Last week's headlines about the United Nation's latest report on global warming were typically breathless, predicting doom and human  damnation like the most fervent religious evangelical. Yet the real  news in the fourth assessment from the Intergovernmental Panel on Climate Change (IPCC) may be how far it is backpedaling on some key issues. Beware claims that the science of global warming is settled.

The document that caused such a stir was only a short policy report, a summary of the full scientific report due in May. Written mainly by policymakers (not scientists) who have a stake in the issue, the  summary was long on dire predictions. The press reported the bullet points, noting that this latest summary pronounced with more than  "90% confidence" that humans have been the main drivers of warming  since the 1950s, and that higher temperatures and rising sea levels  would result.

More pertinent is the underlying scientific report. And according to  people who have seen that draft, it contains startling revisions of  previous U.N. predictions. For example, the Center for Science and  Public Policy has just released an illuminating analysis written by Lord Christopher Monckton, a one-time adviser to Margaret Thatcher  who has become a voice of sanity on global warming.

Take rising sea levels. In its 2001 report, the U.N.'s best high-end  estimate of the rise in sea levels by 2100 was three feet. Lord  Monckton notes that the upcoming report's high-end best estimate is  17 inches, or half the previous prediction. Similarly, the new report  shows that the 2001 assessment had overestimated the human influence  on climate change since the Industrial Revolution by at least one-third.

Such reversals (and there are more) are remarkable, given that the  IPCC's previous reports, in 1990, 1995 and 2001, have been steadily  more urgent in their scientific claims and political tone. It's worth  noting that many of the policymakers who tinker with the IPCC reports work for governments that have promoted climate fears as a way of  justifying carbon-restriction policies. More skeptical scientists are  routinely vetoed from contributing to the panel's work. The Pasteur  Institute's Paul Reiter, a malaria expert who thinks global warming  would have little impact on the spread of that disease, is one example.

U.N. scientists have relied heavily on computer models to predict  future climate change, and these crystal balls are notoriously  inaccurate. According to the models, for instance, global temperatures were supposed to have risen in recent years. Yet  according to the U.S. National Climate Data Center, the world in 2006  was only 0.03 degrees Celsius warmer than it was in 2001 -- in the  range of measurement error and thus not statistically significant.

The models also predicted that sea levels would rise much faster than  they actually have. The models didn't predict the significant cooling  the oceans have undergone since 2003 -- which is the opposite of what  you'd expect with global warming. Cooler oceans have also put a  damper on claims that global warming is the cause of more frequent or  intense hurricanes. The models also failed to predict falling  concentrations of methane in the atmosphere, another surprise.

Meanwhile, new scientific evidence keeps challenging previous  assumptions. The latest report, for instance, takes greater note of  the role of pollutant particles, which are thought to reflect  sunlight back to space, supplying a cooling effect. More scientists  are also studying the effect of solar activity on climate, and some  believe it alone is responsible for recent warming.

All this appears to be resulting in a more cautious scientific  approach, which is largely good news. We're told that the upcoming  report is also missing any reference to the infamous "hockey stick,"  a study by Michael Mann that purported to show 900 years of minor  fluctuations in temperature, followed by a dramatic spike over the  past century. The IPCC featured the graph in 2001, but it has since  been widely rebutted.

While everyone concedes that the Earth is about a degree Celsius  warmer than it was a century ago, the debate continues over the cause  and consequences. We don't deny that carbon emissions may play a  role, but we don't believe that the case is sufficiently proven to  justify a revolution in global energy use. The economic dislocations  of such an abrupt policy change could be far more severe than warming  itself, especially if it reduces the growth and innovation that would  help the world cope with, say, rising sea levels. There are also  other problems -- AIDS, malaria and clean drinking water, for example  -- whose claims on scarce resources are at least as urgent as climate  change.

The IPCC report should be understood as one more contribution to the  warming debate, not some definitive last word that justifies radical  policy change. It can be hard to keep one's head when everyone else  is predicting the Apocalypse, but that's all the more reason to keep  cool and focus on the actual science.