by Mark Bowen
As the realities of ocean engineering and cost reared their heads, the array diminished in size. (One section of Roberts’s history is named “The Incredible Shrinking Array.”) The collaboration came up with MICRO, MINI, and MIDI DUMANDs, the smallest consisting of only nineteen strings.
In its wisdom, the Department of Energy had funded only a feasibility study. In 1982, the department turned down a follow-on proposal for a thirty-six-string array, and offered to fund just one string, which would hang from the deck of a ship rather than sit on the ocean floor. Seven years into the project, in 1987, the collaboration finally deployed this so-called Short Prototype String, comprising seven optical modules, and ran what was basically a glorified version of the experiments Learned had run for his Ph.D. thesis twenty years earlier. They took a total of thirty-eight hours of data at five depths between two and four kilometers and measured the drop in the intensity of down-going, atmospheric muons the deeper they went. They also produced a rough profile of the muons’ angular distribution. DOE deemed this enough of a success to justify the funding of a nine-string array named DUMAND-II in 1989.
* * *
The Russians, meanwhile, made steady progress. At about the time DUMAND was funded, the Soviet Academy of Sciences dedicated a laboratory at Moscow’s Institute for Nuclear Research to the Baikal project and named as chairman Grigorii Domogatsky, the theorist who had handed out rubles in the night to the hippies from America.
He and his lead experimentalist, Leonid Bezrukov, took a conservative approach, beginning with their choice of site. They were dealing with a shallower body of water to begin with, only about seven-tenths of a mile, but they also chose a spot only two and a quarter miles from shore—about a tenth the distance to DUMAND. They would be plagued by more down-going muons at this depth, but the scale of their project would be manageable. Also, and crucially as it would turn out, for the two months of the year that the lake would be frozen over, they would be able not only to walk on their experiment, but drive heavy cranes over it.
Science is still seen as a noble calling in Russia, and Baikal, in its time, was probably the most soulful experiment in physics. Its field headquarters are still located in an abandoned station on a dead-end spur of what used to be the Trans-Siberian Railroad.
There are no roads to this spot. In summer it can be reached only by train or boat. In winter, when the lake is frozen, one can drive across the ice from the town of Listvyanka on the north side of the Angara River, about thirty miles northeast. Until 1997, when their German collaborators installed a satellite link at headquarters, there was no internet. A 1930s vintage railroad telephone provided the only connection to the outside world.
The researchers live in small wooden cabins heated with wood stoves, which burn mainly birch. There is no running water. There’s a wood-heated sauna, furnished with a hand-operated water pump, and every Saturday night the residents gather for the weekly banya, or bath. The whole community strips down and bathes, in shifts, pouring water on the heated stones in the sauna to make steam. Many nights there is singing and guitar playing, and while the food does not receive high praise, generally speaking, delicious bread is baked on the premises daily.
Deployment takes place in March and early April, when the ice is at its thickest. Holes are cut in the frozen surface, the detector strings are pulled up from the bottom of the lake for repairs, tests are run, new hardware is added, and the telescope is re-committed to the depths. The last step is to gather the cables that have been used to lower the strings, attach them to a buoy, and release the buoy into the water, where it will hover safely below the surface through the warm months while the instrument takes data. (Data is transmitted to shore by an electrical cable, which snakes along the bottom of the lake.) At the start of each deployment season, divers in dry suits drop into a hole and swim down to the buoy with a line that is used to pull the strings up.
The scientists commemorate the release of the buoy each spring with a ceremony that borrows something from traditional Russian funerals. One person stands on the ice next to the deployment hole with a bottle of vodka and a small shot glass as the others file past. One at a time, they fill the glass, pour a few drops into the hole for Burkhan, the god of the lake, and drink the rest. This solemnity is followed by a final banya and a party. (I once asked Ralf Wischnewski, one of the German collaborators, whether they take a dip in the lake before the sauna. “Some do, but only by mistake,” he responded.)
