The Telescope in the Ice

by Mark Bowen

  The University of Nebraska had contracted with the National Science Foundation to drill a hole through the two-hundred-foot ice shelf so that a group of scientists could study the ocean beneath. The drill was a so-called flame-jet, which is conventionally used by the mining industry to cut crystalline rock. It consisted of two 10,000-pound compressors, feeding air at a thousand psi to a modified jet engine—a huge Bunsen burner, basically—which was lowered into a roiling, water-filled hole, spitting out flame and partially combusted diesel fuel.

  “Well,” said Bruce, “it makes an awful racket … Lotta noise; lotta smoke; it’s real dirty … But it did drill through the ice shelf and relatively quickly. Provided a hole about eighteen inches in diameter, so the scientists were then able to lower their things down and do their experiments.”

  As with any engine that burns fossil fuel, however, the flame-jet produced prodigious quantities of carbon dioxide, which dissolved easily in the freshwater in the hole and the seawater beneath it, turning both into something like seltzer water. At one point during the drilling, a kink formed in the hose that delivered air to the drill, and when it un-kinked a large air bubble entered the freshwater column. This caused the pressure in the column to drop and the dissolved carbon dioxide to bubble out of the water like the fizz from a champagne bottle.

  “We had four thousand pounds of hose, a fifteen-hundred-pound drill on the end of it, and we were lowering it down the hole with a bulldozer. All of a sudden it quit going down the hole … and then it started coming back out of the hole … and everybody took off! The drill, the hose and everything came out the top, and then we had a geyser about forty feet high and four feet in diameter. And we got salt water! The fact that it was salt water really blew us away, because it meant that we’d bailed the whole hole.”

  So that was his first Antarctic experience. Over the next decade or so, he helped a glaciologist named Charley Bentley from the University of Wisconsin at Madison drill numerous short holes with a hot water drill—basically a glorified garden hose—in various parts of Antarctica, so that Bentley could drop charges of dynamite into the holes for seismic testing. He became adept at the sophisticated art of ice core drilling, in which a hollow, tubular drill with a smooth interior, threads on the outside, and sharp cutting teeth on the bottom is repeatedly lowered into the ice to carve out core segments and pull them up, one meter at a time. He drilled ice cores in Greenland and many locations in Antarctica, including the pole—and also found time to invent the field of high-altitude ice core drilling with Lonnie Thompson. This took him on at least four expeditions to high mountain glaciers in the Peruvian Andes and the Qilian Shan, the range that defines the border between the Tibetan Plateau and the Gobi Desert.

  At the beginning of the 1980s, the National Science Foundation made the dubious decision to turn the group that was carrying out the Ross Ice Shelf project into a small bureaucracy, which they named the Polar Ice Coring Office, or PICO (“pike-oh”). The organization remained at the University of Nebraska until 1989, when the contract was taken over by the University of Alaska, Fairbanks. Bruce and his understanding wife, Ann, moved with it.

  By the spring of 1990, when Francis Halzen and his colleagues met with the NSF officers in Washington, Bruce was pretty much at the top of his game. He had participated in about thirty remote drilling expeditions altogether and was arguably the most accomplished practitioner of ice drilling in all its forms on the face of the planet. He had spent two months the previous austral summer helping Lonnie Thompson drill an ice core on the high spine of the Antarctic Peninsula in generally appalling conditions, and he was about to make his tenth or eleventh trip to Greenland to help get the second Greenland Ice Sheet Project off the ground—or into it. GISP2 would succeed in recovering a two-mile-long ice core at the summit three years later.

  When PICO’s new director, John Kelley, walked into Bruce’s office and asked him if he’d be interested in helping a group of physicists drill some holes in Greenland in order to explore the possibility of constructing a neutrino telescope at the South Pole, he jumped at the chance. “Absolutely!” he remembered saying. “This is the neatest project I’ve ever heard of in my life! I’ll work at night if I have to and just not sleep.”

