The Telescope in the Ice

by Mark Bowen

  “Now PICO’s come a long way.… The relationship is much, much better. But in that first season, it was awful! And, ah—”

  “So were the holes,” interjected Bruce.

  “Hah! Yeah. And so were the holes. The holes were terrible. They were just crooked as hell.… Hole one drifted off as much as six or seven meters from where it was supposed to be.”

  * * *

  A stellar addition to the field team that year was a talented and hardworking experimentalist named Albrecht Karle, who had recently moved to a post-doctoral position at Zeuthen after earning his Ph.D. at the University of Munich. Albrecht’s adviser had been Eckart Lorenz, one of the great pioneers of gamma ray astronomy, and his thesis research had been conducted on an air Cherenkov telescope high on a mountainside in the Canary Islands. This type of instrument employs the atmosphere as the Cherenkov medium, rather than water or ice, and is more reminiscent of the usual telescope in that it sits on the ground looking up.

  The season got off to a slow start because of trouble with the drill head. Every time they dropped it into the water-filled hole it would short itself out after fifty or so meters. “It was the first week of December, second week, third week; it was Christmas, and we still hadn’t any holes drilled,” Albrecht recalls. “[The deployers were] getting really antsy. One was behaving so weird that Pat Mock [who was again heading up deployment] sent him off station.… People had nothing to do.”

  The institutional arrangement with PICO made it difficult to address the problem, as the drillers reported to NSF rather than the scientists, so the situation was ripe for finger-pointing. Steve Barwick, the senior scientist on the Ice, advised Albrecht not to step in to help, since he might be blamed if he failed, but Albrecht demonstrated his leadership skills by teaming up with an electrical engineer from PICO to find a solution. It was already January.

  Once the drilling got started, Jones relegated Bruce to the background, and in this difficult position Bruce did what he could to keep the drill running and prevent a true disaster, such as freezing it in. He got very little sleep during the drilling of a hole, which might take a week, and the deployment, which came right on its heels.

  Since the holes were deeper this year, a couple of the deployments bumped up against the forty-hour deadline when the water in the hole would refreeze. During one such nail-biter, Bruce was wandering around in his tennis shoes on one of his spot checks when he noticed the main cable for the optical modules skipping on its reel as it was being lowered into the hole. It hadn’t been spooled tightly enough, so it jumped periodically to release the slack. He ended up standing by the reel for seven hours, shifting from foot to frozen foot, cushioning the skipping cable by hand.

  Albrecht saw right away that Bruce was the only driller who was interested in “the final purpose.” “He would always point out, you know, the hole is only good when the string is safe. He was really a different, but also in many ways a very unique and marvelous and amazing person.”

  Also in contrast to his PICO counterparts, Bruce kept thinking about the drill in the off-season. It turned out that more than the usual number of optical modules failed during refreezing that year. A few months after the season ended, he was puzzling this over in his office in Alaska, when he noticed that most of the failures occurred at two specific depths: one at the very bottom of the strings, and the other about eight hundred feet above it—the precise length of a hose segment. At the higher of the two depths, the drilling had stopped while the drillers changed reels. “The light went on,” and he realized that the failures must have been caused by changes in hole diameter: When the drillers had stopped to change reels, more than the usual amount of hot water had been released at that particular depth and the hole had developed a bulge. At the bottom of the array, on the other hand, the holes had tapered off: As the head worked its way down into the ice, the plume of hot water it left in its wake would continue to melt the ice fifty or sixty meters above it. And since they had stopped and pulled the drill up from a point just below the spot where they intended to place the lowest module, the hole was narrower there. He realized that they needed to drill sixty to seventy meters below the bottom module in order to ensure a constant diameter for the entire string.