Christian Spiering once showed me a scrapbook of photographs from one of these closing ceremonies. In one, the group poses naked and grinning in the snow by the sauna. In others they stand on the flat ice, with brown, wooded hills in the distance, surrounded by heavy machinery and large spools of cable, or lie on their bellies on the ice, peering through boxy, homemade, handheld “goggles” into the black water. They wear wooly winter clothes and the quintessential Russian fur cap with folded up brim and earflaps, as they wrestle with electronics or hack at computers inside box-like structures that resemble ice-fishing huts. In one picture, Grigorii Domogatsky sends a piercing look toward the camera, under a fur cap with dangling earflaps, looking rugged and wiry even in his winter clothes. Lines of sharp intelligence radiate from his eyes, and a Papyrossi, the loosely rolled Russian cigarette, dangles from his mouth.
* * *
The Russians proceeded methodically. In 1981, they ran preliminary tests with individual phototubes. The next year, they lowered a detector nearly to the bottom and measured the natural luminescence of the lake water—a source of spurious light that limits the ultimate sensitivity of the detector. (This confounding background does not exist in ice.) In 1983—four years ahead of DUMAND—they lowered the equivalent of a Short Prototype String from their stable ice platform and took a few days’ worth of data. The following year they placed a string on the bottom, connected it to shore, and detected their first atmospheric muons. They suffered a couple of mishaps that year, however. There was a leak in the buoy that held the string vertical, so that it sunk to the bottom over the course of about fifty days, and ten days after that a lightning strike fried the electronics.
In 1998, at a workshop on computer simulations for neutrino telescopes that was held in Zeuthen, Germany, I spoke with the leader of the Baikal group at the University of Irkutsk, a warm, soft-spoken man named Nikolaij Budnev. Drawing sketches in my notebook, he described the ingenious method they had used to recover the string that had sunk and been struck by lightning.
They designed a device that looked something like a window fan. It held a propeller inside a small frame that was fitted with fins, canted at an angle so that the unit would “swim” in spirals when it was dropped into the water. The autumn after they lost the string, they sailed a research vessel out to the spot where they had lowered the string the previous April. Attaching a flexible cable to their invention, they switched on the propeller and lowered it into the water by means of a winch. As their mechanical fish swam into the depths, the cable traced out the path of its spiraling descent, and as it approached the bottom it wound the cable around the collapsed string. The scientists then reversed the winch, the cable knotted itself around the string, and they pulled it to the surface.
The stark contrast between the simple, low-tech approach of the Russians and the complicated, high-tech approach of the Americans probably explains why one succeeded and the other failed.
In 1986, the Baikal collaboration produced its first physics. They set a limit on the existence of magnetic monopoles using a string equipped with improved buoys and protected against lightning. (The monopole, first postulated by Dirac in 1931, is an exotic particle whose existence has yet to be confirmed: a subatomic magnet with only one pole. In theory, if such a creature exists, it should produce a huge signal if it happens to pass near a Cherenkov detector. This was one reason the Baikal group thought it worthwhile to search for monopoles with just one string.)
These early studies resolved basic concerns about the feasibilit
y of the lake for a kilometer-scale instrument, the most basic of which had to do with its shallow depth. Since more down-going muons punch through to this depth, it’s harder to pick the up-going muons from the down-going haystack.
In 1987, the year the Americans deployed their Short Prototype String and got the go-ahead for DUMAND-II, the Soviet Academy of Sciences approved Baikal as a long-term research project and increased its funding. The following year, the Russians received a great shot in the arm when a group led by Christian Spiering from the East German Institute of High Energy Physics, in Zeuthen, just south of Berlin, agreed to join the collaboration.
A race was on.
* * *
Nineteen eighty-seven was a watershed year for neutrino astronomy, and, quite unexpectedly, the biggest splash was made by UNDINE, the shell-type Greisen design that wasn’t supposed to have much promise astronomically at all. She had been wooed from “her sunless and solitary abode” about ten years earlier, for an entirely different purpose.
