A review of 20 major news organization’s coverage of a recent study in Science on melting of ice sheets shows some high-quality reporting … but with lots of the coverage falling short on providing deeper context.

One way to look at the late-November publication of the landmark Science study, “A Reconciled Estimate of Ice-Sheet Mass Balance,” is to frame it as a kind of “test” of the emergency broadcasting system — in this case, the climate communications system.

Here was an important finding, based on consensus from 26 different institutions and 47 researchers and published in a top-notch science journal. In addition, the findings bear directly on sea-level rise — perhaps the one aspect of a changing Earth that even low-information members of the public can immediately grasp.

According to the study, the melting of the Antarctic and Greenland ice sheets has contributed an estimated 11.1 millimeters (about .437 in.) to sea-level rise between 1992 and 2011. This represents 20 percent, 0.6 millimeters (.02 in.) on average, of total sea-level rise over that period; the current observed rate of total sea level rise is about 3 millimeters (.118 in.) a year.

And the rate of ice-sheet melting in these regions is now accelerating, collectively taking place at a rate three times that of the 1990s (and with Greenland’s ice sheet melting at five times the earlier rate.)

The study brings a much higher degree of certainty to an unresolved historical question — one of the most important left unanswered in the 2007 IPCC report — though it does not attempt to predict patterns in coming decades. It reconciled hugely varying estimates — from 676 billion tons lost a year to 69 billion gained, according to an accompanying new analysis article in Science; it did so by crunching data from various satellite measurement techniques and quantifying strengths and weaknesses.

“This does not tell us the whole answer for the future of sea level — it is a necessary but not sufficient step — but I believe that this is a spectacular achievement,” respected Penn State geosciences professor Richard Alley, who was not involved in the study, said in an e-mail. “The work is significant in showing the remarkable advances in our observational and modeling capabilities, and the care of the scientific community in getting this right.”

Report co-author Ian Joughin, Antarctica. Courtesy: University of Washington

That said, how did the study “perform” in the media? Or, perhaps more appropriately, how did the media perform in reporting on the study?

Separating signal from noise in a busy communications environment to assess public understanding is of course difficult, certainly in a precise, social science sense. The study was published amid the din of the film “Chasing Ice,” the Doha climate talks, and other important scientific findings in this area.

Study co-author Ian Joughin, of the University of Washington’s Polar Science Center, noted in an e-mail some of the structural difficulties for the news media, and he says he saw a checkered performance. His assessment:

This paper was somewhat difficult for them because a tightening of error bars is often not their idea of a story. The results allow some cause for optimism (the numbers now are not huge), but cause for concern (the numbers are growing). Some in the press got it right and others missed much of the point while trying to hype the story for all its worth.

But the news media are going into a year, 2013, when the fifth IPCC assessment report is expected to begin going public in the September/October time frame. Tightening of “error bars” will likely make up much of the substance that reporters are required to cover in the months ahead.

Content Analysis: More In-Depth Context Needed

So how did this “test” go?

A Yale Forum analysis of coverage from 20 media outlets and news websites that ran the story — which includes many of the “majors,” from NBC, CBS, BBC and NPR to Reuters, the Associated Press, the Los Angeles Times and the Washington Post — reveals divergent, at times sub-optimal, outcomes as mainstream journalists attempted to frame the findings to the public. None of these outlets quoted the usual cacophony of sources attempting to cast all-imaginable doubt on the findings, but many journalists might have done more to provide context as they conveyed the technical facts and figures.

It suggests, in any case, that outlets are not always taking the time to break down the significance of complex data for the general public, using “step-back” paragraphs, for instance, to put the eye-glazing statistics in plain English.

Of course, journalists are given varying — and in these days often declining — amounts of space in which to communicate. But as a reasonable standard, the article by Seth Borenstein of the Associated Press — not overly long at 760 words, and also available early through syndication for use by most news outlets — might be seen as a benchmark.

In terms of reporting “best practices,” key explanatory techniques in the AP story were identified and 19 other news pieces were compared (the stories were gathered from either Google News or Lexis-Nexis.)

