Habitability properties of circumbinary planets
Ivan I. Shevchenko
TL;DR
This paper argues that circumbinary planets naturally fulfill many habitability conditions, making them prime candidates for life, and suggests that Earth's life is an outlier in this context.
Contribution
It introduces the idea that circumbinary planets inherently meet multiple habitability criteria, providing a new perspective on potential locations for life beyond Earth.
Findings
Circumbinary planets automatically satisfy at least seven habitability conditions.
Most biopolymer production in the galaxy likely occurs on circumbinary planets.
Earth's habitability conditions are fulfilled only by chance, making it an outlier.
Abstract
It is shown that several habitability conditions (in fact, at least seven such conditions) appear to be fulfilled automatically by circumbinary planets of main-sequence stars (CBP-MS), whereas on Earth these conditions are fulfilled only by chance. Therefore, it looks natural that most of the production of replicating biopolymers in the Galaxy is concentrated on particular classes of CBP-MS, and life on Earth is an outlier, in this sense. In this scenario, Lathe's mechanism for the tidal "chain reaction" abiogenesis on Earth is favored as generic for CBP-MS, due to photo-tidal synchronization inherent to them. Problems with this scenario are discussed in detail.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Habitability properties of circumbinary planets
Ivan I. Shevchenko
Pulkovo Observatory of the Russian Academy of Sciences,
196140 Saint Petersburg, Russia E-mail: [email protected]
Abstract
It is shown that several habitability conditions (in fact, at least seven such conditions) appear to be fulfilled automatically by circumbinary planets of main-sequence stars (CBP-MS), whereas on Earth these conditions are fulfilled only by chance. Therefore, it looks natural that most of the production of replicating biopolymers in the Galaxy is concentrated on particular classes of CBP-MS, and life on Earth is an outlier, in this sense. In this scenario, Lathe’s mechanism for the tidal “chain reaction” abiogenesis on Earth is favored as generic for CBP-MS, due to photo-tidal synchronization inherent to them. Problems with this scenario are discussed in detail.
Key words: astrobiology – binaries: general – methods: analytical – planetary systems – planets and satellites: dynamical evolution and stability.
1 Introduction
Certain conditions should be satisfied for life, as present on the Earth, to emerge and be sustained on a planet; such conditions concern the insolation level (Huang, 1960), seasons (Olsen & Bohr, 2016), climate stability (Laskar, Joutel & Robutel, 1993), tidal phenomena (Lathe, 2004), protection from XUV radiation and stellar wind (Mason et al., 2013), active tectonics (Van Laerhoven, Barnes & Greenberg, 2014), and presence of water (Oró, Miller & Lazcano, 1990), among others. The importance of each of these factors can be quantified only approximately, but it is at least known that their intensity and variation as present on the Earth are suitable for life.
As discussed below, the apparent fulfillment of all the stated conditions on the Earth is due to an overlap of lucky chances. One may ask: are there planets that exist where these conditions are satisfied automatically (generically)? In a more technical formulation, are there generic biochemical reactors that exist to produce replicating biopolymers, as, e.g., analogously, stars can be considered as reactors for producing metals?
Recently, Mason et al. (2013, 2015a, 2015b) and Zuluaga, Mason & Cuartas-Restrepo (2016), analyzing habitability conditions in circumbinary planetary systems, came to the conclusion that planets orbiting stellar binaries with particular parameters might be exceptionally good, from an astrophysical viewpoint (concerning the protection of life from XUV radiation and stellar wind), as life habitats: mutual tides in the stellar main-sequence binaries with orbital periods greater than 10 d radically suppress the magnetic dynamo mechanism, thus reducing the chromospheric activity and the life-hostile extreme UV radiation and stellar wind. This suppression is favorable for life to emerge and survive on the planets orbiting such stars. Therefore, as follows from these astrophysical arguments, “life may even thrive on some circumbinary planets” (Mason et al., 2015a, p. 391).
Here we enforce this conclusion, arguing that there are, in fact, a number of major life-supporting conditions that arise naturally on many circumbinary planets of main-sequence stellar binaries (abbreviated “CBP-MS” in what follows).
On the other hand, we argue and emphasize that the fulfillment of these conditions on the Earth is purely accidental. Therefore, the CBP-MS are possibly the main “life-breeders” in the Galaxy, while life on Earth is an accidental exception, as the Earth accidentally mimics a typical CBP-MS in several life-favorable respects.
2 Insolation
The observational data on the recently discovered CBP-MS (Kepler-16b, 34b, 35b, and others) testify that most of them move in the orbits closely encircling central zones of orbital instability around their host binaries (Doyle et al., 2011; Welsh et al., 2012, 2014; Popova & Shevchenko, 2013). According to Popova & Shevchenko (2016a, Table 3), the observed planets concentrate in the resonance cells between orbital resonances 5/1 and 6/1 or 6/1 and 7/1 (designated as resonance cells 5/1–6/1 and 6/1–7/1 in what follows) with the binary. Their location is, therefore, quite predictable.
The presence of the central zone of orbital instability (chaotic zone) is automatic for any binary with a large enough mass ratio of companions: in fact, the gravitating binaries with components of comparable masses possess circumbinary zones of dynamical chaos, as massive numerical simulations (Dvorak, 1984, 1986; Rabl & Dvorak, 1988; Dvorak, Froeschlé & Froeschlé, 1989; Holman & Wiegert, 1999) and theoretical celestial mechanics (Shevchenko, 2015) in the framework of the restricted three-body problem tell us.
The extent of the chaotic zone around a system of two gravitationally bound bodies was estimated analytically in Shevchenko (2015), based on Chirikov’s resonance overlap criterion (Chirikov, 1979). The binary’s mass ratio, above which such a chaotic zone is universally present, was also estimated. These analytical results are in agreement with the modern data on the orbits of CBP-MS (Shevchenko, 2015; Popova & Shevchenko, 2016b).
Central cavities of analogous origin are observed in protoplanetary disks, which contain planetesimals, dust, and gas; the gas is present in the initial stages of the disk evolution. In contrast to disks of single stars, circumbinary disks have large central cavities. The existence and possible characteristics of such cavities in gaseous circumbinary disks were first considered analytically by Artymowicz & Lubow (1994, 1996a). The central cavity exists irrespective of the gas content in the disk.
All observed host stars of CBP-MS belong to a particular part of the overall period distribution of stellar binaries: their period range is 7–40 d (Martin, Mazeh & Fabrycky, 2015). The “dearth” of planets around the shorter-period main-sequence binaries is naturally explained as a consequence of the Lidov–Kozai effect (Martin, Mazeh & Fabrycky, 2015; Hamers, Perets & Portegies Zwart, 2016; Muñoz & Lai, 2015): if a third distant stellar companion is present, the Lidov–Kozai mechanism in concert with tidal friction shrinks the inner binary, and this shrinking perturbs the planets. They either escape or fall on the stars, or even their formation is prevented. In fact, the Lidov–Kozai mechanism in concert with tidal friction is considered to be responsible for the formation of most stellar binaries with periods below 7 d (Martin, Mazeh & Fabrycky, 2015; Winn & Fabrycky, 2015). Thus, the planet-hosting stellar binaries with periods less than 7 d are lacking, naturally due to the Lidov–Kozai mechanism in stellar triples.
On the other hand, when CBP-MS are formed, they migrate toward the central stellar binary and stall at the outer border of the chaotic zone around the binary because there is no more matter to cause the migration (Pierens & Nelson, 2007; Meschiari, 2012; Paardekooper et al., 2012). (Formation and migration issues are considered in more detail in Subsection 10.1.)
If the period of the central binary is 10 d, then the minimum period of the stalled planet is 50 d because the resonance cell 5/1–6/1 with the binary is typically the closest (to the binary) stable one: it is located at the border of the central chaotic region around the binary; see Shevchenko (2015). Therefore, for the Solar-like star binaries with periods 10–100 d the border-line (“leading”) CBP, or the planets trapped in outer low-order mean motion resonances with this leading CBP, are automatically placed close to the habitable zone (HZ) or inside it.
This is made as an order of magnitude estimate, using a rough analogy with the location of the Earth in the Solar system. Of course, habitable zone science, in its progress, now allows one to make much more rigorous estimates (Cuntz, 2014; Kane & Hinkel, 2013; Kopparapu et al., 2013; Müller & Haghighipour, 2014). In what follows, we use equations from Kane & Hinkel (2013), in particular.
Most of the observed CBP-MS are indeed close to the habitable zone (Welsh et al., 2014). As we have just seen, this is not just a mere coincidence, but an inevitable consequence of generic dynamical and physical effects. Concerning Earth, its life-favorable surface temperature is, in fact, fortuitous: in contrast to the circumbinary systems, there is no such a straightforward mechanism in the Solar system known to stall a planet at an appropriate orbital distance from the Sun; the presence of the Earth (and marginally Mars) in the Solar habitable zone is accidental, in this sense. However, one should admit that the absence of a known mechanism does not necessarily mean that it is actually absent. In numerical simulations, there do exist planetary formation scenarios where the overall architecture of the Solar system, with rocky planets residing in the habitable zone, is reproduced; see Brasser et al. (2016). We discuss formation issues further on in Subsection 10.1.
