≡ Menu

TRAPPIST-1: The Importance of Age

If life can arise around red dwarf stars, you would think TRAPPIST-1 would be the place to look. Home to seven planets, this ultracool M8V dwarf star about 40 light years away in Aquarius has been around for a long time. The age range in a new study on the matter goes from 5.4 billion years up to almost ten billion years. And we have more than one habitable zone planet to look at.

Adam Burgasser (UC-San Diego) and Eric Mamajek (JPL) are behind the age calculations, which appear in a paper that has been accepted at The Astrophysical Journal. We have no idea how long it takes life to emerge, having only one example to work with, but it’s encouraging that we find evidence for it very early in Earth’s history, dating back some 3.8 billion years. But we also have much to learn about habitability around red dwarfs in general.

Image: This illustration shows what the TRAPPIST-1 system might look like from a vantage point near planet TRAPPIST-1f (at right). Credit: NASA/JPL-Caltech. [PG note to JPL: Please append artists’ names to such images! I want to credit the artist but have no idea who it is].

The good thing about being a somewhat older red dwarf is that flare activity should have slowed over time, a fact that the authors confirm. This doesn’t make it necessarily benign. In fact, as the paper points out, “…despite TRAPPIST-1’s modest emission as compared to other late-M dwarfs, the radiation and particle environment is still extreme as compared to the Earth.”

And because the habitable zone planets (e, f and g) around TRAPPIST-1 orbit as close to it as they do — all seven planets orbit within Mercury’s orbit around the Sun — they have long been exposed to radiation that could have destroyed their atmospheres and caused their oceans to evaporate. The orbital periods here range from 1.5 to 19 days, with orbital semi-major axes of 0.011-0.062 AU. Tight indeed!

The paper notes that based on current estimates, the high energy emissions of TRAPPIST-1 are probably enough to have evaporated an Earth’s ocean of water mass from each of the planets save the outer two over the lifetime of the stellar system. Another ominous note: The process of stripping an atmosphere can go into high gear when the magnetic field lines of a star interact with those of a planet, funneling stellar wind particles directly to the planet’s surface.

But we don’t know enough about these planets to make the call, and there are other factors that come into play, including the possibility of thick atmospheres. From the preprint:

… current estimates of the planets’ densities are generally below Earth’s average density (Gillon et al. 2017; Wang et al. 2017), suggesting volatile-rich worlds that may have ample reservoirs; while ocean evaporation and hydrogen loss could result in an oxygen- and ozone-rich atmosphere that could shield the surface from high UV fluxes (Luger & Barnes 2015; O’Malley James & Kaltenegger 2017). Transit spectroscopy measurements of the atmospheres of these planets are currently insufficient to detect the signatures of all but the lightest elements (de Wit et al. 2016), but the James Webb Space Telescope should have the sensitivity to detect Earth-like atmospheres around these planets, if they exist (Barstow & Irwin 2016).

So the key question is, will we find atmospheres on these planets when we have the technology in place to spot them? TRAPPIST-1 is close enough to the Earth that space-based assets should be able to give us an answer soon. You can see how significant the James Webb Space Telescope is as we look toward the future of characterizing Earth-mass planet atmospheres. A thick atmosphere can shield a planetary surface as well as redistributing heat from the dayside to the dark on these presumably tidally locked worlds. Too much of a good thing, of course, can lead to the kind of greenhouse effects that have so ravaged our neighboring world Venus.

Burgasser and Mamajek’s study is an important one, because age is critical for understanding the evolution of this star’s planetary system. The paper uses a variety of tools, ranging from average density to flare activity, lithium absorption, metallicity, kinematics, rotation and magnetic activity to make the call on age. Also a factor: How fast the star is moving in the galaxy.

