Grandiose schemes to try to repair the earth's climate and delay catastrophic warming continue to proliferate almost as quickly as coal-fired power stations. Until recently most scientists distanced themselves from these controversial proposals, pointing out that none had been tested and it was human interference which caused the problems in the first place.

But in a recent change of tack, some researchers now argue these diverse plans should be scientifically evaluated as soon as possible, in case emissions are not reduced quickly enough.

Philip Boyd, of the Centre for Chemical and Physical Oceanography at the University of Otago in New Zealand, said most schemes have been proposed by individuals and companies, and some have been overhyped, so objective evaluation is required. "An assessment of all the well-established proposals is urgently needed."

The University of Adelaide's professor of climate change, Barry Brook, said that soot and smog from cars and coal burning are cooling the planet, but when these are reduced, warming already "in the pipeline" will quickly become apparent.

"So it's emergency mode, and all options must be evaluated as a priority, including deliberate climate engineering."

Dr Boyd said geo-engineering schemes should first be ranked on factors including cost, risks and how quickly they might work. Those that cannot deliver big benefits fast enough, within a few decades, should be knocked out of consideration. "We need to weed out the wacky ones."

Rigorous examination could then be done on a few of the most promising options, but this would take time, because they couldn't all be tested at once. "We wouldn't know which one was benefiting the planet."

This would lead to a "climate change toolbox" of one or two reliable plans, said Dr Boyd, who outlined his proposals in the journal Nature Geoscience.

He dismissed the argument that treating geo-engineering projects seriously provides an excuse for governments not to slash emissions. "They represent a last-ditch attempt to do something."

A sober appraisal revealing the extreme risks, very high costs, and unlikely success of geo-engineering could inspire a redoubling of efforts to reduce emissions, he said.

Schemes fall into two categories: removing carbon dioxide from the air, or reflecting radiation from the sun back into space.

British researchers, led by Tim Lenton of the University of East Anglia, recently completed a study comparing the potential effectiveness of a range of proposals.

It showed that some, such as planting vast forests and turning agricultural waste into charcoal to bury, could help.

"But geo-engineering alone cannot solve the climate problem," said Professor Lenton, whose research is published in the journal Atmospheric Chemistry and Physics Discussions.

He dismissed as "globally ineffective" an ocean cooling scheme proposed by Professor James Lovelock to place long tubes below the ocean surface to draw up cold water.

Creating a giant sunshade by launching trillions of transparent discs into space has been proposed by Roger Angel of the University of Arizona.

Professor Lenton found that this and other barrier approaches, such as injecting sulphur particles into the upper atmosphere or spraying seawater into clouds by using a fleet of Flettner vessels to make them more reflective, have the greatest potential to cool the planet by 2010.

But they are also the most risky, because warming would be extremely rapid if their deployment was suddenly stopped.

The second-most effective idea after sunshades was found to be the chemical absorption of carbon dioxide from the air, for example, with artificial trees like those proposed by Dr Klaus Lackner of Columbia University in New York.

The gas would still have to be disposed of, and one idea is to pump it into greenhouses to be absorbed by crops.

The Greek islands scheme of painting urban areas white to increase reflectivity "could reduce urban heat islands but will have minimal global effect", he concluded.

Dr Boyd said the effectiveness of fertilising the ocean with iron was probably less than a third of that initially claimed, and "its mitigation timescale of millennia is not useful".

It could also have costly side effects, such as reducing productivity in other areas of the ocean.