The main-sequence binary stars in the Galaxy have a rather wide distribution of periods (Duquennoy & Mayor, 1991), with a median value of 180 yr. Therefore, if one considers the overall population of the main-sequence binary stars, then the suitable lower bound of insolation is not provided automatically for the habitability of CBP-MS. However, there seems to exist a physically distinct stellar subpopulation that does provide such a bound. This is the population of so-called “twin binaries”: the near-equal mass binaries (namely the binaries with mass ratios from 0.8 to 1) forming a statistical excess at short orbital periods (Halbwachs et al., 2003; Lucy, 2006; Simon & Obbie, 2009). For the twins, the median period is 7 d, and the upper cutoff of the period distribution is at 43 d (Lucy, 2006; Simon & Obbie, 2009). Therefore, at the cutoff, the CBP orbiting at the border of the central chaotic zone around the binary (in resonance cell 5/1–6/1 or 6/1–7/1) would have the orbital periods 200–300 d, quite close to the inner border of the habitable zone of a double Solar-like star.
Twins seem to be physically distinct, in their formation process, from other binaries (Halbwachs et al., 2003): the binary components could form in situ (and the process could be followed by accretion from a common gaseous envelope, equalizing the masses of the components), whereas the binaries with smaller mass ratios may acquire shorter periods by migration in the circumbinary disk.
Note that, in twin population statistics, the mass ratio of stellar companions may have a range as wide as 0.8–1.0 (Halbwachs et al., 2003); therefore, the ratios of stellar radii can be as low as 0.8, close to the ratio allowing for the “optimal eclipse” (see Section 9).
Quantitatively, what is the percentage of binaries that may give birth to habitable CBP-MS? Nowadays, only very rough estimates can be made. According to the distribution of periods of nearby G-dwarf binaries given in Duquennoy & Mayor (1991, Figure 7), 13 of 82 binaries in the sample (16%) are short-period ( d). On the other hand, according to the same database, a much larger percentage is observed, e.g., in the Hyades field, where short-period binaries constitute 44% of the total sample. Thus, it might be no exaggeration to say, that, by the order of magnitude, 10% of all main-sequence binaries are suitable, at least among G-dwarfs.
Let us estimate the sizes of habitable zones for twin binaries. According to Underwood, Jones & Sleep (2003) and Kane & Hinkel (2013, Equations (2)–(5)), the stellar fluxes at the inner and outer borders of the habitable zone of a single star are given by
[TABLE]
[TABLE]
and the locations of the inner and outer borders of the habitable zone are given by
[TABLE]
where the radii and are in astronomical units, stellar luminosity in Solar units, and effective temperature in Kelvins.
For a twin binary consisting of companions with luminosity and effective temperature , the locations of the inner and outer borders of the circumbinary habitable zone can be roughly estimated using the hierarchical approximation, by setting and in the given equations. Using data from Kaltenegger & Traub (2009, Table 1), one finds that the circumbinary habitable zone overlaps completely with resonance cells 5/1–6/1 and 6/1–7/1 of a central binary with period 7 d (the median period for twins) if the twin is of M2V type (doubled luminosity , doubled mass in Solar units, K), or M3V type (, , K), or M4V type (, , K). For other spectral classes, there is not even any partial overlap.
For the twins with the cutoff period 40 d, the outer edge of the resonance cell 6/1–7/1 is close to the inner edge of the habitable zone even for Solar-like yellow dwarf twins (, , K).
Due to the general correlation of the planetary and host-star masses (Raymond, Scalo & Meadows, 2007; Raymond, 2008), CBP in M-dwarf systems should be generally much smaller than Neptune-like CBP in the Kepler systems; in fact, they may be Earth-like as, e.g., the Earth-like planets observed in the Trappist-1 single-star system (Gillon et al., 2017). The problem of planetary masses is discussed in more detail in Subsection 10.1.
The observed host stars of Kepler CBP are mostly Solar-like stars, not M-dwarfs. Therefore, let us consider the case of Solar-like star twin binaries in more detail. Today, we are aware of the existence of several Kepler systems that resemble such twins (Kepler-34, 1647), or have at least one component that is Solar-like (Kepler-38, 47, 453). All known CBP are Neptune-like or Jupiter-like giant planets, and most of them reside in orbits close to the central chaotic zone, as mentioned above. The fact that only giant CBP have been discovered up to now seems to be a natural observational bias, and the existence of less massive planets in orbits outside of the observed “leading” CBP is in no way precluded. (Inner planets are precluded as they would reside in the chaotic zone.) What is more, the existence of such objects looks plausible. Indeed, in the Solar system, the outer giant planet, Neptune, shepherds Pluto and plutinos in 3/2 orbital resonance with Neptune, and a number of other trans-Neptunian objects (TNO) in higher-order half-integer resonances. According to Gladman et al. (2012), the population of TNO in 5/2 resonance with Neptune is estimated to be as large as that in 3/2 resonance.
As for the outer (presumably terrestrial) CBP, they can be expected to reside in three-body resonances with the central binary and the leading giant planet, forming stable configurations, as discussed below. (On the three-body resonances in the Solar system, see Smirnov & Shevchenko 2013). Besides, a plausible orbital architecture of CBP can be expected to resemble that observed in the satellite system of the Pluto–Charon binary, where Styx, Nix, Kerberos, and Hydra all move in the orbits close to integer resonances with the Pluto–Charon binary (Buie et al., 2013; Brozović et al., 2015), and the system is dynamically closely-packed.
The inference that various CBP in the habitable zone outside the leading CBP may indeed actually exist follows from the example of the Kepler-47 system, where at least two CBP Kepler-47b and Kepler-47c are observed (Orosz et al., 2012a; Kostov et al., 2013), and there is enough dynamical space between their orbits to harbor more planets (Kratter & Shannon, 2014; Hinse et al., 2015). These latter planets should be much smaller than b and c (because they are not observed at present) and, thus, can be super-Earths or even Earth-like.
Among the binaries hosting Kepler CBP, according to Welsh et al. (2012, Table 1), there are two twin binaries, namely Kepler-34 and Kepler-35. Let us consider first Kepler-34 as a Solar-like star twin binary prototype. Using formulas (3) and observational orbital and physical data from Welsh et al. (2012, Table 1) and Haghighipour & Kaltenegger (2013), one can readily find that the inner and outer radii for this binary’s habitable zone are equal to 1.4 and 2.7 AU, corresponding to orbital periods of 410 and 1150 d. These values are respectively 1.4 and 4.0 times greater than the period of the planet Kepler-34b. Therefore, the outer orbital resonances 3/2, 2/1, 5/2, 3/1, and 7/2 with the observed giant planet are all within the HZ. These are the resonances where habitable rocky planets (Earth-like “plutinos,” “twotinos,” etc.), shepherded by the leading CBP Kepler-35b, may reside. An analogous calculation for Kepler-35 shows that its HZ ranges approximately from resonance 2/1 to resonance 6/1 with the leading planet Kepler-35b, i.e., the HZ starts with “twotinos.”
Among these outer resonances, 2/1 is of particular interest. Indeed, the observed Kepler CBP tend to concentrate in the centers of resonance cells bounded by integer resonances with the central binary (Shevchenko, 2015, figure 4). The centers of the cells correspond to the half-integer resonances with the central binary; therefore, a planet in 2/1 resonance with an observed leading CBP would tend to be additionally in the integer resonance with the central binary. Thus, the whole system would be close to a three-body resonance. This fact is interesting from the viewpoint of the long-term stability of such a configuration; the stability deserves a further study. (On the importance of three-body resonance in closely-packed planetary systems, see Quillen 2011.)
One should outline that it is not only in Kepler-34 and Kepler-35, but also in the overall population of Solar-like star twin binaries that the twotinos of the leading CBP (“twotino tatooines”) are more or less automatically placed in the insolation HZ. This is due to the fact that such stellar binaries have a suitable median orbital period, as discussed above.
An important observational circumstance is that among the seven first discovered Kepler CBP, the planets not belonging to the stellar HZ or its extended versions are all inside the inner edge of the HZ (Welsh et al., 2014, Table 1); therefore, all of these giant planets may shepherd smaller outer planets in the HZ.
In conclusion, at least two habitable niches, where the habitable level of insolation is provided more or less automatically, can exist: (1) the leading CBP of M-dwarf twin binaries and (2) the outer terrestrial bodies shepherded by the leading CBP of Solar-like star twin binaries.
At present, the prospects for an observational discovery of a terrestrial planet belonging to any of these two classes are rather vague because its transit signal is by two or three orders of magnitude less than that of the discovered leading CBP; besides, in the second class, its influence on the transit timing variations of the leading CBP is also small.
Relevant issues concerning the formation and migration of terrestrial planets of single and binary stars are considered further on in Subsection 10.1.
