Solar System Archaeology, Part I

Figure 1. Lunar craters: Von Karman, Leibniz, and Oppenheimer. Credit: NASA/Michael Benson

 
The ongoing (and exciting!) Dawn mission to the Asteroid Belt is just the latest in a long line of efforts to understand the nature and origin of our Solar System. The modern era of system exploration began more than half a century ago, back in the 1950s. That’s when C.C. Patterson of Caltech used measurements of lead isotopes in meteorites associated with Meteor Crater, Arizona, to estimate that the Solar System formed 4.55 billion years ago (Patterson 1956). Within error margins, his date still stands today.

Other crucial evidence came from manned and unmanned missions to the Moon, beginning in 1959 with the Luna program of the Soviet Union and quickly followed by the Surveyor and Apollo programs of the United States. These and later efforts mapped the entire Lunar surface, enabling a detailed study of the Moon’s history of cratering (Figure 1). They also brought back rocks from Lunar impact sites that provided a range of dates, none of which were older than 3.9 billion years. This was the first indication of a discrete, system-wide cataclysm now known as the Late Heavy Bombardment.

Since those pioneering space programs, ever larger amounts of data have been accumulating from robotic missions to various parts of the inner and outer Solar System, as well as from ground- and space-based studies of star-forming regions in our neighborhood of the Milky Way. Beginning in the mid-1990s, data on extrasolar planetary systems have contributed strikingly new perspectives, since most exoplanetary systems bear little resemblance to ours, even though they must have formed from similar building blocks.

Figure 2. The eight Solar planets at their approximate relative sizes.

As a result of all this cosmic archaeology over the past six decades by some thousands of researchers, it’s now possible for a geeky layperson (like myself) to sketch a surprisingly detailed history of our Solar System (Figure 2). This review of current findings is principally informed by the work of Alessandro Morbidelli and the growing number of astronomers and planetary scientists who have collaborated on successive iterations of the so-called Nice model of system evolution. (As a theoretical framework, this model very nice indeed, but its name actually derives from the French city (Nice) that houses the institution (l’Observatoire de la Côte d’Azur) with which Dr. Morbidelli is affiliated.)

The story proceeds in seven sections:

• primordial chaos: Birth environment of the Solar System
• structure emerges: Structure and early evolution of the Sun’s protoplanetary disk
• accretion everywhere: Assembly of planetesimals and planetary cores
• the grand tack: Migration of the four giant planets through the primordial gas nebula
• worlds in collision: Formation of the four terrestrial planets by giant impacts
• late heavy bombardment: Asteroids and comets scatter throughout the system
• life: Biogenesis on Earth and potentially elsewhere in our system

primordial chaos

All stars form in clusters. Our Sun was no exception. All clusters originate in huge dark nebulae (Figure 3) composed of molecular hydrogen swirling in supersonic turbulence (McKee & Ostriker 2007). Such a vast dark cloud, extending for tens of light years, was the primordial chaos out of which you and I and everything else on Earth originated.

Figure 3. Dark molecular clouds in Taurus.
Credit: European Southern Observatory/Davide de Martine

Fragmentation and clumping within the birth nebula produced hydrogen “cores” that collapsed into protostars. Evidence from short-lived radioactive isotopes in the present-day Solar System indicate that our native cloud produced at least one extremely massive O-type star (~25 times Solar mass), which enriched its surroundings either by exploding as a supernova or by producing fierce stellar winds. A star so massive can form only in a relatively large cluster with more than 1,000 members (Adams 2010).

Accordingly, various analyses have suggested an initial membership of 1,000 to 5,000 stars for the Solar System’s birth cluster. This is an unusually large population, since the median membership for birth clusters in the Milky Way is only about 300 protostars (Lada & Lada 2003, Adams et al. 2006). The powerful gravitational influence of member stars would have kept our home cluster bound together for tens or even hundreds of millions of years – a span of ages comparable to those of the Pleiades (100 million; see Figure 7) and Hyades Cluster (625 million). Our huge cohort of sibling stars was originally confined to a volume only a few parsecs in diameter (~10 light years), but the cluster steadily expanded over time as individual stars began to spill out of its tidal radius.

Stars and substellar objects are currently classified into 10 types according to their spectra. They range from the very brightest, hottest, and bluest objects (types O and B, which include the three stars of Orion’s Belt), through bright white stars (type A, represented by such familiar backyard stars as Sirius, Vega, and Altair), through the yellowish Sun-like stars (types F, G, and K, represented by the nearby planet-hosting stars Upsilon Andromeda, our Sun, and Epsilon Eridani, respectively), through the red dwarfs (type M, which includes GJ 581, another nearby planetary host), down to the dim brown dwarfs (types L, T, and now Y), which are not massive enough to sustain nuclear fusion, and thus are not stars.

