The Prediction Predicament

The risks in reporting long-range weather forecasts

New York Times environment reporter Andrew Revkin posted an interesting bit of media criticism (or perhaps context) on his Dot Earth blog Friday. The post was a quick response to an article, published earlier that day in The Independent of London, which asserted that the Arctic’s usually perennial floating sea ice may disappear completely this summer, for the first time in recorded history. But such predictions are risky, according to Revkin:

Given the unpredictable short-term dynamics up there, which make the ice subject to vagaries of Siberian winds and a mix of currents, a lot of polar ice experts tell me it’s pretty much impossible to make such a prediction with high confidence. In fact, The Independent’s story — the opening sentences and headline at least — go way beyond what Mark Serreze of the National Snow and Ice Data Center tells the reporter.

Revkin doesn’t belabor his analysis of The Independent’s work, although he might have. The article’s opening sentences are indeed misleading. First, there is the lede:

It seems unthinkable, but for the first time in human history, ice is on course to disappear entirely from the North Pole this year.

The wording makes it sound like the vanishing act is an incontrovertible certainty, rather than the probability that it actually is. It is not until a few paragraphs down that Steve Connor, The Independent’s science editor, states that there’s about a 50 percent chance the ice will disappear entirely, and includes some more nuanced quotes from scientists. What never materializes, however, is the subhed’s promise: “Polar scientists reveal dramatic new evidence of climate change.”

Connor might be referring to scientists’ opinion that “seventy percent of the sea ice present this spring was single-year ice formed over last winter,” making it thin and fragile. Or he might be referring to recent NASA and native Inuit observations that the sea is already breaking up. Connor notes all of this in his story, but it hardly counts as “dramatic” new evidence.

Sea ice may disappear, and journalists should watch the process closely, but the problem with reporting predictions in an irresponsibly confident or dire manner is that such prognostications are bound to be parroted mindlessly by other outlets. Revkin began his Dot Earth post by noting that The Independent’s article, which the paper billed as an “exclusive,” drew a link from the Drudge Report, a popular news aggregator that gives a strong readership boost to linked articles. Indeed, a number of other media outlets, including Fox News, Popular Science, and CBS, piggybacked on The Independent’s reporting with abbreviated articles that mostly failed to include even the scant nuances of the original piece.

Revkin himself covered the potential loss of Arctic sea ice in another Dot Earth post at the beginning of June, but his approach was much less sensationalistic. Rather than forecasting a complete loss of sea ice in an attempt to grab readers’ attention, Revkin’s headline read, “Most Experts Foresee a Repeat, at Least, of 2007 Arctic Ice Loss.” Then, in his lead paragraph, rather than abstractly referencing to a group of faceless “scientists”, he gave a detailed summary of what different scientific groups have said:

Fourteen research teams studying the impacts of warming on the Arctic Ocean have issued independent projections of how the sea ice will behave this summer, and 11 of them foresee an ice retreat at least as extraordinary as last year’s or even more dramatic. The other three groups that issued a numerical estimate see the ice extent heading back toward, but not equaling, the average minimum for summers since satellites began tracking the comings and goings of Arctic sea ice in 1979. Five other groups chose not to issue a numerical estimate.

In the post, Revkin also included a quote from the Study of Environmental Arctic Change (or SEARCH) initiative, an interagency effort that funds seventy Arctic research projects. Given the nature of Revkin’s post, the SEARCH quote (essentially its mission statement) is somewhat ironic, but it reinforces the idea that reporting what is happening is often more constructive than trying to report what will happen:

“The intent is not to issue predictions, but rather to summarize all available information from ongoing observing and modeling efforts to provide the scientific community, stakeholders, and the public the best available information on the evolution of the arctic sea ice cover,” said the coordinators in a statement.

Given the inherent problems with long-range weather predictions, it is a scientifically reasonable and prudent statement. It also harkens to some of the admonitions directed toward hurricane forecasters in a spate of stories that marked the beginning of the Atlantic hurricane season in June. For years, the media has used the start of the storm season as story fodder, but 2008 saw a marked number of articles that questioned the accuracy and utility of predictions. The Associated Press compared meteorologists forecasting hurricanes to groundhogs forecasting winter.

Predictions are not that worthless, however. Coastal Floridians and others in the storm belt would obviously prefer to have a scientist’s best guess than nothing at all. Like all science journalism, though, reporters must be true to the data at hand and resist the temptation to stretch their conclusions. Along those lines, The Wall Street Journal’s Numbers Guy blog recently made a very intelligent call for forecasters and journalists alike to “embrace uncertainty”:

When climate forecasters estimate the number of hurricanes for the coming storm season before it’s even started, there’s a lot of uncertainty. Yet the news media sometimes reports these predictions with great confidence — then balks when a hurricane season defies forecasts. Now for the first time, two major forecasters are reporting what they don’t know along with what they do know.

Posts like this, and other articles that cast a skeptical eye on long-range hurricane forecasts, are a reassuring phenomenon. Journalists cannot ignore weather and climate predictions — after all, most of the arguments for addressing global warming are founded upon what is likely to happen if temperatures continue to climb — but they must respect the limitations of such efforts by describing them as probabilities rather than certainties.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Curtis Brainard writes on science and environment reporting. Follow him on Twitter @cbrainard.