3 Seasonal variations
Olsen & Bohr (2016) showed that the inherent long-term instability of planetary climates can be quenched if seasonal variations are present. In other words, seasonal variations might be a necessary condition for a habitable climate to be maintained.
The seasonal analogy between the Earth and a typical CBP-MS first of all means that the amplitudes of surface temperature variations are similar. For the CBP-MS, the life-favorable amplitude (10%, as on the Earth) of the temperature variation arises generically, as the following simple calculation for a twin binary shows, even if the planet has zero tilt of the rotation axis.
The relative amplitudes of the surface temperature and stellar radiation flux , related by the equation
[TABLE]
are derived from the dependence of the flux on the phase angle , given by the formula
[TABLE]
where , , is the semimajor axis of the central twin, is the barycentric radius of the planetary orbit, and is the angle between the planet’s barycentric radius-vector and the “star 1 – star 2” direction. Of course, using the Stefan–Boltzmann law in Equation (4) is a kind of oversimplification, but here we make only rough estimates.
Let us see what is the relative amplitude of the surface temperature variations at the orbital radius corresponding to period ratio 6/1 (6/1 resonance between resonance cells 5/1–6/1 and 6/1–7/1). Setting , , from Equation (5) one has 1.50 for the ratio of the maximum and minimum fluxes, and 11% for the relative amplitude of the corresponding temperature variations. For period ratios 5/1 and 7/1, the temperature amplitude is the same, with accuracy better than 2%.
Concerning Earth, the favorable range of its surface temperature variations is conditioned by an accidental, quite high value of the obliquity (most probably caused by a giant impact; see Williams 1993) of Earth’s equator to the ecliptic plane. Note that the generic tilt of a planet in a relatively low orbit is equal to zero, as in the case of Mercury and Venus, because this is a natural outcome of the tidal spin–orbit evolution of orbiting bodies (see, e.g., Wisdom, Peale & Mignard 1984).
However, there exists a difference in the timescale: the seasonal variations on CBP-MS generically occur on a month scale (in contrast to the Earth); however, if a CBP-MS is tilted, the annual harmonic is also present.
Olsen & Bohr (2016) designate three solutions to the heat balance equation: the “snowball-planet point” (low effective temperature), the “habitable-planet point” (suitable effective temperature), and the “desert-planet point” (high effective temperature). The habitable point is inherently unstable. However, it can be stabilized by high-frequency periodic driving, provided (on the Earth) by seasonal variations. It is stabilized in the same way as the inverted rigid pendulum is stabilized by high-frequency periodic oscillations of the suspension point (Stephenson, 1908; Kapitsa, 1951, 1954). For the pendulum, the phenomenon was demonstrated in real physical experiments; see Kapitsa (1951, 1954).
The exact value of the perturbation frequency is unimportant, as long as the frequency is great with respect to the long-term climate oscillation (Ice Ages) frequency, and the perturbation amplitude remains the same by an order of magnitude. Therefore, whether the perturbation is “obliquity based” (with the period of several months) or “binary based” (with the period of several weeks), it has the same stabilizing effect, as long as the amplitudes are the same. However, if both “obliquity-based” and “binary-based” oscillations of comparable amplitudes are present, the combined effect is not obvious and should be explored in numerical experiments.
4 Climate stability
Laskar, Joutel & Robutel (1993) and Néron de Surgy & Laskar (1997) argued that the current obliquity of the Earth is secularly stable due to the presence of the Moon: the Moon-caused precession of Earth’s spin axis is rapid enough to prevent the chaotic diffusion between relevant spin–orbit resonances, as they are more widely separated in the phase space, due to this precession.
If the Moon were absent, the Earth would suffer large variations of its obliquity (between and ), entailing catastrophic variations of climate (Laskar, Joutel & Robutel, 1993; Néron de Surgy & Laskar, 1997). Conversely, the obliquity of the actually Moonless Mars varies in the range 0∘–60∘ (Laskar & Robutel, 1993), and this is at least one of the causes of its sterility, although it is marginally inside the Solar habitable zone. With the Moon, Earth’s obliquity stays within the range from to (Lissauer, Barnes & Chambers, 2012).
The issue, however, remains highly disputable. Lissauer, Barnes & Chambers (2012) and Li & Batygin (2014) argue on the basis of results of massive numerical experiments that the chaotic diffusion rate in the obliquity of a Moonless Earth is low enough for the development of life to be successful, and the long-term habitability is not precluded.
Nevertheless, one may speculate that the presence of a second star may be important as a stabilizing factor for the obliquity of the planetary rotation axis (vital for the climate stability), similar to the role played by the Moon. The binarity of a host star as the attitude stabilizer has not yet been explored. However, one may at least say that it does not induce the attitude instability. This can be argued as follows. The planar rotations-oscillations of an asymmetric satellite in a fixed elliptic orbit around a single primary body are described by the Beletsky equation (Beletsky, 1963). A generalized (circumbinary) version of the Beletsky equation, taking into account the torque exerted on a rotating body by an inner massive satellite of the primary body, has been derived by Correia et al. (2015), in connection with the problem of rotation of minor outer satellites in the Pluto–Charon system. Adapting the results of Correia et al. (2015), one sees that the major spin–orbit resonances overlap and the rotational chaos becomes global if
[TABLE]
where are the principal moments of inertia of the planet, and and are the mean motions (orbital frequencies) of the stellar binary and the planet, respectively; the orbits of the binary and the planet are assumed to be circular and the masses of the binary components comparable to each other. If condition (6) does not hold, the chaotic layers are thin and the motion is mostly (with respect to the initial conditions) regular.
CBP of twin binaries are orbitally stable if –5; see Holman & Wiegert (1999) and Shevchenko (2015). Therefore, according to criterion (6), for such objects, there is no macroscopic rotational chaos in the planar setting of the problem, even if is as great as 0.1. For rocky Earth-like planets, the quantity is expected to be not greater than \sim$$10^{-5}; see Correia et al. (2015) and references therein. Therefore, the circumbinary rocky planets are unlikely to rotate chaotically in this planar setting of the problem. The instability with respect to tilting the spin axis deserves a further study.
According to a recent analysis by Correia et al. (2016), when the precession rate of a CBP’s spin is much less than the precession rate of its orbit (the situation typical for CBP, as argued by Correia et al. 2016), any significant oscillations in its obliquity are not expected, at least in the framework of the full three-body problem.
Apart from the planetary obliquity variations, climates may suffer Milankovitch cycles. Frequent Ice Ages (on the timescale of order 1000 yr) were predicted by Forgan (2016) for the CBP with typical orbital parameters as observed for Kepler systems. The climates of such planets were shown to suffer short-timescale Milankovitch cycles, which might trigger Ice Ages. Frequent Ice Ages do not seem to be favorable for long-term habitability, but this issue deserves a further study. This is a complex problem, and Milankovitch cycles are not necessarily coupled to Ice Ages; see Ghil (1989), MacDonald (1990), and Shevchenko (2002).
5 Tides
The importance of tides for abiogenesis is at least twofold: first, they produce periodic wetting and drying of beaches, and this process is favorable for life in several respects (Lathe, 2004); second, they provoke plate tectonics, also favorable for life in several respects, in particular by allowing important chemicals produced deeply in planetary interiors to come to its surface (Van Laerhoven, Barnes & Greenberg, 2014). According to Lathe (2004), life on Earth, in fact, has a tidal origin.
As discussed in Section 2, most of the observed CBP-MS reside in resonance cells delineated by the chaotic bands corresponding to 5/1, 6/1, and 7/1 mean motion resonances with the central binary. Therefore, the typical ratio of the orbital periods of the planet and the binary is 6. Since the stellar binary components usually have comparable masses, the Fourier expansion of the time-varying gravitational potential acting on the planet possesses a dominating term with the doubled frequency. If the masses are equal, then the period of neap/spring tides on the planet effectively doubles, i.e., the ratio of the tide frequency to the planet orbital frequency is 12, analogous to what we have (at present) on the Earth subject to the Lunar tides.
The repeatedly dried and wetted tidal pools are thought to be a possible place of origin of self-replicating biopolymers, as such pools provide favorable conditions for concentrating organic molecules (see Lathe 2004 and references therein). Lathe (2004, 2005, 2006) proposed a theory of abiogenesis, based on the tidal “boosting” of the biomolecule production in near-shore lakes and ponds. The essence of the theory consists of the hypothesis that periodic concentration/dissociation of complex molecules leads to a “chain reaction” in the production of specific nucleic acids.
An open question with this theory is that it postulates frequent and high tides, not characteristic of the modern Earth and, presumably, for the early Earth as well (Lathe, 2006; Varga, Rybicki & Denis, 2006). Besides, the rates of concentration/dissociation are not determined quantitatively. The rates of evaporation in modern lakes are of the order 1 cm/day maximum (Jensen, 2010). They are controlled by the processes of Solar energy storage in the lakes’ water. Therefore, much more time than just several hours, as in Lathe’s original theory (Lathe, 2004, 2005, 2006) might be needed to concentrate the “bio-soup.”
