Launched on a clear winter day in January 2005, NASA's Deep Impact spacecraft spanned 268 million miles (431 million kilometers) of deep space in 172 days, then reached out and touched comet Tempel 1.

The collision between the coffee table-sized impactor and city-sized comet occurred on July 4, 2005, at 1:52 a.m. EDT. This hyper-speed collision between spaceborne iceberg and copper-fortified, rocket-powered probe was the first of its kind. It was a boon to not only comet science, but to the study of the evolution of our solar system.

The mission of Deep Impact was supposed to conclude within weeks of this July 4 cometary smackdown. Then, NASA approved a mission extension, re-enlisting the Deep Impact spacecraft for two distinct celestial targets of opportunity.

EPOXI, as the mission was renamed, was a combination of the names for the two extended mission components: the extrasolar planet observations, called Extrasolar Planet Observations and Characterization (EPOCh), and the flyby of comet Hartley 2, called the Deep Impact Extended Investigation (DIXI).

The Deep Impact spacecraft, history's most traveled deep-space comet hunter, provided many significant results for the science community. Here are the top five, according to the mission's principal investigator, Michael A'Hearn of the University of Maryland, College Park.

Studies of imagery showed that that the luminous flash created within a fraction of a second after Deep Impact's impactor was atomized by comet Tempel 1 was much fainter than expected.

Comparison with experimental impacts at the Vertical Gun Range at NASA Ames Research Center in Moffett Field, Calif., showed that such a faint flash was consistent only with a surface layer (depth a few times the diameter of the impactor) that was more than 75 percent empty space. This surprisingly high porosity was in contrast with theories that predicted comets were armored with a stronger, solid crust that impeded outgassing.

Observations of comet Tempel 1 by Deep Impact's spectrometer instrument showed that water was arising primarily at longitudes near noon and peaking near the equator, whereas most of the carbon dioxide was arising from far southern latitudes, not too far from comet Tempel 1's south pole.

This could be due to seasonal effects (southern hemisphere just going into winter darkness) or due to differences in the chemical composition in different parts of the nucleus.

During the mission extension, the EPOXI observations of comet Hartley 2 showed that the comet's smooth waist was emitting pure water, while the small end was emitting excess carbon dioxide, regardless of time of day. This was a clear sign that chemical diversity was the important factor in a comet's chemical makeup.

For many years we have known that a handful of comets (fewer than 10 percent) produced more water vapor than should be possible by sublimation of nucleus of water ice, in which the sizes of the nuclei are known.

The flyby of comet Hartley 2 showed a large number of icy grains in the coma are driven out of the nucleus by the outgassing of carbon dioxide. These icy grains are plausibly the source of much of the water coming from the comet.

Observations of Hartley 2 by the Deep Impact spacecraft showed the importance of carbon-dioxide ice relative to carbon-monoxide ice in comets, and led to reexamination of all previous observations of these two ices in comets. The relative abundances in short-period and long-period comets imply that the short-period comets formed under warmer conditions than did the long-period comets.

Thus, the short-period comets must have formed closer to the sun than their longer-period brethren.

This is contrary to popular belief in the astronomical community (for many decades) that the short-period comets formed in the Kuiper belt beyond Neptune, while the long-period comets formed in the vicinity of the giant planets. The new model fits well with measurements by other astronomers of heavy water in Hartley 2, and with the newest dynamical studies of planetary migration.

The excavation of a crater on Tempel 1 was the trigger that allowed the proposal for the Stardust NExT mission to succeed. In addition to searching for the crater formed by Deep Impact, a key goal of that Stardust-NExT mission was to measure changes in the surface of the comet over an orbital period.

This second set of measurements of Tempel 1 surface features showed that much of the evolution was in discrete, large areas, i.e., there was not a small, uniform erosion of the all parts of the surface, but rather large changes in a few places.

Thus, comets evolve in a manner anaologous to erosion - most erosion takes place in discrete events (floods that make large, local changes) rather than as a slow, continuous process.