≡ Menu

TRAPPIST-1: An Abundance of Water?

Too much water helps planetary habitability not one bit. And while we find the availability of surface water a useful way of describing a potentially habitable world, we’re learning that some planets may have water in such abundance that life may never have the chance to emerge. It would be a shame if the numerous worlds orbiting TRAPPIST-1 fell into this scenario, but a multidisciplinary team from Arizona State University is making a strong case for the prospect.

What’s wrong with water? Let Natalie Hinkel (Vanderbilt University) explain. Hinkel worked with ASU’s Cayman Unterborn, Steven Desch and Alejandro Lorenzo on the question of water composition in these worlds. Coleridge’s “Rime of the Ancient Mariner” comes to mind — “Water, water, every where / Nor any drop to drink.” But in this case, there is plenty to drink, which is precisely the problem. Says Hinkel:

“We typically think having liquid water on a planet as a way to start life, since life, as we know it on Earth, is composed mostly of water and requires it to live. However, a planet that is a water world, or one that doesn’t have any surface above the water, does not have the important geochemical or elemental cycles that are absolutely necessary for life.”

Image: A nice visualization of the TRAPPIST-1 planets, here suggesting their relationship to the ‘snowline.’ As we’ll see below, what these planets are made of has implications for where they formed. Credit & copyright: NASA/Tim Pyle and Robert Hurt.

Not a good prospect, then, if the work of these researchers is any indication. What Unterborn et al. are saying in their paper in Nature Astronomy is that the TRAPPIST-1 planets are lighter than we would expect given our measurements of their mass and volume. All seven planets here appear to be less dense than rock. Remember, this is a transiting system, meaning we have constraints on mass and radius for all seven, allowing us to calculate density.

While low density worlds might well have a large gaseous envelope, the TRAPPIST-1 planets turn out to be not massive enough to hold onto the gas they would need to make up what Unterborn calls the ‘density deficit.’ If they somehow did hold onto the needed gas, they would be much puffier planets than what we see. The researchers argue that the low density component must be water, so the question becomes, how much water is there?

The numbers are daunting. Bear in mind as we look at them that the Earth, even with its magnificent oceans, is but 0.02% water by mass. Around TRAPPIST-1, the ‘dry’ inner planets b and c are likely to have less than 15 percent water by mass. Outer planets f and g are consistent with having more than 50 percent water by mass. These numbers will vary as we continue to constrain the masses of the planets, but the trend is clear enough.

“What we are seeing for the first time are Earth-sized planets that have a lot of water or ice on them,” said Steven Desch.

Which gets us to what planets like this can tell us about their formation and evolution. Planets with this much water — assuming water is the explanation for the density issue — should not have formed within the ‘snowline,’ that region within which water exists as a vapor and cannot be incorporated into a forming planet. Unterborn and team are clear on this point: The TRAPPIST-1 planets must have formed beyond the snowline and migrated to their current orbits. Indeed, these planets must have migrated from a position at least twice as far from the parent star as they are now. Have a look at the graph below to get the idea.

Image: This graph shows the minimum starting distances of the ice-rich TRAPPIST-1 planets (especially f and g) from their star (horizontal axis) as a function of how quickly they formed after their host star was born (vertical axis). The blue line represents a model where water condenses to ice at 170 K, as in our solar system’s planet-forming disk. The red line applies to water condensing to ice at 212 K, appropriate to the TRAPPIST-1 disk. If planets formed quickly, they must have formed farther away (and migrated in a greater distance) to contain significant ice. Because TRAPPIST-1 dims over time, if the planets formed later, they could have formed closer to the host star and still be ice-rich. Credit: Unterborn et al. / ASU.

This isn’t the first time we’ve seen migration discussed in relation to TRAPPIST-1. Simon Grimm (University of Bern Centre for Space and Habitability) and colleagues have looked at migration, noting that the resonant orbits here — the planets form a single resonant chain — is an indication of a slow migration consistent with the current perceived stability.

Other researchers have likewise addressed migration, including Chris Ormel (University of Amsterdam) and team, who look at planetary formation at the snowline itself in what they call a ‘resonant convoy,’ with the outer planets ‘pushing’ on the inner ones. So the idea of migration at TRAPPIST-1 is not new. What is new in the Unterborn et al. work is the use of planetary composition to add weight to the overall case for migration, which allows the team to quantify how much migration actually took place.