A high Solar flux sufficient for the rapid-enough evaporation may extinct the emerging life altogether. Therefore, some kind of photo-tidal synchronization is needed. Such a synchronization is automatically maintained on CBP-MS, where “seasonal” variations of insolation are synchronized with neap/spring tides.
It is well known that the Solar and Lunar tides raised on the Earth have rather similar amplitudes. The ratio of the amplitudes of the Solar and Lunar tides is equal to
[TABLE]
as given by formula (4.17) in Murray & Dermott (1999).
It follows directly from Equation (7) that the relative amplitudes of the Solar and Lunar tides on the Earth are roughly the same as the relative amplitudes of the tides raised on a typical CBP-MS by its host stellar binary’s components. Indeed, in the latter case one has
[TABLE]
because the planet’s radial distances from the binary components , and the masses of the main-sequence components .
The absolute amplitudes of the Lunar and Solar tides are equal to m and m, respectively (Murray & Dermott, 1999). For a CBP in the habitable zone of a stellar binary whose most massive component is Solar-like, the absolute and relative amplitudes of the tides raised by the binary components are of the same order as on the Earth if the mass ratio of the companions is 2. Tantalizingly, the condition for “optimal eclipses” (see Section 9) also requires inequality of the stellar masses by a factor of 1.5–2.
In the polymerase chain reaction (PCR), periodic cycling between low and high temperatures (50–100∘C) drives the exponential amplification of the concentration of DNA molecules: at low temperature, association is promoted and complementary strands are synthesized (the molecule number doubles), whereas at high temperature, the duplex strands dissociate. When the temperature is low again, the molecule number again doubles, and so on (for details, see Lathe 2004, 2005). In Lathe’s tidal chain reaction (TCR), the water salinity, not temperature, is periodically driven, leading to the same exponential amplification of nucleic acids (Lathe, 2004, 2005). At low salinity, association is promoted and complementary strands are synthesized, and, at high salinity, the duplex strands dissociate.
In the case of the Earth–Moon system solely the salinity factor operates because one of the tide-raising bodies does not radiate practically. The case of CBP-MS is different: as a tide-raising body (any component of the stellar binary) approaches the planet, both the tide driven by this star and radiation flux from this star rise, and therefore, salinity decreases and temperature increases in concert, and vice versa. Thus, PCR and TCR conditions are simultaneously satisfied, and the whole process is expected to be more effective. However, a problem is that the reaction product is unstable and decays quite rapidly, as discussed in detail in Lathe (2005). Perhaps encapsulation in protocells increases survivability; see Subsection 10.7. Another problem is that the photo-tidal synchronization may result in the UV radiation sterilizing tidal pools during a drying phase. This is a complex problem, depending on the spectra of the host stars and the light-absorption properties of atmospheres. What is more, the periodically variable UV radiation may be, on the contrary, favorable for abiogenetic chemical reactions, as discussed in Subsection 10.7.
Both the Earth’s tilt and Moon’s presence are due to chance, as follows from modern cosmogonical theories (Lissauer, 1997; Lissauer, Barnes & Chambers, 2012). Conversely, on the CBP-MS the synchronism of insolation and tidal variations (the photo-tidal synchronism) with suitable periods is maintained automatically, as shown above, and, what is more, the amplitudes of these variations can be basically the same as on the Earth.
In conclusion, Lathe’s tidal mechanism (Lathe, 2004, 2005) of abiogenesis, if active in reality, seems to be generic for CBP-MS and solely accidental for Earth; then, life on Earth is favored by the complete (both in period and amplitude) similarity of the neap/spring tidal conditions on CBP-MS and the Earth.
6 Active tectonics
Active tectonics facilitate the habitability of Earth-like planets: “terrestrial life depends on heat-driven plate tectonics to maintain the carbon cycle and to moderate the greenhouse effect” (Van Laerhoven, Barnes & Greenberg, 2014, p. 1888). On the primitive Earth, underwater volcanic processes provided energy and material for the synthesis of organic compounds; due to large pressure and temperature gradients, underwater volcanoes or hydrothermal systems favor the survival, in their close vicinities, of the organic compounds formed (Mukhin, 1976; Koonin, 2012).
The compositional diversity of terrestrial exoplanets, according to numerical models, can be rather broad; see Carter-Bond et al. (2012) and references therein. Taking into account the effects of giant planet migration leads to the conclusion that planets with Earth-like composition can be ubiquitous and, what is more, the delivery of water to terrestrial planets can be favored and enhanced (Carter-Bond et al., 2012). Therefore, the tidal dissipation parameter and Love number for terrestrial exoplanets can be expected, in many cases, to be mostly the same as for the rocky planets in the Solar system.
The habitability may require a heat source driven by tidal friction: without such a mechanism, a rocky planet normally cools down on the timescale of 10 Gyr (Van Laerhoven, Barnes & Greenberg, 2014). The presence of a perturber, producing tides, may be required for the heating. For the tidal friction to be active, the perturber should maintain the planet’s orbital eccentricity. Tantalizingly, the orbital eccentricities of CBP are periodically forced on secular timescales, as shown by Moriwaki & Nakagawa (2004).
The central binary induces forced eccentricity in circumbinary orbits; the eccentricity is forced periodically on secular timescales (Moriwaki & Nakagawa, 2004). Thus, the orbits cannot sustain permanent circularity. In the case of zero eccentricity of the central binary, the forced eccentricity is given by the formula
[TABLE]
(Paardekooper et al., 2012), where are the masses of the binary components; are the semimajor axes of the binary and the particle, respectively; and and are their orbital periods. From formula (9), for a stellar twin binary () one has
[TABLE]
For a circumbinary object in the orbital range between resonances 5/1 and 7/1 with the central binary, this formula gives –, similar to the current eccentricity of Earth’s orbit (). Again, Earth’s value is accidental in the sense that it is not an outcome of any known automatic process, contrary to the CBP’s case. The eccentricity similarity means that the heating conditioned by the eccentricity is also similar in both cases, i.e., it is much less than the radiogenic heating on the Earth (where the current heat flow is 0.065 and 0.1 W m*-2* through the continents and the ocean crust, respectively; Zahnle et al. 2007).
The tidal heating flux through the planetary surface is estimated as
[TABLE]
(Jackson, Greenberg & Barnes, 2008; Barnes et al., 2009), where is the gravitational constant, is the stellar mass, and are the semimajor axis and eccentricity of the planetary orbit, is the planetary radius, and , where and are the planetary tidal dissipation parameter and Love number, respectively. In our hierarchical model, . Combining Equation (11) with Equation (10), one has
[TABLE]
where () is the ratio of the orbital periods of the planet and the binary. Note that, quite unexpectedly, the combining of Equation (11) with Equation (10) has cancelled the stellar mass; therefore, formula (12) is valid for any type of the central twin binary, be it composed, say, of M-dwarfs or Solar-like stars.
According to heuristic estimates in Barnes et al. (2009), the heating rates less than 0.04 W m*-2* and greater than 2 W m*-2* imply non-habitability. Taking d (the median period of stellar twin binaries; see Section 2), as for a super-Earth, and as justified in Barnes et al. (2009), one finds that the tidal heating for such a planet provides habitability at –, i.e., approximately in the range of resonances from to with the central binary. This is inside the chaotic zone around the stellar binary because the zone radius for a twin corresponds to (see Holman & Wiegert 1999 and Figure 2 in Shevchenko 2015). If one takes instead of 500, then the habitability range in would be 2.4–4.0; and if, in addition, one takes , then the range in would be 2.8–4.7. In both cases, the zone of suitable tidal heating overlaps with the zone of orbital stability. Taking into account the highly approximate character of the model, one may conclude that the forced eccentricity of CBP can indeed be an important factor in maintaining active tectonics on such planets, and there can even be no need for radiogenic heating to maintain them tectonically active, contrary to the case of Earth.
Finally, note that the equilibrium ratio of planetary rotation and orbital frequencies is equal to and in the constant phase-lag and constant time-lag tidal models, respectively; see, e.g., Barnes (2015). Therefore, the forced orbital eccentricity also solves the problem of life-unfriendly synchronous tidal locking. Such a locking implies that the central illumination source permanently faces one and the same side of a planet; planets in the HZ of single M-dwarfs are especially prone to this effect — see Shields, Ballard & Johnson (2017). Due to the forced eccentricity, for CBP this effect is impossible.
In conclusion, the tidal heating may provide a natural automatic constant internal heating for CBP-MS. However, its effectiveness deserves a further study. Conversely, the heating of the Earth by Lunar tides is accidental, not generic, having resulted from the Moon-forming impact.
7 Protection from stellar wind
Mason et al. (2013, 2015a, 2015b) came to the conclusion that CBP might be exceptionally good, in an important astrophysical aspect, for abiogenesis: in the main-sequence stellar binaries with orbital periods longer than 10 d and less than 50 d, the stellar rotation and binary revolution synchronize with each other, and the tides on the binaries radically suppress the magnetic dynamo mechanism in the host stars, thus reducing chromospheric activity and life-hostile extreme UV and stellar wind. This mechanism reduces “stellar aggression” and favors retaining moist atmospheres for the planets in habitable zones. The planetary atmospheres are protected from mass loss and water loss, and strong magnetospheres are not needed, contrary to the case of planets of single stars.
