Saturday, March 18, 2017

TRAPPIST-1 and the Seven Dwarfs

Figure 1. The seven planets of TRAPPIST-1 are shown at their relative sizes, with colors corresponding to the densities estimated by Gillon et al. 2017. The area shaded in yellow represents the system’s habitable zone. Planet f is the most similar to Earth in radius, but not in mass or density (see Table 1). The density of planet h has not been estimated yet.

The TRAPPIST-1 system has been collecting superlatives since it was first announced just 10 months ago. With a mass only 8% of our Sun, the system’s ultra-cool M dwarf host is the smallest and least massive star ever detected at the head of a family of bona fide planets. Michaël Gillon & colleagues initially reported three Earth-sized objects (b, c, d) transiting this rosy little orb in May of 2016. All have periods shorter than 10 days. With their discovery, TRAPPIST-1 became the nearest star known to host multiple transiting planets (hereafter tranets), as the system is located only 39 light years (12 parsecs) away in the Sun’s back yard. In addition, these objects constitute the first multi-tranet system ever detected by a ground-based telescope – the TRAPPIST instrument in La Silla, Chile. All the other multi-tranet systems known at the time of the discovery were originally detected by the Kepler Telescope, a space-based observatory on an independent orbit around our Sun.

Gillon & colleagues determined that these three tranets did not have hydrogen atmospheres, given their small radii. On theoretical grounds, the astronomers also argued that they were more likely to include a significant icy component than to be purely rocky objects, given our current understanding of the protoplanetary disks surrounding very low-mass stars. The discovery team concluded by expressing their hopes for more precise characterization of the system parameters with more powerful telescopes.

Those hopes have now been partially fulfilled, and the view is astounding. Three weeks ago, Gillon’s team made global headlines that momentarily eclipsed the nonstop buffoonery of the American president. In brief: four additional tranets have been confirmed around TRAPPIST-1, bringing the total to seven. The new discoveries were made on the basis of an extensive observational campaign using the Spitzer Space Telescope (Gillon et al. 2017). Then, just this week, members of the same scientific team reported follow-up data obtained by the K2 program, which uses the Kepler Telescope (Luger et al. 2017). The new findings further constrain the system parameters.

Analyses of the data provided by all telescopes established that at least six of the seven planets experience transit timing variations. These data enabled an estimate of the masses and densities of all but planet h, the outermost. Not only do the six inner tranets resemble Earth and Venus in size and mass – three of them (e, f, g) occupy the system habitable zone (see Figure 1).

As a colorful header in The Sun proclaimed, “Nasa says TRAPPIST-1 solar system could be teeming with ‘exotic’ alien life forms!”

Figure 2. Comparing TRAPPIST-1 to our Solar System
Image Credit: European Southern Observatory

three terrestrial planets in the liquid water zone

If TRAPPIST-1 were a G-type star like our Sun, or a K-type star like Epsilon Eridani or HD 219134, this news would rank among the greatest discoveries in the history of science. As it is, however, the primary of this fascinating system is barely a star at all. It’s a red dwarf of spectral type M8, even smaller and dimmer than Proxima Centauri. It’s so lightweight that it barely crosses the threshold between stars, which are the sites of nuclear fusion, from brown dwarfs, which are not.

But the problem isn’t the diminutive size or puny luminosity of TRAPPIST-1 (see Figure 2 for perspective). It’s the evolutionary history of M dwarfs, as detailed in an earlier blog post. During the first billion years of their development, these stars are hotter and more prone to destructive outbursts, and they emit a much higher flux of X-rays, than during the far longer period of their maturity. Any planet that currently orbits in the habitable zone of an M dwarf – including Proxima Centauri b and TRAPPIST-1e, 1f, and 1g – was subject to high temperatures, high levels of extreme ultraviolet radiation, and frequent stellar flares for hundreds of millions of years, potentially stripping away its reservoir of water and much of its atmosphere.

