At 3:22 p.m. Eastern time on January 21, 1968, Captain John Haug radioed the control tower at Thule Air Base in northwestern Greenland with a problem no pilot wants. His B-52G Stratofortress, call sign HOBO 28, had a fire on board. Within five minutes the extinguishers were empty, the electrical power was gone and smoke was so thick in the cockpit that the pilots could not read their instruments. The base was 90 miles to the north, and it was the middle of the polar night. Haug told his crew to get ready to leave the airplane.
They waited until the spare pilot, Major Alfred D’Amario, confirmed they were over the lights of the base. Then six men ejected into the Arctic dark. The seventh, co-pilot Leonard Svitenko, had given his ejection seat to D’Amario. He tried to bail out through a hatch in the lower deck and suffered fatal head injuries. The empty bomber flew on to the north, swung around in a wide left turn and, at 3:39 p.m., hit the frozen surface of North Star Bay about seven and a half miles west of the base.
In its bomb bay were four B28 hydrogen bombs, each rated at 1.1 megatons, roughly 70 times the power of the bomb that destroyed Hiroshima. None of them went off as a nuclear weapon. But the ordinary high explosive packed around their cores did detonate on impact, and in the fire that followed, plutonium and uranium were smeared across the ice like the contents of a giant dirty bomb. What followed was an eight-month cleanup in the dark and cold, a hunt for a piece of a bomb that was never found, and a political scandal in Denmark that took almost 30 years to come out.
A bomber circling a radar
HOBO 28 was exactly where it was meant to be. Since 1961, the U.S. Strategic Air Command had kept nuclear-armed B-52s in the air around the clock under a program later known as Chrome Dome, so that a Soviet surprise attack could never catch the whole bomber force on the ground. One of those bombers had a special job. Every day, a B-52 flew to a waypoint over Baffin Bay and then circled Thule in a figure-eight at about 35,000 feet. The Air Force called it the Thule monitor mission, or “Hard Head.”
The reason was the radar. In 1960 the United States had switched on a Ballistic Missile Early Warning System site at Thule, four steel antennas, each wider than a football field is long, that looked across the North Pole toward the Soviet Union. If the radar’s link to North American Air Defense Command suddenly went dead, the generals needed to know right away whether the site had been destroyed or a cable had simply failed. A bomber crew circling overhead could just look. That happened for real in November 1961, when an operator’s mistake at a relay station in Colorado cut off the warning network, and a B-52 near Thule confirmed the radar was still standing.
Thule was already a Cold War curiosity. About 150 miles to the east, the Army had built its nuclear-powered tunnel town, Camp Century, inside the ice sheet. By 1966 Defense Secretary Robert McNamara wanted to cut the airborne alert, since missiles had made it less important and ending it would save $123 million a year. The Air Force pushed back, and a smaller force of four alert bombers a day survived. One of them kept going to Thule, even after a B-52 collided with its tanker over Palomares, Spain, in January 1966 and dropped four hydrogen bombs on and near a Spanish village. Strategic Air Command decided that civilian officials in Washington did not have a “need to know” about the Thule assignment.
Four cushions on a heater vent
The fire that ended Chrome Dome started with seat cushions. The bomber belonged to a wing at Plattsburgh Air Force Base in New York. Before take-off, D’Amario had tucked three foam cushions on top of a heating vent beneath a seat in the rear of the lower deck, and another was stowed there after take-off. Hours into the flight the crew was cold, so D’Amario opened a valve to pull hot air from an engine into the cabin heater. A fault meant the air reached the ducts almost as hot as it left the engine, and soon someone smelled burning rubber. The navigator found the fire behind a metal box and emptied two extinguishers on it without success.
The crash itself was strangely gentle for a bomber, a shallow 20-degree dive into the sea ice. That did not help the bombs. The high explosive in all four weapons went off, flames climbed an estimated 850 meters into the dark, and smoke drifted toward the coast. An aerial survey afterward found only six engines, a tire and small bits of wreckage on a blackened patch of ice. The Pentagon gave the accident its code for a nuclear weapons accident that does not risk war: a Broken Arrow.
Dog sleds in the polar night
The survivors were scattered across miles of dark ice. Haug and D’Amario came down on the base itself. Because of the darkness, the cold and the broken ice, the search relied heavily on Jens Zinglersen, the Danish official at Thule for the Royal Greenland Trade Department, who organized local Greenlandic hunters and their dog sled teams. Three men who landed within a mile and a half of the base were found within two hours. The gunner, Staff Sergeant Calvin Snapp, was not so lucky. He came down about six miles out and spent 21 hours on an ice floe in temperatures around 23 degrees below zero, surviving by wrapping himself in his parachute.