Back in 1974, two theorists from Harvard, Howard Georgi and Sheldon Glashow, had proposed one of the first credible theories of grand unification, the quest to unify the four fundamental forces of nature. This is a lofty goal. Albert Einstein spent the last thirty years of his life trying to unify two of them, electromagnetism and gravity, and failed more or less completely.
Georgi and Glashow’s theory, known as SU(5) in physics parlance, was an attempt to unify the weak force, the strong force, and the electromagnetic force; it ignored gravity. The name is shorthand for “special unitary group of degree five,” a term from group theory, which is a branch of mathematics that makes elegant use of all forms of symmetry and is one of the most powerful and aesthetically pleasing tools in modern physics. Wolfgang Pauli employed group theory, for example, in proving his theorem of CPT or “strong reflection” symmetry in the 1950s.
It took a few years for the implications of SU(5) to sink in. The most surprising was its prediction that the proton, which had hitherto been thought to be absolutely stable, might actually decay after some unimaginably long time—or, as Glashow once put it, that “diamonds might not be forever.” This would also have the larger implication that all the matter in the universe might eventually dissolve into a vast sea of elementary particles. By 1979 (the same year in which Glashow, Steven Weinberg of Harvard, and Abdus Salam of Imperial College, London, shared the Nobel Prize in Physics for their previous success in unifying the weak and electromagnetic forces), the new theory had been used to estimate the proton’s lifetime at somewhere between 1027 and 1032 years. This poses little threat to the quality of your diamond ring or your health and well-being, in case you were wondering, since even the lower bound works out to about 100 million billion times the age of the universe. Amazingly, however, the entire range was within reach experimentally.
“This was accessible,” says John Learned. “If it was right, we would have found it with a low-energy detector. So we pulled the [UNDINE] designs out, and that’s why there’s this interesting tie.”
Over the next few years, several people who had attended the seminal 1976 DUMAND workshop in Hawaii submitted proposals to build UNDINE or Greisen-style detectors to search for proton decay.
“Now, Saburo Miyake claims that he told the design for that detector to [Masatoshi] Koshiba, who then proposed the Kamioka detector in Japan,” Learned continues. “So, the amusing thing is that DUMAND, or the idea for generic DUMAND, was in some way the mother or father of this dual chain of experiments that goes on ’til today.… So, that’s kinda neat. I like to point that out to people.”
* * *
Ahead of the game, as usual, Dave Cline held the first-ever meeting on proton decay in Wisconsin in 1978. Fred Reines, who attended the meeting, took his eggs out of DUMAND at that point and put them in this basket.
In May 1979, Reines’s group in Irvine joined forces with groups from the University of Michigan and Brookhaven National Laboratory to propose a 10,000-ton detector in the shell-type or UNDINE style, two thousand feet deep in a Morton-Thiokol salt mine on the shores of Lake Erie. They named their collaboration IMB, after the names of the institutions involved.
As with neutrino detection, the size of the detector was key: if the proton had a lifetime of 1032 years, then 1032 protons, which is roughly the number in a hundred tons of water, should produce an average of one decay event per year, and a 10,000-ton tank should produce a hundred events per year. Theory held that the proton would decay most often into a positron and a neutral pion, which would fly off in opposite directions. The positron would send a cone of Cherenkov light toward one wall of the detector, while the pion would swiftly decay into two gamma rays, each of which would produce electron-positron pairs, and these would produce multiple Cherenkov rings on the opposite wall. This characteristic light pattern would serve as a signature of proton decay and help distinguish this kind of event from atmospheric muons passing through the detector from above (which were minimized by placing the detector underground) and from the occasional neutrino interaction inside the tank. Neutrinos, in other words, were a hindrance to this experiment; they added to the background. At one point Reines compared the teasing of the signal from the background in IMB to “listening for a gnat’s whisper in a hurricane.”
The Japanese project, led by Masatoshi Koshiba, was named Kamiokande, the Kamioka Nucleon Decay Experiment. It was located one kilometer deep in a zinc mine owned by the Kamioka Mining and Smelting Company, in the center of a mountain in the Japanese Alps. In the end, it was smaller and therefore less sensitive than IMB: 3,000 tons versus 8,000. IMB also got off to a faster start.
