Since the 1970s, most scientists have thought that the Moon formed in a young solar system littered with planetoids on unstable trajectories. One of those planetoids crashed into Earth, heaving into orbit molten chunks that eventually coalesced to form the Moon. An updated computer model by two astrophysicists has added details to the story, suggesting that, unlike conflicting scenarios from previous models, the Moon formed near the very end of Earth’s formation when Earth sustained an oblique hit from a planetoid the size of Mars.
Collisions were relatively frequent at the time of the proposed collision, 50 to 70 million years after the first meteorites formed, says Robin Canup of the Department of Space Studies of the Southwestern Research Institute in Boulder, Colo. Canup, along with Erik Asphaug of the University of California at Santa Cruz, describes the model and its results in the Aug. 16 Nature. Ten or so collisions of this magnitude likely took place, but modelers have found it difficult to simulate the ejection of a Moon’s mass of material into stable orbit. Either the entire impactor becomes part of Earth, or the ejected material travels faster than escape velocity.
What, then, were the conditions that made this collision different? Computer modeling of potential Moon-forming impacts may hold an answer, although such a model is complicated. It must combine shock physics, melting and vaporization, and the mutual gravitational interactions of heated fluids. Most such models use a method called smooth particle hydrodynamics (SPH). In SPH, Earth-asteroid interactions are simulated by splitting the bodies into small computational chunks called particles and then computing the interactions between those particles. The more particles, the more accurate the simulation. Canup and Asphaug used at least 20,000 particles, whereas previous simulations of the scenario they investigated used about 3,000. Those earlier simulations showed that an impactor much larger than Mars created the Moon, and suggested that the collision occurred when Earth was half or two-thirds formed. Canup and Asphaug were unconvinced. If Earth was not yet fully formed, it would have continued accreting after the impact. But the Moon, geochemically similar to dehydrated terrestrial mantle, has a miniature iron core forming about 3 percent of its mass. If the Moon accreted even one-tenth of the proportional material of a young Earth, it would have much more iron than it actually does.
Furthermore, impactors smaller than Mars would produce iron-rich disks because too much of the impactor’s iron core would remain in orbit. Larger hits would likely carve too much iron from Earth, which at this point had already divided into a core and mantle, according to Asphaug.
Next the researchers assumed constant angular momentum from the time of lunar formation to present day. “It is hard to lose angular momentum,” Asphaug says. “The whole system has the same angular momentum over geologic time.” The impactor’s blow had to be a size and angle that would impart present-day spin.
Canup and Asphaug used these assumptions to tweak their model. Collision simulations were considered successful if they produced bodies of the right size and required little or no modification of Earth-Moon dynamics after impact, Asphaug says.
While their results are valid, they are by no means the final word. One controversy is over the equation of state used. The equation of state is a thermodynamic relation between internal energy, density and pressure. “The equation they used lacks clear distinction between solid, melt and vapor phases,” writes Jay Melosh of the University of Arizona. He discussed the model in the same issue of Nature. And, he says, “the equation is not well known for the complex silicates of which the Earth and Moon are mainly composed.” Al Cameron, also of the University of Arizona, first conceived of the impact theory of formation and has run several simulations. He says that this particular equation of state, called the Tillotson equation, is flawed because it assumes that when bodies collide they merge into one body. Cameron would like to see a simulation with even more particles run for more time, although that means running fewer simulations. “So it’s a trade-off,” he says.
Even if the researchers are correct in their results, many questions remain. “The process that created the Moon was more typical than we thought,” Asphaug says. “The part that we don’t know is how long it takes that disk to accrete and form the Moon. And it might be difficult to answer why moons aren’t more common — especially moons like ours.”
Emily D. Johnson