Recording A Century of Night Skies Through A Scanner Darkly

Telescope in Peru, 1897: The 24-inch Bruce doublet telescope, one prolific source of astronomical plates, was installed in Cambridge in 1893. It moved to Harvard Boyden Station in Arequipa, Peru, in 1896. This image shows Boyden Station in 1897, with the Bruce building in the center.  Harvard College Observatory

DIGITIZING HARVARD’S LEGACY

The plate collection, like Harvard College Observatory itself, looms large in the history of astrophotography. In July 1850, daguerrotype photographer John Adams Whipple captured the first-ever picture of a star, Vega, using the observatory’s mahogany-and-brass Great Refractor. By the late 1880s, observatory director Edward Charles Pickering had endeavored to photograph the entire sky, collecting photographs from the northern and southern hemispheres. The college shipped a 24-inch telescope to Arequipa, Peru, in 1896, and followed with telescopes in South Africa and other locations. During the next three decades, astronomers slipped glass plates into the observing tubes of their telescopes, making exposures of the entire sky, and then gathered them and shipped them back to Cambridge.

There they would be examined by “computers,” women recruited by Pickering because they were good at math, who scrutinized the plates through magnifying loupes and calculated stars’ brightness levels by studying their diameters. Leavitt, one of the more talented computers, discovered Cepheid variables in the Large Magellanic Clouds, which became crucial for determining astronomical distances. Without Leavitt’s discovery, Edwin Hubble could not have discovered the expanding universe.

“It’s such an old-fashioned way to do science,” Doane says. “Younger scientists, even middle-aged ones, don’t know how to look at the plates. They don’t have the developed skill to look at something that is the size of a point, and see that has changed in size by a very small amount to understand that it’s getting brighter.”

Modern computers are doing the work automatically, calculating stellar magnitude once the plates have been scanned. Grindlay used this data to make his Type Ia discovery, he says. He noted a 14th magnitude star (a very dim star, several orders of magnitude dimmer than the faintest stars you can see with the naked eye) that quickly brightened to 12th magnitude (two orders brighter, but still relatively dim). This brightening happened in an unusual manner, with a rapid increase and slow decay over a decade’s time, Grindlay says. Astronomers checked it against star catalogs and determined it was an M giant, a big red star, in a binary system. So why did it produce this bright flash? A white dwarf is orbiting the red giant, accreting mass from its larger companion, Grindlay explains. The mass accretion ignited thermonuclear burning on the surface of the smaller star. And this happens to be how Type Ia supernovae are born.

“This is something that we astrophysicists have been looking for, for years,” Grindlay says. “What we think we have found, thanks to 100 years of data, and one star doing this amazing thing, is a very likely long-sought mechanism for how you add mass to a white dwarf.”

Other astronomers are using heritage plates for different types of new science. Elizabeth Griffin, an astronomer at the Herzberg Institute of Astrophysics in Victoria, Canada, has been dusting off near-ultraviolet negatives from 1905 up to the 1980s to study the ozone layer. She examines spectral lines, in which starlight is split up into its constituent parts so astronomers could determine the star’s composition. The atmosphere imprints its own signature on the negatives, and in some cases it interferes directly with astronomical measurements. Ozone, for instance, gets in the way of the element beryllium, which can be studied to determine a star’s age.

“[For several decades], people had hardly heard of ozone. They thought it was an awful nuisance,” says Griffin. Astronomers determined the ozone layer’s spectral signature and learned how to filter it out; now Griffin is looking for that signal and ignoring the rest, trying to determine how the ozone layer’s thickness changes over time. “I did learn that ozone is very variable from night to night and place to place,” she says. “It was a novel application for heritage data.”

Still, the Harvard collection is much larger, and is the only collection to represent both hemispheres of the sky. It will help astronomers in several fields — perhaps even astronomers using the Kepler Space Telescope, for whom light-curve information is especially crucial. Grindlay says the so-called Kepler field, an area between Vega and the northern cross in Cygnus, is almost done being scanned. Laurance Doyle, an astrophysicist with the SETI Institute who used Kepler data to find the first circumbinary planet, said light curves can teach researchers plenty about stellar history. “There’s almost an endless amount of stuff you can do with a light curve. Just using a light curve, you can detect circumbinary planets three ways,” he says. “The time realm has been under-explored in astronomy.”

Who knows what else is in those slides, just waiting to be found.