We’ve lucked out when it comes to nearby red dwarfs. TRAPPIST-1 will clearly be a primary source of data for red dwarf planets as we address the issue of habitability that their density and formation history implies. And then there’s that intriguing planet around Proxima Centauri…

The paper is Unterborn et al., “Inward migration of the TRAPPIST-1 planets as inferred from their water-rich compositions,” Nature Astronomy 19 March 2018 (abstract). The Grimm paper is Grimm et al., “The nature of the TRAPPIST-1 exoplanets,” in press at Astronomy & Astrophysics (preprint). The Ormel paper is Ormel et al., “Formation of Trappist-1 and other compact systems,” Astronomy & Astrophysics Vol. 604 (August 2017) (abstract).


Comments on this entry are closed.

  • Brett March 27, 2018, 11:07

    Water is a tricky issue when it comes to habitability. It seems to be important to have dry land that can exchange with the seas through a hydrological cycle, but getting enough water to have oceans but not so much that dry land is impossible appears to be pure chance (there’s no feedback systems that control the level of surface water on rocky planets).

    • Nicky March 27, 2018, 18:45

      Some studies suggest a negative feedback of volatile cycle on Earth. The amount of surface water tends to evolve to a steady state, which the water dagassing rate through subduction into mantle and outgassing rate through magmatism are equal, like Earth. This deep-water cycling leaves planet surface in partly wet (ocean) and partly dry (land).

      You can find more information in this paper.
      Cowan, N. B., & Abbot, D. S. (2014). Water cycling between ocean and mantle: Super-Earths need not be waterworlds. The Astrophysical Journal, 781(1), 27.

      • John walker March 28, 2018, 14:45

        Thanks for the link.
        From the paper, “We conclude that a tectonically active terrestrial planet of any mass can maintain exposed continents if its water mass fraction is less than ∼0.2%”
        Less than 0.2%. 10 times Earth’s water mass ratio. But still a tiny fraction.

  • Spaceman March 27, 2018, 11:28

    Hi Paul,
    Interesting post. Is this study saying that some of these TRAPPIST-1 planets could be covered by giant, sterile oceans of liquid water? What type of atmospheric conditions and/or composition would be needed to maintain a waterworld in its liquid state over billions of years?

    • Paul Gilster March 27, 2018, 13:17

      Good question on maintaining a global ocean, and I don’t have a good source for this, but maybe one of the readers will weigh in.

  • Ron S. March 27, 2018, 12:01

    “Earth, even with its magnificent oceans, is but 0.02 water by mass.”

    Minor typo here. It should read “0.02%”.

    • Paul Gilster March 27, 2018, 13:15

      Thanks for catching that!

      • Robin Datta March 27, 2018, 23:27

        That takes it from two-hundredths to twe-ten thousandths!

  • Tulse March 27, 2018, 13:57

    “a planet that is a water world, or one that doesn’t have any surface above the water, does not have the important geochemical or elemental cycles that are absolutely necessary for life”

    Europa, Enceladus, and the various other “ocean moons” have been strongly touted as possible locales for life. What would make the Trappist-1 planets different from these moons in terms of potential for life? Isn’t the presumption for the moons that the necessary geochemical processes are happening below the surface, at the water/rock interface?

    • Nicky March 27, 2018, 18:54

      Icy satellites pose less threat to life than Earth-size oceanworlds due to lower mass and gravity. Upper water body can create pressure onto the lower water body and turn it into high-pressure ice. On Earth-size oceanworlds, ocean bottom experiences much larger pressure and forms much thicker pressure ice, and this ice shell would separate the water body and underlying silicate rocks. That would be a lifeless ocean.

      You can find more information in this paper.
      Noack, L., Höning, D., Rivoldini, A., Heistracher, C., Zimov, N., Journaux, B., … & Bredehöft, J. H. (2016). Water-rich planets: How habitable is a water layer deeper than on Earth?. Icarus, 277, 215-236.

      • David H March 27, 2018, 21:40

        Very informative and fascinating. I was aware of pressure ice, but I didn’t think about it in connection with alien oceans.

  • Alex Tolley March 27, 2018, 14:28

    If water worlds are unsuited to life, then why the intense interest in icy moons? These are effectively water worlds with an ice cover. I’m not seeing any reason why we should hope to find life on Europa or Enceladus if life is unlikely for warmer water worlds. Conversely, should our icy moon[s] prove to be living, resulting from a local genesis, then water world exoplanets should be equally likely to be alive.

    • Nicky March 27, 2018, 18:38

      Detectable biosignature is the most important point. There might be life on ocean planets and icy satellites, but their biosignature would not detected by current and future telescopes because of low productivity. Not even in situ observation can confirm the existence of life on these planets. Icy satellites are perfect examples. We have to focus on more Earth-like planets, which harbor active and detectable biospheres.

