In 2001, the White House Climate Change Technology Program quietly convened a conference called Response Options to Rapid or Severe Climate Change.
Those in attendance talked about possibilities for giant technological antidotes to global climate warming.
Such ideas have been percolating for years — giant mirrors in space to act as sunshades, giant clouds of sulfur particles in the upper atmosphere to reduce the amount of light that reaches the ground. Others have proposed burying carbon dioxide underground, or sprinkling the oceans with iron to grow algae, which would sop up carbon dioxide.
If they worked, their advocates said we could have our cake and eat it too. No need to cut down on coal and oil use; instead, we could just find ways to counteract the heat-trapping effect of carbon dioxide in the atmosphere.
But now, a group of scientists has examined some of them, and say they're potentially dangerous if done wrong.
"Employing geoengineering schemes with continued carbon emissions could lead to severe risks for the global climate system," wrote Damon Matthews and Ken Caldeira in this week's issue of Proceedings of the National Academy of Sciences.
Matthews added, in an e-mail to ABC News, "I think this is motivated more by increasing concern over dangerous climate change impacts rather than an increased confidence in the effectiveness of geoengineering as a climate-control strategy."
The scientists picked one scenario: emitting large amounts of sulfur or other fine particles into the upper atmosphere, effectively acting as a worldwide filter so that less of the sun's heat gets trapped in the atmosphere. They ran it through a computer simulation of Earth's climate system to see what would happen.
It would not be that difficult to change the atmosphere. Several major volcanic eruptions in recent centuries — the most recent was Mount Pinatubo in the Philippines in 1991 — have sent massive clouds of material into the upper air, and caused measurable global cooling.
If a single volcano can cool the world's climate, certainly human beings can do better, right? Among others, the Dutch-born scientist Paul Crutzen, who shared a Nobel Prize in chemistry for proving how man-made chemicals were damaging the planet's ozone layer, has pronounced himself an advocate of the idea.
"Our calculations using the best models available have shown that injecting 1 million tons of sulfur a year would cool down the climate so the greenhouse effect is wiped out," Crutzen said in an interview with Reuters.
But Matthews and Caldeira say it could be risky business. In their computer model, they found they could cool the atmosphere quickly, but what if you overdo it? And what if you suddenly stop doing it?
"Should the engineered system later fail for technical or policy reasons, the downside is dramatic," wrote Peter Brewer of the Monterey Bay Aquarium Research Institute, in a commentary that accompanied the research. Brewer wrote that the atmosphere "quickly bites back, leading to severe and rapid climate change with rates of up to 20 times the current rates of warming."
Computer models have their limitations, scientists concede, but they've been getting more and more precise in their simulations of how the climate behaves. For this experiment, the researchers used a computer at the University of Victoria in Canada.