  * * *

  In August 1990, Bob Morse and a fourth Buford Price student named Tim Miller embarked for Greenland to conduct the first known ice fishing for muons—their fishing line consisting of three photomultiplier tubes that Morse had scavenged from Carlo Rubbia and Dave Cline’s failed HPW experiment.

  Tim and Bob flew to the summit in an LC-130 operated by the same wing of the New York Air National Guard that supports the science work in Antarctica. They were met in that high, cold place by Bruce Koci and Bill Barber, a tall, good-natured, unflappable, and incredibly strong Brit. This is where the physicists had their first encounter with glaciology—a field they would know better than they had ever wished by the time their telescope was built.

  The topmost, so-called firn layer of a glacier consists of opaque snow, which becomes increasingly dense the farther you go down. Below the firn, compression from the overburden and the passage of time have transformed the dense-packed snow into bubbly ice. Very deep in a glacier the individual air molecules in the bubbles migrate into the ice to form a crystal, known as a clathrate, in coordination with the ice molecules, and even the bubbles disappear. Ultimately, the array for a neutrino telescope must be placed in this lowest and clearest, bubble-free region, but since the goal in Greenland was simply to look for down-going muons, the bubbly layer just below the firn would do. Basically, since down-going atmospheric muons give off light, if all three of the phototubes were to light up within a very short time span—thirty billionths of a second or so—the odds would be very good that they had detected a muon rather than three simultaneous pulses of random noise.

  Bruce had prepared an ice core hole he had drilled the previous summer for Ellen Mosley-Thompson, Lonnie Thompson’s wife and research partner. Since ice core holes will collapse over time from shear forces in a glacier, he had reamed this one out to a depth of 217 meters, about a hundred meters below the firn/ice transition.

  The physicists dropped their fishing line into the dry hole and took their first series of measurements. Then they decided they wanted to increase the optical coupling between the phototubes and the ice, so they asked the drillers if they had any liquid on hand that they could pour down the hole. Unfrozen liquid is scarce on the summit of Greenland, but there happened to be a tremendous amount of butyl acetate on hand, for use in keeping the GISP hole from collapsing. They poured enough of it into the hole to cover the string and took another round of data.

  “I don’t know why we felt like we had to haul the string back up,” says Bob, “but we did haul the string up, and all of a sudden we saw this blue sludge all over everything.… The butyl acetate had completely dissolved the outer jacket of the wires, and it turned all the snow and all the liquid in the vicinity this beautiful purple-blue color.… We wondered if we had any light transmission at all, because everything seemed to be, heh.… And we took pictures of the tubes going down, and then we took pictures of the tubes coming up, and the picture of the tube coming up made it look like a nice big grape snow cone. [Bruce called it a “blue slushy.”] There was blue in my gloves, blue in my clothes, blue in my face.… It was the dye or whatever it was that’s normally in the cable.… Electrically, it worked fine. It was just, optically we weren’t sure what the hell we had done.”

  * * *

  Bob had begun suspecting that there was something strange about Bruce when he realized he was sleeping outside in an unheated tent, even though he was surrounded by numerous heated tents all over the summit encampment. On the day he, Bruce, and Tim flew out, Bob’s suspicions were confirmed. Two LC-130s left the summit that day. The one Bob and Tim rode in was heated, while the other was not, because it was loaded with ice core segments that had to be kept frozen. Bruce elected to ride in the unheated p
lane.

  “I thought Bruce was really insane then,” Bob recalled, some years later. “Now I know he’s insane.”

  * * *

  Back in Madison, Bob showed Francis the picture of the blue slushy, and Francis told him to hide it and never, ever show it to anyone again! He couldn’t have been all that serious, though, because not only did they not hide it, they showed it to their funding officer, John Lynch.

  Just enough data had survived the slushy incident. The blue dye had blocked out the light, but it had seeped into the surrounding liquid gradually, so that enough light had reached the detectors at the beginning of the run to demonstrate that the ice in Greenland was transparent to the blue light given off by traveling muons.

