More than one way to smash the Earth and build our Moon

The Moon is a bit of an enigma. In some ways, it's nothing like Earth, as its minerals contain few volatile chemicals and it has a relatively tiny core. But in other ways, it's nearly our twin, with many elements having isotopic signatures that are almost identical.

Currently, our best model for Moon formation involves having a Mars-sized object smack in to the early Earth. This could create a Moon that has some similarities to the Earth, but ends up with most of the iron from the impact being deposited in the Earth's core. The only problem with this is that anything as big as Mars probably originated from elsewhere in the Solar System, and thus would have a very distinct isotope ratio.

Today, Science is releasing two papers that take very different routes to tackling this problem. One models what would happen if, instead of a large size difference between the Earth and its impactor, the two bodies were of roughly equal size. Another models two differently sized bodies colliding, but assumes the proto-Earth was spinning much faster than it is now, with "days" on the order of 2.5 hours long. And, in a dilemma that may interest planetary scientists, both models produce the sort of distribution of materials we currently see.

Models of Solar System formation suggest that its rocky planets were built sequentially, with planetesimals condensing from the dust and debris, and then merging to form protoplanets. Over time, these protoplanets underwent a series of collisions and mergers, building planets like Venus and the Earth. (A few objects, like Mars and Vesta, may have sat out most of these later mergers, leaving much smaller objects behind.)

A collision of this sort could help explain some puzzling aspects of the Moon. If the debris left behind were dispersed enough, some of it could condense into a separate object, explaining how two large bodies could end up in such close proximity. And heavy elements would preferentially end up in the larger one, explaining the Moon's small core.

The simulations that show how these collisions would work involve smacking together two spheres full of particles, with each particle having a distinct identity and location—heavy metal particles in the core, silicate rocks in the crust, etc. By tracing these particles through the collision, it's possible to follow the iron from the impactor, and watch it drop into the Earth's core, and so on. The problem is that they also show that the Moon should end up with a crust that's largely composed of material from the impactor. And that's tough to square with the fact that the different forms of many elements, called isotopes, vary with a body's location in the Solar System.

So, unless the impactor started right next door, we'd expect that the isotope ratios of the material it brought wouldn't look like the ones on Earth. And, as we look more carefully at the material in the Moon, that just doesn't seem to be the case.

One of the new papers, from the Southwest Research Institute's Robin Canup, takes a look at what would happen if two large bodies combined. Typically, the Earth is modeled as being nearly its current size, the product of multiple planetoid mergers; its impactor as being about a tenth of its size. Canup ran a series of models in which the two were much closer to equal size, starting with an impactor that was half the mass and moving up to one that was about 90 percent. Models where the impactor was about 80 percent of the size of the pre-Earth created debris disks that could form a Moon; both it and the resulting Earth ended up with very similar material in their crusts.

A second paper, from Sarah Stewart of Harvard and Matija Ćuk (now with the SETI institute), went about things completely differently. They focused on the fact that all the collisions involved in building a planet are expected to leave it spinning very rapidly. So they modeled a normal Mars-sized impactor, but had the Earth spinning really fast, with days lasting anywhere from 2.3 to 2.7 hours. These collisions produce a debris disk that was composed primarily of material from the pre-Earth's mantle, which would explain the Moon's present similarity to the Earth.

An example of one of the runs of this model is seen here. The smash leaves the impactor's matter evenly distributed, and a sufficient amount of debris far enough from the Earth to form a separate body.

The problem with both of these models is that they leave the Earth-Moon system spinning very rapidly. Since the angular momentum of the system has to be conserved, we need some way of getting rid of some of this spin. Tidal forces can do some of that, but the authors of the second paper go on to show that there's a resonance between the lunar orbit and the system's orbit around the Sun. This effectively moves some of the angular momentum out of the Earth-Moon system, and into the system's orbit around the Sun. Combined with tidal forces, this can put a strong enough brake on the system.

Which model will win out? Right now, neither paper really addresses the others' model, so it's a bit hard to say. It's possible that, as the details are filled in, one or the other model will provide a better match to the data we already have. Or one of them could identify data we don't have yet. The ongoing GRAIL mission, which is mapping the density of the Moon's crust, may also provide some further information that will help us understand the Moon's formation.

Science, 2012. DOI: 10.1126/science.1225542, 10.1126/science.1226073  (About DOIs).