But Dave Cline, Carlo Rubbia, and Jim Gaidos of Purdue (another Madison alum) beat them both to the punch with a smaller, easier-to-build experiment, which they placed two thousand feet deep in a silver mine in Park City, Utah. Like IMB, this effort was named for the institutions involved: Harvard, Purdue, and Wisconsin. All of Cline’s colleagues in CCFMR participated in HPW at one point or another … including Bob Morse, who would later become the principal investigator of AMANDA. In fact, this was Bob’s introduction to cosmic ray physics. He would carry the DNA and even some of the equipment from HPW into AMANDA. (He also carried the lineage of Kenneth Greisen. For Bob’s doctoral adviser, William D. “Bill” Walker, had obtained his Ph.D. at Cornell under the great man. “That’s important, you know, your lineage in this business,” says Bob. “So Morse is traceable to Walker, traceable to Ken Greisen at Cornell, and Greisen, of course, walks on water.”)
“In typical Carlo/Dave fashion,” Bob continues, “they were going to build a small, cheap and dirty detector, about eight hundred tons of water, and if the proton decay lifetime was hovering around ten to the thirty-two [1032] years, we would’ve nabbed it. It was a high risk/high payoff experiment. We took a chance on it, and we missed.”
Part of the reason they missed was that their detector had a serious design flaw. The phototubes weren’t placed only in a shell around the detection volume, they were placed inside too, and there were mirrors in there as well. This was supposed to help catch every bit of light from a decay event, in order to measure its total energy, but since it also made it nearly impossible to reconstruct Cherenkov light cones, all directional information was lost.
I once asked Bob if HPW had done anything useful. He thought for a long time … “I don’t think so,” he finally said, “other than figuring out that there’s a lot of goddam thorium in that mine. Thorium is radioactive, hah! hah!… Between the wire chambers and the mirrors on the inside of the tank, the reconstruction was a mess!”
None other than Sheldon Glashow pronounced the experiment “an unmitigated failure.”
* * *
Neither HPW, IMB, nor Kamiokande ever detected a single credible proton decay. (Maurice Goldhaber from Brookhaven once remarked that IMB “had some candidates, but they weren’t elected.”) They took a step forward, however, nevertheless.
On a philosophical level, it i
s impossible for science to prove that something does not exist—“absence of evidence is not evidence of absence,” as Carl Sagan used to say. But if you run an experiment and obtain a null result you can use the sensitivity of the experiment to estimate just how unlikely it is that whatever you were looking for does exist or will occur. So, while these experiments did not find proton decay, they still made progress by setting new limits for the minimum lifetime of the proton.
Limit-setting is a common game in particle and astrophysics. We will see a lot of it when AMANDA and IceCube begin generating data. The code for the setting of a new limit is to use the word search in the title of your journal article, meaning you have searched but not found. You would use evidence or observation if you had.
Limit-setting can also be used to rule out theories, and in fact, IMB and Kamiokande set high enough limits for the proton’s lifetime to rule out SU(5)—a significant contribution. Then, sure enough, SU(5) was extended by supersymmetry, and the theoretical limit for the proton’s lifetime was bumped up to 1036 years—beyond the reach of those experiments. Kamiokande was then replaced by the 50,000-ton Super-Kamiokande, or Super-K, which holds the present record of 1034 years. (“The Kamiokande people have just beat that one to death,” notes Morse.) But an UNDINE-style detector would need to weigh in at about a million tons to cover the next two orders of magnitude and put supersymmetry to the test. So there’s a Hyper-K in the offing … and so it goes.
A proton decay experiment also becomes more sensitive the longer it runs. IMB and Kamiokande started searching for proton decay in 1982 and 1983, respectively, and ran all through the eighties.
This was a marvelous coincidence, because in 1987 a near-miracle occurred: the brightest supernova in almost four hundred years.