All That Does Not Glitter
Astrophysicist Throws Light on Dark Matter

Gazing at the heavens, men and women have long been stirred by the swarm of lights processing solemnly through the "fabulous, formless darkness" (Yeats). Renaissance playwright Christopher Marlowe wrote of the evening air "[c]lad in the beauty of a thousand stars." In modern times, astronomer Carl Sagan enlightened the public on the romance and science of "billions and billions" of stars that inhabit what is, in essence, a night universe.

Darkness, it turns out, makes up a vastly greater portion of the universe than the lights that enchant us. Today, scientists who study the star fields of space agree there is far more out there than meets the eye, but they really don't know what all that non-star stuff is. They call it dark matter—dark to the eye, dark to the mind that hungers to understand it.

we all go into the dark
"We've discovered that we don't know what the universe is made out of," says Charles Alcock, the Reese W. Flower Professor of Astronomy and Astrophysics. "As a purely scientific question, it's one of the most compelling of the age."

Lay persons are apt to characterize the cosmos as mind-boggling and immeasurable, but astronomers and cosmologists who strive to comprehend it, who try to roll the universe into a ball and hold it in their minds, are about the business of measuring and counting up all the contents. Scientists may not know what the dark matter is, but they do have a pretty good idea how much of it there is. "It appears that the universe is probably 99 percent dark matter," Alcock claims, "or possibly 'dark energy.' And there is between 10 and 100 times more mass in the Milky Way [our galaxy] than we can actually see."

The reason he knows this has to do with observations and measurements of the Milky Way and what Einstein's theory of General Relativity tells us about mass, motion, and gravity.

To the eye, a galaxy is a giant configuration of countless stars that spins across the cosmic night. The Milky Way's galactic wheel turns much faster than it ought to, given its mass of stars, which astronomers have taken the trouble to count. The sum of their masses cannot account for the gravity that keeps the spiraling star structure from flying apart, so scientists believe there has to be more mass somewhere—a lot more. The star counters have summed up the mass of the whole visible universe and come up short in that equation too.

Theorists argue among themselves about what all this shadowy stuff could be. "There are many different ideas about what the dark matter might be made out of," Alcock explains. Candidates range from a black-out blizzard of particles, each thousands of times smaller than an electron, to black-hole leviathans that could swallow the Earth as though it were plankton. "Each idea can be tested observationally in different ways," he adds. "We set out to test the idea that, whatever the dark matter is made out of, it's made out of massive objects—objects that weigh as much as a planet or a small star."

macho man
Alcock came to Penn last year after 15 years with the Lawrence Livermore National Laboratory, where he headed the Institute of Geophysics and Planetary Physics. As a physics major at Auckland University in New Zealand, he read about the discovery of pulsars (energetic radio stars) by a Cambridge graduate student in the late 1960s. "It was like an adventure story," he recalls, "and I thought, 'That's what I want to do: I want to become an astronomer and make discoveries.'"

At Livermore, he started talking with a physicist friend about doing an ambitious sky survey that had been first suggested by a Princeton scientist but dismissed as impossible by most everyone in the field. "We wanted to do something unusual, a little risky. Most of our colleagues thought we were crazy. People basically thought it wouldn't work because it involved making measurements of millions of stars every night for years at a time. The technology didn't exist to do that. We looked at it and said, 'You know, I bet we could do that.' So we started the MACHO Project."

The acronym stands for MAssive Compact Halo Objects. Scientists theorize that one explanation for the Milky Way's rapid rotation is a halo of dark matter just beyond the reach of the galaxy's spiral arms. The invisible halo might be made of things like planet-size bodies or burnt-out carcasses of earlier generations of stars. The Princeton paper maintained that MACHOs, if they are there, could be seen indirectly by their "gravitational microlensing effect."

Einstein showed that the fabric of space is warped around massive celestial bodies and that light passing through the gravity-wrung space gets bent along its contours. If a MACHO passed between a background star and an observer on Earth, the dark object's gravitational field would act like a lens, bending the starlight and making the star shine brighter. One of the biggest problems for detecting this phenomenon is that, at any given moment, about one star in maybe two million is being lensed. The project would need to measure—and remeasure for years—spectacular quantities of stars just to have a chance at catching a lensing event and "seeing" a few MACHOs. "We estimated that 'enough' meant exceeding the total number of photometric measurements made in the history of astronomy by two orders of magnitude," says Alcock.

Ask him why he thought he could pull off such an unlikely undertaking, and he pauses briefly before replying with a precise and modest and matter-of-fact British accent: "Hubris."