  Analyzing the data on the blackboard in Bob’s office, he and Francis came up with estimates for the transparency of the ice and the effective area of their phototubes for sensing muons. This area is larger than the geometric size of the tubes, because a muon doesn’t have to hit one in order to be detected. Since Cherenkov light can travel some distance in the ice before being absorbed or scattered, a muon could pass about a meter away and still be sensed. In later years, when they learned more about the ice, they would realize that these back-of-the-envelope calculations were off by a bit, but this was a detail. There was little doubt that they had detected down-going muons in polar ice.

  * * *

  By this time, Steve Barwick had become an assistant professor at UC Irvine (where there was a vigorous program in neutrino physics, thanks to Fred Reines) and had come up with the name AMANDA (Antarctic Muon And Neutrino Detector Array) for the new experiment. (Francis doesn’t like giving instruments female names; he thinks it’s sexist.) The fledgling AMANDA collaboration, consisting of Doug Lowder, Tim Miller, Buford Price, Andrew Westphal, Steve Barwick, Francis Halzen, and Bob Morse, submitted a letter to Nature, which was published the following September. Francis believes this “letter launched the experiment” by showing that the idea of using polar ice for neutrino detection “was still crazy but not that crazy.”

  One reads that “the hole was filled with butyl acetate, an organic liquid chosen for its low freezing point and optical clarity.” There is no mention of blue slush. “We find these results very encouraging, and are planning more extensive experiments at the South Pole during the coming austral summer.”

  This, warts and all, is how experimental physics is done. As Bob Morse writes,

  Greenland is a really beautiful example of an experiment quickly thrown together to meet a window of opportunity—mistakes were made—and where flawed or less than perfect data is also very useful, as failure can be when in the right hands.… This little pre-AMANDA experiment has all of the features of the later AMANDA and IceCube experiments. The later successes … were simply a matter of getting the bugs out of the deployments and data retrieval systems—not a trivial task.… This is a rare example where the funding agencies had more faith in the data (flawed as it might be) than many of the experimenters pushing on the DUMAND-in-ice idea.

  Francis adds that “it’s pretty clear we had no idea what we were doing, and so this was real research, right?”

  He suspected “that many people had had this idea, knew more about glaciology than I did, and obviously concluded it could never work.” “If we really had [known] what we were doing we would probably not have done it. And, in fact, it turns out that a lot of the things we should have known turned out not to be true.”

  In lectures to young scientists today, he sometimes uses the early days of AMANDA as an example of the dictum, “Don’t read books. Do things. There’s nothing better than to be ignorant and lucky.” (It goes over well.) It is usually the young, unaware of accepted knowledge, who make original discoveries. He believes that the only reason he managed to do something original in his late forties was that he was “like a young person again” in the sense of being naïve. “It’s only when you’re ignorant and you haven’t read all the books yet that you can do something original and new.”

  He was now taking a clear step into experimentalism, which is not ruled by the pristine logic of theory. Not only do numerous practical and strategic considerations come into play, oftentimes the wisest course is to stop thinking for a while and just do it.

  Had he read what was then the definitive textbook on the optics of water and ice, for example, he would have “learned” that the absorption length of blue light in pure ice—the distance over which about two-thirds of the light will be absorbed—was about eight meters. That would have been a show-stopper. They would have given the phototubes back to Cline and Rubbia and gone home. If Cherenkov light really was absorbed in that short a distance, it would take something like two million phototubes to fill a cubic kilometer of ice, and the tubes alone would cost about $6 billion. Luckily, the book turned out to be very, very wrong. They obtained an estimate of eighteen meters from the Greenland data, and even though it, too, turned out to be wrong, it was a step in the right direction.

  A few years later, when they were still struggling to understand the ice but had seen signs that the absorption length was actually much longer than even eighteen meters, the library in Madison mistakenly delivered the textbook to Francis’s office; it was meant for someone else. He started browsing through it, naturally, and when he came to the line about the eight-meter absorption length a chill ran up his spine.

