Geoengineering—using technology to purposefully change the climate—is the only option for reducing the risk of climate change from greenhouse-gas emissions in the next few decades, says David Keith, a professor of public policy and applied physics at Harvard University. And he says that if it’s done in moderation, it could be much safer than some experts have argued. In fact, says Keith, effective methods of geoengineering are so cheap and easy that just about any country could do it—for better or worse.

Keith, speaking this week at MIT Technology Review’s annual EmTech conference, says it is already too late to avoid climate changes by reducing carbon emissions alone. The carbon dioxide that’s been released into the atmosphere by burning fossil fuels is already likely to cause significant harm, such as raising temperatures enough to hurt crop yields in many places. “If you want to, say, really stop the loss of Arctic sea ice or stop heat-stress crop losses over the next few decades, geoengineering is pretty much the only thing you can do,” he says (see “Why Climate Scientists Support Geoengineering Research”).

Keith’s preferred method of geoengineering is to shade the earth by injecting sulfate particles into the upper atmosphere, imitating a similar process that happens with large volcanic eruptions, which are known to temporarily cool the planet. The technique could be effective even if far less sulfate were injected than is currently emitted by fossil-fuel power plants. A million tons per year injected into the stratosphere would be enough—whereas 50 million tons are injected into the lower part of the atmosphere by coal plants, he says. (In the lower atmosphere, the sulfates are less effective at cooling because they stay airborne for shorter periods.)

One of the main objections to geoengineering is that the measures that might be taken to cool the planet won’t exactly offset the effects of carbon dioxide, so they could actually make things much worse—for example, by altering patterns of precipitation. Keith says recent climate models suggest that injecting sulfate particles into the upper reaches of the atmosphere might not affect precipitation nearly as much as others have warned.

“I propose that you start in about 2020, and you start very, very gradually increasing your amount of sulfate engineering so that you cut about in half the rate of warming,” he says. “Not eliminate it, but cut it about in half. Cutting it in half is a big benefit.”

One of the benefits could be increased crop production. Though some critics have worried that geoengineering would alter monsoon patterns that are key to agriculture in India, Keith says moderate geoengineering could actually boost crop productivity there by 20 percent, in part by reducing temperatures.

Keith and some of his colleagues recently hired engineers to estimate how much one approach to sulfate injection might work, and how much it might cost. It could be done at first with existing airplanes—certain business jets can fly high enough to inject the particles into the upper atmosphere. Eventually we would need new planes that can fly higher. All in all, once the procedure is scaled up it would cost about a billion dollars a year and require about 100 aircraft. That’s cheap enough for most countries to pull off on their own.

The fact that it’s easy isn’t necessarily a good thing, Keith says. There’s the potential that if one country does it, another might blame that country—rightly or wrongly—for ensuing bad weather (see “The Geoengineering Gambit”).

And there are also real concerns about the impact sulfates might have on the atmosphere (see Geoengineering May Be Necessary, Despite Its Perils). It’s known that sulfates can be involved in reactions that deplete the ozone layer. As the earth warms, water vapor levels are increasing, which could exacerbate the problem. Keith is proposing a test to discover quantitatively just what the effect of the injections could be. He would introduce small clouds of sulfate and water vapor into the stratosphere using balloons, and then carefully measure the reactions that take place.

And Keith acknowledges a concern many have had about geoengineering: that using it to offset problems from climate change will reduce the incentive to tackle the greenhouse-gas emissions at the root of the problem. Even if geoengineering is employed, reducing emissions will still be important. Sulfate injection does nothing to address the ocean acidification associated with increased levels of carbon dioxide in the atmosphere. And if emissions continue to grow, ever-increasing amounts of sulfate will be needed.

But Keith thinks the potential benefits might be worth the dangers. “We don’t know enough yet to start,” he says. “But the current balance of evidence is that doing this really would reduce risks. And for that reason, we’ve got to take it seriously. It really would be reckless not to look at something that could reduce risk like this could.”