The given “Binary Habitability Mechanism” (BHM), as coined by Mason et al. (2015a), is effective for the stellar binaries with periods from 10 d up to 50 d (Mason et al., 2015a). It is easy to calculate that the upper bound corresponds to the orbital period 250–350 d for a planet in resonance cells 5/1–6/1 or 6/1–7/1, at the border of the chaotic zone around the central binary. Such a period is similar to Earth’s orbital period, and thus the planet turns out to be close to the habitable zone of Solar-like host stars. Conversely, taking the lower bound 10 d for the period of a BHM-effective stellar binary, one finds that it corresponds to the period 50–70 d for a planet at the chaos border. Therefore, the planet appears to be tentatively inside the habitable zone of M-class host stars, as, according to Quintana et al. (2014), Kepler-186f is.
While the protection of CBP-MS from stellar wind is provided, due to the BHM, almost automatically, Earth’s permanent magnetospheric protection might have been granted by an accidental event in its early history, namely, by the impact of a planetary embryo: this impact enriched the planet with iron, and a substantial metallic core was formed (see Taylor 1993 and references therein). On the other hand, the same impact is thought to have been responsible for the formation of the Moon.
However, note that some doubts exist on whether the relevant properties of binary stars are indeed favorable. In Johnstone et al. (2015), it is argued that the protective magnetospheres of CBP may decrease in size because the planets may suffer strong shock waves several times per orbital revolution.
8 Delivery of water
The problem of appearance of water on the Earth is most disputable, and, consequently, any inferences on this habitability condition can only be rather speculative. In one of the scenarios, water is thought to have been delivered to the Earth only after its formation, as the volatiles are lost during the formation (Morbidelli et al., 2000). According to the Nice model (Gomes et al., 2005), the delivery of water to the after-molten Earth was partially due to a fortuitous event: the passage of a migrating Jupiter and Saturn through the 2/1 mean motion resonance. This passage provoked disturbances in the planetesimal debris disks, and swarms of icy minor bodies entered the inner realms of the Solar system, bearing water to the already-cooled Earth.
For CBP-MS, an analogous process may emerge due to the passage of a migrating planet through the 6/1 or 7/1 or a higher-order mean motion resonance with the central binary, before the migration stalls the planet at an appropriate resonance cell on the border of the central chaotic zone. This passage provokes disturbances in the outer parts of debris disks, and icy bodies may enter the inner realms of the circumbinary system, bearing water to already-cold planets. The process is automatic, in contrast to that in the Solar system.
However, in the scenario believed at present to be the most plausible, most of the water is delivered to the Earth by a few planetary embryos from the outer asteroid belt at the final stage of Earth’s formation, and only a small fraction (less than 10%) is delivered by later bombardments (Morbidelli et al., 2000). This scenario is verified by the observed isotope ratios (the D/H ratio first of all) in Earth’s seawater and in meteoritic and cometary material. Data retrieved from studies of the Jupiter family comet 67P/Churyumov–Gerasimenko by the Rosetta spacecraft shows that its D/H ratio is three times the terrestrial value (Altwegg et al., 2015).
The percentage of water delivered to the Earth by later bombardments does not exceed 10% (Morbidelli et al., 2000). That is why the more or less automatic mechanism of water delivery from outer reservoirs (stellar Oort clouds) by the Galactic tide cannot alone provide the water amounts necessary for habitability. Therefore, mechanisms acting during planetary formation and/or migration in the system are needed. Clearly, statistical preferences for the delivery of water to CBP (in comparison to planets of single stars) require further study.
9 Ecliptic variations
Neap/spring tides arise due to the beating between the Solar and Lunar tides. This beating is substantial because the amplitudes of the Solar and Lunar tides are of the same order of magnitude. Thus, a new, longer timescale is introduced in the tidal variations. As outlined by Benn (2001), for the tidal beating to be present on a planet, the angular sizes of the perturbing bodies, as seen from a planet, should be roughly equal.
Consequently, the existence of the tidal beating on CBP-MS is generic, and, what is more, in coplanar systems it is accompanied by prominent stellar eclipses — periodic rapid changes of the stellar flux received by the planetary surface.
Therefore, there exists an ecliptic analogy between CBP-MS and the Earth: in both cases, the stellar eclipses are common. However, there exist two substantial differences. (1) In the former case, the ecliptic shadows are typically global: the planetary star-side surface can be shadowed by a stellar companion totally, whereas the Moon’s shadow traces the Solar side of the Earth only partially. This means that the ecliptic factor in abiogenesis, if any, may play a greater role in CBP-MS than in Earth. (2) On CBP-MS, the flux never reduces to zero, as the eclipsing body also radiates. The reduction magnitude depends on spectral classes of the binary components. The relative minimum is achieved when the eclipse is complete but the disk of the transiting star, as seen from the planet, is not much greater than the disk of the eclipsed star, so that the luminosity of the transiting star is minimal. For a planet in resonance cell 5/1–6/1, the eclipse of a G2V (“Sun”-like) component of the binary by its K5V counterpart (“61 Cyg A”-like; note that 61 Cyg A has radius 0.665, luminosity 0.153, and mass 0.67 of the Solar values, Kervella et al. 2008) would imply a fourfold reduction of the total flux at the planetary surface. Besides, the spectrum of the received radiation varies during the eclipse quite radically.
As the eclipses provide sharp and fast changes in the surface temperature, these frequent and periodic phenomena may modify the photo-tidal driving mechanism of PCR-type considered in Section 5.
Photoperiodism and photosynthesis conditions on CBP were numerically explored in Forgan et al. (2015); a rich variety of “forcing timescales for photosynthesis” was recovered, which is appealing for further studies of possible biospheres “rich in rhythms and cycles.” Frequent and periodic eclipses induce an even greater diversity. CBP-MS, whose orbits are coplanar with orbits of their parent stellar binaries, are more able to provide such a diversity in this respect than Earth because the eclipses are frequent and generic on them; in contrast, on the Earth they are rare and non-generic, being conditioned by a chance orientation and size of the Lunar orbit.
Quite a perfect alignment of the orbital planes of a CBP and its parent stellar binary is needed for this ecliptic mechanism to work. However, note that the alignment in the most Kepler circumbinary systems is indeed perfect: the mutual inclination of stellar and planetary orbits typically does not exceed 0.5∘ (Welsh et al., 2012; Orosz et al., 2012a, b).
10 Discussion
In this section, open and controversial issues in the proposed scenario are discussed.
10.1 Formation issues
The observed CBP-MS all have masses in the range from 0.1 to 0.5 Jovian masses (Doyle et al., 2011; Orosz et al., 2012a, b; Welsh et al., 2012). Therefore, they are Neptune-like or Jupiter-like. They cannot be Earth-like or cannot even belong to the class of super-Earths. However, it is clear that the observed large-mass preference does not refute, due to an obvious observational bias, the potential existence of smaller rocky planets in the same systems because the transit signal from a super-Earth or an Earth-like planet is 100–1000 times smaller than that from a Jupiter-like planet. As discussed in Section 2, the smaller planets can reside in the orbits outer to the observed “leading” CBP, but basically not inner to them because the inner orbits are mostly unstable.
Other possibilities include the following. (1) Large satellites of the observed CBP-MS may be habitable (Quarles et al., 2012). However, this opportunity requires a strong revision of the whole scenario, in application to satellites’ physics and dynamics. (2) The observed CBP-MS can be non-habitable, but the co-orbital material that they shepherd can be habitable. The co-orbital material may include any rocky objects of suitable masses; a particular class of such objects was considered as “Trojan exomoons” in Quarles et al. (2012). The dynamical stability of the material co-orbital with the planet sharply increases if the host star is binary (Demidova & Shevchenko, 2016).
Since super-Earth or Earth-like class CBP-MS have not yet been discovered, let us consider whether the modern theories of planet formation favor their actual existence.
Though a number of problems still remain enigmatic, the terrestrial planet formation in single-star systems is a profoundly explored subject (Raymond, 2006, 2008; Raymond, Quinn & Lunine, 2004; Raymond, Barnes & Kaib, 2006; Raymond, Quinn & Lunine, 2006; Raymond, Scalo & Meadows, 2007; Raymond et al., 2009; Davies et al., 2014). Conversely, the terrestrial planet formation in circumbinary systems is quite a novel subject, given that actual CBP were discovered only recently.
Terrestrial planets and cores of giant planets have a common origin: they form in protoplanetary disks containing dust and gas (Safronov, 1969; Lissauer, 1993). The formation has three distinct stages (Safronov, 1969; Lissauer, 1993; Raymond, 2008): (1) dust grains in the gas–dust protoplanetary disk coalesce into km-sized planetesimals; (2) the planetesimals form thousand-km planetary embryos via pair-wise accretion; and (3) the embryos form full-sized planets via merging impacts.
