedly mention diffusion rate limitations of oxygen as a limiting problem in evolution, but seldom mention nutrient density problems. In this regard it is worth noting that H2 diffuses four times faster than O2, and that most planets that have H2 rich atmospheres will have far more than 20,000 Pa of this gas at their surfaces, thus almost certainly providing high potential for active life than the modern Earth does.
    3) On a planet where food production for plants is three times easier, it is not at all obvious that plants might tend to overcompensate for the lower energy density of the food, and their might typically be more total energy available to animals.

    Also worth noting is that if Houtkooper is right and the Viking results were produced by Martian life using a eutectic mix of hydrogen peroxide and water as their internal solvent, then this opens the possibility of concentrated oxidants being the basis for high energy storage, and that gas diffusion limitations having no importance whatsoever for more typical biospheres than our own.

  • Rob Henry August 20, 2012, 22:22

    Zanstel, I completely agree, but I additionally see something lurking in the background that you do not mention.

    Note how Venus is assumed to have lost a similar amount of water to Earth. Photolysing 1000m of water would produce 900 tons of it per square metre, or about 8 million Pascals. To oxidise it all would require tens of thousands of tons of ferric rich rock erupted over the Cytherean surface FOR EACH SQUARE METRE of its surface, or a massive non-Jeans outflow of O+ that stretches my be belief in the orthodox explanation to the limit. I have never seen this situation explained in detail, and have this horrible suspicion that we have got something here very wrong.

  • Eniac August 20, 2012, 23:04

    For what it’s worth, this figure:


    seems to indicate that weathering makes up only 0.1% of the carbon flux, roughly. Perhaps it would be dominant on lifeless worlds, but that claim better be well supported before it can be taken seriously. Assuming the figure is accurate and I am interpreting it correctly.

  • Eniac August 20, 2012, 23:41


    And from what I have read you have. Notice how wonderfully stable the climate of Venus is, and how stable that of Snowball Earth should be. Between these two stable configurations for Earth climate, exists a series of metastable configurations for Earth – that seem to be made truly stable by some powerful and not fully understood negative feedback mechanism(s).

    We really do not know how “wonderfully stable” the climate of Venus is, if at all. We know even much less about this “Snowball Earth”. All this sounds like pure speculation to me, and hardly enough to discount the well-known facts about the carbon cycle I was referring to. I still do not see where I am off, here.

    The stability of Earth climate can be quite readily explained by a simple negative feedback mechanism: As CO2 is removed, Earth will get colder, which will reduce life activity, which will reduce CO2 removal, which will cause warming. No mysterious “series of metastable configurations” is required, at all.

    Eventually, the increasing solar output should overwhelm this stabilizing effect when it gets so hot that life activity is decreased, leading to the reappearance of CO2 in a positive feedback loop and a sudden climate catastrophe. Life could then be reduced to thermophiles, and photosynthesis might cease.

    Another long-term factor that may not have been appreciated enough in geological history is that the Earth must be getting steadily drier, and may have been much wetter in earlier periods. Similarly, the atmosphere must be getting thinner and probably was much thicker earlier on. A large part of this thinning is probably due to CO2 removal, but loss of other gases, including water, must also have played a role.

    If you think about it, once you remove the CO2 from the atmosphere of Venus, what remains is much more Earth-like in density and composition, although still with substantial differences. Most notably there is still a substantial lack of water, albeit somewhat less pronounced. Perhaps Venus is what the Earth will look like after the aforementioned climate catastrophe, in a billion years or two.

  • Rob Henry August 21, 2012, 17:04

    Eniac, notice how the perspective of we landlubber humans is so distorted that we call our planet Earth. With that same fault I would first bring it to your attention that global warming might increase the desert coverage and, if this is sufficient, might actually result in a decrease of carbon sequestration on land.

    That would have been a red herring though because it is far more significant that our oceans become significantly more productive the closer they come to 4C, because this promotes mixing of surface layers with mineral rich bottom layers of water. The effect is so dramatic that our Artic and Antarctic seas are more productive than the tropical oceans despite their paucity of sunlight.

    So the feedback loop from biological C sequestration should be positive – though there might be a few narrow sweet spots in the range where it is negative.