  • John walker March 27, 2018, 14:57

    I bring this up every time planetary water ratios is mentioned in combination with biology. The situation with exoplanetary water content has been an astrobiological red herring in the popular consciousness for too long. While it is commonly assumed that liquid water will be of benefit to the evolution of life, it seems to me that the thought that often follows is: the more the merrier. What is overlooked is how a comparatively small amount of water can complicate the building of an ecosystem. With very little more water than the 0.02% we have on Earth, ice phases other than our familiar ice I will become a negative factor. A planetary body in the classical habitable zone with moderate gravity and single digit weight% water will not only have no land but will have a bathysphere so deep that the sea floor will be sealed with km thick ice VII (or VIII or X)reducing the habitability to nearly zero.

    At 100km depth or 10e9 Pa sea floor temperatures would need to be around 300K to avoid this scenario. At 200km to 300km this rises to 100°C and more. Oceanic depths of this magnitude will likely be the norm for all objects with even comparitively “low” water weight to mass ratios.
    From an astrobiological perspective it seems more sensible to look for water content numbers on planetary mass bodies well south of 1% rather than 2, 5 or 15.

    • David H March 27, 2018, 21:46

      Agreed on all points. Occasionally, when smart people argue about candidates for the Great Filter, the collision that produced our moon and dried out the Earth gets mentioned. In any case, it’s nice to see evidence that increases the likelihood of the Filter being behind us.

    • Brett Bellmore March 28, 2018, 12:24

      But it seems to me you could still have volcanism generating pockets of liquid water under the ice, even if the ice was too thick to penetrate. You might get isolated black smoker type ecosystems buried under the ice.

      Also, volcanism might cause ice volcanoes to carry minerals to the surface, eventually mineralizing the ocean.

      Obviously a thick layer of ice would be a problem, but it’s not a game over scenario.

  • andy March 27, 2018, 15:48

    The arXiv version of the Unterborn et al. (2018) paper was out before the Grimm et al. (2018) results on the planetary masses were out. At the time, the most up-to-date values for the masses implied extremely low densities (hence extremely high volatile contents for the planets), later results have increased the densities somewhat, though most of the planets (with the possible exceptions of c and e) are still implied to be very volatile rich. Not sure if the Nature Astronomy version has been updated to discuss the Grimm et al. mass determination though.

    • Nicky March 27, 2018, 18:33

      Nature Astronomy received Unterborn et al paper in 2017, so it represents the best TRAPPIST-1 composition study in 2017. It will not be updated since it is published already. Unterborn’s team might analyzed the data in the updated Grimm et al paper, but we just have to wait.

      • Harry R Ray March 28, 2018, 9:21

        I think the only MAJOR change would be that the f and g planets would be NO WHERE NEAR 50% water, but the both would still have global oceans hundreds of kilometers deep.

  • Michael March 27, 2018, 15:51

    If they are just water worlds could the U.V from the star have created highly oxygenated atmospheres.

  • Geoffrey Hillend March 27, 2018, 15:56

    The mass is what decides the gravity and escape velocity of the planet, not the density. If you have planet which has the equivalent mass or slight more mass than the Earth, the escape velocity we be about the same as the Earth.

    The density tells us what an exoplanets chemical composition. A low density does imply a lot of water. The question is whether or not the Trappist 1 planets have held onto a lot of atmosphere with the potential for solar wind stripping if they have no magnetic fields. If a planet does not spin quickly and is low density, it probably does not have a strong magnetic field, so it’s atmosphere must endure stripping.

    Ultra violet radiation does penetrate into water to a certain depth, but if life could start without direct sunlight, then it could begin on the night side or twilight parts and produce enough oxygen make ozone. If life needs sunlight to make the necessary chemical reactions to make it, then these are probably sterile worlds or the amount of life won’t ever be enough to produce detectable biosignature spectra.

  • galacsi March 27, 2018, 15:57

    The article says “All seven planets here appear to be less dense than rock.”
    but densities of planets b,c and e are 0.73 , 0.88 and 1.02 when earth density is 1.00 and Mars density is 0.71. So the above conclusions are very speculative and depend of this chain migration thing.
    In my opinion the discussion is not finished yet.

    • Nicky March 27, 2018, 18:29

      The article is based on the study of Unterborn et al, which used outdated masses from Wang et al, because at the time of writing of Unterborn, new mass data was just waiting to be analyzed. Briefly, the conclusion given in this article is not valid.

      As you have pointed out, new masses data suggest a rather Earth-like composition and density.

