Was Moon Born From Planet's Crash Into Earth?

By Ben Harder
for National Geographic News
August 20, 2001
The moon is not made of green cheese, as myth suggests. But the real story of the moon's creation may hardly be more probable.

Many scientists have thought for years that the moon was formed during the early days of the solar system when another planet collided with Earth, ejecting fragments of rocky material that condensed into Earth's only satellite.

The effect would have been as though a lousy cosmic golfer tore up a giant chunk of turf and sent it hurtling into orbit.

For more than two decades, scientists have sought to determine how large the mysterious intruder planet must have been and exactly how its cataclysmic crash could have helped form the moon. But none of their models have offered a completely satisfying explanation.

Now, new research offers a scenario that may work. It suggests the impact may have come from a much more modestly sized foreign body than previous research has proposed.

Robin Canup, a researcher at Southwest Research Institute in Boulder, Colorado, and her colleague have fashioned an improved model using a sophisticated computer-modeling technique. It explains the size, composition, and orbital properties of both Earth and the moon.

"We determined that a Mars-sized impactor would work the best," said Canup. She and co-author Erik Asphaug, a scientist at the University of California at Santa Cruz, proposed their scenario in a study that appeared last week in the scientific journal Nature.

Colossal Impact

"Giant impact" theories explaining the moon's formation were first proposed in the mid-1970s. A decade later, researchers ruled out a Mars-sized object as the source of the impact and began to model larger and larger impacts. The two best models that emerged, however, both had inherent problems.

In one model, the mass of the Earth was right, as was the composition of the moon. But the Earth's rotation rate after the collision was unrealistically fast. An improbable second impact would have been required to slow the Earth's spin.

A second scenario suggested that the impact occurred when Earth was only half formed. That idea better explained the Earth's modern rate of rotation and the moon's orbit, but it required Earth to continue accumulating matter after the impact. That material would have been rich in iron, which composes 30 percent of Earth's mass. But the moon, which contains almost no iron, would have simultaneously absorbed similarly iron-rich rock. The model offers no way to explain the moon's confounding dearth of iron.

Canup and Asphaug have proposed that the impact came from an object that was smaller than in the previous models, but was nonetheless substantial. At one-tenth the mass of the Earth, it was about the size of Mars, the two researchers say.

The collision occurred 4.5 billion years ago, only 50 million years after the solar system formed. The colossal impact must have nearly rent the young Earth apart.

"It didn't break the Earth up, but it came pretty close," Canup said.

"The Earth was distorted into an oblong shape before it gravitationally rebounded" over the course of several hours or a day, she said. Some of the material flung into space settled into orbit and eventually clumped together to form the moon.

Better Modeling

Canup and Asphaug were able to re-test the discredited mid-1980s hypothesis of impact by a Mars-size object thanks to greater computing power. They used a technique called smooth particle hydrodynamics to simulate interactions among the many rocky fragments that would have been created by the impact.

Using several powerful computers, the two scientists produced simulations involving 20,000 virtual fragments of the Earth and of the smaller foreign planet that collided with the Earth. Earlier simulations of similar impacts had been done with only 3,000 particles, which limited the realism of the simulations.

The researchers ran many simulations, adjusting the key variables—the size of the object that caused the impact, the angle of its course, and the mass of the Earth—to see which combination produced the best result.

The scenario involving a Mars-size object won out. That was when the researchers realized "the resolution makes a big difference," said Canup, referring to the number of particles that were used in the simulations. Three thousand particles, it turns out, is not enough realistically to simulate a collision between planet-sized objects.

In a companion article in Nature addressing Canup and Asphaug's study, planetary scientist Jay Melosh of the University of Arizona in Tucson noted, "Encouraging as these new results are, they are not the final word."

One major question is the accuracy of the mathematical equation underlying the new impact model. That equation, developed in 1962, doesn't distinguish well the behavior of ejected solids, liquids, and gases in the hours following the impact.

Treating these states of matter differently in the simulation could explain another peculiar aspect of the moon's composition: its dearth of easily vaporized "volatile" compounds such as water.

A newer, more sophisticated modeling equation has been developed, but Canup and Asphaug did not use it because it was known to have some imperfections. Since their study, Melosh has reworked that equation. Now he is teaming with Canup and Asphaug to test their new model with the more sophisticated equation to see if the results are consistent with their present findings.

© 1996-2008 National Geographic Society. All rights reserved.