Stage (1) is a complex physico-chemical process, whose details are not all established yet, but there is no known reason to believe that it is more difficult in circumbinary disks than in disks of single stars. Stage (3) proceeds effectively in circumbinary disks, producing varieties of multiplanet orbital configurations (Quintana & Lissauer, 2006, 2010). Quintana & Lissauer (2006) performed numerical simulations of this stage for a circumbinary disk surrounding short-period binary stars; they found that building terrestrial planets, starting from Moon-sized planetesimals, in circumbinary environments is not much more difficult than in single-star systems, and a diversity of circumbinary terrestrial planets can be generated, given that large building blocks are present.
It is stage (2) that looks most controversial. Before this stage, km-sized planetesimals emerge, large enough for the gravity forces to become dominant in their dynamics in comparison with the aerodynamical drag in the gas disk. For CBP formation, the main problem is just with stage (2), in which the pair-wise accretion of the km-sized planetesimals takes place. Moriwaki & Nakagawa (2004) showed analytically that an extended (with respect to the central cavity) circumbinary zone is hostile for pair-wise planetesimal accretion. The barycentric radius of this unfriendly zone is typically by an order of magnitude greater than the radius of the central chaotic zone. The accretion is precluded because the stellar binary drives the eccentricity and differential (radius-dependent) precession of the planetesimal orbits; this implies destructive collisions of planetesimals. Only quite far from the barycenter, the collisional velocities become small and start to allow accretion. In a gas-containing disk, the eccentric orbits forced by the central binary would be aligned apsidally due to the gas drag, but the drag depends on planetesimal size, and this introduces differential orbital phasing, increasing the collisional destruction of bodies of different sizes (Thébault et al., 2006).
Within the accretion framework, the following scenario for the formation of CBP is favored (Pierens & Nelson, 2007; Meschiari, 2012; Paardekooper et al., 2012): the planetary core forms in the outer accretion-friendly zone in the protoplanetary disk; then it migrates inward and stalls at the border of the central cavity. Paardekooper et al. (2012) performed simulations of planetesimal evolution in circumbinary disks, including the formation of km-sized planetesimals; it was found that Kepler-16b, 34b, and 35b are unlikely to have formed in situ.
As Bromley & Kenyon (2015) note, in the outer regions of circumbinary disks, planets form “in much the same way” as their analogues in single-star systems. The circumbinary environment may even be friendly to planet formation. However, migration through viscous interactions with a gas-containing disk seems to be a necessary factor in forming final architectures. The location of the orbits of the observed giant planets in Kepler systems are reproduced at certain parameters of viscous drag and disk profile models (Pierens & Nelson, 2013; Kley & Haghighipour, 2014, 2015).
Nelson (2003) was first to study the migration of CBP in disks; his hydrodynamical simulations showed that a Jupiter-like giant can be captured in the 4/1 resonance with the central binary. Pierens & Nelson (2007) showed that lower-mass CBP (with masses from 5 to 20 in Earth units) may stall at the border of the disk cavity and predicted that Neptune-like CBP can be discovered there. This was confirmed by discoveries of the Kepler CBP. An important role of orbital resonances was considered in Pierens & Nelson (2008a, b) in hydrodynamical simulations of planetary migration in circumbinary gaseous disks. Such resonances as the 4/1 and 5/1 with the central binary may serve as traps for migrating planets and generally influence the formation process and orbital evolution of giant Saturn-mass planets embedded in the disk. These results, favoring inward migration, were confirmed in Kley & Haghighipour (2014), where a detailed balance of viscous heating and radiative cooling of the disk was taken into account, and in Lines et al. (2016), where planetesimal growth was considered taking into account perturbations due to the precession of the eccentric disk.
Pierens & Nelson (2007) were first to explore the outcomes of type I migration of low-mass planets in gaseous circumbinary disks. It was found that the migration can be halted at the edge of the central cavity. For a gaseous disk, this halting is explained by counterbalancing the negative Lindblad torque by the positive corotation torque; this “planet trap” mechanism (considered below in more detail) operates when there is a sharp positive gradient of the disk surface density (Masset et al., 2006). Pierens & Nelson (2008a) performed hydrodynamical simulations of multiple-planet systems embedded in circumbinary disks. In a system consisting of two planets, when the largest one has settled at the cavity border and the smaller one continues to migrate inward from outwards, the outcome usually represents an equilibrium configuration with the planets locked in orbital resonances. (If the smaller planet is first to settle at the cavity border, the usual outcome consists of scattering of the planets.) In a simulation of a five-planet system, Pierens & Nelson (2008a) revealed a stable final outcome representing a totally resonant three-planet system. This is in accordance with the possible existence of smaller planets outer to the observed “leading” giants, as proposed in Section 2.
All mentioned results of massive hydrodynamical simulations are in agreement with analytical predictions of the so-called “planet trap” concept, developed in Masset et al. (2006) to overcome the problem with type I planet migration, which can be too fast relative to the rate of planet build-up, thus eliminating the planet totally. As already mentioned above, when there exists a sharp positive gradient of the disk surface density at some radial distance from its center, the negative Lindblad torque can be counterbalanced by the positive corotation torque, and the planet-forming material can be trapped at this location (Masset et al., 2006; Meschiari, 2014). Masset et al. (2006) identify possible locations of the traps: apart from the edge of the central cavity, these locations may comprise the outer borders of resonant ring-like zones cleared from the gas by giant protoplanets (the single-star case was considered, but the circumbinary case is analogous and is even more representative). The smaller planets outer to the observed “leading” giant CBP can, in principle, be trapped at these local borders.
In the framework of the planetary trap concept, Meschiari (2014) suggested that large planetesimals form at the cavity edge, where the pressure is maximum and the surface density of solid material is enhanced; this may give rise to CBP formation in situ. Circumbinary planet formation in systems of close binaries may be even easier than planet formation around single stars (Martin et al., 2013). Indeed, it is usually assumed that the circumbinary disk is fully ionized and turbulent, but in fact, it is typically layered, consisting of a radially extended “dead zone” (a non-turbulent midplane layer) and two surface layers (strongly turbulent). The dead zone is a place favorable for planet formation where solids may settle to the midplane. Therefore, according to Martin et al. (2013), in poorly ionized disks, CBP may form in an easier way than planets in disks of single stars.
The radial locations of the planet traps in single-star disks are ill-constrained (Morbidelli et al., 2008), and this fact, as outlined in Pelupessy & Portegies Zwart (2013), is the principal problem with the concept of planet traps in single-star disks. The circumbinary case is much more clear-cut in this respect because circumbinary protoplanetary disks possess a definite scale defined by the size of the binary. Therefore, the radial location of the sharply positive radial gradient of the surface density is perfectly defined. As opposed to single-star disks, circumbinary disks do not need excessive fine-tuning of parameters to provide suitable radial locations of forming planets.
For single stars, Morbidelli et al. (2012) emphasize a great diversity in the evolved orbital architectures of terrestrial planets, given that the accretion process for the forming terrestrial planets depends sensitively on the orbital configuration and evolution of giant planets, which form first in the protoplanetary disk. Can an even greater diversity be expected in planetary systems of binary stars? For the S-type (circumcomponent) planetary systems in wide binaries, this can be indeed so, but for the P-type (circumbinary) systems, the inference is not obvious at all. Formally comparing a circumbinary planetary system architecture with that of the Solar system, one sees that the former may be considered as a “truncated version” of the latter: the inner rocky part of the system simply does not exist because it is in the circumbinary chaotic zone. In most known CBP-MS, this zone is closely orbited by a giant planet, a Jupiter analogue. If this “truncated system” picture is valid, then the orbital architecture of the zone outer to the leading giant can be rather simple, resembling that of the Kuiper belt, with icy/rocky objects tending to reside in orbital resonances with the shepherding giant. (The latter’s role in the Solar system is played by Neptune.) This architecture is in accordance with the possible existence of terrestrial CBP outer to the leading giant CBP, as proposed in Section 2.
An observational evidence for the existence of planets outer to those trapped at the edge of the central cavity is provided by the discoveries of such planets as Kepler-47c, Kepler-1647b, and OGLE-2007-BLG-349L(AB)c (Orosz et al., 2012a; Kostov et al., 2013; Hinse et al., 2015; Kostov et al., 2016a; Bennett et al., 2016). The discoveries of such objects demonstrate that not all CBP halt migration at the cavity edge, but some settle at much greater distances from the host binary. In the case of the Kepler-47 multiplanet system, there is a lot of space between Kepler-47b and Kepler-47c to harbor possible dynamically stable low-mass planets (Hinse et al., 2015), i.e., the observed two-planet system is in no way closely-packed dynamically. The systems Kepler-1647 and OGLE-2007-BLG-349L(AB) may as well contain low-mass planets in orbits inner to the observed giant planets because the orbits of the former ones are large in size in comparison with the central chaotic zone.
