Monday, April 21, 2014

In the Realm of Crimson Twilight

Figure 1. Red twilight here on Earth
Kepler did it again! Exactly one year after the announcement of the first truly Earth-like extrasolar planet, our beloved space robot has provided good evidence for a second one. In scientific terms, this means a rocky object about the same size as Earth, orbiting at the right distance from its host star to support surface bodies of liquid water, given the appropriate atmospheric pressure and composition. In layman’s terms, it means a planet that might have life.

This time the lucky star is Kepler-186, a ruddy M dwarf located about 493 light years (151 parsecs) away. The lucky planet is Kepler-186f, the outermost of five companions traveling in an orbital space smaller than the semimajor axis of Mercury. Our new sibling’s radius is between 1.11 and 1.17 times that of Earth (1.11-1.17 Rea) – too small to maintain a hydrogen envelope, according to recent studies (Lopez & Fortney 2013, Weiss & Marcy 2014). If planet f has a composition similar to Venus or Earth, as seems likely, its mass will fall in the range of 1.4 to 1.6 times Earth (1.4-1.6 Mea). Astrobiologically speaking, these are extremely sexy stats.

Figure 2. Artist’s view of Kepler-186f, with its inner companions visible near the host star (one in transit). This version of the planet finds a happy medium between the “blue versus gold” approaches of many other images in circulation – see Figures 4 through 7, below. Credit: NASA
The architecture of the Kepler-186 system is surprisingly similar to that of Kepler-62, home of the Earth analog announced last year. Both Kepler-186 and Kepler-62 are known to support five low-mass planets on co-planar orbits, such that all five can be observed in transit across the central star. All planets in both systems are smaller than 2 Rea, and each system supports a tightly packed configuration of three or four planets inside 0.15 astronomical units (AU), along with one or two planets on wider orbits in the system habitable zone. None of the planets in either system experience transit timing variations or produce a measurable change in the host star’s radial velocity. Therefore, planet radii provide the only clue to their masses. In both systems, the most Earth-like planet orbits near the outer edge of the habitable zone, receiving about as much energy as Mars in the Solar System. Notably, both host stars are much cooler and less massive than our Sun, so that Mars-like temperatures are available much closer to the star than in our system.

Figure 3. Orbital architecture of the Kepler-186 system. Planets are shown at their relative sizes.
As an M dwarf, Kepler-186 is very cool indeed – and its coolness gives rise to the principal difference between the two systems. With a spectral type of M1, Kepler-186 is half as massive as our Sun – probably somewhere between 0.48 and 0.54 Solar masses (0.48-0.54 Msol). Its effective temperature is about 3750 K, and its luminosity is less than 5% Solar. By contrast, Kepler-62 is a K2 star with a mass of 0.69 Msol, an effective temperature of 4925 K, and a luminosity 21% Solar. Kepler-62f receives about as much heat as Mars in an orbital period only 39% as long (267/687), whereas Kepler-186f has an even shorter period – just 130 days, or 19% of a Martian year – and it receives even less heat and light than Kepler-62f.

The planet’s proximity to its cool host means that its rotation is likely to be synchronous or “tidally locked” (Selsis et al. 2007). In other words, one hemisphere of Kepler-186 always faces the star, while the other always faces away. Under this regime, day and night are geographic features rather than measures of time. On the day side, the planet’s red sun would appear motionless in the sky, neither rising nor setting, and shadows would never change angle or length. On the night side, darkness would be eternal, while the stars would float forever in a slow-motion parade across the sky, with a distinctly different set of “summer” and “winter” constellations, just as on Earth. Tidally locked rotation can pose serious challenges to the emergence and survival of life, as we will see below.

Table 1. Planet parameters for the Kepler-186 system
Source: Kepler Mission Site. Column 1 shows the planet name. Column 2 shows the planet orbital period in days. Column 3 shows the planet radius in Earth units (Rea). Column 4 shows the planet semimajor axis in astronomical units (AU). Column 5 shows the planet equilibrium temperature (Teq) in Kelvin (K), assuming an albedo of 0.3 and no greenhouse effect.
The detection of Kepler-186f was a major news story the day it broke (Thursday, April 17, 2014). Science published the discovery paper (Quintana et al. 2014), and the Internet was rapidly flooded with renditions of the planet’s possible appearance (check out some nifty videos: this one, this one, and this one on YouTube). Visions of the planet reveal a basic split in assumptions about its geography and climate: the blues versus the golds.