The conclusion that TRAPPIST-1 is a far older star than previously thought has implications for the stability of the planetary system:

… N-body simulations presented in Gillon et al. (2017) showed the planetary system to be consistently unstable on timescales < 0.5 Myr, with only an 8% chance of surviving 1 Gyr. This is refuted by the much older age we infer for the TRAPPIST-1 star. However, recent simulations show that the resonant configuration of these planets is in fact highly stable through disk migration on timescales of 50 Myr (1010 orbits), with or without eccentricity dampening. That this system appears to have persisted for over 5 Gyr, despite dynamical interactions that are readily detectable through transit timing variations (Gillon et al. 2017; Wang et al. 2017), suggests that the resonant configuration is indeed inherently stable.

Addendum: I was unclear about the issue of disk migration and stability, a matter about which Dr. Burgasser was kind enough to comment in an email. He refers to a 2017 paper called “Convergent Migration Renders TRAPPIST-1 Long-lived” by Daniel Tamayo and colleagues (abstract here), and goes on to say this:

In brief, their simulations show that the T-1 system must have migrated in slowly in order to end up in its compact system, and in effect uses the disk to damp out any eccentricities that arise between the moving planets. Such systems are not “immune” to instabilities, but the authors of this paper were able to show for a range of initial conditions a planetary configuration that looks similar to T-1 and lasts for a large number of orbits (10 billion!).

The current work has been done with data from the Spitzer space telescope, the continued use of which should help to tighten up estimates of the TRAPPIST-1 planet densities, which is also a factor as we try to determine their compositions. Further work with Hubble and, of course, with JWST should help us learn whether there are indeed atmospheres in this planetary system.

The paper is Burgasser & Mamajek, “On the Age of the TRAPPIST-1 System,” accepted at The Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Rob Flores August 15, 2017, 15:38

    Assuming that the M star is as old as stated,
    Then why make the assumption that ALL of them started out as
    near terrestrial planets. What if the Outer Planet (H) was in fact
    a mini-neptune. Could we be looking the CORE of a small jovian Giving rise to the possibility that it has recently entertained conditions of a much cooler near earth twin at .75 RE. This would still mean that any surface liquid and atmosphere will be blown away eventually, Sterilizing the planet, But for a limited window it might marginally be able to give rise life.

    • andy August 16, 2017, 14:00

      The Wang et al. mass estimates suggest that these planets are all volatile-rich (i.e. non-rocky) even at the present day, with the possible exception of TRAPPIST-1c. According to this study taking into account the formation and migration of the system, even b and c are far more water-rich than the terrestrial planets in our solar system, while the outer planets f and g contain around 50% water by mass.

      • Rob Flores August 17, 2017, 12:26

        If these are water rich, does that mean they will
        be analogues Jupiters moons. Most of those Icy moons
        are estimated to be 50% water/ice and 50% silicate and other
        type solid materials.

        Back of the envelope calculation needed here.
        How long would it take to boil off, calisto’s water/ice, if it were
        placed at an orbit between Earth. Without an atmosphere? and
        tide locked to the sun? maybe would it be a very cloudy world?
        My Swag, 5 billion years.

        • andy August 17, 2017, 17:56

          According to Lehmer, Catling & Zahnle (2017), the ice on Ganymede-sized moons can survive for a long time provided they remain outside the runaway greenhouse distance. Europa-sized moons are not so fortunate. They only consider pure water-vapour atmospheres, so the surface never gets above freezing point without also going into a runaway greenhouse state.

          As for TRAPPIST-1, the situation looks less promising. Quarles et al. (2017) (which used the earlier, higher estimates for the planetary masses) predict that the base of the water envelope on planet f is likely too hot for high pressure ices to form (rather like Uranus and Neptune), with the result that “it is no more likely to be habitable than any other gas or ice-giant with water clouds in its atmosphere”.

  • Alex Tolley August 15, 2017, 17:02

    I’m not sure time has a lot of value. The Earth will be inhospitable to surface multicellular life in 0.5 – 1 billion years as the sun continues to emit more energy.

    Confirming atmospheres would be good to eliminate the possibility that they have been stripped, but if protist-like life exists, it could survive in the lithosphere as long as there is some water in the rocks. That is assuming life ever got started or seeded.

