The Help of Computer Simulations
During the 1990's, a new generation of planetary scientists, including Robin M. Canup, now at the Southwest Research Institute in Boulder, Colorado, developed the theory further. They made improved calculations with better computers and more advanced programs that more realistically simulated cosmic collisions and their consequences. By 1996, Canup and her associate Larry W. Esposito of the University of Colorado in Boulder had demonstrated that the giant-impact theory required a much more massive impacting object than previously suspected. They found that the object had to be at least twice as massive as Mars, or even larger, for the debris from the collision to coalesce into the single moon that we have today. Otherwise, the calculations indicated, there would be several moons circling the Earth.
Though Canup and Esposito's calculations were more sophisticated than those of the past, they still were not precise enough. The team used a simplifying assumption, called gas dynamic theory, that represented the processes in the formation of the moon after the giant impact on Earth as though the debris were all gas. In reality, even though the material hurled up by the collision was so hot that it took the form of vapor, it soon cooled and condensed. So the actual conditions must have been much more complicated than those represented by the gas dynamic calculations, with clumps of rock cooling, solidifying, and colliding. In fact, the various proponents of the giant-impact theory had shown only that an enormous collision could have produced the necessary conditions to form the moon. But they had not proven that such an event had actually occurred.
In 1997, the Japanese geophysicist Shigeru Ida at the Tokyo Institute of Technology, together with Canup and Canup's University of Colorado colleague Glen R. Stewart, presented a more advanced simulation of the impact. They used a mathematical technique called the N-body method. This model recreated the processes inside the impact debris cloud as a series of interactions between a large number (“N”) of individual clumps or objects that collided with one another and were affected by one another's gravitational forces. Thus, the N-body method represented real processes in space much better than the gas dynamic technique or other methods did. The Ida team found that the debris from a giant impact with the Earth would indeed have formed a large moon and that this process might have taken as little as one year. Refined calculations by the team and others showed that the most likely mass for the body that collided with Earth was three times that of Mars.