Johns Hopkins University

Spring 2008
Vol. 5, No.2

FEATURES

>Amateur Hour

Inspired Intersections

The More Things Change...

A Decade of Discussion

adjust type size + -

Amateur Hour

In a modern twist on the centuries-old tradition of the backyard astronomer, thousands of galaxy gazers around the world have plugged in to become keepers of the Galaxy Zoo.

illustration

Jan vandenBerg is a geeky guy, complete with the skinny glasses, the sexy hardware, and the impenetrable tech speak. As an information technology manager for the Bloomberg Center for Physics and Astronomy, he runs a "web farm" of "competent iron" —a rack of expensive computer servers that keep dozens of astronomy websites up and running.

This humming computer bank can crank out a fair amount of heat, depending on how much work the machines are doing; more people surfing its sites means more heat. But on a normal day the computers are only working at about 10 percent of their full capacity, so he doesn't worry too much.

On July 10, 2007, though, vandenBerg was awakened in the middle of the night by the urgent beeping of his laptop. His "snazzy web service gear" at the Bloomberg Center had just gone completely dead. He hurried to the Homewood campus, where he opened up his hardware to find melted wires and a burned-out fuse. "It totally blew my mind," he remembers. Something had massively overheated the system.

The culprit turned out to be a cute new website called Galaxy Zoo, a small side project vandenBerg was hosting as a favor to a friend-of-a-friend at Oxford University. In one day, it had generated thousands of times more traffic than vandenBerg's academic outfit usually sees in a month: 15 million hits, a number that only makes sense for sites like AOL or Google. Even as he patched in a couple of new "nodes" (computers) to handle the crazy load, vandenBerg couldn't believe the computer logs. How could this goofy astronomy site generate such a surge in its first day? It had to be a glitch in the system, he thought, maybe a loop in the code or a bug that was multiplying hits.

Better Than the "Telly"

On the other side of the world in England, where the sun was shining, Alice Sheppard clicked her mouse and sent one of those 15 million hits racing toward vandenBerg's doomed computers in Baltimore. The traffic wasn't a glitch; it was an invading horde of British amateur astronomers.

Sheppard was a student at the time, training at the University of Sussex to become a teacher, and she had always had a casual interest in astronomy. She had heard about the Galaxy Zoo site at a talk given by one of its creators—Chris Lintott, an astrophysicist at Oxford University. Lintott and his fellow astronomers, she had heard, were trying to sort millions of newly discovered galaxies into categories. The galaxies had been photographed by a small telescope in the U.S.—at Apache Point, N.M.—as part of the Sloan Digital Sky Survey, the largest effort in history to photograph the night sky (and a collaboration involving Hopkins researchers). But there were simply too many galaxies and too few astronomers to look at them. So the pros created Galaxy Zoo, an experimental website designed to train the public to help them.

Eager to help the cause of science (and to put off doing her schoolwork), Sheppard opened a web browser and pulled up the site. It was pretty, a great big, high-resolution image of a bluish galaxy plastered onto the background, framed by transparent gray and orange text blocks. She clicked on "About the Science" and learned that galaxies are a bit like fossils for astronomers, cosmic specimens living in different times in the universe's history. Sorting galaxies into categories gives astronomers a starting point to study how these celestial objects form and evolve over time.

A 10-minute tutorial taught her how to classify a galaxy. Some galaxies, she learned, are elliptical, shaped like fuzzy footballs. Others are spiral—like whirlpools—with twirling streaks that spin either clockwise or counter-clockwise. Her task seemed simple enough; the shapes were easy to recognize. So she fired up the site's galaxy analysis tool, which splashed a fuzzy image onto her screen and asked her to click a button identifying it. It was like playing a computer game, and it was surprisingly fun.

"I've never been all that good at computers," she says. "I didn't use the Internet all that much before this." But on that first day, Sheppard spent four hours at the computer and clicked through hundreds of galaxies. "I much prefer it to watching the telly," she says.

illustration of galaxy gazers

Part Science, Part Eye Candy

Sheppard wasn't alone in her fascination. Galaxy Zoo was quickly picked up by BBC Online News, becoming the second most e-mailed story of the day (narrowly beating out an article titled "Garlic may cut cow flatulence"). It spread to the Associated Press, Slashdot (a source for technology-related news), and the blogosphere as a whole. Between June and September, drawn by this publicity, more than 118,000 members of the public visited Galaxy Zoo—not just Brits, but people from across the globe. Collectively, this army of astronomy enthusiasts has looked at 40 million images and classified more than a million galaxies (each galaxy is verified by at least 30 people). It's one of the largest experiments in "human computing" in history, a vast collection of human brains networked together like a giant computer.

