Stern told 2,000 climate scientists meeting here that they had failed to clearly tell humanity what it faces if global temperatures reach the upper range of forecasts made by the UN's Intergovernmental Panel for Climate Change (IPCC).

"There has been lots of scientific information on 2.0 and 3.0 degrees Celsius (3.6 and 5.4 degrees Fahrenheit), but you have to tell people loudly and clearly just how difficult 4.0 or 5.0 would be," he said.

The IPPC's 2007 report for policymakers predicted an increase by 2100 of 1.1 to 6.4 C (2.0 to 11.5 F) compared to a century earlier.

New findings show that these projections were vastly understated, scientists here said.

"Recent observations confirm that, given high rates of observed emissions, the worst-case IPCC scenario trajectories -- or even worse -- are being realized," the three-day conference, organized by top universities worldwide and the Danish government, concluded Thursday in a closing statement.

Stern, whose 2006 Stern Review has become the benchmark for calculating the economic cost of tackling climate change, conceded that his report had also fallen short in assessing the potential consequences of global warming.

Greenhouse gas emissions are growing faster, and the planet's capacity to absorb them is weaker, than was understood only a few years ago.

"The costs of delay are very deep," he told the conference in Copenhagen, which will host critical United Nations climate talks in December.

"Climate change is not like a WTO negotiation where, if it falls apart, you can pick it up five years later be more or less in the same position. If you wait, you will be in a significantly worse position."

Even smaller increases in temperatures, the IPCC has said, could unleash a devastating maelstrom of violent storms, drought, expanding disease and hunger over the coming decades.

A "five degree world" -- well within the range of IPCC predictions -- would cause an almost unimaginable level of disruption and suffering.

The last time Earth was four or five degree hotter than it is now, some 30 million years ago, alligator-like creatures navigated swampy primeval forests at the North Pole.

"Sea levels, in the long run, would rise by 50 meters. You would have to redraw the map of Europe," and every other continent, said John Schellnhuber, director of the Postdam Institute for Climate Impact Research in Germany.

"The carrying capacity of the planet would fall to one billion people or less," Schellnhuber told the conference.

"This is not a 'black swan'," said Stern, invoking a term used by philosophers to describe an event beyond the realm of normal expectation.

"This is not a small probability of a rather unattractive outcome. This is a big probability of a very bad outcome."

Faced with this unacceptable scenario, decision makers and the people they govern should be willing to buy some insurance, he said.

"Would you pay one-to-two percent of GDP for this kind of risk reduction, thinking about the cost of inaction? I think people will understand," he said.

Katherine Richardson, head of the Danish government's Commission on Climate Change Policy and a co-organizer of the meeting, agreed that scientists had not done a perfect job in getting the message out.

"Most of us have been trained as scientists to not get our hands dirty by talking to politicians. But we now realize that what we are dealing with is so complicated and urgent that we have to help to make sure the results are understood," she said.