    And I admit it, Venus and Snowball Earth might have been said to be in very stable configurations but I have never seen figures, and doubt that that contention has ever been put to the test with detailed modelling. It still feels right to me though.

  • Eniac August 21, 2012, 22:09


    … it is far more significant that our oceans become significantly more productive the closer they come to 4C, because this promotes mixing of surface layers with mineral rich bottom layers of water.

    This is a good point, but I would submit that increasing ice cover will decrease productivity, which reverses the feedback to be negative, again. The degree to which the arctic and antarctic ocean are covered by ice would suggest that this type of feedback may be dominant. The ice ages may be seen as periodic bumps against the limits of this feedback. It gets cold, until the growing ice cover reduces primary productivity so much that CO2 starts building up leading to a bounce-back via the greenhouse effect. Of course, this is also all just speculation, but it seems right to me, somehow.

  • amphiox August 22, 2012, 0:23

    We know even much less about this “Snowball Earth”.

    We know Snowball Earth occurred at least twice, and, most critically, ended twice.

    We also know that during its history earth spent far less time as a Snowball Earth than as a not-Snowball Earth.

    These factoids alone demonstrate that Snowball Earth (specifically, as opposed to a snowball-generic-earth-like-planet) must, in fact, NOT be all that stable a climactic configuration.

    Because if it was, we would see it. Right now. Right here.

  • Rob Henry August 22, 2012, 16:49

    Amphiox, why did geologists ignore the evidence for Snowball Earth until very recently? I put it to you that their problem was that they thought that knew such an event would result in a new permanent stable climate, and so could never have occurred on our particular planet.

    Ways to escape this require us to go to extremes, such as 30,000 Pa CO2 (compared to the current *elevated* level of 35 Pa). In that case it would be a wonder that Earth’s volcanism could supply this, and so such mechanisms (whatever they actually turn out to be) might not apply to more typical life-bearing planets.

    Observations of Earth’s past are biased by our existence as observers.

  • amphiox August 22, 2012, 21:08

    But, Rob Henry, no matter how “extreme” the escapes from a Snowball Earth scenario are, the fact is that they happened, not just once, but at least twice, in the history of just one planet.

    And means that either these “extreme” escapes are not so extreme after all, or we are wrong and there are other less extreme escapes from the Snowball Earth.

    Either way, that means that Snowball Earth was NOT a stable state.

    Otherwise you are suggesting that the Copernican Principle does not apply and Earth is somehow special, and you will need positive evidence for that assertion.

    And even if that were true, that STILL means that Snowball EARTH was not a stable state, it just means that Snowball Typical-Habitable-Planet is a stable state, and earth is not a typical habitable planet upon which the snowball state is NOT very stable.

  • Eniac August 23, 2012, 8:20

    Rob: According to here:


    Volcanic emissions of CO2 are 65 to 319 MT per year. The atmospheric reservoir (today) is around 800 GT. In 2-5 thousand years, therefore, atmospheric carbon could be replaced entirely. The 30,000 Pa atmosphere you mention would take 2-5 million years to accumulate. So, perhaps that is how snowball Earth ended: Accumulation of volcanic CO2 after the biological sinks for it were shut down by all the ice.

  • Dave Moore August 23, 2012, 18:05

    As far as I’m aware silicate weathering is a very effective CO2 removal mechanism. When rocks break down, they release Calcium ions. These combine with aqueous CO2 to precipitate out Calcium Carbonate. CaCO3 is highly insoluble so CO2 in solution scavenges nearly all Calcium ions.

    Atmospheres with high levels of CO2 in them, such as Earth after several million years of being in the snowball state, produce very acidic rain (carbonic acid), which dissolves rock rapidly thereby sequestering the CO2 very effectively.

    The end of Earth’s snowball periods are marked with enormous limestone deposits.

  • Rob Henry August 23, 2012, 22:39

    Oh, so that explains it Eniac. If Dave Moore is right, rock weathering might not currently sequester much C vis-à-vis biology, but it could do if thing heat up.

