Failure to make difficult choices to cut greenhouse gas emissions exposes humanity to an increasingly dire set of climate scenarios. But there is a way to buy time: Geoengineering.

The idea of tinkering with planetary controls is not for the faint of heart. Even advocates acknowledge that any attempt to set the Earth's thermostat is full of hubris and laden with risk.

Some ideas are the stuff of science fiction: 15 trillion mirrors positioned in orbit to shield the planet from the sun's rays; a fleet of blimps 20 kilometers up feeding a constant stream of sulfur into the stratosphere; a navy of robot-controlled ships prowling the world's oceans, spraying seawater skyward to generate reflective clouds.

Others are more mundane: Plant trees to soak up carbon dioxide or paint roofs white to reflect sunlight. Most are unproven. All have major drawbacks. None offset ocean acidification.

But the concept is gaining more traction as politicians, confronted with the ugly reality of trying to wean economies off fossil fuels, cast about for a strategy that will work if climate changes quickly or in nasty ways.

"Most analysts who examined the options closely had concluded that it would be reckless to mess with the planet," said David Victor, a professor of international relations at the University of California, San Diego, who specializes in energy and climate policy. "That is changing."

It's changing, in large part, because the chances of any sort of international agreement on radical emissions cuts are plummeting even as scientists find evidence that these emissions have the potential to destabilize the Earth's climate to a degree unforeseen in human history.

If those predictions come true, scientists fear any hand-wringing over the consequences of planet-wide mitigation will pale in comparison to the inconsolable pleas of populations facing rising seas, searing dust storms and savage famines, scientists warn. The world needs to know of geoengineering's pitfalls before desperate leaders turn to an untested technology.

Practical applications date to the Cold War, when both Russian and American military started seeding clouds in an attempt to induce rain. When President Lyndon Johnson was briefed about the dire effects of global warming, geoengineering was the only solution prescribed by his scientific advisors.

For years, however, it was taboo - on the fear that, if climate control was seen as a viable option, pressure on world leaders to reduce emissions might ease.

That changed in 2006 with the publication of a seminal essay in the journal Climatic Change by Nobel laureate Paul Crutzen, emeritus professor at the Institute for Marine and Atmospheric Systems at Utrecht University in the Netherlands.

Emissions cuts are the first priority, Crutzen repeated throughout his eight-page essay. But given climate change's catastrophic implications for ecosystems - and the "grossly disappointing" international political response to necessary emissions cuts - geoengineering must be explored as a potential escape route.

"Its possibility should not be used to justify inadequate climate policies," he wrote, "but merely to create a possibility to combat potentially drastic climate heating."

Crutzen and other climate scientists draw a bright bold line between the need to undertake geoengineering research and the decision to engage in actual geoengineering.

The former is crucial, scientists say: Society must know the limits and impacts of geoengineering schemes before crisis hits and governments, pinched by a desperate citizenry, cast about for relief. The latter is a political decision.

That decision - fraught with philosophical, moral and practical questions - is increasingly viewed as a key component of any climate plan.

"The reason is the uncertainty," said David Keith, director of the energy and environmental systems group for the University of Calgary's Institute for Sustainable Energy, Environment and Economy. "Even if we cut emissions to zero tomorrow, we would still have climate risk."

"The hard reality is that doing geoengineering might be better for nature than not doing it. It would've been better to not put CO2 into the air, but that's been done now."

Even studying the science has its own peril. Scientists have no way to field test a geoengineering scheme at a scale sufficient to provide reliable data without also exposing vast regions of the globe to the risk of side effects.

To Martin Bunzl a philosopher at Rutgers University, that quandary makes geoengineering right now less an ethical question than a methodological one.

"It's going to be extremely difficult to engage in experimentation that would adequately answer whether the balance of risk and benefits makes it worth it to try geoengineering," said Bunzl, who has studied geoengineering and was part of an ethics panel convened by the Royal Society to explore its moral implications.

Yet we need to try, argue several scientists. The ability to geoengineer is as necessary for reducing climate risk as the need to cut emissions.

Many researchers fear industrial emissions have already pushed the Earth outside the so-called "tolerable window" for climate warming. Average global temperatures have already risen 1 degree Celsius higher than pre-industrial times, with another degree worth of warming likely bought and paid for but not yet delivered, given time-lags in natural cycles. Atmospheric concentrations of carbon dioxide are already 40 percent higher than anything experienced in the past 2.1 million years, the Arctic is warming faster than anyone predicted, and major species declines have been reported by scientists worldwide.

To keep climate within this window humanity has known for much of its existence, scientists say emissions from industrialized countries need to drop 80 percent or more within the next few decades. The United Nations' climate talks in Copenhagen this December are meant to build a framework for that reduction, but climate analysts see scant signs any agreement will emerge soon.

But planetary engineering will likely remain part of society's future irrespective of Copenhagen's outcome. First, Bunzl said, there's a good bet that even with strict enforcement of a plausible international treaty, levels of atmospheric carbon dioxide will end up higher than agreed upon - a so-called "overshoot" that could put the climate outside that tolerable window, with disastrous consequences. Second, he added, "even if there is an agreement, people are extremely naive if they think the world will actually stick to an agreement."