Even the writers for The Sun know a thing or two about M dwarfs and the problem of extrasolar habitability. The article I cited above also engaged in a little analysis:

Now [Nasa’s] top scientists are trying to work out whether these worlds are abundant with extraterrestrial beings – or as dead as a terrestrial doorknob. Nasa boffins suggested the planets “could harbour exotic lifeforms, thriving under skies of ruddy twilight.”

However, they could also be barren wastelands because their parent star is a red dwarf, a relatively cool type of sun that could wipe out early lifeforms before they have a chance to evolve into sentient beings.

“A bumper crop of Earth-size planets huddled around an ultra-cool, red dwarf star could be little more than chunks of rock blasted by radiation, or cloud-covered worlds as broiling hot as Venus,” Nasa warned. (Hamill 2017)


Table 1. TRAPPIST-1 system parameters

Tags: Period = orbital period in days; a = semimajor axis in astronomical units (AU); Mass = mass in Earth units; Radius = radius in Earth units; Density = density in Earth units; Teq = equilibrium temperature in Kelvin. All data derive from Gillon et al. (2017) and Luger et al. (2017).

system architecture

Now let’s get behind the headlines and put this discovery in context (see Table 1). A key detail that was omitted by the mainstream media is the system’s unique orbital architecture. Among the 3,593 exoplanetary systems in the current census, just three are reported to host as many as seven planets. Those are Kepler-90, HD 10180, and now TRAPPIST-1, which immediately becomes the odd man out. Unlike TRAPPIST-1, the first two systems center on Sun-like stars: Kepler-90 has a mass of 1.13 Solar masses (Msol), HD 10180 of 1.06 Msol. These two stars also host a greater diversity of planets than TRAPPIST-1. Each supports at least one gas giant and six smaller planets with likely masses in the range of Super Earths, gas dwarfs, and tweens. By contrast, all seven planets orbiting TRAPPIST-1 have estimated masses within 50% of Earth and Venus, suggesting that all are probably rocky, with at most a small contribution from volatile constituents. This is the most uniform family of planets among known systems with at least six of them.

TRAPPIST-1 is also the most tightly packed collection of planets discovered to date. The outermost planet of HD 10180 has a cool semimajor axis of 3.4 AU, while the outermost planet of Kepler-90 orbits at about 1 AU (the same as the distance of Earth from the Sun). Among the known six-planet systems, the outermost planets of Kepler-11 and Kepler-20 have semimajor axes of 0.46 AU and 0.35 AU, respectively. By contrast, the semimajor axis of the seventh planet (h) of TRAPPIST-1 is only 0.06 AU, similar to the orbit of a Hot Jupiter. In this regard, the system’s nearest rivals are Kepler-444, with five subterrestrial planets orbiting within 0.08 AU, and Kepler-80, with five terrestrial and gas dwarf planets within a similar radius. Both of the latter systems center on K-type stars, demonstrating that ultra-compact architectures are not unique to ultra-cool M dwarfs.

Because we can observe all seven planets in transit, we know that the TRAPPIST-1 system is completely “flat” or co-planar: all seven orbit in the same spatial plane. All also appear to be engaged in orbital resonances, since the periods of each pair of adjacent planets can be expressed as the ratio of two small integers. From planet b to planet h, the ratios are 8:5 (b/c), 5:3 (c/d), 3:2 (d/e), 3:2 (e/f), 4:3 (f/g), and 3:2 (g/h) (Gillon et al. 2017, Luger et al. 2017). These commensurabilities translate into relationships extending to sequences of three and four TRAPPIST-1 planets, both adjacent and non-adjacent.

The best-known example of an orbital architecture with at least three objects in a resonant chain appears in Jupiter’s satellite system, where the periods of Io, Europa, and Ganymede conform to a ratio of 4:2:1. This relationship is known as a Laplace resonance, in honor of the first astronomer to describe it (Pierre-Simon de Laplace) This and similar instances of interlocking resonances are likely to enhance the long-term stability of any system architecture (Gillon et al. 2017).