For the blues, Kepler-186f is an ocean planet (Figures 4 & 5), with vast stretches of open water and jagged island continents reminiscent of the Indonesian or Japanese archipelagoes. You can almost make out “argosies of magic sails” guided by “pilots of the purple twilight.”

Figure 4. Kepler-186f imagined as a warm ocean planet. The artist whimsically adds a moon.
Figure 5. Size comparison of Kepler-186f and Earth: two blue marbles. Credit: Planetary Habitability Laboratory, University of Puerto Rico, Arecibo.

These blue images encourage us to picture Kepler-186f as a slightly overgrown Earth twin. The gold party, however, present a much dryer world (Figures 6 & 7), with landlocked seas and vast plains that recall Canada, Siberia, or the Sahara. Such golden visions assume that the planet’s rotation is rapid, like Earth or Mars, rather than tidally locked. Only a planet that experiences a day/night cycle could sustain the extensive polar ice visible in these renditions.

Figure 6. Kepler-186f imagined as a less humid world suitable for “amber waves of grain,” like the plains of Asia, Africa, and North America. Credit: The Mars Underground
Figure 7. Size comparison of Earth and Kepler-186f: on this golden version of the planet, a large fraction of the water seems to be frozen. Credit: CNet

Now for a closer look at the science behind these beguiling extrasolar visions. Simultaneous with the official discovery paper by Elisa Quintana and colleagues, a more detailed analysis of the formation and physical characteristics of Kepler-186f appeared as a preprint (Bolmont et al. 2014). I highly recommend this careful, nuanced article.

Mindful that our understanding of exoplanets is strongly conditioned by the accuracy of our data on their host stars, Bolmont and colleagues considered a range of values for the mass and radius of the red parent star, Kepler-186. All stellar mass values fall between 0.48 and 0.54 Msol, but overall this group prefers numbers that yield slightly larger planets than the ones in Table 1 above. Nevertheless, their various models agree that Kepler-186f is about 50% more massive and 15% larger than Earth. This is an excellent size for a life-bearing planet, consistent with plate tectonics and long-term retention of an Earth-like atmosphere.

They also conducted numerical simulations to investigate the potential formation history of all five planets in the system. Results emphasized the oddness of the wide gap between the orbits of planets e and f (see Figure 3, above). The study team found that an additional planet would fit comfortably in this gap. As long as its orbit was slightly inclined to those of the other five planets, such an object would not be observed in transit – yet it would very likely share the habitable zone with planet f. Dynamical analyses suggest that this planet, if it existed, would be less massive than Earth, possibly a hotter sibling of Mars at ~0.1 Mea.

Bolmont and colleagues also examined the potential habitability of Kepler-186f. They note that the planet receives only about one-third the insolation of Earth, so its nominal equilibrium temperature is even lower than that of Mars. However, their analysis finds that a rocky planet of 1.5 Mea is almost certain to support a substantial atmosphere (plausibly composed of nitrogen and carbon dioxide). This combination would produce a greenhouse effect strong enough to raise the planet’s temperature above 273 K and thereby sustain liquid water – just as Mars would, if only it were massive enough. These results are consistent with both the blue and the gold visions of the planet.

Their models of tidal evolution assume that Kepler-186f originally had a fast rotation, like the outer planets of our Solar System, but stellar tides have steadily reduced its state of spin. As they conclude, “The age of Kepler-186 is thought to be higher than a few billion years, so assuming an Earth dissipation […] Kepler-186f should be in a pseudo-synchronous rotation” (Bolmont et al. 2014). In other words, if Kepler-186f is old enough for evolution to have produced complex organisms (i.e., about as old as the Earth), it is also old enough for its rotation to become tidally locked. Such a state would have profound effects on its environment.

As decades of research have shown, a tidally locked planet has two temperature poles corresponding to its two hemispheres: a hot pole near the substellar point on the bright side, where the parent star appears directly overhead, and a cold pole near the anti-stellar point on the dark side, where the sun never shines and temperatures vary little from the minimum. If a tidally locked planet has a thick enough atmosphere, especially with a significant amount of water vapor, heat will circulate between hemispheres and prevent a complete freeze-out of volatiles on the dark side. On relatively warm planets, a global ocean might persist. On colder planets, a regime known as “Eyeball Earth” might evolve (Angerhausen et al. 2013).