Dr. Freezelove
Within hours the Air Force had activated a disaster control team under Major General Richard Hunziker. Its job had an official code name, Project Crested Ice, and an unofficial one that everybody used: Dr. Freezelove. Temperatures at the site averaged minus 40 degrees, a point where Fahrenheit and Celsius meet, and fell as low as minus 76 degrees Fahrenheit, with winds up to 89 miles an hour. Batteries gave out quickly, so technicians carried them inside their coats. The sun would not rise until mid-February.
By January 25 the team had drawn a “zero line” around a one-by-three-mile area where alpha radiation from plutonium could be measured. Lawrence Livermore National Laboratory sent two men with a brand-new radiation detector called FIDLER, built after Palomares for exactly this kind of emergency. They flew north on a KC-135 tanker that was also carrying 1,500 pounds of raw horsemeat for the sled dogs. At the site they found a wooden shack lit by Coleman lanterns. It soon grew into Camp Hunziker, with a heliport, igloos, huts, a decontamination trailer and two ice roads back to the base.
The Americans and the Danes saw the mess differently. U.S. scientists, including the Los Alamos plutonium expert Wright Langham, argued that the risk was tiny if the contamination melted into the bay and spread through its water. Danish scientists wanted “zero contamination.” About 200 Inuit families hunted birds, seals and walruses around North Star Bay, and the Danes worried that the fuel on the ice would melt into a plutonium-laced slick and wash up on Saunders Island, where the animals gathered in summer. One Dane, as Langham later told a closed session of Congress, put it this way: “Don’t dump your garbage in someone else’s doorway and walk away if you get caught.”
Denmark won the argument. Under a “gentleman’s agreement” reached in Copenhagen in February, the United States promised to scrape up the blackened ice and send it, and every bit of the wreckage, out of Greenland. Graders peeled off the top few inches of snow and ice, which was boxed, trucked to a staging area near the base called the Tank Farm and sealed into steel tanks. Airmen walking in long lines swept the ice for debris, from pieces as big as wings to bits the size of flashlight batteries. In all, about 550,000 gallons of melted, contaminated ice were shipped to the Savannah River site in South Carolina.
About 700 people from the two countries worked on the cleanup, often without proper protective clothing. Crested Ice ended on September 13, 1968, when the last tank went aboard a ship, at a cost of about $9.4 million. Hunziker reported that 93 percent of the contamination had been removed, a far more thorough job than the Navy had managed after Operation Crossroads, when radioactive ships proved almost impossible to clean.
The piece nobody found
While the ice was being scraped, a quieter problem was growing in Washington. Weapons experts were sorting the recovered fragments to work out which pieces came from which bomb. By late February they had the tritium reservoirs from the tails of all four bombs, but no positively identified parts of one bomb’s secondary, the thermonuclear stage of an H-bomb. A July 1968 analysis by the Atomic Energy Commission said the recovered secondary components amounted to 85 percent of the uranium and 94 percent, by weight, of three secondaries.
In August the Air Force brought in Star III, a small research submersible built by Electric Boat, to search the bottom of the bay, roughly 600 feet down at most. The Danes and the press were told it was a survey of the sea floor under the impact point. A State Department telegram spelled out the real purpose: “to assure that no visible and recoverable sensitive parts of the nuclear devices are on the bottom.” The dives were dogged by a broken manipulator arm, a faulty camera and a tangled tether. The search found plenty of aircraft wreckage but, in the words of an AEC report, “no parts of a weapon secondary,” and it was called off at the end of August.
Forty years later, in 2008, the BBC reported on declassified documents it had obtained under the Freedom of Information Act and concluded that the United States had abandoned a nuclear weapon beneath the ice. “It would be very difficult for anyone else to recover classified pieces if we couldn’t find them,” William Chambers, a former Los Alamos weapons designer who had worked on the accident, told the reporter. Denmark’s foreign minister ordered an independent review.
That study, by the Danish researcher Svend Aage Christensen in 2009, put its conclusion in its title: “There is no bomb, there was no bomb, they were not looking for a bomb.” All four weapons broke apart, he showed, and the hunt was for one component, the uranium-235 fissile core of a secondary, often called the spark plug. Holding perhaps 8 kilograms of uranium, it would have been a cylinder about 50 centimeters long, so Christensen called it the marshal’s baton. Nobody knows whether it survived the crash in one piece, and a Danish sonar survey of the bay in 2003 found only an old iron frame. The United States had chased secrets on the sea floor on a far grander scale in Project Azorian. At Thule it gave up after a few weeks.