    • Alex Tolley March 27, 2018, 21:02

      I tend to agree that even the Unterborn paper doesn’t seem to make that claim for all planets. The Grimm paper seems to indicate that 1e is compositionally like Earth or Venus. Both papers indicate that the values are uncertain for all the planets. Both papers emphasize that the planets have migrated inwards, with several beyond the snowline, indicating that they should be volatile rich. So I suppose more data is needed to try to lock down the measurements to get more accurate characterizations.

  • Anton M. March 27, 2018, 16:04

    Well, it is not necessarily a bad news by me. If a planet STILL has so much volatiles now (rather than being a dry husk) means that the star flashes failed to rip these gases away. Even Venus lost its water, while these planets did not. Also, a massive ocean might solve the climatic woes associated with tidal lock. It is not so bad again. The bad news, maybe life on Earth occurred first on wetlands, estuaries, things like that, no chance for this on a waterworld. Also, any geothermals are proably at the crushing, sterilizing pressure. But again, the other day we thought about the lifeless, volatiles ripped wastelands and the ocean world does not seem nearly as hopeless to me.

  • Gary Wilson March 27, 2018, 17:49

    I tend to agree with galacsi. The density of 1e suggests a rocky planet doesn’t it? Otherwise 7 water worlds in one system? It seems highly unlikely doesn’t it? At what density is a planet thought to have no dense core for example? How can that even be determined given that the core could be very dense but very small? And if it was similar to Earth’s core in terms of composition wouldn’t it produce a strong magnetic field to protect the planet? Otherwise if these are 7 water worlds wouldn’t it imply that Trappist 1 originally was surrounded by a planet forming disc containing absolutely incredible amounts of water? Please understand that my training is in biology so I’m struggling with these concepts.

  • Michael C. Fidler March 27, 2018, 20:06

    1. A little different scenario: Starting with a simple example, taking Trappist 1 M dwarf star and planet b. What is happening is the hill space by this planets is carving out a much larger and faster area around this star then is taking place around our sun. In Trappist b case the comets would be encountering its hill sphere and have their orbits being change on a much shorter and more often time period. What you have is a planet with a mass similar to the earth in an orbital period that is 242 times shorter then earth’s, that is influencing a much larger gravitational area by being so close to the star. Next: 2. Adding the rest of the planets around Trappist1.

    • Michael C. Fidler March 28, 2018, 11:07

      2. Adding the rest of the planets around Trappist 1: As in the case of planet b the rest of these planets would also influence comets but creating more short period comets that would create lower velocity impacts on the inner planets.(Jupiter family of comets is an example.) Instead of the atmospheres being stripped by these impacts they should keep a healthy quantity of volatiles on their surfaces. We see an extended hydrogen exospheres around b and c that could be caused by comet impacts but also water and other volatiles being released from a large deep tectonic system that is continually recycling them. Similar to the earth’s spreading oceanic ridge system, that would hold onto the volatiles for a much longer time period and release them in volcanic eruptions. Not only would these impacts create panspermia in these system but also rings and possible temporary moons. The impacts should also be on the leading side of these tidally locked planets and that would affect the atmospheric circulation and temperatures.
      Next: 3. Polluted dwarfs and Rogue planets.

      • Michael C. Fidler March 29, 2018, 10:33

        3. Polluted dwarfs and Rogue planets. Since a large number of comets would be causing impacts on Trappist 1 planets, the M dwarf star will also be having its share of cometary impacts. Both H2O and CO molecules have been observed in the Near-infrared spectroscopy of M dwarfs. This is caused by a higher cometary impactor to mass ratio in these systems. Other elements and molecules related to comets should also be looked for in the spectrum of these polluted M dwarfs resonant chain systems.
        These cometary Oort clouds can have a mass of anywhere between the mass of earth to larger than Jupiter depending on initial birth and evolution of the clouds. The comets have several different ways that their orbits can be changed to send them into the inner planetary system. Passing nearby stars, planets in orbit beyond the cloud or inside the oort cloud and a new possibility; the large number of rogue planets that are between star systems. There would also be an accelerated period of impacts caused by when these stars pass thru the Galactic midplane about every 35 million years.

        Cometary impactors on the TRAPPIST-1 planets can
        destroy all planetary atmospheres and rebuild secondary
        atmospheres on planets f, g, h.


        Near-infrared spectroscopy of M dwarfs. I. CO molecule as an abundance indicator of carbon.


        Near-infrared spectroscopy of M dwarfs. II. H2O molecule as an abundance indicator of oxygen.