  Bruce and Bob’s hand-waving explanation for these failures was that irregularities in the diameter would cause uneven shear pressures to develop as the water in the hole refroze. This would produce a net force on Benthospheres that housed the phototubes, which were essentially large corks, or pistons, bobbing in the hole. The spheres would move, however slightly, and they would snap their connections to the fixed cable. In subsequent years, the drillers avoided changing reels in the middle of the array and drilled seventy or more meters below the bottom module, and the number of failures dropped dramatically.

  * * *

  They deployed the string that was intended to be in the center of the pentagon first, and then began working their way around the perimeter. By the time they had three strings in the ice, they got the laser flashers up and running and began looking into the quality of the ice. The burning question, of course, was whether they had gotten below the bubbles.

  Ironically, it turns out to be more difficult to attach hard numbers to the optics of a relatively clear medium than a medium with strong scattering. A few students back in the real world applied crude Monte Carlos to the problem, Bob and Per Olof did some hand calculations on the Ice, and it soon became clear that the scattering dropped dramatically between one and two thousand meters. Buford was right! The bubbles were gone! Joyful e-mails flew north, and Francis called John Lynch to pass along the good news.

  But it wasn’t all good news, because the flasher data also told them that there was more scattering in the deep ice, indicating more dust than Buford had predicted. This meant light wouldn’t travel as far as they had hoped, and there was a possibility that their array might be too sparse. They had deployed the first three strings in the shape of a triangle, between sixty and eighty meters from each other, but they now estimated that the attenuation length of the light in the ice, absorption plus scattering, was only twenty or thirty meters. If they persisted with their original design, they ran the risk of leaving “holes” in their detector: the light from a passing muon might not be able to travel the full distance between strings, and this would make it difficult if not impossible to reconstruct its direction.

  The delays in the drilling also led to another consideration: if they kept to their original plan and continued to work their way around the perimeter of the pentagon and then ran out of time or the drill broke down, they would end up with a weirdly shaped, open-sided array that would also complicate reconstruction.

  On the fly, then, as Bob Morse sent e-mails north pleading for a decision and as the crane lumbered out to set the tower for what would turn out to be the final hole of the season, the collaboration decided to place it in the center of the triangle. This produced a nice, solid detection volume with strings close enough together to prevent blind spots.

  This four-string configuration became known as AMANDA-B4, and with it AMANDA edged ahead of Baikal—equipment-wise anyway—for about two months’ time. In April, the Baikal collaboration caught up by adding a fourth string to their array. And so, by the spring of 1996, there were two teams in the exclusive four-string club of neutrino astronomy.

  At the Venice conference in February, straight on the heels of the Pole season, as it always was, Buford Price announced triumphantly that that bubbles were “absent below 1,500 meters,” while Grigorii Domogatsky demonstrated just how far ahead Baikal was altogether by presenting their reconstructions of the first-ever up-going muons, detected more than a year earlier with their three-string array. These results were tentative, since three short strings aren’t really up to the task of identifying up-going muons unequivocally, but Domogatsky nevertheless suggested that he and his collaborators may have detected one or two actual neutrinos.

  * * *

  John Lea
rned spoke on “The Many Experimental Problems Connected with the UHE [Ultra-High-Energy] Detectors,” and in retrospect his title is fraught with meaning.

  The Division of High Energy Physics at the Department of Energy had recently formed the Scientific Assessment Group for Experiments in Non-Accelerator Physics, aka SAGENAP (“sage nap”), to review its commitments in particle astrophysics. And in addition to evoking an image of a group of wise men dozing off, the committee’s name indicates just what the high-energy physics community thought of this fledgling field. The lion’s share of the money in DOE’s high-energy physics budget was and still is spent at the accelerators, and most of its physicists worked at them then and still do; thus everything else was designated condescendingly as “non-accelerator physics.” Admittedly, several particle astrophysicists, including Tom Gaisser and Steve Barwick, sat on the committee, and its reviews were helpful, but the SAGENAP exercise still highlights the fact that neutrino astronomers, air shower scientists, and other cosmic ray physicists have lived off the crumbs from the rich man’s table ever since the accelerator era began. Some, and especially Learned, rail about this, but Francis believes it’s a blessing in disguise. Lower budgets tend to keep you out of the spotlight, and until recently, most cosmic ray physicists wouldn’t have known what to do with $1 billion anyway.

