Formation of Habitable Planets
John E. Chambers
Carnegie Institution for Science, 5241 Broad Branch Road, NW, Washington, DC 20015.
Characteristics of a habitable planet
What is a habitable planet? There is no formal definition at present, but the term is generally understood to mean a planet that can sustain life in some form. This concept is of limited use in practice since the conditions required to support life are poorly constrained. A narrower definition of a habitable planet is one that shares some characteristics with Earth, and hence one that could support at least some of Earth’s inhabitants. A commonly adopted minimum requirement is that a planet can sustain liquid water on its surface for geological periods of time. Earth is the only body in the Solar System that qualifies as habitable in this sense. One advantage of this definition is that it can be used to categorize hypothetical and observable planets in a relatively straightforward manner, and we will use it in the rest of this chapter. However, one should bear in mind that not all life-sustaining environments will be included under this definition. Tidally heated satellites of giant planets, like Europa, are likely to possess oceans of liquid water beneath a layer of ice (Cassen et al., 1979), but these objects would not be ``habitable’’ according to the conventional usage definition.
Planets that can support liquid water at their surface must have an atmosphere, and surface temperatures and pressures within a certain range. These planets will occupy a particular range of orbital distances from their star that is commonly referred to as the star’s habitable zone (HZ). For a planet undergoing crustal recycling, the width of the HZ is enhanced by the carbon-silicon cycle (Walker et al., 1981). The weathering of silicate rocks in the presence of liquid water can be encapsulated by the reaction
(1)
which proceeds more rapidly with increasing temperature. However, weathering removes carbon dioxide from the atmosphere, ultimately depositing it in the mantle, which lowers the temperature due to a weakened greenhouse effect. This negative feedback loop keeps the surface temperature in the range for liquid water over a wide range of orbital distances. Kasting et al. (1993) estimate the Sun’s habitable zone extends from 0.95 to 1.4 AU, while more recent calculations by Forget & Pierrehumbert (1997) suggest the outer edge may be > 2 AU from the Sun. Stars tend to grow more luminous as they age, so the HZ moves outwards over time. This has led to the concept of a continuously habitable zone, which is the range of orbital distances that lie in the HZ over a particular interval of time.
The precise extent of the habitable zone depends on a planet’s albedo, which depends on the degree of cloud cover and whether the surface is covered in ice. For example, Earth currently has two stable climate modes: the current climate with a low albedo, and ``snowball Earth’’ in which surface water is frozen at all latitudes and the planetary albedo is high. Habitable-zone calculations implicitly assume the climate is in the former mode. The location of the HZ depends on the distribution of continents and oceans on a planet’s surface, and thus can change over time as a result of plate tectonics (Spiegel et al., 2008). Additional greenhouse gases, such as methane and sulphur dioxide, can also raise the surface temperature, extending the outer edge of the HZ (Pavlov et al., 2000; Halevy et al., 2007).
The habitable zone concept is a useful starting point, but a planet must have additional characteristics in order to support life. The calculations used to determine a star’s HZ assume that a planet possesses an atmosphere and a reservoir of water to begin with. It is currently unclear how planets like Earth acquire their water and other volatiles, and how these inventories change over time. Venus is almost dry today but probably possessed more water in the past, while Earth may have acquired far more water during its formation than exists in its oceans today (Abe et al., 2000). Earth is depleted in all atmosphere-forming elements compared to the Sun, and several of these same elements (C, N, H, O) are essential to terrestrial life. The sources and removal mechanisms for these elements during the accretion of terrestrial planets are poorly understood at present, but the abundances of these elements will have a profound impact on a planet’s habitability.
A planet must be geologically active in order for the carbon-silicon cycle to operate. The nature and degree of crustal recycling probably depends on a planet’s size and composition. Small planets like Mars lose their heat rapidly, and are likely to be less active than larger bodies. Small planets are also less able to retain their atmospheres due to their weaker gravitational fields, and the lack of crustal recycling means that atmospheres are less likely to be replenished once lost. There may be a critical lower limit to the mass of a life-sustaining planet, somewhere between the mass of Mars and Earth, perhaps 0.2-0.3 Earth masses (Williams et al., 1997; Raymond et al., 2007a). A planet’s mass and thermal evolution determine whether a magnetic dynamo operates in its core, generating a strong magnetic field that helps reduce atmospheric erosion caused by interactions with the surrounding stellar wind (Hutchins et al., 1997; Stevenson, 2003).
The climate and habitability of a planet are affected by its obliquity e (axial tilt relative to its orbit) and orbital eccentricity e (degree of non-circularity), since these quantities affect the spatial and temporal distribution of light from the star on the planet’s surface. An increase in either of these quantities increases the degree of seasonal variations on the planet. Surface temperature variations of > 100 K are possible for planets in the HZ with e near 90 degrees (Williams & Pollard, 2003). Both e and e will vary on million-year timescales if giant planets are present in the same system. For Earth, these variations are small, because Earth’s obliquity evolution is dominated by gravitational interactions with its large moon (Laskar et al., 1993), and because its orbit is not near an orbital resonance with Jupiter or Saturn. However, terrestrial planets in some systems may undergo large changes in e and e.
The habitable zones of low-mass stars lie sufficiently close to the star that habitable planets on circular orbits will be tidally locked, with one face permanently lit and the other in darkness. However, the presence of a thick atmosphere should render such planets at least partially habitable (Joshi, 2003). Planets moving on eccentric orbits around low-mass stars are unlikely to be tidally locked. These planets can move significantly closer to their star on billion-year timescales, due to stellar tides, affecting their location with respect to the star’s HZ (Barnes et al., 2008).
