Wennberg and Andersons' research addresses the fundamental question of whether Boeing, or any firm, can build such vast fleets of supersonic aircraft without doing irreparable damage to the lower ozone layer.

To complete their study, Wennberg and Anderson flew instruments in a converted U-2 spy plane 20 kilometers (70,000 feet) above the earth's surface. Taking part in the Stratospheric Photo chemistry, Aerosols and Dynamics Expedition [SPADE], the researchers were able to make many measurements simultaneously. This procedure is "unprecedented" according to Ross J. Salawitch, a research associate in Atmospheric Sciences and a co-author of the paper.

"Most of the nitrogen and hydrogen radicals are produced by natural activities," says Salawitch. "But the halogen radicals, primarily bromine and chlorine, are primarily due to industrial activity."

The U-2 jet was designed to reach extremely high altitudes for reconnaissance during the Cold War, and was converted to become the NASA ER-2, a major part of the SPADE program.

But Anderson acknowledges a weakness in their research. "We only have comprehensive measurements from 13 to 20 kilometers," Anderson says. "The region above is unexplored in hard detail. We know there is a point above 20 kilometers where nitrogen radicals control the depletion of ozone."

The scientists will be unable to determine exactly where this reversal occurs using NASA's ER-2 jet.

"The ER-2 is limited to a ceiling altitude of 71,000-72,000 feet [22 kilometers] unloaded," says James R. Podolske, an atmospheric scientist at the NASA Ames Research Center. "To go higher, you either have to go faster or have a lighter plane."

But some researchers, including Wennberg and Anderson, have been able to take partial measurements at greater altitudes using balloons. "Balloons have been used for years to go to 120,000 feet," says Podolske, "but there are three limitations. You have no control over where the balloon will go, you often have to wait for the right winds and you only get a vertical profile [because the balloon goes up and down]."

"Balloons have short flights with no inherent repeatability," Wennberg says. "Aircraft eliminate these limitations."

The SR-71 Blackbird, originally designed in the 1960s for US Air Force strategic reconnaissance, set the record for flying at Mach 3 in 1976.

"The SR-71 can go to at least 85,000 feet [26 kilometers] and can fly at as fast as Mach 3," says Podolske. "But, as the air goes through the shockwave, it heats up," and .

Another option, flying slower, has its own problems.

"The air is so thin above 21 kilometers, you can't use conventional gas engines," says Podolske. "Either you have to 'supercharge' the air, or carry your own liquid oxygen to support combustion."

John S. Langford, president of Aurora Flight Sciences of Manassis, Va., thinks he has the solution to this problem. "Atmospheric scientists have long had to settle for the hand-me-down platforms from the military," Langford says. "[Aurora is] developing the Perseus and Theseus aircraft, taking advantage of a convergence in technologies, to fly as high as 25 kilometers, or more."

Perseus and Theseus are both "flying robots," Anderson says, unmanned planes that are guided by remote control.

Langford hopes to keep the Perseus and Theseus aircraft inexpensive enough that researchers can afford to use it. "In numbers, the Perseus will cost about $500,000 and each flight will cost $10,000 to $20,000," Langford says.