Nonetheless, the unparalleled intricacy of the architecture of TRAPPIST-1 takes this configuration to a whole new level (Figure 3), with several critical implications for our understanding of the system. First, in situ formation could not have produced such a clockwork mechanism. These seven planets must have formed at a greater distance from the star – probably outside the current system ice line, which lies at a radial distance of about 0.06 AU (equivalent to the present orbit of planet h). Subsequent interactions with the primordial nebula caused them to migrate inward to their present locations very early in system history.

Figure 3. A triple transit of TRAPPIST-1
This figure illustrates the light curve of TRAPPIST-1 on a day when three of its planets – c, e, and f – transited almost simultaneously. Given the system’s multi-resonant orbital architecture, simultaneous transits must be frequent. Image credit: Gillon et al. 2017.

Second, these tightly packed orbits give rise to powerful gravitational interactions that cause transit timing variations (TTVs), as in the classic case of Kepler-11. The TTVs observed in TRAPPIST-1 enable estimates of the masses of the six inner planets, and thus of their physical compositions. Outside the Solar System, comparable data are in extremely short supply.

Third, the dynamical history of TRAPPIST-1 must have been extremely peaceful, with none of the perturbations or scattering events that are theorized for the Solar System and many extrasolar systems. Such instances of interplanetary violence would have disrupted the interlocking orbital structure we observe today.


Our Sun is known to host 29 naturally spheroidal objects, comprising 8 planets, 2 dwarf planets (Ceres and Pluto), and 19 regular moons. Only 6 of these 29 spheres consist primarily of rock, with a minimum of volatile content: Earth’s only Moon, Jupiter’s moon Io, and the 4 inner planets from Mercury to Mars. Earth is the densest object among these six (1.0 in the scale used by Gillon & colleagues, or about 5.5 g/cm-3), while our Moon is the most rarefied (just 61% of Earth’s density).

As shown in Table 1 and Figure 2, the six well-characterized planets of TRAPPIST-1 are remarkably similar in size and mass to Earth and Venus, and remarkably similar in density to the six rocky spheres of our Solar System. Both TRAPPIST-1c and -1g are more massive than Earth by a factor of about one-third; the others are intermediate in mass between Venus and Mars. TRAPPIST-1c is also a bit denser than Earth (117%), possibly because of enrichment in iron. The densities of the other five planets range from TRAPPIST-1g at 94% of Earth (virtually identical to Venus) to TRAPPIST-1f at 61% of Earth (virtually identical to the Moon). Therefore, it is quite possible that all six inner planets are rocky.

On theoretical grounds, however, Gillon & colleagues prefer a mixed composition of rock and ice for planets b through g. They argue that these objects most likely formed in the outer system, where ices would be more abundant than refractory elements. We can get a helpful perspective on their potential compositions by examining the spherical icy moons in our own Solar System (three around Jupiter, seven around Saturn, five around Uranus, and one each around Neptune and Pluto). Their interior structure includes rock and ice in varying proportions. The three largest – Ganymede, Titan, and Callisto – are each about one-third ice and two-thirds rock (Hussmann et al. 2015). With the notable exception of Europa, which is only about 7% ice, the mid-size moons in this sample tend to be at least 40% ice (Hussmann et al. 2015). Thus, we might expect planets as large as the TRAPPIST seven to accrete about 30%-50% of their original mass in ice. Later, after migrating into their present orbits, they would have endured intense irradiation over geological timescales, reducing their overall volatile content.

More massive planets on cooler orbits would be expected to retain more volatiles than warmer, less massive objects. Nonetheless, even though we see some association between mass and density in the TRAPPIST family – for example, the two most massive planets (c and g) are also the densest – we also see anomalies. Planet b is hotter and less massive than planet c, but it is also much less dense. One possible explanation is that it is a purely rocky planet with enrichment in silicates relative to iron, like Io and our Moon.