In conclusion, from the viewpoint of planet formation scenarios, a major preference for the occurrence of CBP in locations suitable for habitability seems to exist indeed. In brief, in circumbinary systems, the major planet trap forms at the strictly defined (in radial location) border of the central cavity in the protoplanetary disk. Due to the observational preference for particular values of the orbital periods of twin Solar-like binaries (this preference also has a theoretical explanation, as described in Section 2), any planets moving just outer to the leading giant CBP are preferentially placed inside the insolation HZ. In other words, there exist a natural radial scale and pattern for the overall planetary architecture in circumbinary systems, whereas there are no such clear-cut scale and pattern in single-star systems, at least as known at present.
10.2 Non-ubiquity of double planets
Terrestrial CBP, as discussed above in Subsection 10.1, seem to be generic, whereas, on the other hand, double planets seem to be non-generic: the probability for an Earth-like planet to acquire a large Moon-like satellite, from the cosmogonical viewpoint, is extremely low (Lissauer, 1997).
The theme, however, remains disputable, as recent massive numerical experiments taking into account fragmentation and bouncing collisions show. In Quintana et al. (2016), model statistics of giant impacts during the late stages of terrestrial planet formation around a Solar-like star was obtained, according to which the impacts forming Moon-sized satellites are rather typical in the protoplanetary systems of Solar-like stars. Therefore, double terrestrial planets are not necessarily very rare. Clearly, the problem of ubiquity of double terrestrial planets in habitable zones of single stars requires further quantitative study, including estimates of the abundance of such objects.
10.3 Exomoons
Habitable exomoons may exist under appropriate conditions (Kaltenegger, 2010). Exomoon habitability is constrained by orbital stability, illumination, and tidal heating (Heller, 2012; Heller & Barnes, 2013). Magnetic shielding of exomoons can be provided by the magnetospheres of the host planets (Heller & Zuluaga, 2013). Hinkel & Kane (2013) considered habitability properties of exomoons residing either far from their host planet, close to the Hill stability border, or conversely, close to the planet, near the tidal locking threshold. Exomoons can be quite different from the moons present in the Solar system; in particular, they can be much larger in size. As Heller & Pudritz (2015) found in numerical simulations of the cosmogony of satellites of super-Jupiters, Earth-size moons may form at appropriate (permitting the habitability) orbital distances from the host planet.
Like any other planets, CBP-MS may have moons inside their Hill spheres. Since all known CBP-MS are giants, these moons can be large and can, in principle, be habitable. Exomoon habitability conditions were considered in Heller & Pudritz (2015). These conditions are modified by the binarity effect of the host star. Quarles et al. (2012) considered hypothetical moons of Kepler-16b as possible habitable niches, as Kepler-16b itself is a Saturnian-type giant planet and may hardly be habitable. Forgan (2014) suggested that while Kepler-47c (a planet in the HZ of Kepler-47) is a giant one (not terrestrial), it may have a habitable terrestrial satellite, with the climate subject to strong variations.
Generally, the expected strong day/night, seasonal, and climatic variations on possibly habitable terrestrial exomoons on various timescales related to the orbital motions of the host planet and the moon itself can make the survival of their biospheres questionable. For exomoons of CBP, where additional variations related to the orbital motion of the host stellar binary are present, the picture becomes even more complicated. As Forgan (2014) notes, if a giant CBP in the stellar HZ, such as Kepler-47c, has an Earth-like moon, the latter’s biosphere can be sustained only if it is able to endure strong climatic variations.
10.4 Spin–orbit coupling
An analysis by Correia et al. (2015) demonstrated that stable spin–orbit coupling may well exist for circumbinary bodies in circular orbits in the hierarchical three-body problem. However, as shown by Correia et al. (2015), for gas giants, the coupling is unlikely due to their expected almost perfect axial symmetry, similar to that of Jupiter or Saturn; for terrestrial CBP, the coupling may be plausible because the asymmetry is greater.
Spin–orbit resonances generally do not preclude habitability. Brown et al. (2014) studied photosynthetic conditions in a model of a CBP in 3/2 spin–orbit resonance and concluded that life on such a planet may well survive, though, contrary to the Earth, it would have geographically longitudinal, not latitudinal, areas of habitat. Capture of a CBP in low-order spin–orbit resonances, such as 3/2 or 1/1 (synchronous) ones, is not expected to affect the photo-tidal synchronization; such resonance would only mean elimination of fast (day/night) insolation and tidal harmonics. Thus, the photo-tidal synchronization would operate in its sheer form. The interplay of fast and slow photo-tidal harmonics, when CBP rotates rapidly, deserves a separate study.
10.5 Photosynthetic efficiency
Not only the total flux of stellar radiation at the surface of a planet matters, but the radiation spectrum as well. If the companion of a yellow dwarf in a binary is a red one, this can favor a variety of primordial chemical reactions on planetary surface, since the spectrum of the light falling on the planet is significantly widened. As mentioned by Mason et al. (2013), an excessive photosynthetic radiation flux provides an additional excellent condition for habitability of CBP.
It is well known that the absorption spectrum of chlorophyll is double-peaked, and the peaks do not coincide with the maximum of the Solar spectrum; see, e.g., Shields, Ballard & Johnson (2017) and references therein. Since there exist evolutionary (rather complicated) scenarios explaining this spectral mismatch (see a review in Raven & Wolstencroft 2004), it would be premature to interpret the double-peaked absorption spectrum of chlorophyll as a “fingerprint” of a double star. However, note that if the components of a binary are even only slightly different in spectral class, the widening of the composite spectrum can increase the photosynthetic quality of the radiation received at the surface of a CBP with Earth-like photosynthesis.
The photosynthetic conditions at some known CBP-MS were modeled in detail in Forgan et al. (2015); they explored the variations of the photosynthetic spectral quality of radiation with time and surface position. The photosynthetic conditions at CBP provide a much wider field for evolutionary diversification than on Earth. Combinations of G and M stars provide opportunities for both familiar and exotic forms of photosynthetic life, such as infrared photo-synthesizers (O’Malley-James et al., 2012).
10.6 Habitability of CBP of M-dwarf binaries
As noted already in Section 2, M-dwarfs can be especially suitable as host stars of habitable CBP. Habitability properties of planets of M-dwarfs were extensively studied in the last years; see reviews by Tarter et al. (2007) and Shields, Ballard & Johnson (2017). While M-dwarfs are numerous and have long lifetimes in comparison with stars of other spectral classes, two major disadvantages for the potential habitability of their planetary systems exist (Cuntz & Guinan, 2016), namely, (1) small sizes of stellar habitable zones and (2) frequent flares, including superflares. Note, however, that for circumbinary systems the situation can be different. The circumbinary HZ is more extended, as a twin binary produces twice as many photons as any of its components, and of the same spectrum (see Section 2); what is more, the magnetic activity can be suppressed via BHM, discussed in Section 7.
Planets of M-dwarfs have on average smaller sizes than planets of Solar-like stars (Martinez et al., 2017). This observational fact may provide another answer to the problem of formation of planets of terrestrial type, again pointing out that abiogenesis can be concentrated on CBP of double M-dwarfs.
M-dwarfs comprise more than 70% of the Galactic stellar population, and more than 50% of the stellar population are in binaries (Duquennoy & Mayor, 1991; Bochanski et al., 2010). The mass function of M-dwarfs peaks at classes M3V–M4V (Bochanski et al., 2010, Figure 23) — just where resonance cells 5/1–6/1 and 6/1–7/1, typically occupied by planets, overlap with the circumbinary habitable zone (see Section 2). This means that if abiogenesis is indeed concentrated on CBP of double M-dwarfs, then the production of replicating biopolymers in the Galaxy can be a massive phenomenon.
10.7 Abiogenesis apart from TCR
In this article, we have focused almost entirely on Lathe’s tidal mechanism for abiogenesis because it has directly verifiable astronomical aspects. Let us briefly discuss the astronomical aspects of other modern approaches to the problem of life origin.
The photo-tidal periodic driving could also be important in the “everything-at-once” abiogenesis scenario, where metabolism, replication, and compartmentalization all arise simultaneously. The scenario is described in Powner, Gerland & Sutherland (2009), Sutherland (2015), and Patel et al. (2015). According to it, the first life forms are thought to arise in pools generated by hot metal-enriched water streams in tectonically active environments. Analogously to Lathe’s mechanism, an intermittent cooling/heating is needed to provide association/replication cycling of the proto-replicating material, pumping its concentration. Heating induces the strands of the replicating material to come apart.
An important point is that the compartmentalization of proto-replicating material in protocells may provide extended timescales for keeping the produced biomaterial from destruction; therefore, the problem of a necessarily short period of the association/replication cycling in Lathe’s mechanism (discussed in detail in Lathe 2005) can be avoided. However, the period of the proto-replication cycle should be somehow adjusted to the astronomical photo-tidal cycling period.
Therefore, Sutherland et al.’s geochemical scenario seems to be subject to the same general photo-tidal conditions as those in Lathe’s mechanism, and therefore, it may operate on CBP-MS in the same way as it hypothetically operated on Earth.
Of course, apart from the photo-tidal ones, other conditions should be satisfied, such as active tectonics, UV radiation, and in particular, the presence of particular metals (copper, iron, sulphur, magnesium; Powner, Gerland & Sutherland 2009; Patel et al. 2015).