        Systematic Search for Rings around Kepler Planet Candidates:
        Constraints on Ring Size and Occurrence Rate.


  • Michael March 28, 2018, 6:01

    If these worlds have migrated inwards from twice their current distance a serious amount of energy must have been dissipated, could this heat have inflated their atmospheres to lower their overall densities? I would not be surprised with the closeness of these worlds to each other that they are massively volcanic.

    • Michael March 28, 2018, 15:02

      Tidal heating of these worlds is significant some are as high as Io the moon that orbits around Jupiter. Perhaps a massive steamy atmosphere?


    • Nicky March 29, 2018, 1:01

      Planetary migration takes place in gaseous disk, driven by surrounding gases or planetesimals, it does not provide heat to drive atmospheric expansion.
      What you are talking about here is orbital tidal evolution, a completely different mechanism. Tidal dissipation from eccentricity damping is small for the outer TRAPPIST-1 planets (d-h), because in such tightly packed system high eccentricity would cause dynamical instability.

      • Michael March 29, 2018, 13:27

        Planetary inward migration can also occur if an orbiting world is traveling faster than the rotation of the central star as some of trappist-1 worlds do. Some of these worlds are near a hundred times closer to the central star than earth is to our sun. This closeness will raise tides on the central star which decays the orbit of some of these inner worlds. These inner worlds decaying orbits and the resonances with the outer worlds may be pulling the whole lot of them in. These orbital dances and central star interaction via tidal forces will be huge and generate a lot of heat, there is also the possibility of tidal love number resonances having occurred in the past to generate even more heat.

        • Nicky April 1, 2018, 3:25

          That is exactly what I was saying, orbital tidal evolution, but the migration talking about in the article is based on the interaction with early disk not stellar tides, because the paper focused on the location of snowline in the early disk.
          Regardless what the planets have experienced in the past, the current tidal dissipation is too low to drive atmospheric expansion and too insignificant to lower the bulk densities.

          • Michael April 1, 2018, 15:18

            Migration inward and outward can take place after the disc has long gone and it can be significant for close in planets. There are two tidal force bulges at play but both are the same, one on the star and the others on the planet’s themselves. As each planet orbits the star the planet’s pass each other closely, they will distort as each other as they pass. The world will rock back and forth as each distorted bulge is attached to another. It is this rocking and change in shape that will heat the world up just like with Io even if the eccentricity is very low. Io has an eccentricity of 0.004 which is less than the earth but it gets really hot because of other orbiting moons. Trappist world are much larger and heavier that is why they can get very hot. If this heat can’t get out it could inflate an atmosphere, planet d may be getting a lot more by been trapped between two large worlds.

            • Nicky April 1, 2018, 19:02

              I do not deny the fact of orbital evolution induced by tides, but it is widely acknowledged that the configuration of TRAPPIST-1 is a result of migration in protoplanetary disk, and tidal evolution cannot cause this scale of inward migration.
              Both tidal dissipation studies arxiv.org/pdf/1712.05641.pdf and arxiv.org/pdf/1707.06927.pdf show that only the inner two planets are tidally heated enough to melt the inside, but these two are already receiving enough stellar flux to drive atmospheric expansion, therefore tidal heating is unneeded to explain their low densities. For example, 1d is already at the runaway greenhouse limit, stellar flux itself can expand the atmosphere. For the outer planets (f-h), a large amount of condensed volatile must present.

              • Michael April 2, 2018, 13:35

                “..but it is widely acknowledged that the configuration of TRAPPIST-1 is a result of migration in protoplanetary disk, and tidal evolution cannot cause this scale of inward migration.”

                Inward migration on low mass stars is very significant, these tidal forces are hundreds to thousand of times what the earth experiences.


                “Both tidal dissipation studies… show that only the inner two planets are tidally heated enough to melt the inside…”

                If you heat a planet from inside it can have a dramatic effect on the density of the rock and water on the surface. Water for instance at 300 C is thirty percent less dense than water near 0 C so a molten surface is not needed.

                “For example, 1d is already at the runaway greenhouse limit, stellar flux itself can expand the atmosphere.”

                Trappist d does not appear to be near the greenhouse runway threshold, however if it has degased greatly to provide a thick atmosphere it could be.


                “For the outer planets (f-h), a large amount of condensed volatile must present.”

                I believe all of these Trappist worlds where near or well into the snowline and collected a lot of water. They are been pulled inwards by the two inner worlds through intense tidal interactions. The two inner worlds are subject to powerful tides pulling them inwards and the other five are been pushed out but the tidal forces doing that are less. It is the resonance that is causing the whole systems inward migration and they cant break free without a massive disruption.