  SAGENAP held its first-ever review meeting in Washington, D.C., a few days before the Venice conference. Learned stopped off on his way to Europe to deliver a status report on DUMAND, and it didn’t go well. The committee would not release its report until spring, but he must have seen the writing on the wall.

  We last looked in on the project during the disastrous deployment attempt in 1993, when the buoy at the top of their one and only string had failed to release from its sacrificial anchor and the arched string had stopped giving signals about two minutes later. A junction box on the sea floor had failed two months after that. Sometime in 1995 the collaboration had attempted to repair it with an undersea robot—and failed again.

  “The overall impression was that of a collaboration encountering a never-ending series of technical problems,” the SAGENAP committee wrote. “As each was understood and fixed another seemed to arise.” The committee noted “a general lack of rigid quality control, and a freewheeling style toward new ideas and improvements.”

  “Essentially every member of the SAGENAP group expressed the view that the present DUMAND management had displayed a lack of detailed planning and careful attention to quality assurance deemed essential for the success of a project like DUMAND. This fatal failure was laid squarely on the project leadership, whose managerial skills and commitment to the project were strongly deprecated. Many reviewers said that the DUMAND ‘Project Director’ should have been a strong Engineer/Manager of the NASA project variety.”

  The leader of the project was Learned. Many felt that the reviewers went overboard in their criticism of John and that some of their comments were unnecessarily personal. They hurt him deeply.

  And then came the judgment.

  “Further support of the DUMAND project is not recommended.… The project should be terminated, and its operations phased out in an orderly manner.”

  Thus the pioneering project in high-energy neutrino astronomy died, and its leading visionary, John Learned, was essentially banished. He has participated in—but not led—several important “non-accelerator” projects in the meantime, but they have all been in the low-energy arena.

  Nevertheless, Learned is still the pied piper of neutrino astronomy. “His heart and soul are with neutrinos, and he’s been remarkably steady in that love for over thirty years,” says Bob Morse. John probably has the broadest view of non-accelerator neutrino physics of anyone alive, and he continues to keep his finger on the pulse of neutrino astronomy. Twenty years later, he is a friend to IceCube and welcomed into their closed collaboration meetings. Francis Halzen says John “was really the shining light in that collaboration, and then, you know, he got blamed.”

  With DUMAND’s demise, neutrino astronomy became a two-horse race. (Although, how serious could it be with the Germans playing on both sides?) In April and May 1996, just after it was installed, the new four-string array in Siberia detected the first-ever “gold-plated,” that is, irrefutable, up-going neutrinos. One could say that this made it the world’s first working neutrino telescope, although it was far too small to make an astrophysical discovery: the three neutrinos in Baikal’s 1996 data sample pretty much had to have been atmospheric neutrinos, born on the far side of the planet rather than in the cosmos beyond.

  12. Glory Days

  The children have built something out of an orange crate, something preposterous and ascendant.

  —JOHN CHEEVER

  Bob Morse remembers these as the glory days of AMANDA: It “was very frustrating at times and very demanding and caused you a lot of pain and anguish, but, you know, it was also a really wonderful, happy time in, I think, many people’s lives.… It’s hard to be bored when you’re scared witless.” With all the hard work they did year-round and the adventures and good times they shared at Pole, the collaboration coalesced into a kind of family. Many of those who traveled to the Ice for the 1996–97 season went on to play key roles in the crucial developments of the next several years, and a good percentage are still with the project two decades later.

  Their group of forty was small for the task at hand; they were pushing the boundaries on every front. Their anarchistic “war party” approach was probably the only way they could have done what they needed to do.

