The Antarctic wind near the bleak Adelie Coast last week was blowing at 46 mph. The temperature was hovering at 13 degrees below zero -- forbidding weather for most fliers.
But to pilot Henry Perk, this was the kind of "break" he needed to get airborne in a place where the weather is often brutal. Mr. Perk had been hired to fly his twin-engine De Havilland Twin Otter to a remote spot 140 miles inland from the French coastal station at Dumont d'Urville, where a Johns Hopkins University solar telescope had parachuted to Earth after flying for more than 19 days beneath a helium balloon 25 miles above the ice.
The telescope had survived despite a lost antenna, a wrong turn over the Ross Sea and a storm that struck after it landed. Now it was Mr. Perk's job to find it, retrieve its data tapes and a computer drive containing its scientific findings, and fly them back to scientists.
There was little time. The storm had ended, but the Antarctic winter was fast approaching, threatening to ground aircraft for months with temperatures of 100 degrees below zero and winds of 100 mph and more.
It was yet another extraordinary challenge for a project launched in one of the world's most difficult scientific outposts.
"You're not down there for a vacation or to go see the penguins," said Dr. David M. Rust, 55. An astrogeophysicist at Hopkins' Applied Physics Laboratory (APL), he is the project's principal investigator. "It's a tough place to be, and nobody in his right mind would do it unless he's a scientist determined to get his data."
Called Flare Genesis, the $16 million unmanned observatory had been launched Jan. 7 from McMurdo Station, the United States' Antarctic outpost on the edge of the Ross Sea. It was developed by APL and the Air Force. Its 32-inch mirror is the most powerful ever flown for solar research.
Dr. Rust's goal was to use the Antarctic summer and the remote continent's peculiar winds to conduct the longest, most detailed observation of the sun ever.
Record of solar flares
He wanted a 10- to 14-day photographic record of solar flares and gigantic magnetic storms -- the titanic outbursts of solar energy that can disrupt communications and electrical distribution on Earth and threaten the lives of astronauts in space.
The Antarctic summer offered 24 hours of uninterrupted sunlight, and the balloon's high altitude placed it above 99 percent of the Earth's atmosphere.
And because high-altitude winds blow in a circle around the South Pole, balloons will usually circle the continent and return close to their launch site. There, scientists can signal the payloads to cut loose from the balloons and parachute to Earth with their data for convenient recovery.
But no expedition to the planet's harshest outpost is hazard-free. This one was no exception.
Weeks before the project's scheduled Dec. 20 launch, high winds literally blew away a temporary steel and fabric building at McMurdo where the telescope was to have been readied for flight.
The ground crew -- funded by the National Science Foundation (NSF) -- quickly built a replacement, this time of wood. But it meant a two-week delay, and time was running out. If the telescope were not launched by Jan. 10, the onrushing winter weather would make recovery of its data impossible. And the NSF would insist on pulling the team out before transportation became impossible.
It was Jan. 7 before the nine-member Flare Genesis team could complete launch preparations. They filled the balloon with 28 million cubic feet of helium -- enough to loft the 3,000-pound telescope to 125,000 feet.
But just as the payload was about to lurch into the sky, a technician walked up to chief engineer Kim Strohbehn with a foot-long tube in his hand.
"He comes up and says, 'Here, this fell off,' " Dr. Rust recalled. "He [Mr. Strohbehn] looked and said, 'Oh my God.' "
It was an antenna. Scientists needed it to sample the astronomy data being recorded on board, and to keep the telescope sharply focused on the sun. It had apparently shaken loose during transport to the launch site.
Somebody suggested trying to reattach it, but there wasn't time. The balloon was rising, and the payload was about to follow. There was no stopping it. An unlucky technician might find himself yanked into the stratosphere.
"Oh gosh, we were just terribly depressed," Dr. Rust said. The telescope had been focused prior to shipment from the United States, but even temperature changes could throw it off.
There was one other possibility. The telescope can be focused in flight, sort of, through a backup system that requires line-of-sight radio communications.
Several days after the launch, the team climbed aboard a four-engine LC-130 Hercules cargo plane and flew until they were cruising at 12,000 feet, beneath the balloon. From there, they were able to check on the observatory for about three hours. They also tried to focus the telescope.
"It's slow and cumbersome," Dr. Rust said. "We did it, but I'm not sure whether the images were ever in focus."
There was another glitch. They found that the telescope's computer had hung up. A day's data were lost, he said, "but we got it going again."
The team returned to McMurdo and waited, following the balloon's flight with reports from a satellite system. And with growing alarm, they watched as the polar winds began to defy experience and carry the balloon north toward the ocean and disaster.
"I thought it was gone," Dr. Rust said. "We never dreamed it would head out over the ocean." "It got out over the Ross Sea well off the coast of Antarctica," he said. But then it turned west again and flew back over land.
At McMurdo, the team rushed back to the Hercules and took off. Minutes later, they sent the signal that cut the payload loose. In 40 minutes, the telescope had parachuted to the ice south of Dumont d'Urville.
The Navy pilot found the landing site, but it was a region of "sastrugi" ice -- ice that has been blown into 2- and 3-foot waves. He tested the surface with a daring touch-down, but the ice was too hard and threatened to tear up the plane's landing skis. He aborted the landing.
It would be another week before the weather would clear enough for Henry Perk and his Twin Otter to attempt it.
A private contractor, Mr. Perk flies in the Antarctic summer, then moves to the Arctic to fly during the northern summer. Somehow, he managed to land.
There, in the wind and cold, he and a crew member repaired a landing ski broken on the ice. Then they labored for two hours with wrenches to remove the telescope's data systems, which scientists hope will contain as many as 120,000 images of the sun.
Mr. Perk also noted, before flying out, that the telescope's costly 32-inch mirror had survived the landing. "That means probably the payload is in good condition and can be retrieved next October or November," Dr. Rust said.
The data tapes and the hard drive are now en route back to the United States via New Zealand.
Dr. Rust arrived home earlier after surviving a blown tire, an engine failure, and then an emergency landing during a commercial flight from New Zealand. He hopes to get his first look at his Flare Genesis data in several weeks.
Pictures and more information are available on the Internet at http://sd-www.jhuapl.edu