The secret over Greenland
For Denmark, the most dangerous thing in the wreckage was political. Since 1957 the country had said it would allow no nuclear weapons on its soil in peacetime, including Greenland. Yet here was an American bomber with four H-bombs off Greenland’s coast. On January 29, 1968, the Danish government declared that there were no nuclear weapons stored in Greenland and no nuclear-armed flights over it, using wording suggested by U.S. Secretary of State Dean Rusk, while Washington stuck to its policy of neither confirming nor denying where its weapons were.
In secret, the two governments then added a supplement to their 1951 defense agreement. Settled in an exchange of notes on May 31, 1968, it said the United States would not store nuclear weapons in Greenland or fly over it with them, though U.S. negotiators added a spoken warning about a “grave and sudden threat.” Denmark had a strong hand, because Washington badly wanted to keep its radar at Thule. Most of those documents were published only in 2025, by the National Security Archive.
The fuller truth came out in 1995, when documents made public in Denmark showed that Danish Prime Minister H.C. Hansen had quietly given the Americans room to operate in 1957. In a famous letter, Hansen noted that the question of “supplies of munition of a special kind” had not come up in their talks, and said he had nothing to add. The Danish press called the scandal Thulegate. A two-volume official report in 1997 confirmed that nuclear-armed flights over Greenland had been routine and that the United States had stored nuclear weapons there until 1965. It also revealed Project Iceworm, the Army’s plan to hide hundreds of missiles in tunnels under the ice.
What the crash left behind
The cleanup workers carried their own questions for decades. Many Danes who had worked at the Tank Farm, the port and the motor pool later developed cancer. A Danish study in the 1980s found higher cancer rates among Crested Ice workers than among those who served at the base before or after, but could not show that radiation was the cause. In 1995 the Danish government paid 1,700 of them 50,000 kroner each. Their fight became a 2015 Danish film, The Idealist.
The plutonium mostly went where the Americans had predicted, into the bay, where repeated expeditions have found it buried in the seabed with little passing into the water or animals. Not all of it stayed at sea. Surveys by Denmark’s Risø laboratory from 2003 to 2008 found patches in the soil at Narsaarsuk, on the coast, about 100 grams of plutonium in all. In 2011 Danish health authorities concluded that even under extreme assumptions, the dose to anyone visiting the area was below 1 millisievert a year, about what a person gets from natural background radiation.
The accident also changed the bombs. Investigators concluded after Palomares and Thule that the explosives in nuclear weapons were not stable enough to survive an aircraft crash. Livermore and Los Alamos went on to develop insensitive high explosives such as TATB, which can take a fire or a hard impact without detonating; the weapons designer Ray Kidder later judged that the explosives at Palomares and Thule would probably not have detonated if they had been the insensitive kind. The airborne alert program, already shrinking, was ended almost at once. It was another hard lesson about how unforgiving nuclear materials can be, one that the scientists who handled the demon core had learned two decades earlier.
Thule is still there, under a new name. In 2023 it became Pituffik Space Base, and its radar still watches the polar sky for missiles. Greenland’s place in American strategy has rarely been so much in the news. But the bomber that circled the radar every day for seven years is a reminder of how close the Cold War kept its weapons to the people who lived beneath them. The hunters of North Star Bay never asked for four hydrogen bombs on their sea ice. For a few weeks in the winter of 1968, they were the ones who went out on their dog sleds to bring the crew home.
Further reading
Command and Control: Nuclear Weapons, the Damascus Accident, and the Illusion of Safety by Eric Schlosser
A Pulitzer finalist’s gripping history of how close America came, again and again, to nuclear disaster by accident, from a Titan missile explosion in Arkansas to bombers like HOBO 28 flying on airborne alert.
Strange Glow: The Story of Radiation by Timothy J. Jorgensen
A Georgetown radiation scientist explains, in plain language, what plutonium, uranium and radiation really do to the body, and how to think about risks like the ones the Thule cleanup crews faced.
GQ GMC-300S Digital Geiger Counter Radiation Detector
A pocket-sized, well-reviewed Geiger counter that reads beta, gamma and X-ray radiation in counts per minute and logs the data, so you can check background levels at home or on the road.
Dr. Strangelove, or: How I Learned to Stop Worrying and Love the Bomb (The Criterion Collection) [Blu-ray]
Kubrick’s 1964 satire about a B-52 on airborne alert gave the Thule cleanup its nickname, Dr. Freezelove. Criterion’s restored edition comes with interviews and documentaries about the film and the Cold War.