Remarkably, star clusters tend to produce objects at the full range of spectral types (except for type O, which as we have seen is confined to larger clusters). True stars are born in numbers that correspond inversely to their mass and temperature, so that red dwarfs are the most numerous and O stars are the rarest. This correlation does not extend into the substellar regime, however, as brown dwarfs are apparently much less common than red dwarfs (Kirkpatrick et al. 2012).

Figure 4. The Iris Nebula (NGC 7023), a stellar nursery undergoing
dispersal. Credit: Wikimedia/Hunter Wilson

Within the depths of our primordial cloud, star formation proceeded in waves – sometimes called starburst episodes – extending over thousands or even millions of years. As soon as the brightest stars ignited, their intense radiation and stellar winds began to disperse the parent nebula (Figure 4). Complete dispersal of the gas envelope required 1 to 10 million years (Lada & Lada 2003, Throop & Bally 2008), whereupon the cluster of newborn stars was unveiled for the universe to see. By this time the process of star formation had ended and cluster evaporation had already begun.

Generations of newborn stars arrived in cocoons of hydrogen and dust known as protoplanetary nebulae or protoplanetary disks (“proplyds”). Like the vast, encompassing parent cloud, these smaller clouds also dispersed within a few million years, with dissipation proceeding faster around hotter, more massive stars and slower around smaller, cooler stars. Typical lifetimes of protoplanetary disks fall in the same range as the lifetimes of molecular clouds – i.e., 1-10 million years – with a median age between 2 and 3 million years (Williams & Cieza 2011). Stars older than 6 million years rarely preserve their primordial disks (Williams 2010).

structure emerges

Despite the Copernican principle of mediocrity, which asserts that there is nothing special about Earth, many factors indicate that we humans occupy a favored piece of real estate. First, we originated in an unusually large birth cluster with a substantial cohort of very bright stars. Second, our native star belongs to spectral class G2, meaning that it is hotter and more massive than 90% of the stars in the Milky Way. Notably, stars of higher mass (such as our Sun) are more likely to host planets than stars of lower mass (Johnson et al. 2010). Third, our Sun has a higher than average enrichment in elements heavier than hydrogen (known to astronomers as metals). All members of our birth cluster must have shared this endowment, which like high mass encourages the evolution of planets, especially gas giants like Jupiter and Saturn (Johnson & Li 2012). Thus, our native environment showed clear promise of future greatness.

 Figure 5. Proplyds in the Orion Nebula.
Credit: NASA/ESA/Luca Ricci

The protoplanetary nebula surrounding the newly-formed Sun was probably fluffy and irregular at first, like many of the proplyds in Figure 5. In short order, nevertheless, the angular momentum of our rapidly spinning parent star caused this puffy cloud to flatten into a thin disk of gas and dust, typically known as the Solar nebula. The disk midplane was aligned with the Sun’s equator, and its spin naturally followed the Sun’s rotation (Figure 6).

The ratio of gas to solids in the Solar nebula was originally about 100 to 1 (Andrews & Williams 2005), and its overall mass must have been about 1% to 3% of the Sun’s (Williams 2010, Adams 2010), corresponding to 10-35 Jupiter masses (Mjup; the Sun is 1048 Mjup). Since the total mass of all surviving planets, dwarf planets, moons, asteroids, and comets in the present-day Solar System is less than 2 Mjup, the evolutionary process was markedly inefficient: 80% to 95% of our original patrimony of gas and dust has vanished.

Considerable evidence indicates that the primordial disk of solids in our system had a sharp outer edge at a radius of 30-35 astronomical units (AU; Levison et al. 2003, Tsiganis et al. 2005, Adams 2010). This is notably smaller than the observed dimensions of many nearby dust disks, suggesting that our native nebula was truncated in some way, perhaps by sibling stars in the birth cluster. A likely culprit would be a nearby O star, since observations of the Orion Nebula demonstrate that such massive luminaries can burn off the outer layers of the protoplanetary nebulae of smaller stars that venture within a few light years of their highly energetic photospheres (Williams & Cieza 2011).

Figure 6. Artist’s impression of a young protoplanetary disk.
Credit: NASA/JPL-CalTech

Our own Sun’s magnetospheric activity created an inner cavity in the disk at an approximate radius of 0.05-0.10 AU (Laine et al. 2008). Between this inner edge and the outer limit at ~30 AU, the temperature, surface density, and composition of the disk varied systematically. At small radii, close to the blazing Sun, only the refractory particles (metals and silicates) suspended in the gas cloud could survive, while icy particles (water, methane, ammonia) dominated the mix on wider orbits.