It’s one of the
largest experiments in “human computing”
in history, a vast
collection of human brains networked together like a giant computer.

For some, like Bill Bliss, the site has become a part of their daily routines. Bliss is a retired helicopter pilot for the Navy. He lives in rural Colorado, where he helps his wife raise and train sheepdogs. "I'll wake up in the morning and feed the sheep," he says. "Then I grab a cup of coffee and watch the news. When I get tired of the news, I sit down at the computer and classify a hundred galaxies or so."

For others, like Els Baeten, it's an addiction. By day, Baeten works as a secretary in a Belgian insurance company. But when she goes home and the sun goes down, she pulls up Galaxy Zoo and gets to work. She has no background in science, but she loves the idea that her efforts are helping to further real research. "In the beginning," she says with a laugh, "I would spend five or six hours a night classifying. There were nights when I'd promise myself I'd go to bed early, and I'd be up until one or two in the morning." To date, Baeten has personally classified 200,000 objects.

Hopkins astronomer Alexander Szalay, one of the original builders of the Sloan Digital Sky Survey, knows well the passion—and advances—that amateur astronomers have brought to the field over the centuries: tracking asteroids, discovering new comets, sighting supernovae. But even Szalay, Alumni Centennial Professor of Astronomy in the Zanvyl Krieger School of Arts and Sciences, didn't expect the site to be such a hit. "No one predicted it," he says. "But then again, could you have predicted the popularity of Google or YouTube?" Szalay wrote the software that plucks the images for Galaxy Zoo out of the Sloan database and, though he is known for his database wizardry, it took him nearly a month and a half to troubleshoot the problems created by the unexpected traffic. He speculates that the colorful site appeals both to those nostalgic about science and to computer gamer types. "It's a balance between science and eye candy," he says.

Jordan Raddick, an education and public outreach specialist in the school's Department of Physics and Astronomy, agrees: "There's no single thing that explains why people do it." He's working on a survey to figure out what drives the obsession. And to help fan the flames and draw in new volunteers, Raddick is recruiting other astronomers to translate the site into French, Portuguese, and a variety of other languages.

Like artists at a gallery show or fishermen mounting their catches, the zooites proudly post the "most gorgeous
ellipticals" or "sexiest edge on spirals" they've come across.

The mania and mix of motivations come out on the site's online forum, maintained by a particularly devoted subset of volunteers who call themselves "zooites" (they've nicknamed the professionals "the zookeepers"). In the "Science Questions" section, for example, zooites ask questions about astronomy ranging from basic ("What is space?") to the most intricate and the esoteric ("Is our galaxy secretly merging with a dwarf galaxy?"). And the responses they get—both from other amateurs and from the pros—are some of the most polite and even-tempered posts you're likely to find on an Internet chat forum. It's an unusual form of science education on the web, driven purely by the curiosity of the public.

In the "Stunning Sights!" section, though, science takes a backseat to aesthetics. Like artists at a gallery show or fishermen mounting their catches, the zooites proudly post the "most gorgeous ellipticals" or "sexiest edge on spirals" they've come across. More imaginative users can also visit the "Weird and the Wonderful" and try to find celestial objects shaped like animals or letters of the alphabet—cloud watching on a cosmic scale.

Over a Pint at a Pub

Like many ideas in the world of science, Galaxy Zoo began with an overworked graduate student and the dissertation that destroyed his social life. Near the end of 2006, Kevin Schawinski, a PhD candidate in astrophysics at Oxford, spent most of his waking hours searching for a particularly rare kind of galaxy: the blue elliptical. Blueness in a galaxy means that new stars are forming. But football-shaped elliptical galaxies are generally red, filled with old dead stars. So the blue elliptical is a paradoxical—almost mythical—object that could give astronomers clues about how stars form within galaxies.

Schawinski had countless Sloan Digital Sky Survey images—billions of photographs of the cosmos, 1,500 gigabytes of data originally managed by the same servers at Hopkins that now host Galaxy Zoo—but he didn't have a good way to search through them for his rare ellipticals. He couldn't use computer software; pattern recognition algorithms have trouble making sense of the complicated details and variability in the fuzzy galaxies. But the human eye is really good at this kind of pattern recognition, so Schawinski built a little program to display images of the galaxies for him on his computer, one by one. "I literally spent all day, every day clicking through galaxies," he recalls. "It's not something I recommend." After one week classifying 50,000 galaxies by type—elliptical or spiral—he decided that, for the sake of his sanity, he needed some help.