  * * *

  While drilling and deployment were the most arduous tasks, they faced another challenge in the need to invent an electrical technology that could sense the minute impulses from single photons as they hit detectors more than a mile deep in the ice, carry them to the surface, and determine when the impacts occurred to within several billionths of a second. All the detectors in the instrument—in IceCube there would be thousands—also needed to be calibrated to at least the same timing accuracy relative to each other. This was applied physics again. The invention of an adequate technology would take about five years, as they came up with new gadgets to test in the ice pretty much annually.

  Per Olof had tested the first new idea on the final string of AMANDA-B4, the one that had been dropped on the fly into the middle of the triangle.

  So far, they had been relying on the standard cable used in physics labs in the real world, a so-called coaxial cable, to deliver the high voltage that drove the photomultipliers and to bring their signals back to the surface. But coaxials are thick and inflexible, so this limited the number of optical modules they could put on a string. At the bottom of string four, as it was called, Per Olof arranged to add an extra six optical modules, driven by “twisted pairs”: two thin strands of insulated copper wire twisted into a spiral or double-helix. Twisted pairs are thinner and more flexible than coaxials, and it was obvious the minute they were tested that they also improved signal quality. They never used coaxials again.

  But twisted pairs were not a magic bullet. When a photon hits the active surface of a photomultiplier tube it produces an asymmetrical pulse that rises to a peak in about two billionths of a second (nanoseconds) and decays back to zero in roughly another four nanoseconds. The twisted pairs were better than coaxials at carrying these pulses to the surface, but they still distorted them significantly. In a dispersive effect analogous to the one that distorted the laser flashes in the bubbly ice of AMANDA-A, the original six-nanosecond pulse from a phototube would be stretched out to about three hundred nanoseconds. This made it difficult to detect two photons arriving within, say, ten nanoseconds of each other, which often occurred, and it also made it hard to pin down the moment when a photon hit a detector, which, again, they needed to know to within a few nanoseconds.

  In the off-season after AMANDA-B4, Albrecht Karle came up with an option that promised to eliminate dispersion altogether. His idea was to use a fiber-optic rath
er than an electrical cable to carry the pulse to the surface. A light emitting diode would be attached directly to the electrical output of the photomultiplier. It would produce a light pulse that had the same shape as the electrical pulse, and it would be coupled right there to a fiber-optic that ran to the surface, where the light pulse would be converted back into an electrical signal. Since virtually no dispersion takes place in a fiber-optic, the shape of the original pulse would be preserved.

  * * *

  Meanwhile, with some help from Buford Price, Francis Halzen’s criminal optimism had worked some magic on David Nygren of Lawrence Berkeley Laboratory. Dave joined AMANDA in 1996.

  He had been watching neutrino astronomy from a distance ever since his grad school days at the University of Washington, when John Learned was also a student there. He followed DUMAND “as a kind of spectacle,” he says. “That’s too strong a word or too derogatory, but it was very heroic.”

  Dave is a tall, rangy man, now approaching eighty, with wispy, often shaggy, silver-white hair, who could be mistaken for a perpetual grad student in his button-down shirt and jeans. Although he’s a quiet presence at collaboration meetings, Dave can change the thinking in a room with one or two sentences. He has made a career of inventing what he calls “disruptive technologies” or game-changers: he likes to pull rabbits from hats. And he’s something of a legend. In the mid-seventies, he revolutionized particle physics with an invention he named the time projection chamber.

  One of the leading problems of the day was the difficulty of observing the many particles that are produced simultaneously in the collisions of very high-energy particles in accelerator beams. The state of the art was to do various tricks with two-dimensional images, similar to photographic plates, but these techniques were limited to about ten emerging particles. The time projection chamber made it possible to analyze as many as ten thousand particle tracks at the same time—a disruptive technology indeed, and it remains state of the art a quarter of a century later. Pretty much solely on the basis of this invention, he was elected to the National Academy of Sciences in 2000. Among other things, he has also invented an imaging technology for mammograms that reduces the risk of breast cancer by reducing the amount of X-rays needed to produce a good image by a factor of three or four.