In September the United Nations Intergovernmental Panel on Climate Change (IPCC) released the final draft of Climate Change 2013: The Physical Sciences Basis. The report's "Summary for Policymakers" flatly states: "Warming of the climate system is unequivocal, and since the 1950s, many of the observed changes are unprecedented over decades to millennia. The atmosphere and ocean have warmed, the amounts of snow and ice have diminished, sea level has risen, and the concentrations of greenhouse gases have increased."

Pretty much everyone concerned with this issue agrees that those are the facts. But what is causing the planet to warm up? Here is where it gets interesting.

The report's "Summary for Policymakers" declares it "extremely likely that human influence has been the dominant cause of the observed warming since the mid-20th century." Whether that is so can be probed by comparing observed temperature trends with the simulations of the U.N.'s computer climate models, which assume that human influences are driving climate change. According to the IPCC researchers, "There is very high confidence that models reproduce the general features of the global and annual mean surface temperature changes over the historical period, including the warming in the second half of the 20th century" (emphasis in original). So far, so good: Both the model's projections and actual temperatures rose during the latter half of the 20th century.

As evidence that the models "reproduce the general features" of actual temperature trends, the new report provides a handy graph comparing projections made in the panel's previous report with three different temperature records. The report says "the trend in globally-averaged surface temperatures falls within the range of the previous IPCC projections."

But is that so? Most temperature records show that since 1998 the models and observed average global temperatures have parted ways. The temperatures in the models continue to rise, while the real climate has refused to warm up much during the last 15 years.

The IPCC report acknowledges that almost all of the "historical simulations do not reproduce the observed recent warming hiatus." Not to worry, it assures us; 15-year pauses just happen, and you can't really expect the models to simulate such random natural fluctuations in the climate. Once this little slow-down passes, the report maintains, "It is more likely than not that internal climate variability in the near-term will enhance and not counteract the surface warming expected to arise from the increasing anthropogenic forcing" (emphasis in original). In other words, when the warm-up resumes temperatures will soar.

John Christy, a climatologist at the University of Alabama in Huntsville, has come to a different conclusion. Christy compared the outputs for the tropical troposphere of 73 models used by the IPCC in its latest report with satellite and weather balloon temperature trends since 1979. "The tropics is so important," Christy explains in an email message, "because that is where models show the clearest and most distinct signal of greenhouse warming-so that is where the comparison should be made (rather than say for temperatures in North Dakota). Plus, the key cloud and water vapor feedback processes occur in the tropics."

When it comes to simulating the atmospheric temperature trends of the last 35 years, Christy found, all of the IPCC models are running hotter than the actual climate. The IPCC report admits that "most, though not all, of [the climate models] overestimate the observed warming trend in the tropical troposphere during the satellite period 1979-2012." To defend himself against any accusations of cherry-picking his data, Christy notes that his "comparisons start in 1979, so these are 35-year time series comparisons"-rather longer than the 15-year periods whose importance the IPCC downplays.

Why the discrepancy between the IPCC and Christy? As Georgia Tech climatologist Judith Curry notes, data don't speak for themselves; researchers have to put them into a context. And your choice of context-say, the year you choose to begin with-can influence your conclusions considerably. While there may be nothing technically wrong with the way the IPCC chose to display its comparison between model data and observation data, Curry observes, "it will mislead the public to infer that climate models are better than we thought." She adds, "What is wrong is the failure of the IPCC to note the failure of nearly all climate model simulations to reproduce a pause of 15-plus years."

As noted above, the IPCC report concedes that "almost all historical simulations do not reproduce the observed recent warming hiatus." It argues that the difference "could be caused by some combination of (a) internal climate variability, (b) missing or incorrect radiative forcing, and (c) model response error." That is to say, the projections are being thrown by pesky natural climate fluctuations, and possible errors in calculating how much warming a given greenhouse gas will produce.

The IPCC report obliquely refers to an August study in the journal Nature Climate Change finding that the observed rate of warming during the last 20 years was half of what a representative sample of the models relied upon by the IPCC projected. Looking at just the last 15 years, the models were four times hotter than the actual trend in the average global temperature.

But the IPCC is confident that warming will soon resume at a pretty fast clip. Back in 2007, other modelers were similarly confident about their forecasts for future warming. At the U.N.'s annual climate change conference in Bali, the U.K.'s Hadley Centre predicted that between 2004 and 2014 the global average temperature would rise by around 0.3 degree Celsius. Instead, the Nature Climate Change article reports, the trend during the last 15 years has amounted to an increase of just 0.05 degree Celsius per decade-one-sixth the Hadley Centre's predicted rise.

The Hadley Centre also predicted that half of the years after 2009 would be hotter than the record year at the time, 1998. So far, judging from the Hadley Centre's latest data, which were adjusted upward last year, only one year after 2009 has been hotter than 1998, and then only by 0.02 degree.

The IPCC now reports that the observed global mean surface temperature increased at a rate of 0.12 degree Celsius per decade from 1951 to 2012, for a total increase of about 0.72 degree during that period. At that rate, the global average temperature by the end of this century will be more than one degree higher than it is now. An increase of just one degree more is unlikely to be catastrophic.

The computer climate models are supposed to give policy makers reliable data regarding future trends in man-made global warming. By failing to predict a flat 15-year period, the U.N.'s new report does not inspire the kind of confidence that could justify a trillion-dollar climate policy bet.