On more than one occasion in recent weeks, President Obama’s science advisor, John Holdren, has said that he supports research into geo-engineering, a controversial approach to addressing climate change that would involve large-scale engineering projects designed to cool the earth in the event that efforts to cut carbon dioxide emissions fail to curb global warming.

It’s not clear whether Holdren’s personal views will prevail at the White House, but coordinated federal research on geo-engineering would be a marked change from current policy. Very little money is currently spent on this kind of research, and there is no coordinated effort to assess the potential benefits or risks of the various approaches that have been proposed. In part, this is because so many experts have ruled out geo-engineering entirely, citing the potential for unforeseen side effects. Holdren’s position, however, reflects that of a growing number of researchers who say that a continued growth in carbon dioxide emissions and a lack of effective political response to global warming could make geo-engineering necessary.

Geo-engineering schemes fall broadly into two categories: those designed to remove carbon dioxide from the atmosphere and those designed to shade the earth and reflect sunlight back into space to cool the planet. Some researchers, for example, have proposed seeding the oceans with iron particles to fertilize carbon-dioxide-consuming algae. Others, including the Nobel laureate Paul Crutzen, have suggested injecting sulfurous particles into the upper reaches of the atmosphere, where they would block a small fraction of sunlight that reaches the earth. Other proposals range from the extremely simple–painting roofs white to reflect sunlight–to the extremely costly and elaborate: assembling sunshades in space.

To be effective, these schemes would have to be done on massive scales, and so far, researchers lack the experimental data and computer models needed to determine how they might affect ecosystems or weather patterns worldwide. The uncertainty is compounded by the fact that scientists have a poor understanding of how natural systems deal with carbon dioxide. About half of the carbon dioxide emitted by burning fossil fuels and other human activities is absorbed by plants and the ocean, but scientists don’t know precisely how this works or whether it will continue.

Without understanding how the natural systems work, it’s difficult to predict how engineering schemes could change them–a fact acknowledged even by proponents of geo-engineering. John Latham, a senior research associate at the National Center for Atmospheric Research, in Boulder, CO, says that more research is needed to understand the unintended consequences of all proposed geo-engineering approaches, including his own. Latham’s plan would entail spraying fine mists of seawater from wind-powered boats; the mist would increase the reflectivity of low-lying clouds, thus shading the earth. But he admits that it could also cause changes in precipitation, potentially leading to droughts. Latham says that large-scale experiments and better computer models are needed to better understand the potential effects of his idea. If these experiments and models suggest that there will be problems, “we should drop the scheme, unless we can find a way out of it.” But so far, the necessary tests haven’t been possible. “The problem is,” Latham says, “with one or two tiny exceptions, there’s been no funding.”