The CBC provided us with an interesting case study in science reporting on Monday as its “community team” blog trumpeted “UN climate change projections made in 1990 ‘coming true.’”

Climate change projections made over two decades ago have stood the test of time, according to a new report published Monday in the journal Nature.

The world is warming at a rate that is consistent with forecasts made by the UN’s Intergovernmental Panel on Climate Change 22 years ago.

Climate scientists from around the world forecasted the global mean temperature trend for a 40-year period, from 1990 to 2030—and at this halfway point the report authors have found the projections “seem accurate” after accounting for natural fluctuations.

These are absolutely all the numbers you are going to get out of this news item. And if you peruse the new assessment of the 1990 IPCC predictions, which was actually published on the Nature Climate Change website, what you find is a more nuanced picture than the CBC’s “They nailed it, no worries” interpretation implies.

David Frame and Dáithí Stone write that the 1990 IPCC report predicted a rise in global mean temperatures of between 0.7 degrees C and 1.5 degrees C by the year 2030; on a linear interpolation, we might have expected half the increase to have occurred by now. The actual observed warming during the past 20 years (almost all of it taking place in the first ten) has been in the vicinity of 0.35 degrees C to 0.39 degrees C, “on the borderline” of the range given in 1990. In other words, the IPCC’s point estimate was high, and the overall warming has been consistent with the outer confidence bounds of their stated prediction, but barely.

Frame and Stone think, with some justification, that this is a pretty good performance given the simplicity of the climate models available at the time. It’s especially good, they think, because the models could not predict what would happen in the economy, or below the planet’s crust. Their story is that the Earth caught a series of lucky breaks despite the substantive failure of greenhouse gas reduction efforts.

The highlighted [IPCC] prediction assumed a business-as-usual scenario of GHG emissions; three other scenarios were considered and in fact Scenario B (which assumed a shift to natural gas, a decrease in the deforestation rate, and implementation of the Montreal Protocol, all independent of global climate negotiations) was closer to the mark as of 2010, especially with respect to methane emissions… Of course, [even these Scenario B] predictions were based on idealized future scenarios that did not foresee the eruption of Mount Pinatubo, the collapse of the Soviet Bloc industry, or the growth of some Asian economies, so one could argue that the prediction is right for the wrong reasons.

The authors conclude by noting that predicting the future is a lot harder than predicting the past—and, unfortunately, the resolving power of crystal balls has not improved much since 1990.

…the 1990 prediction following [the IPCC's] business-as-usual scenario covered a full 0.4ºC range due solely to uncertainty in the climate sensitivity that has not narrowed substantially so far, whereas a larger range was implied by the examination of further scenarios of emissions and a larger range still should have been considered owing to uncertainty in the evolution of natural forcings and internally generated variability.

Believers in and skeptics of the threat from anthropogenic climate change will both find promising fodder in this paper for conversion into mountains of delicious hay. (Mind the carbon emissions, though.) I’ll resist the temptation to join in that exercise, but it is very clear that the authors’ “Well done” message to the IPCC carries a sizable asterisk. If the CBC is going to report on a scientific paper, why not show some indication somebody has read it?