Luger & colleagues remind us that tidal heating is yet another consequence of the interlocking resonances that structure the TRAPPIST-1 system. It is also likely to be a factor in the thermal evolution of these seven planets. Luger’s group argues that planet b could have a tidal flux similar to that of Io (which also participates in a multi-resonant chain), potentially generating intense volcanism. Thus, it might be no coincidence that the density of planet b (66% Earth) is also very similar to the density of Io (64% Earth). Planets c through e experience lesser tidal effects, but still much stronger than the heat flux caused on Earth by our homeworld’s radioactive core. These planets might also experience volcanism comparable to Io’s. The more distant planets f through h appear to have avoided such a sizzling history.

All seven planets might still support atmospheres, even if their original envelopes were stripped by stellar irradiation at primordial times. Various sources of outgassing, including volcanism, would readily regenerate a lost atmosphere around a rocky planet (see discussion in Barnes et al. 2016).

exotic alien life-forms?

So here we have a system with not one, not two, but three Earth-size, Earth-mass planets whose blackbody equilibrium temperatures are consistent with surface bodies of water. Unfortunately, we have no information on the actual surface temperatures of these worlds, nor do we know whether water in any form is present on any of them. Our knowledge of the evolutionary history of M dwarfs predisposes us in the direction of pessimism, as illustrated by the excerpt from The Sun quoted above. Nevertheless, members of the original discovery team for TRAPPIST-1 conducted an analysis of the thermal evolution of this system, returning a more optimistic conclusion: “Depending on their initial water contents, [the planets of TRAPPIST-1] could have enough water to remain habitable” (Bolmont et al. 2017).

However, their conclusion did not take into account the tidal heating that is likely generated by the interlocking resonances in which the system’s seven planets participate. This omission is perfectly understandable, since the analysis offered by Bolmont & colleagues was based on the three-planet model presented in the original discovery paper (Gillon et al. 2016). That paper simply noted, “In some cases tidal heating could trigger a runaway greenhouse state.” The more recent analysis by Luger & colleagues (2017) presents exactly the case in which tidal heating would produce this unhappy outcome.

As a result, I remain agnostic about the possibility of life-bearing environments in the TRAPPIST-1 system. I don’t reject it out of hand, especially given the prospect of three terrestrial planets in the system habitable zone (not to mention the potential for various factors to lift the temperatures of planets g and h into habitable levels). But so many other hostile factors are in play that it seems inappropriate to sing a cheery song about the inevitability and ubiquity of life.

Dead or alive, these planets remain fascinating objects of study. Because they transit so frequently and are situated so close to home, we can look forward to more and more conclusive data on their surface conditions over the next few years.

celebrity septets

Given the remarkable nature of this discovery, especially its potential to create buzz in various desirable audiences, NASA invited the general public (or at least people with Twitter accounts) to suggest names for the seven planets of TRAPPIST-1. This was an unusual move, since the standard practice is to use robotic catalog numbers with alphabetic suffixes, such as Kepler-62f and HD 95872 b. In this case, the fact that seven names are involved should inspire creativity, since so many cultural phenomena involve groups of seven: the seven seas, the seven deadly sins, the seven classical planets, the seven days of the week, the seven hills of Rome, the seven swans a-swimming, the seven stars of the Pleiades, the Seven Samurai, the Seven Against Thebes, the Seven Sisters of Academia, the seven wonders of the world, and of course the very first septet that popped into my head:

So how about a system with planets named after Tydeus, Hippomedon, Amphiaraos, Polyneikes, and the other ill-fated warriors in Aeschylus’ Theban septet? Alas, those choices are probably too recherché and hard to pronounce for typical consumers of modern infotainment. Sloth, Greed, and Lust would present no such problems, but might they be too racy? Could Aventine, Viminal, Caelian, Capitoline, and so forth strike an appropriately classical note without seeming too recondite?