    Anyhow modern biological C capture is easy to measure, but its long-term sequestration due to these processes hard (eg what rate is peat actually getting buried in bogs at), so we tend to just look at the first well-known figure, and that might be very misleading.

  • Dave Moore August 24, 2012, 14:25

    Unfortunately, until now I haven’t had time to read the paper, but now that I have I notice that they have not taken into account the effects of the Oxygen build up.

    A planet in moist greenhouse phase will lose Hydrogen and experience a build up of Oxygen in its atmosphere. Now, if the Oxygen is not sequestered, then the surface atmospheric pressure will increase, which also means the surface temperature will in increase through the adiabatic lapse rate. And, if the surface temperature increases this will lead to more water vapor in the air and an increase in the greenhouse effect. Even without this, a sufficient increase in atmospheric pressure will raise the ocean surface temperature above boiling point at some point.

    Vast amounts of Oxygen can be sequestered through oxidation of the crustal and mantel rocks. This is what is though to have happened on Venus, but on an ocean planet the rate is limited by the amount of seafloor spreading and weathering. If this rate is exceeding, there will be a build up of many bars of O2 in the atmosphere.

    Another point to remember is that on an Earth mass planet, once the Ocean depth exceeds 42 miles, you get Ice VI forming on the ocean floor and therefor no ocean floor weathering.

  • Rob Henry August 24, 2012, 17:55

    Dave Moore, wouldn’t it take a truly colossal amount of oxygen to increase temperature just “through the adiabatic lapse rate”. In fact wouldn’t it take so much that the extra pressure raises the boiling point of the water faster than it does the temperature? I admit ignorance here, but the opposite result sounds rather counterintuitive.

    That brings me back to the intriguing case of Venus. Above I calculated that if it started off with about half Earth’s inventory of water it would have to rid itself of about 900 tons of oxygen for every square metre of surface. It would be an impressive sort of rock that reduced more than a 100g of oxygen for every kilogram of its own weight, so I was wondering where this 10,000 odd tons per square metres of reduced Cytherean rock came from?

  • Eniac August 24, 2012, 19:00

    Rob and Dave: Yes, this is a good point. It seems weathering may be a very effective way to remove CO2 from atmospheres rich in it, while biological mechanisms dominate when there are only trace amounts, as now.

    Dave: I find your remark about limestone deposits after snowball Earth very enlightening. It seems to confirm my suspicion that these periods are marked by increasing atmospheric CO2 concentrations, which then become limestone once the ice cover shrinks and CO2 removal begins. The ice cover could act both by suppressing life activity and by inhibiting weathering, I suppose.

    I also find very interesting your suggestion that many bars of O2 can build up in the atmosphere by water dissociation, inorganically. Is that speculation, or is it well-founded? I would appreciate any references. Would that mean that planets without life, but breathable air are a distinct possibility?

  • Rob Henry August 24, 2012, 21:54

    Eniac, here is the remaining problem that to my mind no one seems able to answer. The D/H ratio on Venus is 100 times that on Earth. This implies strongly that Cytherean atmospheric hydrogen was dominated by Jean’s loss, and not any other mechanisms.

    More specifically it implies that the atmosphere at the time of the hydrogen loss had an exobase temperature around its current one of 400K rather than one more typical of an oxygen dominated atmosphere such as Earth’s which is currently around 1000K (a temperature where both hydrogen and deuterium rate loss should just equal the diffusion rate at which these isotopes can reach the exobase).

    What is happening to all that oxygen! And why can I never find published work that attempts to explain it!

  • Rob Henry August 24, 2012, 22:07

    Above I should have given as my final sentence; Why can I never find work that concentrates on quantitative rather than qualitative explanations for the dioxygen loss.

  • Ronald August 25, 2012, 4:56

    Dave, Eniac: if O2 can build up in an atmosphere inorganically by photodissociation of water, this would mean that the presence and spectroscopic detection of it in an exoplanet’s atmosphere would not necessarily be a biosignature. Annoying.
    However, I suspect that the circumstances and characteristics of such an planet and its atmosphere would be different from one with O2 from biological origin. A much warmer planet? Much higher O2 levels? Only waterworlds? …?