Theories for Planet Formation
Information about how planets form comes from a variety of sources, including astronomical observations of young stars, cosmochemical analysis of meteorites, robotic space missions to planets, asteroids and comets in the Solar System, laboratory experiments, and computer simulations. These data have given rise to a standard model of planet formation in which planets begin life as dust grains in orbit around young stars. These dust grains grow as a result of mutual collisions, aided by gravity, until a handful of planetary-mass bodies remain (Lissauer, 1993). This model is widely accepted for the formation of terrestrial planets—rocky bodies such as Earth and Mars. It is less clear how giant planets form. The popular ``core-accretion’’ model posits that solid cores form in the same way as terrestrial planets, and once these cores become large enough they accrete gaseous envelopes from their surroundings (Pollack et al. 1996).
The standard model was developed in order to account for the origin of the Sun’s planets. At the time of writing, several hundred planets have been discovered orbiting other stars. Most of these objects are believed to be gas giants on the basis of their mass and density. The orbital distribution of these extrasolar planets suggests the standard model may need to be modified somewhat in order to explain the variety of systems observed. In particular, it appears that some giant planets migrate substantial distances towards their star during their formation, something that apparently didn’t happen in the Solar System (Armitage, 2007).
This chapter will focus on the formation of terrestrial planets since these appear to be the most likely to support life. I will place particular emphasis on processes that might affect planetary habitability. To date, Earth-like planets have not been firmly identified orbiting other Sun-like stars. Three terrestrial-mass objects have been discovered in orbit around a pulsar (neutron star), but unfortunately little is known about these objects or how they formed (Wolszczan & Frail, 1992). As a result, the discussion in this chapter will be driven to a large extent by what we know about the terrestrial planets in the Solar System, with the obvious caveat that the standard model may be revised in light of future discoveries.
Formation of terrestrial planets
Protoplanetary disks
Most young stars are surrounded by disks of gas and dust. The presence of dusty material in orbit around young stars, and the flattened nature of disks, which mimics the planar arrangement of the orbits of the planets in the Solar System, suggests these disks provide the environment in which planets form. These are commonly called ``protoplanetary disks’’ as a result. The Sun’s own disk is referred to as the solar nebula. The minimum mass of the solar nebula can be gauged from the amount of material of solar composition needed to form the planets, which is 1-2% of a solar mass, although this provides only a lower limit.
Infrared observations of protoplanetary disks show that they contain huge numbers of dust particles with sizes ranging from < 1 mm to > 1 mm (Natta et al., 2007). Emission lines in the spectra of young stars indicate that they are accreting gas from their disks at rates of roughly 10-7 to 10-9 solar masses per year (Gullbring et al. 1998). The mechanism driving this viscous accretion remains unclear, but it is likely to involve turbulence in the disk gas. A disk with a composition similar to the Sun will contain roughly 99% gas by mass, mainly hydrogen and helium, and 1% dust. The dust fraction depends on the stellar ``metallicity’’ (logarithm of the Fe/H ratio), so metal-rich stars have dust-rich disks, and may be more likely to form terrestrial planets as a result. A typical disk has a radius of order 200 AU (Andrews & Williams, 2007), which suggests the maximum radial extent of most planetary systems will be similar to this. Stars older than about 10 Ma (Ma = million years) appear to have lost their disks (Haisch et al., 2001, Pascucci et al., 2006; Cieza et al., 2007). Disks probably dissipate due to a combination of several factors: accretion onto the star, acceleration of gas away from the star by interactions with ultraviolet photons (photoevaporation), and accretion of material by planets. The final stage of disk dispersal appears to be especially rapid, taking a few times 105 years (Cieza et al., 2007).
Temperature and pressure in a protoplanetary disk generally decrease with distance from the star. In the inner disk, only refractory materials such as silicates, oxides, metal, and sulphides (collectively ``rock’’) remain in the solid phase. Further from the star, water ice is also present, first appearing at a distance that is referred to as the ``snow line’’. Highly volatile ices such as methane and carbon monoxide can exist in the outermost regions of a disk. In the cooler regions of the solar nebula, the rock-to-water-ice ratio would have been very roughly 1:1 (Lodders, 2003). Temperatures decline over time as the disk loses mass and the central protostar becomes less luminous (Kennedy & Kenyon, 2008). As a result, volatile materials may be progressively depleted compared to refractory elements, since the former remain in the gas phase for longer (Cassen, 2001). The make up of solid material in any given region of a disk will naturally influence the composition of large bodies that form there. To a first approximation, this explains why the inner planets of the Solar System are composed mainly of rock, while comets and the satellites of the outer planets contain large amounts of water ice.
The dynamics of dust grains is controlled by the gravity of the central star and aerodynamic drag forces from the surrounding disk gas. Gas tends to orbit the star more slowly than solid objects due to the outward pressure gradient. As a result, solid objects experience a headwind, losing angular momentum and drifting inwards. Inward radial drift is especially rapid for m-sized bodies, which can move inwards by 1 AU in 102-103 years (Weidenschilling, 1977). Particles also tend to settle towards the disk midplane due to the vertical component of the star’s gravity. In a turbulent disk, vertical and radial motions are partially opposed by turbulent diffusion. A combination of turbulence and radial drift of solid particles can lead to large spatial and temporal variations in the solid-to-gas ratio, chemical composition, and oxidation state of the disk, and these variations may be reflected in planets and asteroids that form subsequently (Ciesla & Cuzzi, 2006). Solid material can accumulate near the snow line or at local pressure maxima, possibly making these preferred sites for planet formation (Stevenson & Lunine, 1988, Haghighipour & Boss, 2003).