In light of the predictable objections, I’m favoring homier alternatives – maybe Happy, Dopey, Sleepy, and the rest of the dwarf mining crew, or maybe an even simpler schema based on the days of the week – Monday for tranet b through Sunday for tranet h, with Thursday, Friday, and Saturday assigned to the lucky trio in the habitable zone. (As in, “We’re spending a month on Saturday to ski the tropical glaciers, then rocketing off to Thursday for windsailing in the twilight zone.”)

With so many rich possibilities, the response to NASA’s invitation has been even more childish and absurd than I anticipated. I clearly overestimated the intelligence of the Twitterverse. Respondents tweeted names like Planet McPlanetface and Moony McMoonface, plus a set of seven taken from the principal characters in an American sitcom, Friends (which I confess I’ve never seen), plus another septet based on the titles of the first seven movies in the Fast & Furious franchise (ditto).

Alongside many whimsical suggestions to use the names of the Seven Dwarfs, which I heartily endorse, a few serious options emerged. One involves the names of seven of the eleven official Trappist breweries, from Achel to Westvleteren. That suggestion makes sense because TRAPPIST (Transiting Planets and Planetesimals Small Telescope), the acronym that names the telescope as well as the planetary system, was intentionally crafted to reflect the research group’s fondness for Trappist beer. Another suggestion was to name the planets after the seven astronauts who died in the crash of the Challenger. You might still be able to check out the March 2 issue of The Telegraph and the March 3 issue of The Daily Mail to see a selection of these candidates.

As it turns out, however, NASA was just kidding. It looks like we’ll have to use those old catalog numbers after all. Far better that outcome than Planet McPlanetface or Furious 7!

Barnes R, Deitrick R, Luger R, Driscoll PE, Quinn TR, Fleming DP, Guyer B, McDonald DV, Meadows VS, Arney G, Crisp D, Domagal-Goldman SD, Lincowski A, Lustig-Yaeger J, Schwieterman E. (2016) The habitability of Proxima Centauri b I: Evolutionary scenarios. In press. Abstract: 2016arXiv160806919B
Bolmont E, Selsis F, Owen JE, Ribas I, Raymond SN, Leconte J, Gillon M. (2017) Water loss from Earth-sized planets in the habitable zones of ultracool dwarfs: Implications for the planets of TRAPPIST-1. Monthly Notices of the Royal Astronomical Society 464, 3728-3741. Abstract: 2017MNRAS.464.3728B
Gillon M, Jehin E, Lederer SM, Delrez L, de Wit J, Burdanov A, Van Grootel V, Burgasser A, Triaud A, Opitom C, Demory B-O, Sahu DK, Bardalez-Gagliuffi D, Magain P, Queloz D. (2016) Temperate Earth-sized planets transiting a nearby ultracool dwarf star. Nature 533, 221-224. Abstract: 2016Natur.533..221G
Gillon M, Triaud A, Demory B-O, Jehin E, Agol E, Deck KM, Lederer SM, de Wit J, Burdanov A, Ingalls JG, Bolmont E, Leconte J, Raymond SN, Selsis F, Turbet M, Barkaoui K, Burgasser A, Burleigh MR, Carey SJ, Chaushev A, Copperwheat CM, Delrez L, Fernandes CS, Holdsworth DL, Kotze EJ, Van Grootel V, Almleaky Y, Benkhaldoun Z, Magain P, Queloz D. (2017) Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542, 456-460. Abstract: 2017Natur.542..456G
Hamill J. “Life, But Not As We Know It.” The Sun, 24 February 2017.
Hussmann H, Sotin C, Lunine J. (2015) Interiors and evolution of icy satellites. In Treatise on Geophysics, Volume 10: Physics of Terrestrial Planets and Moons, ed. G. Schubert. Elsevier B.V.
Luger R, Sestovic M, Kruse E, Grimm SL, Demory B-O, Agol E, Bolmont E, Fabrycky D, Fernandes CS, Van Grootel V, Burgasser A, Gillon M, et al. (2017) A terrestrial-sized exoplanet at the snow line of TRAPPIST-1. In press.