Robin Canup

Her intricate computer simulations re-create the birth of our moon, among other ancient dramas.

A week after finishing her dissertation on the formation of the moon, Robin Canup danced the lead in Coppelia with the Boulder Ballet. "At the time, it felt like I had a wonderfully full and busy life," she says, "but I can't believe now I did it all." Canup, 35, stopped dancing professionally five years ago. "By that age you're an old dancer but a young scientist," she says. Still, there's an unexpected harmony to her career: Now she studies how moons glide around planets in space.

A key breakthrough came in 2001, when she provided crucial support for the so-called giant impact hypothesis. Ever since the Apollo missions, astronomers had suspected that the moon originated from bits of the Earth that were knocked off in a collision, but they didn't know how the collision had occurred. Canup's computer simulations showed that the hypothesis makes sense over a wide range of scenarios and explains some of what we know about the moon and Earth today. For example, scientists had wondered why the moon has much less iron than Earth. Canup's models offered strong evidence that another large protoplanet, about the size of Mars, hit Earth when it was almost fully formed, sending up a spray of debris. Heavier elements sank into the molten Earth, and much of the rest coalesced into the moon.

Canup spends her days tweaking computer code and watching colored particles spin on cue. For her moon simulations, she creates an Earth and a protoplanet, and gives each a specific mass, velocity, temperature and other characteristics. Then she programs a collision and sits back to watch the show. The math is so complex, it takes a 2GHz processor a week to compute each scenario. Recently, by running a simulation backward, Canup found that most of the moon probably came not from the Earth but from a specific area on the protoplanet that hit it. Her next step is to add "particle splitting": Partway through a simulation, she'll program it to ignore humdrum information (say, the particles at Earth's core) and focus on the interesting stuff (particles that might clump together as the moon).

Canup has begun developing new models to simulate the formation of other planet-moon systems. She thinks the spiral of gas that fed Jupiter during its birth carried debris that clumped up to make its moons; she plans to model it soon. Someday her simulations may predict which planets outside our solar system contain moons like ours. At any rate, as astronomers learn more about planets orbiting nearby stars, it's a good bet that Canup will be revealing how those planets and their moons dance together.