                Just my observation of course.

                • Nicky April 2, 2018, 23:25

                  In fact, the first reference given in the reply does not support your view. Tidal migration happens on the *very* close-in planets, for M-dwarf the orbital rotation is less than a day (fig.2 and section 3, ultra-cool dwarf would be even shorter), let alone pulling the planets at 0.05 AU inward. The paper also states:”we have also neglected the contributions from interactions with a disk. Planet migration in a disk through Lindbald resonances is generally thought to be more efficient than the effects considered in this work”.

                  Tidal evolution models of low-mass star (0.1 solar mass) system show that it would be efficient on very close-in planets (<0.01 AU), but pulling planets from outside of snowline is just simply impossible for tidal migration. See the evolution models, outside 0.01 AU the semi-major axis of the planet barely change.

                  Bolmont, E., Raymond, S. N., & Leconte, J. (2011). Tidal evolution of planets around brown dwarfs. Astronomy & Astrophysics, 535, A94.
                  Bolmont, E., Raymond, S. N., Leconte, J., & Matt, S. P. (2012). Effect of the stellar spin history on the tidal evolution of close-in planets. Astronomy & Astrophysics, 544, A124.
                  Turbet, M., Bolmont, E., Leconte, J., Forget, F., Selsis, F., Tobie, G., … & Gillon, M. (2018). Modeling climate diversity, tidal dynamics and the fate of volatiles on TRAPPIST-1 planets.

                  Molten and solid mantles with same composition and mass do not change the planet radius at an observable scale, because they are refractory not like volatile. A molten Earth would just result the same density as today.

                  TRAPPIST-1d does lie in the optimistic habitable-zone defined by Recent Venus limit in Kopparapu et al (2013), which is derived empirically. Multiple sophisticated 3D General Climate Model rather show TRAPPIST-1d lies outside the runaway greenhouse limit.

                  Wolf, E. T. (2017). Assessing the habitability of the TRAPPIST-1 system using a 3D climate model. The Astrophysical Journal Letters, 839(1), L1.
                  Turbet, M., Bolmont, E., Leconte, J., Forget, F., Selsis, F., Tobie, G., … & Gillon, M. (2018). Modeling climate diversity, tidal dynamics and the fate of volatiles on TRAPPIST-1 planets.
                  Gillon, M., Triaud, A. H., Demory, B. O., Jehin, E., Agol, E., Deck, K. M., … & Bolmont, E. (2017). Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature, 542(7642), 456.

                  • Michael April 3, 2018, 14:08

                    Let’s agree to disagree.

                    • Nicky April 3, 2018, 21:07


  • Roon John March 28, 2018, 7:02

    Even Venus lost its water, while these planets did not. Also, a massive ocean might solve the climatic woes associated with tidal lock. It is not so bad again.

  • Michael March 28, 2018, 10:45

    Water loss from U.V maybe not be that bad at least on more massive worlds. If water scales in proportion with the mass of material forming the planet i.e. double the mass double the water the area exposed to the U.V does not, it is a quarter.

  • Joe McCam March 28, 2018, 11:14

    We have seven planets in close proximity, locked into harmonious orbits that are mutually stabilizing. This highly improbable configuration is suspicious enough. Now we find that the planets have very low densities. I don’t think it’s totally out of line to consider one more idea: These could be planetary Dyson Spheres. That seems like a plausible step before attempting to encase an entire star, and would explain the low densities and orbital configuration.
    I’d still apply Occam’s Razor & give this a low probability, but ruling it out all together is trying to dictate reality rather than explore it.

  • Michael C. Fidler March 28, 2018, 11:36
  • Harold Shaw March 28, 2018, 11:42

    This may be a naive assumption but underwater volcanic activity should partially compensate for the supply of minerals, gasses etc, from the weathering of land features. If abiogenesis can occur at hydrothermal vents, then I don’t think we can rule out water worlds as living worlds.

    This may also be a naive assumption, but wouldn’t lifeless water worlds have strikingly clear oceans; life needs dissolved stuffs to start and makes dissolves stuffs as it lives? If we are ever able to measure the clarity of an exoplanets oceans, how clear the oceans are could indicate whether the oceans contain life.

  • andy March 28, 2018, 13:08

    A potential problem for the habitability of planets with substantial water layers is that the water layer needs to cool sufficiently for oceans to form. The interior of the planet will be hot from accretion and radioactive decay, so the water should stay heated for a long time. You then end up with a supercritical envelope that lasts for billions of years rather than an ocean.

