A month spent on a barren island perched off the western coast of Africa allowed NASA scientists to fly a sensor-laden airplane into storms blowing off the continent. Because such storms can gather strength from warm ocean waters and spin up into hurricanes, this was not simply a daredevil mission but a scientific expedition to track such storms from birth to death. With most of the team back in the U.S., one conclusion proved obvious to them: despite the sensors and volume of data collected, it is difficult to tell which storms would ultimately gather hurricane strength.

"As soon as storms get into the eastern ocean they undergo a brief period of intensification. Most of these decay the second and third day that they are in the ocean, and a few of them seem to completely disappear," explains meteorologist Edward Zipser of the University of Utah, who was chief mission scientist for the NASA effort based in the Cape Verde Islands. "Or we thought that they had disappeared but then they were sort of reborn in the central or west Atlantic and became named storms."

For example, Ernesto, which menaced Cuba and soaked the eastern seaboard of the U.S., had disappeared from satellite imagery until it reached the mid-Atlantic, where it gained fresh power. "The circulation [of the winds] never really disappeared, but it took until it had passed a good chunk of the Atlantic to get its act together," Zipser adds. Nevertheless, mission scientists had sampled the storm in its infancy, flying a DC-8 aircraft fitted with sensors to measure water vapor, dust particles, wind speed, rainfall, temperature and other factors in several layers of its structure. Combined with information acquired by satellites passing overhead, the researchers gathered the most complete picture of such hurricanes yet.

Although the basics of hurricane formation remain unchanged--converging winds partner with warm seas and high humidity to create potentially catastrophic storms--the scientists did sample a number of oddities, including a storm that appeared weak at ground level but had hurricane force higher in the atmosphere. "This storm only looked like a hurricane between 5,000 and 25,000 feet," Zipser recalls. "Below that it looked like a weak tropical storm, winds of maybe 40 knots, but at 10,000 feet it had winds of 60 knots. It was definitely an unconventional storm."

More important, the team was able to sample several storms as they interacted with dust blown west off the Sahara. Some theories argue that such dust should kill storms, basically by drying them up. "One day we would fly over a system, and there would be a very weak vortex that you could look down through the center of it. Next day, it would be filled in with dust. Man, where did that dust come from?" recalls mission scientist Robbie Hood of NASA's Marshall Space Flight Center. "I still can't tell you whether dust hurts or helps."

Preliminary results, such as information about the odd storm, could be presented in the next few months, but it will take years to sort through all the data and develop a coherent picture of the life cycle of hurricanes. Some preliminary conclusions are becoming clear: "I've learned that it's just a bigger geographic space you have to look at," Hood says. "You have to take what we saw in the local area and put that in context with what was happening in the whole Atlantic Ocean."

"It's apparently a greater challenge to understand which storms are going to intensify," Zipser adds. "Every storm is a little different. The surrounding environment is a little different. It's a pretty complicated problem." But one that needs solving.