The boundary between the two regimes of ice and rock is known as the ice line or snow line – the distance from the Sun where free-floating molecules of water and other volatiles condense into ice (Ida & Lin 2004, Kennedy & Kenyon 2008). This occurs at a temperature of 170 K, which in the contemporary Solar System corresponds to a radius of 2.7 AU. At primordial times, however, the situation was different. On the one hand, our Sun was about 30% cooler than it is now, so it added less heat to the circumstellar environment; on the other hand, the viscous flow of gases through the Solar nebula created a powerful alternative source of heat. The net result is that our system’s ice line may have been located around 4 or 5 AU when the Sun was ~100,000 years old, shrinking to 2 AU by the time the Sun reached its one-millionth birthday (Kennedy & Kenyon 2008).

Another key structural feature of the Solar nebula was its flared profile, such that its vertical dimension increased along with its radius. This configuration is typical of protoplanetary disks (Williams & Cieza 2011), and is visible in the proplyd at the very center of Figure 5.

Figure 7. The Pleiades star cluster. Image by Antonio Fernández Sánchez.
Credit: NASA/courtesy of nasaimages.org

accretion everywhere

The Solar System’s primordial ice line has been compared to a standing blizzard (Lin 2008). Just beyond this limit, the presence of ice particles increases the density of solid materials by a factor of 3 or 4 (Kennedy & Kenyon 2008). This “bump” in density and solid mass has crucial implications for the evolution of planetary systems.

The abundance of icy particles, along with the favorable collisional velocities obtaining on long, chilly orbits, combined to make this region the “sweet spot” of planet formation (Mordasini et al. 2012). Conditions for the coagulation of solids were more favorable here than anywhere else in the disk, so it is no surprise that our system’s most massive planet, Jupiter, assembled just beyond this boundary.

The formation of gas giant planets has been described as a threshold phenomenon, as it requires a specific set of initial conditions – sufficient density of solids in the protoplanetary nebula to assemble large planetary cores, sufficient persistence of nebular hydrogen to support runaway gas accretion – that apparently obtain around less than half of all stars (see How Weird Is Our Solar System?). As Andrew Mann and colleagues have argued, “a substantial fraction, and probably the majority of stars do not host [gas] giant planets” (Mann et al. 2010). Low-mass stars are especially likely to be giant-free (Laughlin et al. 2004).

The Solar nebula was moderately massive, with a modest enhancement of metals. These initial conditions enabled it to generate two relatively small gas giants. Although Jupiter has triple the mass of Saturn, it is still a relative lightweight in the exoplanetary population. The median mass of extrasolar gas giants is 1.6 Mjup, and many systems host giants of 5 or even 10 Mjup. Regardless of their mass, however, gas giants seem more likely to come in pairs than in triplets or quadruplets; only four systems (47 Ursae Majoris, HD 37124, HIP 14810, Mu Arae) are known to contain three, and only Upsilon Andromedae has so far been proposed as the home of four.

Yet proto-Jupiter and proto-Saturn were not alone on the cold orbits of the outer nebula. At least two and possibly three or more objects in the range of the gas dwarfs formed on the far side of the sweet spot – Uranus and Neptune for sure, and perhaps a smaller gas dwarf or Super Earth nicknamed “Amphitrite” (Desch 2007, Desch & Porter 2010). Their cores assembled somewhere between 5 and 10 AU.

Figure 8. The South Pole of Vesta, oldest rocky sphere in the Solar System.
Credit: NASA/JPL-Caltech/UCLA/MPS/DLR/IDA

A single favored region of the Solar nebula, then, produced all of the most massive planets in our system. Nevertheless, accretion – the process of growth by gradual accumulation – was active throughout the nebula as soon as it flattened into a disk. Accretion is the route by which primordial dust grains conglomerated into planetesimals, the building blocks of planets. Current theories propose that a steady, stepwise process of growth was punctuated by the much more rapid collapse of refractory particles into full-sized planetesimals, with an approximate range in diameter of 1 to 100 kilometers (Johansen et al. 2007, Morbidelli et al. 2009b).

In the inner Solar System, accretion proceeded quickly because of the relatively high velocities of orbiting solids. The result was most likely a large population of so-called planetary embryos traveling in concentric orbits. The presence of nebular gas damped the eccentricity of their orbits – i.e., prevented any deviations from circularity – so that their growth stopped once they exhausted their local “feeding zones.” Isolation masses for these rocky embryos have been estimated at 1%-10% of Earth mass, equivalent to a population of Moon- to Mars-sized objects (Morbidelli et al. 2012).

Some sense of the time required for embryo gestation is provided by the rocky asteroid Vesta (Figure 8), which evidently attained its mature mass and spherical shape just 2 to 3 million years after the condensation of the Solar nebula (Russell et al. 2012). Vesta must have had numerous siblings of roughly similar mass and composition in the young Solar System, though very few survive as isolated objects today.

. . . . continued in next blog post, with list of references . . . .