Figure 8. Schematic view of an Eyeball Earth. Credit: Beau.TheConsortium

In the classic model of an Eyeball Earth, the dark hemisphere is an icy, starlit wasteland, something like a mash-up of Antarctica and the Arctic Ocean in deep winter. The bright hemisphere is mostly dry and barren, perhaps like the Gobi Desert. Where the two hemispheres meet is a long, narrow country where the parent star is either just above or just below the horizon, a land forever on the brink of twilight. That’s where we can imagine seas, rivers, and vegetation.

Kepler-186f is an excellent candidate for an Eyeball Earth, so I’m surprised that none of the artist’s views in circulation have represented it as such. Whether its red sun would remain motionless in the sky as seen by an observer in the bright hemisphere would depend on the planet’s orbital eccentricity: the more eccentric the orbit, the greater the visible motion of the local sun. Since Bolmont’s group estimated that all five planets in the system must have eccentricities of 0.05 or less, we can expect relatively little movement above or below the horizon on a tidally locked version of Kepler-186f.

Still, the tidal locking scenario has a silver lining of sorts, since recent work shows that a terrestrial planet can maintain a magnetic field (provided it evolves one) even in a state of synchronous rotation. Thus Kepler-186f fulfills one widely cited criterion for habitability: the potential for a magnetosphere that can protect its atmosphere from erosion.

Alongside their calculations of the planet’s insolation, Bolmont and colleagues note that the wavelengths at which an M dwarf radiates are less suitable for photosynthesis than the light of a G-type star like our Sun. This drawback has been observed by previous investigators, who nevertheless conclude that life might still evolve and thrive under such conditions (Tarter et al. 2007). One workaround involves darker pigmentation for photosynthetic plants, as illustrated by a feature in Scientific American back in 2008.

Figure 9. Hypothetical vegetation on an M dwarf planet. Credit: Scientific American

Now that the sample of extrasolar Earth analogs has two members, it's almost not too soon to start generalizing. But here goes anyway. Both host stars, Kepler-62 and Kepler-186, are dimmer and less massive than our Sun, and they represent spectral types - K and M - that are much more numerous in the Milky Way than G-type stars. Both systems have very similar architectures, and their design is not unusual among the multi-planet systems revealed by radial velocity and transit searches. GJ 581, HD 40307, GJ 667C, Kepler-11, Kepler-20, Kepler-169, Kepler-215, and Kepler-351 all have a similar distribution of orbits. Even though most of these stars are hotter than Kepler-186, and none has an Earth-size planet in the right place, they demonstrate that well-packed systems of small planets are as common as crows in Seattle. Kepler data in particular have shown that planets between 1 and 2 Rea are the commonest of all. In brief: prospects for Earth-like planets across our Galaxy just dramatically improved. 


Angerhausen D, Sapers H, Citron R, Bergantini A, Lutz S, Queiroz LL, Alexandre M, Araujo AC. (2013) HABEBEE: Habitability of Eyeball-Exo-Earths. Astrobiology 13, 309-314.
Bolmont E, Raymond SN, von Paris P, Selsis F, Hersant F, Quintana EV, Barclay T. (2014) Formation, tidal evolution and habitability of the Kepler-186 system. In press; full text available at
Lopez ED & Fortney JJ. (2013) Understanding the mass-radius relation for sub-Neptunes: Radius as a proxy for composition. In press. Abstract:
Quintana EV, Barclay T, Raymond SN, Rowe JF, Bolmont E, Caldwell DA, Howell SB, Kane SR, Huber D, Crepp JR, Lissauer JJ, Ciardi DR, Coughlin JL, Everett ME, Henze CE, Horch E, Isaacson H, Ford EB, Adams FC, Still M, Hunter RC, Quarles B, Selsis F. (2014) An Earth-sized planet in the habitable zone of a cool star. Science 344, 277-280.
Selsis F, J. F.Kasting, B. Levrard, J. Paillet, I. Ribas, and X. Delfosse. (2007) Habitable planets around the star Gliese 581? Astronomy & Astrophysics 476, 1373-1387.
Tarter JC, Backus PR, Mancinelli RL, et al. (2007) A re-appraisal of the habitability of planets around M dwarf stars. Astrobiology 7, 30-65. Abstract:
Weiss LM & Marcy GW. (2014) The mass–radius relation for 65 exoplanets smaller than 4 Earth radii. Astrophysical Journal Letters 783, L6.