In effect, future generations are stuck with geoengineering because society's main strategy for controlling emissions isn't working well and the climate could prove far more sensitive to emissions than scientists and policy makers think.

"The politics of actually stopping global climate change by mitigating emissions are nasty, brutish and endless," Victor wrote in a recent article in the Oxford Review of Economic Policy. "Geoengineering, by contrast, offers prompt benefits with seemingly small costs."

Those benefits are deceptive. And, Victor stressed, there are many risks.

Today most schemes focus on enhancing the globe's albedo, or reflection, in part because it's technologically feasible, cheap and fairly easy to reverse or undo.

More cloud cover reflects light, so some propose spraying ocean water skyward hoping the aerosols will enhance cloud formation. But that could change rainfall patterns, with attendant risk of drought. Others suggest painting every roof white. But that would offset at most two years' worth of emissions, Bunzl said.

The most favored option today is the injection of sunlight-reflecting sulfur particles high into the atmosphere, if only because, as Victor notes, every few decades volcanoes validate scientists' theories. Also attractive: The option could be deployed quickly and, scientists surmise, would lead to a rapid climate response.

The eruption of Mount Pinatubo in 1991 shot 10 to 20 million tons of sulfur dioxide and other fine particles as much as 40 kilometers high in the atmosphere. A year after the explosion more than half the particles remained aloft in the upper atmosphere, reflecting sunlight and cooling the planet by 0.5 degree C.

Bunzl cites estimates that for $1 billion a fleet of 10 retrofitted 747s could inject sufficient quantities of sulfur in the stratosphere to approximate those effects. Other estimates are higher - closer to $50 billion. Either way, Bunzl said, "it's trivial in terms of time and investment."

There are limits, however. Those particles will fall out after a year or two, requiring constant maintenance. The sky whitens as sulfur concentrations increase, and the atmosphere can only hold so much before ozone depletion and acid rain become problematic.

And the true costs are insidious.

In virtually every geoengineering approach some populations benefit while others suffer. Studies suggest efforts to increase the planet's albedo by infusing the upper atmosphere with sulfur could expose hundreds of millions of people, mostly in the tropics, to drought as rainfall and river flow patterns change.

"Eight billion people may be better off while one billion people may be worse off. That's an interesting moral dilemma," Bunzl said. After the Pinatubo eruption, researchers at the National Center for Atmospheric Research found large changes in rainfall and river patterns in many regions, especially the tropics. The 1783-84 eruption of Laki fissure in Iceland altered the Asian and African monsoons sufficiently to trigger famine in Africa, India and Japan.

"It's a problem, but I don't consider it a killer objection," Bunzl added. "I don't mean to be trivial, but there's a lot of detail: How much worse would they be? Are there any offsetting efforts you could make? And many, many other issues."

[In October the International Food Policy Research Institute released a report finding that, if nothing changes, some crop yields in the developing world could be half of their 2000 levels by 2050.]

And that perhaps is the primary concern about geoengineering: Any tinkering will likely beget problems that will beget more tinkering and so on until a complex and ultimately indigestible casserole of solutions is in place. Once an effort begins, even opponents may be hard-pressed to justify abandonment, as failure to keep a geoengineering mask in place could lead to even more rapid and disastrous warming. "There are no magical silver bullets," Victor said.

Nearly all the geophysical analysis to date has focused on simple geoengineering options, such as tuning the planetary albedo, that are unlikely to be deployed in practice. A more realistic geoengineering mask, Victor added, will likely be extremely complex: Albedo modification along with active efforts to offset ecological side effects, ocean acidification and other harms that a simple primary geoengineering system cannot rectify.

Wrapped up in that dilemma is a central question: Who sets the thermostat?

"My biggest fear is that we're getting into the controls of the planet," said Calgary's Keith, "where one part of the world wants to run the planet different than another.

"If one tweaks the knob a different way than another - or adds one knob atop another - it could be a real disaster."

But society is already elbow-deep in various planetary controls, Keith noted. Atmospheric carbon is nearing a level not seen since the Arctic was ice-free and alligators roamed Northern California. Invasive species are running roughshod over entire ecosystems. Genetically modified plants produce biofuels and more robust food crops.

"There are few places you can walk around on this planet that don't bear a very heavy hand of humanity," said Keith.

"That's not to say we should do more (engineering), but you need to have some perspective."

In this light, Bunzl and Keith don't worry so much about the consequences of geoengineering. It is really just a stop-gap, Bunzl said - something to buy at most 50 years' of time to decarbonize the economy.

The real worry, they say, is what happens if society fails to cut emissions and has no way to buffer the planet from the effects of all that gas.

"We face a fork in the road," Bunzl said. "We have to decarbonize. We can decarbonize with the option to geoengineer, or we can decarbonize without the option."

"It's far from obvious - and that's putting it politely - to think that we'll have clean energy in the amounts needed on the timetable envisioned by China and India," he said.

"We're going to face a crisis on international governance on who does what."

This article originally appeared at The Daily Climate, the climate change news source published by Environmental Health Sciences, a nonprofit media company.