  • Dave Moore August 25, 2012, 19:43

    To Rob/Ronald/Eniac

    My source for the idea was a description in Ward & Brownlee’s “The Life and Death of Planet Earth.” He was discussing what Earth’s future would look like under the moist greenhouse effect, and mentioned that Oxygen levels would build up again, only this time inorganically.

    Ward and Brownlee’s books are well annotated. “The Life and Death of Planet Earth” lists approximately 180 sources. Unfortunately they are alphabetically listed by author, not even by chapter, so it is impossible to source a given remark. They did however mention they are drawing on the work of James Kasting for that chapter and cite 7 of his papers, so if anyone has dealt with effects of photodissociative O2 buildup, it would be him.

    Surface weathering of rocks would remove a lot of Oxygen from the atmosphere. Mars has no Oxygen in its atmosphere but it’s surface is intensely oxidized, so we know what happened to the photodisassociated Oxygen there, but we are talking about planets that are losing Hydrogen at a far greater rate than Mars. And it is possible to overwhelm this mechanism (see Earth.) So, I would think that the sort of planet would be a smaller one that Earth were thermal escape would be more prevalent. It would probably start off with a higher volatile component, and it would be not too geologically active by the time the land surface becomes exposed.

    Chemically speaking, I’m thinking this type of planet with its completely oxidized surface and oceans and no reducing compounds will have a build up of Nitrous oxides from lightning (I suspect the NO2 will have a good spectral signature) resulting in an ocean of dilute Nitric Acid.

    Incidentally, Ward & Brownlee quote a mean surface temperature of 70 C before rapid water loss occurs on Earth.


  • Rob Henry August 25, 2012, 20:27

    Above it now seems that I overplayed the high D/H as a problem. On reading literature on this problem itself (rather than that of how the atmosphere and climate of Venus has changed over the last few billion years) I realise that I missed the following.

    It may be that no other means can show as high a preference for H escape over D as the Jeans mechanism, but that does not mean that it can’t still show a very high preference, as in the 10:1 range. Thus this *distillation* might have begun from Earth-like ratios over as little as 1000 times Venus’ present inventory, even if loss was dominated by non-Jeans mechanisms. This is still only about 1% Earth inventory levels, making it possible that this ratio evolved only after the first 99% of H was lost (assuming that Venus was once similar to Earth).

    Also on reflection, a little new water must added to the system from bound water in deep primordial hydroxide minerals through volcanism. I have even seen it postulated that the Earth still holds several times more water in this form in its mantel than in its ocean. That would seem to mitigate the chance that any information that we can glean from this ratio held information from that postulated time of O2 build-up.

  • Rob Henry August 26, 2012, 0:18

    Let me try again for that Venus mystery. Here goes…

    If Venus lost most of its O2 by atmospheric mechanisms, why is it that its atmosphere still retains about 70 times more non-radiogenic neon than Earth.

    If it was mainly by geochemical processing, how did its mantel convey sufficient ferrous rock (replenished from the boundary layer in contact with the core perhaps?) to the very surface of the planet. What is the best guess as to how fast such a conveyer could work once the initial surface material is oxidised to bring up, what amounts to a layer of rock that is at least 2 to 10 km thick. How much more volcanically active than Earth would this make Venus?

  • Rob Henry August 26, 2012, 16:58

    Dave, 70C is obviously an evaporation, not a boil off that I imagine would only happen much closer to the critical temperature of 647K (374C). Brownlee and Ward must postulate that H loss mechanisms go into overdrive here (perhaps the cold trap moves higher than the ozone layer at this point??).

    Actually at 70C we would only have 30,000 Pa of water pressure, or 3000kg/sqm. Since all ocean is vapour when this nventory is 3000tons/sqm I can’t help wondering if they confused their units when quickly turning an overview into a descriptive text that was just for public consumption.

  • Dave Moore August 26, 2012, 17:09

    Just a point with working with lapse rates using rule of thumb: the average value of the adiabatic lapse rate for Earth is 6.4 deg. C. It’s 10 deg C for dry air/5 deg for saturated air.
    Earth’s atmosphere halves in pressure every 6 km, so if Earth had a 2 Bar atmosphere, its surface temp would be 38 deg. C higher than it is now (average 53 deg C vs 15 deg C).

