By James Glave
Pretty pictures will get you only so far," says George Djorgovski, the Caltech astronomer leading the Palomar Digital Sky Survey. "To do anything scientific, you need data."
Using the Palomar Observatory's 48-inch telescope and artificial-intelligence software codeveloped with the Jet Propulsion Lab, Djorgovski and his team have created 3,000 photographic plates, each containing roughly one gigabyte of data and collectively representing almost all the northern sky (astro.caltech.edu/~george/dposs/dposs_pop.html). The resulting 3-terabyte database will contain not only the location and brightness of several billion objects, but dozens of other celestial variables, such as size, ellipticity, and orientation.
The project's ultimate goal: to develop advanced data-mining tools to pluck pearls out of the binary void. "You want to have a more intelligent way to pick the few objects you really want," says Djorgovski. "It may be a few Web sites, or a few galaxies."
ELECTRIC WORD
Search Engine of the Stars