    I expect in a few decades we might well have good statistics on exoplanet atmospheres and robust biosignatures for multi-cellular life. However, if that proves rare or even non-existant, then we are faced with the same problem we have with our neighbor, Mars. Did life emerge on Mars, and is there sub-surface life still there? Or is Mars sterile?

    In the end, we will need to send probes to find out. That will prove expensive and very time consuming, so if the galaxy is filled with microbial life, we may not know it without observation in in situ.

    A galaxy largely devoid of life might be scientifically less interesting, especially for [astro]biologists. OTOH, it gives human civilization a good opportunity to colonize those worlds, terraforming them to conditions approximately suitable for humans and our descendant species. It might take a million years to do that. But what a fantastic human legacy that would be.

  • andy August 15, 2017, 17:37

    It would be good to get a firmer estimate for the planets’ masses, are there any realistic prospects for RV detection? I see there are astrometric limits on the presence of outer gas giants, but I would guess that such a planet would be a disruptive influence on the tightly-packed inner system so it would be surprising to find one there.

  • Karen August 16, 2017, 11:42

    Forgive my rudimentary memory of basic science; am I correct in thinking that, if the estimates of the planets’ densities are correct, the planets must have an atmosphere / be gaseous? In other words, is there any way that a planet with a low density could still be “rocky”? I would imagine that there could be some compositions which would be less dense than the earth but still “rocky”, but do we know the density at which a planet can no longer reasonably still be lacking an atmosphere?

    • DJ Kaplan August 16, 2017, 14:45

      I think we might be limited in not knowing exactly what the radii of these bodies might be.

    • andy August 17, 2017, 17:46

      The masses are more of an issue than the radii here, there are various factors that can throw estimates from transit timing variations off.

      As for whether a low-density planet could be rocky, less massive planets would have less compression in their interiors so would generally be less dense even if they have the same composition. Terrestrial planets with a lower proportion of iron in their composition would also be less dense, but the TRAPPIST-1 planets (with the possible exception of TRAPPIST-1c) have mass estimates so low that even a completely iron-free composition would be too dense.

      Otherwise, it might be possible for a planet that formed inside the snowline to end up being a mainly rocky planet with a thick hydrogen atmosphere and low ice content, something like a version of Uranus or Neptune with a rocky rather than an ice/rock core. It is possible that some extrasolar “hot Neptunes” may be this type of planet thanks to the fact that you can match the same values of mass and radius with different possible compositions: adding more ice to the mix means you need less hydrogen to bulk up the radius, and vice-versa. Nevertheless even if such “rocky Neptunes” exist (I suppose you might call these “rock giants” to mirror the term “ice giants” as used for Uranus and Neptune, not particularly keen on the term though), they aren’t what we typically mean by “rocky” or “terrestrial” planets.

  • DJ Kaplan August 16, 2017, 14:43

    Is it thought that, due to its low apparent rotational velocity, we are viewing TC from one of its poles?

    • Ivan Vuletich August 16, 2017, 22:04

      If we were viewing TC from near its poles, wouldn’t that imply that the inclination of the planets orbits are close to 90°? Which is way higher than our own solar system and except for Uranus, higher than any of the moon systems.

      What effect would orbiting over the poles of a star have on habitability?

    • andy August 17, 2017, 18:15

      I don’t think so. The projected rotation velocity v*sin(i) is given as 6 km/s, which appears to ultimately come from Reiners & Basri (2010), where it is given as 6.0±2.0 km/s. The rotation period and radius are given as 3.295±0.003 days and either 0.117±0.004 or 0.121±0.003 solar radii, which gives an equatorial rotation velocity of about 1.8 km/s. This is lower than the measured value, which suggests some systematic error has crept in somewhere (sin(i) cannot be greater than 1). Certainly there doesn’t seem to be any evidence for low values of the inclination which would indicate a pole-on viewpoint.

  • Brett Davidson August 20, 2017, 0:12

    Is it possible that tidal interactions between such close planets could generate magnetic fields?