Meanwhile, in another part of the Oxford astrophysics building, postdoctoral fellow Kate Land was also looking for something strange in the heavens. She studies the cosmic microwave background, a glow in the universe left over from the Big Bang. Current theories about the Big Bang predict that this radiation should be relatively uniform. But Land had found a strange pattern in this glow, an asymmetry in the cosmos that was weird and confusing and threatened her reputation as a respectable astronomer. She called it the "axis of evil" and had started looking around for other findings that might support her results, which were just on the edge of believability.

Working with Slovenian astronomer Anze Slosar, she came across a small paper, largely ignored by the scientific community, about spiral galaxies. Spiral galaxies can spin either clockwise or counterclockwise, but, according to modern theories, the direction shouldn't matter—it should be random, a 50/50 split either way. But Japanese scientists had counted 3,000 spiral galaxies and found a strange asymmetry: more counterclockwise than clockwise. Maybe an unknown force was at work in the universe, thought Land and Slosar, lining up both the cosmic radiation and the spinning galaxies. What they needed was a larger group of spiral galaxies and someone to count them.

The connections between Schawinski's blue ellipticals and the spirals of Land and Slosar turned out to be a pub, a pint, and a mutual colleague, Chris Lintott. Over a bit of beer one night, Lintott told them about Stardust@home, a neat NASA project that had faced a similar problem of too much data and no good way to count it. The Stardust mission had flown around the solar system for seven years, collecting samples of dust in a special kind of gel. Some of those bits of dust, scientists guessed, might have come from other solar systems—but only a tiny fraction in the countless bits collected. So they started up a website asking amateur astronomers to look at slices of dust-laden gel and identify unusually shaped particles.

Lintott suggested a similar approach to their galaxy troubles. "If they'll look at ugly bits of dust, we can certainly get them to look at these beautiful galaxies," he told them. Maybe they could recruit a few thousand amateur astronomers to sift through all those photographs of galaxies and help classify them by shape (for Schawinski) and spin direction (for Land and Slosar) over the course of the next few years. It could be a fun side project.

The young Brits and the Slovenian didn't have any money, so they started calling in favors through a network of high-profile friends that included Bob Nichol, a British astronomer who has appeared on The Late Show with David Letterman, and Brian May, astronomer and former guitarist for the rock band Queen. They found volunteers in Silicon Valley to build their web interface, and then they came to Szalay at Hopkins to find a home for the site.

An Astounding Amount of Data

illustration

"With all of the volunteer labor, I think that the whole project has cost us around $32," says Lintott. But this shoestring budget has produced an astounding amount of real data and led to four new research papers that came out this spring.

Kevin Schawinski now has an impressive group of rare blue ellipticals to study: 2,000 in all. But he's not the only one who has tapped the human brain's ability to pick out strange objects. Bill Keel, an astronomer at the University of Alabama, asked Galaxy Zoo volunteers on the forum to keep an eye out for galaxies that are far away from each other but happen to overlap in the photos. He's using the responses to study how galactic dust turns into stars.

Then there's the Dutch volunteer Hanny, who noticed an unusual blue blob in one of the images. But when she posted the image and asked the Galaxy Zoo team what it might be, they had no idea—it was unlike any object that astronomers have ever seen. They call it "Hanny's Voorwoop" (Dutch for "object") and are working on pointing more telescopes at it to figure out what it could be.

Kate Land and Anze Slosar were looking for an asymmetry to the universe, and they found it—instead of equal numbers of clockwise and counterclockwise galaxies, the project found a larger number of counterclockwise galaxies, just like the earlier Japanese study. But they still had their doubts, so they tried flipping all of the galaxy images (which reverses the direction) and had volunteers reclassify them. If the bias was real, now they should find more clockwise galaxies. But, once again, the numbers showed more counterclockwise galaxies. So the team believes that they may have stumbled upon some strange phenomena that deals with how our brains perceive spirals. They're now talking to psychologists to see what it could mean. "I suspect that Galaxy Zoo was not about astronomy at all," one volunteer jokes online, "but [was actually] a giant psychology experiment."

For each of these new research findings, the Galaxy Zoo team is giving credit where credit is due; they are listing all 118,000 volunteers as co-authors on every paper they publish (by including a link to a site with all the names). Oxford's Lintott and Hopkins' Raddick have become a two-man recruiting team, looking for new ways to reach out and find volunteers to plug into their human computer.

It's a 21st-century update to a centuries-old concept; amateur astronomers have rediscovered their role in advancing professional research. If you, too, would like to become a published scientist, it's not too late: visit www.galaxyzoo.org.

 

Devin Powell is a master's candidate in the science writing program of the Writing Seminars.