One Researcher's Plan: Fight Storms With Storms

And Emanuel, a top U.S. atmospheric scientist, is not an Alamaro fan: "I hate to sound pessimistic, but Moshe's strategy requires many orders of magnitude of energy more than what he's talking about, and the backfires would have to be almost as strong as the hurricane itself," Emanuel said. "I think, unfortunately, it falls into the category of nutty ideas."

Alamaro noted that in the 1970s the Soviet Union formed clouds on several occasions by using jet engines on land, a more difficult feat than in the tropical sea, and he dismissed Emanuel's criticism. "I say that Kerry Emanuel is not impressed by any idea but his own idea."

Regardless of the method, most atmospheric scientists would probably agree that manipulating hurricanes is not only an uncertain scientific undertaking but a potential catastrophe in several ways, if intended results are not achieved.

If scientists promise to save New Orleans and fail, they will be sued. Or if they hit Mississippi instead, they will be sued. If they promise to save New Orleans and hit Mexico, they will create an international incident. "Who takes the blame?" Emanuel asked. "It's a horrific political problem."

Hurricanes form in the eastern Atlantic as clusters of thunderstorms that pick up heat and moisture from surface water in the tropics. As the vapor rises, it cools and condenses into clouds and rain, releasing heat that causes the air to rise, repeating the cycle.

Air and moisture are drawn into the system in larger and larger volume, causing the storm to grow and begin to turn in sympathy with Earth's rotation and to move west, driven by trade winds. Alamaro seeks to duplicate this process on a mini-scale.

Modeling hurricanes "is a fine art," said Ross N. Hoffman, a principal scientist at the Lexington, Mass.-based firm Atmospheric and Environmental Research Inc. "The models vary in complexity. For water, one model may describe only relative humidity, while another may use hundreds of variables to describe all the different sizes of water and ice particles."

The key to successful modeling is to start a forecast with the most accurate data available, said Hoffman, who developed a method that constantly refines the opening benchmark in light of later information.

Modeling it, however, is not the same as doing it, and Hoffman offered no guidance on whether the future will be one of free jets, monolayer films, cloud seeding, spraying the atmosphere with soot to shield the sea surface from the sun's rays -- or something else.

But what Hoffman probably has shown, Emanuel said, is that it is better to perturb a hurricane earlier, when slight changes can be introduced relatively cheaply, rather than later, when change becomes impossible except at outrageous cost.

"The trouble is that there's a trade-off between energy and information," Emanuel said. "The further in advance you do it, the smaller the energy you need but the more unpredictable the effect."