The solution came from Ronald Reagan's Star Wars initiative. Working at Livermore, Alcock knew about the startling new technologies that could track lots of fast-moving objects. With collaborators from around the world, the MACHO team devised an innovative camera system linked to powerful computers. At the time, the digital cameras were the largest ever built for astronomical observation. The imaging system was mounted on a large telescope in Australia.

"We were able to image and measure the brightnesses of as many as 600,000 stars at a time," says the dark-matter hunter. For almost eight years, the MACHO Project searched the sky for signs that something invisible was interposing itself between the telescope and 69 million tiny pricks of light.

In 1993, the project reported its first sighting of gravitational microlensing. It was the first indisputable discovery of dark matter. Since then, the group has registered over 400 lensing events. "We got very good after a while," notes Alcock, "but the interpretation of those events is still controversial.…If all of the dark matter in the Milky Way were made out of MACHOs, we know how many events we should see, and we definitely saw fewer events than that. So we know that MACHOs do not make up most of the dark matter in the galaxy. That's a firm result." He estimates that as much as 20 percent of the galaxy's dark matter is made of MACHOs, but some scientists contest even that percentage.

Last year, the American Astronomical Society awarded Alcock its Beatrice Tinsley Prize for his "innovative and original work." The citation lauded the MACHO venture as "one of the most challenging astronomical projects ever undertaken." This year Alcock, the project's lead investigator, was elected to the National Academy of Sciences, one of the highest honors accorded to American scientists.

on the cusp
The MACHO Project is perhaps a small step toward solving the what of the dark-matter question, but it's a giant step toward pioneering how astro-science will be done in the future. The observational phase is complete, but analysis of the database generated by the lensing survey is still underway.

"MACHO is almost eight terabytes," Alcock notes. "Projects that I'm getting into now are going to collect five times as much data, so they'll be 40 terabytes. Projects that we start a few years from now will collect 1,000 terabytes of data." That's about 16 quadrillion bits of information, which takes us well beyond Sagan's billions and billions. Comparing these immense datasets to stars in the sky is no longer just a literary device; the claim is fast approaching a statement of fact.

Over the last two decades, computers have brought about back-to-back revolutions in astronomical science. The first occurred in the mid-80s, when researchers would go to their telescopes to collect data and then bring it back to a desktop workstation for analysis. Innovative software allowed them to process more and new kinds of information in highly productive ways. The follow-up revolution was driven forward by the MACHO Project and others like it, which had automated the collection of data. "Now," says Alcock, "you collect so much data that the scientist, even working at a workstation, can't possibly look at it all."

To quantify the vast distances spanned by the cosmic void, astronomers draw on metrics—the light year—that use time to measure space. Alcock uses time to elucidate the vast dimensions of these unprecedented datasets. "Supposing a scientist were trying to look at our data by bringing it up on a computer screen and analyzing it. How long would it take one scientist to get through all of the data on the MACHO Project? It would take several thousand years—and it's only going to get worse."

Gleaning the heavens for points of light and putting all the digitized observations into a storehouse of data points has created a cyber universe filled with astronomical discoveries waiting to be found. Alcock now wants to sift through existing datasets, and generate new ones, looking for planets outside the Solar System. His new study will probe the cyber universe for the "transit effect," the dimming of a distant star when one of its planets passes in front and blocks the light. "What we're trying to do at Penn," says the cyber astronomer, "is look at all the other science that can be done with these data. In some sense, it's like developing a new instrument for a telescope—we're developing instruments for operating the cyber telescope."

The seat-of-the-pants code writing done by the MACHO scientists yielded software that was passable for the purposes of that project, but for other modes of analysis, it presents formidable barriers. Alcock is now spearheading the Cyber Universe Survey Project (CUSP) at Penn, which will search for new planets and also map the dark matter in the universe by measuring the "weak lensing" of galaxy clusters in deep space. Penn cosmologist Bhuvnesh Jain leads the dark-matter mapping project. CUSP will also attempt to perfect the data-processing and management technology that will provide the tools for the exploration—often called virtual observation—of the fast expanding cyber universe.

"We got into CUSP because we want to do the kind of science that this technology allows," he explains. But a major part of the effort will go towards developing the information technology to count, collate, crunch, compare, and calculate the densely stippled data massed inside the cyber universe. That experience should teach the CUSP team how best to design new and more powerful surveys.

"The technology," he prognosticates, "will probably turn out in the long run to be more influential [than the science], if we're successful—which I'm sure we will be."

What makes him so sure?

"Oh, probably hubris," shrugs the MACHO man.