Concerning active tectonics, on CBP-MS it can be provided automatically, as we have seen in Section 6, at least for a subclass of CBP-MS. What is more, on CBP-MS, the available UV radiation is subject to a periodic variation with multiple harmonics, some of which might be suitable for abiogenesis.
In conclusion, Lathe’s TCR mechanism (Lathe, 2004, 2005) for abiogenesis on the Earth is favored as generic for CBP-MS due to the photo-tidal synchronization that is inherent to them. For the same reason, CBP-MS can also be suitable for Sutherland et al.’s “multi-stream” abiogenesis.
Martin and Russell’s abiogenesis scenario (see Koonin 2012 and references therein), in which life is born in deep-sea hydrothermal vents, is indifferent to the presence of the photo-tidal cycling, at least in its “photo” constituent. Since CBP-MS do not provide any theoretical preference here, one may expect that any verification of this scenario also checks the importance of the photo-tidal cycling for abiogenesis.
10.8 Superhabitability
Heller & Armstrong (2014, p. 50) outline “the possible existence of worlds that offer more benign environments to life than Earth does,” and define such worlds as superhabitable.
According to Heller & Armstrong (2014), terrestrial planets with masses from 2 to 3 Earth masses are potentially superhabitable due to a number of circumstances, namely due to (1) the prolonged tectonic activity, and as a consequence, the prolonged existence of the carbon–silicate cycle, (2) the enhanced magnetic shielding owing to a greater mass of the iron core, (3) the greater surface area resulting in greater biomass production, (4) a smoother surface providing shallower seas, and (5) the ability to maintain thicker atmospheres.
On the other hand, K-dwarfs, being only a little bit less massive and luminous than the Sun, have much longer lifetimes, thus favoring long-term habitability of their planets.
Combining these considerations with our inferences for CBP, one may expect that the CBP belonging to the super-Earth class and/or orbiting K-dwarf binaries can superhabitable, in the definition of Heller & Armstrong (2014). However, this conclusion should be put in correspondence with our inferences made above for Solar-like stars and M-dwarfs as host stars; we leave it for future work.
10.9 Intragalactic transport
Finally, it is interesting to note that circumbinary systems may give birth to rogue planets transporting the masses of produced biopolymers or any other biogenic chemicals elsewhere. On cosmogonical timescales, the process of escape of planets from host stellar binaries seems natural because the mass ratio and size of the binary inevitably change with time. This evolution is due, in particular, to the stellar mass loss and mutual tides (Hurley, Tout & Pols, 2002). Therefore, a CBP-MS, initially automatically stalled at the edge of the central chaotic zone, may enter it and, consequently, escape. This would entail the destabilization and reconfiguration of the overall planetary system, including the possible scattering and ejection of its other planets. A production of rogue planets by stellar binaries was considered by Veras & Tout (2012) in similar dynamical scenarios.
The final escape of planets seems to be natural in the case of circumbinary systems, but not for our Solar system, which is known to be extremely stable. Indeed, only Mercury may become rogue, as revealed by Laskar (1994), on a timescale of several billion years.
In fact, single-star and binary-star populations can both loose their planets by a number of dynamical and physical mechanisms (stellar fly-bys, planet–planet scattering, supernova explosions, etc.; see Veras & Tout 2012; Veras et al. 2014; Kostov et al. 2016b). No comparative planetary escape statistics is available up to now, due to the complex and multifaceted nature of the phenomenon. However, CBP (at least the CBP orbiting at the edge of the circumbinary chaotic zone) seem to be much more prone to any destabilizing factors. Sutherland & Fabrycky (2016), based on massive numerical simulations, find that the planetary escape process from circumbinary systems may efficiently fill the Galaxy with rogue planets. Most CBP are ejected rather than are destroyed via planet–planet or planet–star collisions (Sutherland & Fabrycky, 2016; Smullen et al., 2016).
The numerical study by Kostov et al. (2016b) of the long-term dynamical evolution of the Kepler circumbinary systems, taking into account the stellar evolution of the host binaries, reveals that the expansion/contraction of CBP orbits can be adiabatic, coherent with the evolution of the host binary. This reduces the chances for CBP ejection. However, at least two Kepler binaries (namely Kepler-34 and Kepler-1647) are doomed to free their planets because in the course of stellar evolution, they experience a double-degenerate supernova explosion. Tantalizingly, these two binaries belong to the class of twin binaries, i.e., the stellar companions are quite similar in mass in each of the systems.
In conclusion, the intragalactic migration and dissemination of the planetary chemicals can have two sharply different modes, corresponding to the multiplicity of host stars. CBP-MS can be mostly ejected in the course of the stellar evolution of their host binaries and become rogue and then migrate freely, whereas planets of single stars are generally orbitally superstable (as the Earth is) and thus are doomed to migrate radially slowly together with their host stars (on stellar migration, see Sellwood & Binney 2002; Minchev & Famaey 2010; Shevchenko 2011; Grand et al. 2016).
10.10 Methods
Some basic methods used in this article concern the analysis of resonances in Hamiltonian dynamics. Namely, the extents of the chaotic zones around gravitating binaries are estimated (in Section 2) using methods based on Chirikov’s resonance overlap criterion (Chirikov, 1979; Shevchenko, 2015). Theoretical resonant and chaotic Hamiltonian dynamics form the basis for the obtained inferences on the climate stability conditions (Section 4), delivery of water (Section 8), and the final fate of CBP (Section 10). In Section 2, the Lidov–Kozai effect again concerns resonant Hamiltonian dynamics; on the resonant nature of this effect, see Shevchenko (2017).
11 Conclusions
Circumbinary planets are generic: indeed, a lot of circumbinary planetary systems have been discovered up to now, and cosmogonical simulations show that the formation of such stable systems is a natural process, as discussed in Subsection 10.1. As we have seen, striking analogies exist between the habitability conditions on CBP-MS and on the Earth. In fact, in favoring the habitability conditions, the Earth seems to mimic a typical CBP-MS. CBP-MS of particular classes seem to be generic in providing such conditions (i.e., the conditions arise automatically), whereas the Earth is not (i.e., the conditions arise accidentally). Therefore, the multiple analogies revealed between CBP-MS and the Earth may indicate that life on Earth is a low-chance outlier of a generic global chemical process (massive production of replicating biopolymers) concentrated on CBP-MS.
With respect to the insolation condition of habitability, at least two habitability niches, where the needed insolation is provided more or less automatically, can exist. These are the “leading” CBP of red-dwarf twin binaries and the “outer” (as defined in Section 2) CBP of Solar-like star twin binaries.
In the considered scenario, Lathe’s mechanism (Lathe, 2004, 2005) for the tidal “chain reaction” abiogenesis on the Earth is favored as generic for CBP-MS, due to the photo-tidal synchronization inherent to them. For the same reason, CBP-MS can also be favorable for Sutherland et al.’s “multi-stream” abiogenesis. On the other hand, the scenario is indifferent to the deep-sea hydrothermal-vent abiogenesis hypothesis.
In the Solar system, only one planet in the insolation HZ is known to be biogenic. In the proposed scenario, this is accidental and is mostly due to its duplicity, the Earth being an analogue to a CBP-MS in several aspects. In circumbinary systems of MS twin binaries, the HZ is broader than the HZ of single stars of the same class, and, what is more, the planets need not be double to be habitable; therefore, all super-Earths or Earth-like planets in the insolation HZ can actually be habitable. Consequently, in the proposed scenario, CBP-MS seem to be the main carriers of biogenic chemicals in the Galaxy, whereas planets of MS single stars produce biogenic chemicals only accidentally.
Acknowledgements
The author is grateful to the referee for useful remarks. This work was supported in part by the Russian Foundation for Basic Research (project No. 17-02-00028) and the Programme of Fundamental Research of the Russian Academy of Sciences “Experimental and theoretical studies of objects of the Solar system and planetary systems of stars.”
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Altwegg et al. (2015) Altwegg, K., Balsiger, H., Bar-Nun, A., et al. 2015, Science, 347, 1261952
- 2Artymowicz & Lubow (1994) Artymowicz, P., & Lubow, S. H. 1994, Ap J, 421, 651
- 3Artymowicz & Lubow (1996 a) Artymowicz, P., & Lubow, S. H. 1996 a, in Disks and Outflows around Young Stars, ed. S. V. W. Beckwith, J. Staude, A. M. Quetz, & A. Natta (New York: Springer), 115
- 4Barnes (2015) Barnes, R. 2015, Int. J. Astrobiol., 14, 321
- 5Barnes et al. (2009) Barnes, R., Jackson, B., Greenberg, R., & Raymond, S.N. 2009, Ap J, 700, L 30
- 6Beletsky (1963) Beletsky, V. V. 1963, The Motion of an Artificial Satellite about Its Center of Mass (New York: Plenum)
- 7Benn (2001) Benn, C. R. 2001, Earth, Moon and Planets, 85–86, 61
- 8Bennett et al. (2016) Bennett, D. P., Rhie, S. H., Udalski, A., et al. 2016, AJ, 152, 125