    • Dave Moore March 28, 2018, 19:49

      Actually, I think that the planet will cool relatively rapidly on geological timescales. Once the top of the mantel reaches 1200 deg C, it will crust over reducing the heat flux from the interior considerably.
      Then once the top of the steam atmosphere goes below the condensation point of water, you will start getting thunderstorms and violent convective overturn, which is very efficient at bringing heat up from the lower atmosphere and radiating it from the cloud tops.
      This would particularly be the case if you had an atmosphere with a substantial proportion of CO2 in it. Suppose you had an atmosphere that was 75% steam 25% CO2. It would have an average molecular weight of about 25. Once the steam rained out of the air column its average molecular weight would shoot up towards CO2s (44). This heavy column of air would then precipitously descend into the atmosphere greatly increasing the strength of the convection.

  • ole burde March 28, 2018, 15:50

    What density would Venus seem to have when measured from afar ? Much of what we talk about here is still open to any number of interpretations , but observing Venus from the orbit of pluto might add a more solid datapoint . Relative to Earth , Venus have a much thicker atmosphere , and knowing how this would influence transit opservations comparing the two planets could be usefull . Some might believe to know the answer , but experience show that one solid datapoint can narrow down a lot of possibilities . Hopefully this could soon be enhanced by additional transit opservations from a greater distance

  • Dave Moore March 28, 2018, 15:56

    What exactly do we get with a deep ocean planet. I did a little research last night looking at the high pressure, high temperature phase diagrams of water. If we take an Earth mass planet and assume a near freezing ocean, then at about 40 miles down Ice VI forms (6300 bar). At about 160 miles down Ice VII will form (2.2 kilobar). If you increase the temperature of the water column (superfluid over 375 deg C), the pressure at which ice forms increases steadily. By the time you reach mantle-like temperatures (say 1000 deg C), 1 Mbar is required to form the solid form of water. (This is the equivalent of 6300 miles down, which means you’ve run out of planetary radius.) So, if you have a cold water planet like the outer Trappist ones, you’ll have a deep ocean, a high pressure ice layer that will melt where it is in contact with the planet’s mantle. This water/superfluid will have considerably less density that the overlying ice so it will force it’s way up through the ice into the overlying ocean. Being in contact with a mantle that is fully hydrated, it will be saturated with minerals, which will precipitate out much like they do on Earth’s mid-oceanic ridges.
    I would expect the planet’s atmosphere to be more oxidizing from the disassociation of water and subsequent loss of hydrogen. The vent minerals will be reducing, so there may be a chemical gradient to drive life.

    • Michael March 28, 2018, 16:49

      Due to tidal heating of these planets there must be huge convection cells in any ices, a bit like a lava lamp and as you say can transport minerals into an ocean. I have a feeling these worlds are very hot under their atmospheres much too high for life.

    • Nicky March 28, 2018, 23:42

      Hi Moore, ocean does not directly contact with high-temperature mantle, instead there are thick crust and silicate lid separating them. Planetary thermal evolution models suggest the surface temperature should not be higher than 400K, and the maximum thickness of full-time liquid ocean is under 200km.

      Noack, L., Höning, D., Rivoldini, A., Heistracher, C., Zimov, N., Journaux, B., … & Bredehöft, J. H. (2016). Water-rich planets: How habitable is a water layer deeper than on Earth?. Icarus, 277, 215-236.

  • Tom March 28, 2018, 21:11

    This paper might not be as conclusive as it sounds. An ESO study that came out on the arXiv in February (https://arxiv.org/pdf/1802.01377.pdf) drew different conclusions about the TRAPPIST-1 planets’ densities and concluded that they are most likely <5% water. If true, this would make the system much more friendly to life.

  • Harry R Ray April 23, 2018, 9:47

    Paul Gilster: Could you ask the authors of ArXiv: 1804.07537 if the same analysis they did for Jansson could ALSO be done for the TRAPPIST-1 planets? “Mass, radius, and composition of the transiting planet 55 Cnc e: using interferometry and correlations.” by Aurelien Crida, Roxanne Ligi, Caroline Dorn, and Yveline Lebreton. If so, we MAY know the LIKELIHOOD(NOT the CERTAINTY) of atmospheres with Earth-like surface pressures and clear(or NOT)skies IN ADVANCE of JWST secondary eclipse observations, which are NOW not likely to happen until late 2021 AT THE EARLIEST, and most likely early 2022(or even later if the March 2020 launch window is NOT met). The reason I am posting this comment is: “…we find the radius of the gaseous envelope is 0.08+/- 0.05Rp…”

    • Paul Gilster April 23, 2018, 10:13

      Harry, can you clarify your reference to Jansson? The name does not appear in this paper.

