Columbia University geophysicist Klaus Lackner, meanwhile, envisions a "forest" of structures that look more like billboards than trees. The 100-by-200-foot steel rectangles would have surfaces that soak up carbon dioxide—simulating photosynthesis—then "exhale" the CO₂ in a concentrated stream that would be stored in underground chambers. One "tree" could, over one year, vacuum up the annual emissions equivalent of 15,000 cars. Yet if the energy required to operate the trees comes from fossil fuels, say critics, the carbon savings would be minimal.

Over at NASA, researchers are looking at solutions like "seeding" clouds with seawater, to make them more reflective, and creating a giant sun shield in space. The 60,000-mile-long structure would be launched from the earth in billions of pieces; once in place, it could block 2 percent of incoming sunlight. The estimated cost for the first 50 years: more than $3 trillion. It's a long shot, but NASA has awarded a $75,000 grant to University of Arizona astronomer Roger Angel, who's developing the sun-shield blueprint.

Even private investors are getting in the game—spurred on, perhaps, by this year's offer from Virgin mega-mogul Richard Branson to award $25 million to the first person who comes up with a viable CO₂-busting scheme. In tests begun early this summer, a Silicon Valley startup called Planktos planned to dump 50 to 70 tons of powdered iron, an essential nutrient for the growth of plankton blooms, into the South Pacific. Theoretically, the resulting bloom will soak up 20 million to 30 million tons of CO₂ before it dies and sinks to the deep ocean. Skeptics say the plants may rerelease most of that carbon dioxide when they decompose and are eaten by other creatures, minimizing the carbon offset while having unknown effects on the ocean's health.

It's easy to understand why few environmental leaders are cheering all this research. It does, after all, involve messing with nature on a scale that makes the Hoover Dam look like a Tinkertoy. Every plan has the potential to wreak eco-havoc, even in best-case scenarios.

"We have to be sure we don't try to solve one ecological disaster by creating another one," says John Passacantando, executive director of Greenpeace USA.

Another concern is that geoengineering debates could distract policymakers from the more immediate challenge of passing greenhouse-gas-reduction regulations.

"Why sit around figuring out how big the sponge should be when we haven't even begun to contain the size of the mess," says Dan Becker, director of the Sierra Club's global-warming program. "It's a waste of precious time and resources."

Free-market supporters argue that technology may prove more efficient than regulatory measures. "I'm not convinced that some of the behavior-based solutions pass the cost-benefit test," says Jonah Goldberg, a National Review contributing editor. "Historically, when humans have met with challenges—some caused by them, some by other things—ingenuity has saved the day."

"Playing God is dangerous," counters Myron Ebell, the controversial director of energy and global-warming policy at the Washington, D.C.–based Competitive Enterprise Institute. "We don't know what the ideal climate is. Cooling things down too much could be more problematic than warming."

Ultimately, most everyone agrees, we need multiple strategies to deal with the climate crisis: emissions caps, alternative-energy research, better methods for capturing and storing carbon dioxide, and, yes, a contingency plan in the event that we face otherwise irreversible warming.

Contemplating far-out geoengineering schemes is certainly chilling. But we've already spent more than a century playing this game. What is global warming, after all, but an inadvertent experiment in altering the climate?