Image of exoplanet discovered around star HR8799 by sifting through old Hubble photos

Example #3497 of why keeping as much information as possible in the public domain is a good idea:

An exoplanet hidden in the Hubble Space Telescope's archival images has been revealed by data miners using a new technique for spotting the satellites of distant stars.

In search of more information about a known exoplanet orbiting the star HR8799 about 130 light-years from Earth, astronomers turned to the catalog of images Hubble has been amassing for more than 15 years. Using an algorithm that can block the bright light of observed stars allowing the much fainter exoplanets circling them to be seen, the team spotted the planet in an image from 1998.

The same technique could be used on 200 similar datasets from Hubble, as well an unknown number of archival images from ground-based telescopes...The Hubble data archive is a paragon of scientific efficiency. Its easily accessible and very search-friendly information architecture allows scientists to wring more discoveries out of the money we invest in the milestone space telescope.

and:

Other scientific experiments and tools have generated mountains of information, but much of that data is stuffed away in hard copy formats or on hard-to-access databases. Most of that data actually goes unanalyzed.

"The numbers that are typically quoted are three to ten percent of [scientific] data is analyzed," Fox told Wired.com in January. "That means almost none of it is reanalyzed. Is that a good investment in public funds? Probably not."

Space.com also has an article on the same subject here (and a few dozen comments below as well).

There are quite a few places where even non-astronomers can participate in sifting through masses of data such as Galaxy Zoo and Stardust@home.