      • Harry R Ray April 24, 2018, 9:35

        Jansson is now the official IAU name for 55 Cnc e.

        • Paul Gilster April 24, 2018, 10:06

          Thanks, Harry, I had missed that.

  • ljk May 15, 2018, 12:49

    Francois Luus

    January 25, 2018

    TRAPPIST-1 Interplanetary Eavesdropping on IBM Cloud


    The first interplanetary eavesdropping SETI observation between two TRAPPIST-1 planets (e & f) in conjunction with Earth was conducted on April 6, 2017. The hypothesis is that an advanced civilization could have established radio-frequency communications between planets e & f, and if these planets line up with Earth (conjunction) we could “eavesdrop” on that interplanetary transmission with a sensitive radio telescope.

    In this post we are bringing SETI to the cloud by using IBM® Data Science Experience to compute signal spectrograms and autocorrelation plots to look for signs of possible ETI transmissions during a TRAPPIST-1 conjunction.

  • Harry R Ray May 20, 2018, 16:44

    Paul Gilster: I have posted a couple of comments about TRAPPIST-1d NOT being spherical, but instead, being LOPSIDED(i.e. oblate on one AND ONLY ONE side due to VERY EXTREME WEATHER CONDITIONS. This, right now is a conjecture AT BEST and NOWHERE NEAR to being even a HYPOTHESIS. However, a recent preprint on ArXiv(ArXiv 1805-06722 “Polarization of TRAPPIST-1 by the Transit of its Planets.” by Sujan Gupta, MIGHT catapult my conjecture to hypothesis status in the IMMEDIATE FUTURE! The premsie is as follows: TRAPPIST-1 is of a sufficiently LOW temperature that condensate clouds could form, and that dust particles in those clouds would scatter light from TRAPPIST-1, allowing for polarimitry studies. The problem is that NO current instrument available today is SENSITIVE ENOUGH to make a viable detection due to TRAPPIST-1’s assumed SYMETRICALLY SPHERICAL SHAPE. Dr Sengupta has found a BRILLIANT way around this problem by “ASYMETRICISING” TRAPPIST-1 sufficiently during the INGRESS and EGRESS phase of a transiting planet so that EXISTING instruments COULD detect the polarization! Please read the PDF, which is CURRENTLY on page one of the Exoplanet.eu website bibliography section list page. I am assuming that Dr. Sengupta ALSO assumes that exoplanets themselves are ALSO symetrically spherical, thus producing symetrical(sinusoidal?)light curves at both ingress and egress. After reading the PDF(and paying PARTICULAR ATTENTION to the image of the PROJECTED TRAPPIST-1d ingress/egress light curves, if you think that a lopsided planet transit might produce slightly ASYMETRICAL(non-sinusoidal)light curves, ask try to find out if Dr. Sengupta CONCURS and post a “yes” or a “no” as a reply to this comment. I know that this will probably take some time which you may or may not have immediately, but please reply ASAP. Thanks.

    • Paul Gilster May 20, 2018, 16:56

      At the moment I’m afraid I’m swamped, Harry, and will not be able to do this, but you might want to write Dr. Sengupta yourself, since you’re already familiar with the paper, and post here what you find out.

  • ljk September 21, 2018, 12:39

    Planet-Planet Tides in TRAPPIST-1

    Tides are complicated and hard. When a planet orbits a star, the star raises a bulge on the planet’s surface. As the planet rotates, this bulge moves across the surface, and because rocks and oceans and atmospheres are viscous this dissipates energy. The energy comes out of the planet’s rotation, which spins it down. This is likely why Venus and Mercury rotate so slowly.

    I think that the TRAPPIST-1 system is exhibiting yet another effect that we haven’t seen before: planet-planet tides. The planets in this system are very tightly packed, and quite massive (much more than the Solar System moons are). It’s likely that eccentricity tides have made their orbits circular (they have very low eccentricity) and that they are synchronously rotating with the star (or stuck in a spin-orbit resonance of some kind, like Mercury is).

    But if that’s true, then when neighboring planets lap each other, the strain caused by their mutual tides are actually of similar magnitude to the strain caused by the eccentricity tides with the star. In fact, if our estimates of the planet masses and eccentricities are right (they’re pretty uncertain, so this is not necessarily a safe assumption) then planet f causes significant tides on planet g that are actually more important than the very small variations in that planet’s tides raised by the star.