Every year, the United States suffers attacks on American soil so brutal, our military can do little more than rebuild our wrecked cities, and console the wounded once the enemy has withdrawn.This enemy is the Atlantic hurricane system, and the price of its damage, in dollars spent and in lives lost, rivals that of man-made war. Hurricane Katrina, which totaled nearly $100 billion and 1,800 dead in 2005, cost only slightly less than a year of the occupation of Iraq, and killed more Americans in a day than the Iraq war claimed in over two years. Last year, Hurricane Ike claimed only 177 lives, but still wreaked $31 billion of damage.If this enemy were human-imagine, if you can, a rogue Canadian government-we would long since have funded a massive military and civilian project to defend our border, raid enemy bases, and reduce Ottawa to puddle of hot slag. But since hurricanes are inanimate, we resign ourselves to the inevitable destruction.We can do better.For decades, meteorologists have studied ways to strangle hurricanes. Their efforts have not been much rewarded: colleagues shun them, tending to eschew the voodoo-meteorology involved in weather tinkering. But the anti-hurricane scientists are serious, and their efforts, while underfunded, have produced an ingenious array of new tactics.Hurricanes develop when hot air near the sea's surface rises to meet the cold air above. If the rising hot air differs enough in temperature with the cold layer, cold gas rises in spirals and churns, and a hurricane is born. To reduce the heat near the ocean surface, some scientists propose dipping enormous buckets deep into the ocean and hauling frigid seawater up to cool the surface. They've also considered scattering materials into the ocean to reduce or change the sea-spray, which may be a factor in the violence of the churn. And higher up in the atmosphere, scientists propose to scatter huge quantities of carbon-black-a substance so dark it can absorb enough solar radiation to heat up the cold upper reaches of the nascent hurricane.Skeptics claim that these schemes won't work. To date, weather modification has managed only a few modest victories: we can get rid of fog (cold airports, such as Thule Air Base in Greenland, do this regularly), and if conditions are right, we can seed clouds and marginally increase the chance of rain. But to try to change a hurricane is to enter an Olympic steeplechase when we've barely learned to toddle. Other quixotic government-backed science crusades have failed more often than they have triumphed. (Witness Nixon's "War on Cancer.") Even the most eager anti-hurricane crusaders, like Moshe Alamaro of MIT, acknowledge that their work is a long-shot, and fraught with dangers.But it's a bargain compared to other war efforts, and it will yield benefits even if it fails to beat the hurricane. The Manhattan Project cost about $24 billion, in today's dollars; hurricane fighters say the bulk of their work could be done for a very small fraction of that sum. More powerful computer models alone, they claim, would drastically improve our ability to predict hurricane behavior. Higher-precision forecasting would save huge sums, since we'd know which cities to evacuate and when. And many of the proposed interventions into the hurricanes themselves would be quite cheap. Seeding nascent storms with smoke particles-a process for which Daniel Rosenfeld, a distinguished Israeli meteorologist, published a patent this summer-requires only 10 cargo planes full of smoke.And outright success isn't the point. Think of the moon missions. They cost five times as much as the Manhattan Project, and produced almost no knowledge about the moon that NASA couldn't have found out more cheaply with unmanned spacecraft. Nevertheless, Apollo 11 remains the signature scientific accomplishment of the last century, and in pursuit of a moon-shot science learnt a great deal about rocketry, missiles, computers, and physics.The phrase "Manhattan Project" has long descended into cliché-what we need is a New Orleans Project. Ideally, this would yield a way to stymie an actual hurricane, and pay for itself with increased property values from Galveston to Miami. But even if it failed, it will have produced vast leaps forward in our understanding of weather systems. This is how government-funded science progresses-in fits and starts, framed by benchmark schemes with crazy-sounding goals. These schemes start as mad science, slowly morph into fringe science, and eventually become standard practice. Harnessing nuclear energy for clean power generation didn't start with tentative inquiries into how uranium might turn electrical turbines. It started with a grand attempt to make the largest explosion in the history of man.Inspiration is the beginning of scientific progress. At the moment, there are few academic journals more inspiring than the Journal of Weather Modification.Photos from the NOAA and flickr users Army.mil, laffy4k, and Bart & Jill.