  * * *

  So they were headed to Antarctica. John Lynch helped them again by finagling a way to get them started on the Ice with no need for a grant proposal. AMANDA became an official NSF project, with little need for direct money, because PICO paid for the drilling, travel to Pole was subsumed under the foundation’s huge Antarctic program (standard operating procedure), and Morse was going south for his gamma ray project, GASP, anyway. Madison, Berkeley, and Irvine kicked in small sums for the fabrication of two short strings of phototubes, and the scientists earned their salaries through other channels. Whatever cash was required came from GASP, which was about to be shut down anyway. Morse was still arguing for it, but Lynch kept telling him that “GASP was on its last gasp” and he ought to put his energy into AMANDA. (At around that time, the GASP group discovered a fatal design flaw in their instrument and realized it would never work.)

  “I’m not in favor of diverting funds to work on a NASA project or to send your daughter to college,” says John, “but I am in favor of diverting funds from an iffy idea to one that [has] a lot of promise.”

  * * *

  At this point, although they didn’t quite know it, AMANDA became a project in applied physics and glaciology. Their lofty goals in astro- and particle physics receded to the background while they taught themselves how to turn the ice at the South Pole into an enormous particle detector. It would take about ten years.

  The critical unknown was how deep to drill in order to get below the air bubbles, which would scatter the light from muons and erase all memory of their direction. And there were two obvious places to look for the answer: an ice core the Americans had retrieved in 1968 at the Byrd station in West Antarctica and the Russian core from Vostok, both of which had reached about 2.2 kilometers into the ice.

  Antarctica may appear to be one big chunk of ice when you look at it on a map, but it is actually divided in two by the Transantarctic Mountain Range. Vostok and Pole are located on the larger of the two pieces, the East Antarctic Ice Sheet, on one side of the range, while Byrd is located on the much smaller West Antarctic Ice Sheet, on the other. One could argue, therefore, that Vostok is more relevant to Pole. But the AMANDA scientists did the convenient thing and talked only to their fellow Americans.

  The expert on the bubbles in the Byrd core was Anthony Gow from the Cold Regions Research and Engineering Laboratory (CRREL), a division of the Army Corps of Engineers, located in Hanover, New Hampshire. His work suggested that the bubbles should begin to disappear about 800 meters down and vanish more-or-less completely at about 1,100 meters. Recall that in 1988 Moiseĭ Markov
had informed John Learned by telex that the bubbles at Vostok disappeared at 1,300 to 1,400 meters. This was another case in which ignorance may have been a virtue, for if the AMANDA scientists had realized at that early stage that they had to drill the better part of two kilometers into the ice, they might have lost heart.

  When he looks back, Francis sometimes wonders just what they were thinking. They resolved to drill the first holes to only a thousand meters, above which even Gow predicted the ice would have bubbles. Perhaps they compromised because Bruce Koci told them they’d have a hard time drilling even to that depth. Oddly, too, on their first few strings, they placed the detectors only a few meters apart—as if they had read the magical textbook. But they were basing almost everything they did on the pioneering work of the DUMAND collaboration, which had realized years earlier that the spacings need to be large in a Markov-type, plum pudding design. So, yes, what were they thinking?

  * * *

  The University of Wisconsin demonstrated its formidable capabilities from the outset. The Madison campus runs a facility in the farm country near town called the Physical Sciences Laboratory, or PSL, which builds large instruments for high-energy physics experiments and the like. Since PSL had some experience building hot water drills for glaciologist Charley Bentley, PICO contracted with PSL to build a drill for AMANDA.

  Although Bruce could do nothing to prevent this (nor would it have been his way) he realized that this arrangement allowed PICO to duck responsibility. They had very little skin in the game, so they low-balled it. They charged PSL with modifying what he called “an existing small drill that we had left over from the Crary Ice Rise,” a site on the Ross Ice Shelf where he had drilled several years earlier. Morse recalls that the specifications consisted of “a bunch of pencil scratches on napkins”—typical Bruce. They named the drill Bucky-1.

  * * *

  When you travel to the South Pole with the U.S. Antarctic Program you always pass through Christchurch, New Zealand, and McMurdo on your way south. And the only time you’d be doing such a thing would be in the Antarctic summer, which makes for an interesting contrast, since New Zealand will be experiencing spring or summer as you prepare to fly to the Ice.