    If I remember what I read about Venus, it is postulated that once the runaway greenhouse occurred, you got a 1000 Bar atmosphere of mainly steam, which had a higher temperature than the melting point of rock at its base. The circulating magma would have no trouble reacting with the oxygen and absorbing it over geological time.

  • Rob Henry August 26, 2012, 21:00

    Dave, I love that solution to the mystery of Venus’ disappearing O2, especially since I have not heard of a primordial ocean of rock on its surface tens of kilometres deep. It is such a beautiful answer that it seems a pity that ferric salts tend to be significantly less dense than their ferrous equivalents, and wreck its simplicity. I am wondering how fast oxygen can diffuse through this stagnant magma ocean to great depths.

    And your back-of-the-envelope calculations for adiabatic lapse rate temperature rise is also interesting. If doubling pressure really does raise temperature by about 40K then this compares with a rise in water’s bp of only 20K.

  • Eniac August 28, 2012, 21:57

    It may be that a hydrogen rich atmosphere also contains a lot of methane, in which case any free oxygen would be immediately bound into CO2. We would see water and methane decrease, and CO2 increase with time, as hydrogen escapes.

  • Eniac August 31, 2012, 0:51

    Rob: As LJK relates in his latest post, it appears Venus has lots of vulcanism, which I believe is quite capable of explaining how oxygen can be completely absorbed into the lithosphere, over long time periods.

  • Rob Henry September 1, 2012, 22:25

    Eniac, to me this is the blind leading the blind. I had hoped for criticism from a geologist. Note that it was making the figures work that really concerned me.

    Firstly I had boldly stated that it was difficult to imagine a (common) mineral that reduced more than 100g of O2 per kg rock. I was hoping some geologist would mention pyrite, where upon I could show that lifting the requisite quantities of a mineral that was twice the density feldspar or quartz to that height could be all but ruled out by energy considerations alone.

    To me it really has to be ferrous minerals (not sulphides) that are responsible, and the rate that they are transferred from the mantel (not recycled from the crust) is the only relevant figure here. That is a hard figure to come by.

    The crust of Io is an order of magnitude more Volcanic than Earth, but erupts about three orders of magnitude more material at 1-10mm resurfacing per annum. Finally I did the back-of-the-envelope calculation and saw that at Io’s rate it would take only in the order of a million years, and at Earth’s a billion if none of the volcanic ejecta was recycled (from the base of the crust) and most of it was of just the right sort of reduced material. The new problem was that if Earth and Venus are such geological twins, this now implied that biology on Earth could have only have ever produced about 1% of the O2 (through sequestered carbon) that our volcanoes could reduce – absolute tops!

    All that is why I was so attracted to that melted crust possibility that Dave brought up. To me this crust would be a true liquid and not a solid that flowed and allowed S wave transmission like the mantle. Unfortunately that would imply very tight stratification by density, and is why I thought it such a pity that ferric compounds are invariably denser than their ferrous counterparts (simple ones about 10% so).

    Eniac, I hope that you can now see that no simple model seems to fit without aid of some new information hitherto unknown to me. You might well think that because geologists invariably discuss the matter as if they understand it, that they do indeed understand. I have found from past experience that making that sort of assumption is a colossal error.

  • Rob Henry September 2, 2012, 7:10

    Oops I meant to write ferric compound are LIGHTER of course, the problem being that they float to the top.

    Also I should give figures for my 1% assertion. Typical figures for inorganic carbonate in Earth’s crust, put 160 tons of carbon in it per square metre. Typical figures for current crustal organics (including associated carbonates) give these around 28 tons/sqm C. The average oxidation number from living C is very close to 0, meaning that this buried C represents 75 tons of O2 released. For Venus we wanted its hydrolysed water (which I actually placed at only 1/3 Earth’s level) borne 900 ton/sqm inventory to be reduced in a billion years. Over 3 billion years of photosynthetic life on Earth this implied at least 2700 ton/sqm of O2 could have been reduced. OK that gives us 3% of the necessary figure, but that feels too generous to me. Venus and Earth must be very different in some geological way, but what?