"There's no question that large-scale climate engineering is untested and dangerous," said David Keith of the University of Calgary, Canada. "But at some point – which could already be today or it could be 10, 20, 30, 40 years in the future – we may have put enough carbon dioxide in the atmosphere that we will have unacceptably huge responses to that carbon dioxide. And at that point, stopping emissions doesn't solve the problem."

Keith supported small-scale experiments that deploy particles of H2SO4 (sulphuric acid) in the stratosphere from the back of aeroplanes to create a thin sun barrier, like giant tinted windows. He called it a "cheap, fast and imperfect" contingency plan that should be tested as soon as possible.

Alan Robock, an expert in climate prediction and aerosols at Rutgers University, US, disagrees. "If you implement [geoengineering] at a smaller level, you have to do it much longer before you get a signal," he said. "If it takes 50 years to test it, that's too long for society." Robock, who has been a vocal critic of many forms of geoengineering, said that short-term small-scale tests would be nearly useless when applied to the entire planet.

This may not be the case with all geoengineering strategies. Kenneth Coale of the Moss Landing Marine Laboratory, US, presented data from various iron fertilization efforts to encourage blooms of phytoplankton in the ocean to pull carbon dioxide out of the air. This strategy, used by the controversial Lohafex project last year, addresses the underlying problem of high atmospheric carbon dioxide concentrations and at the same time potentially raises the ocean pH away from low, coral-bleaching acidities. As an example, he pointed to the 2002 Southern Ocean Iron Fertilization Experiment, which produced 270 tons of carbon per day and a 10-fold increase in chlorophyll over 39 days.

"There's been a striking transition in our thinking from finding out how the world works to asking the question how to make the world work for us," said Coale. "That has completely turned geochemical thinking upside down."

There was a downside to the 2002 experiment, however. The algal bloom also encouraged a diatom called pseudo-nitzschia, which releases a potentially dangerous neurotoxin called domoic acid. Domoic acid has been blamed for sudden marine mammal and seabird die-offs (one such poisoning of seabirds inspired the film, The Birds). Although this data has not yet been published, Coale said the 2002 experiment increased the domoic acid level locally by 100,000–1,000,000 times.

In the end, debates about the best form of climate manipulations may be moot because geoengineering remains publicly unpopular. Ortwin Renn, a sustainable technology expert at the University of Stuttgart, Germany, cited recent small-scale studies that suggest the more people learn about geoengineering, the less likely it is that they will endorse it.

As if to confirm the public's continued confusion, protesters gathered outside the meeting; several managed to attend and pepper the scientists with questions about secret government climate-manipulation plots.