• Ice to sea level: Many outlets touted the big tonnage of ice that had melted, but only 13 of 20 made the conversion of ice-sheet tons to precise contributions to sea-level rise. The AP put the phenomenon this way: “Because the world’s oceans are so big, it takes a lot of ice melting about 10 trillion tons to raise sea levels 1 inch. Since 1992, ice sheets at the poles have lost nearly 5 trillion tons of ice, the study says, raising sea levels by about a half inch.” For full context, it may also have been worth noting — though few outlets did — that melting these ice sheets entirely would lead to about 60 meters (about 197 feet) of sea-level rise, but at the current rate of melting that would take millennia. Again, the Science study did not make future estimates.
• Other causes: Most of the news outlets reviewed noted that the melting of these ice sheets accounted for 20 percent of total sea-level rise over that period. But seven of the 20 news organizations failed to explain the other 80 percent. The AP explains the other two drivers simply, as follows: “Warm water expands, contributing to the rise along with water from melting glaciers outside the polar regions.”
• Other melting news: The public has had broad access to numerous recent news reports about polar melting, but few outlets — just seven of 20 examined — put the Science study in the context of other recent melting news. This left low-information news audience members potentially confused as to how this puzzle piece fits in. Again, a simple nod to other news might suffice: The AP explained, for example, that the “numbers don’t include the summer of 2012 when Greenland experienced a melt that hadn’t been seen in more than a century, researchers said.”
• Reconciling confusion: Fifteen of 20 outlets explained that the Science study reconciled conflicting data in the past, and many noted that the matter was left unresolved in the 2007 IPCC Fourth Assessment. In terms of explaining the scientific process, these are crucial details. (Otherwise, the proverbial person-on-the-street is left to think, “Didn’t we already know the poles were melting?”)
• Global warming and humans: Nearly every outlet managed to use “climate change,” “global warming,” a “warming world,” or similar language. But the wider global phenomenon was only vaguely alluded to in many stories. By contrast, see the AP’s lead: “Fueled by global warming, polar ice sheets in Greenland and Antarctica are now melting three times faster than they did in the 1990s, a new scientific study says.” It should be noted that the Science study did not attribute the cause of the melting — a narrowness that is customary in such technical research. Only eight of 20 outlets made explicit reference to humans and/or emissions being the cause of warming. A good media debate could be had about whether this context should be stated in every article. (How much can a reporter or editor assume their audience already knows?) Given surveys showing an estimated 60 percent of the public remain doubtful or confused, it may be necessary to provide fuller journalistic context …. but do so in a way that avoids the MEGO (“My Eyes Glaze Over”) response.
• Sandy/storm surge: Eight of the 20 outlets made mention of the recent Hurricane Sandy or used related issues of greater potential storm surge to put ice-sheet melting and sea-level rise in practical context and spell out the significance. Nothing in journalism rule books would say such a reference was necessary, per se. But citing such events/threats may provide a more recent, tangible, and accessible reference point for many lay readers/viewers.

‘Darts, Laurels’* … and Some Creative Comparisons

Surveying errors or confusing language in these pieces also helps highlight some perennial media pitfalls, which include fumbling units of measurements, transposing numbers, and blurring averages and aggregate figures.

Calving iceberg in Disko Bay, Greenland. Credit: Ian Joughin, University of Washington/NASA

Among the 20 pieces of news coverage examined, problems included: the Guardian initially reported that the ice-sheet melting had caused 11 meters, not millimeters, of sea level rise (since corrected); a Climate Central post states that 36 laboratories were involved (it’s 26); the Ottawa Citizen article states “the rate of sea-level rise is accelerating, rising by nearly one millimeter per year now …” (it’s actually 3 mm, about 0.2 in., overall; 0.95 mm (.003 in.) is the current amount caused by melting in Greenland-Antarctica); and Bloomberg Businessweek led its story with “Oceans rose by an average of about 11 millimeters (0.43 inches) from 1992 to 2011 as ice sheets near both poles melted …” (the oceans rose a total of 11 millimeters, not average, over that period because of Greenland-Antarctic melting.)

Bryan Walsh of Time (see here) and Mark Guarino of the Christian Science Monitor (see here) took the laudable step of explaining that all the melting of polar sea ice in the Arctic — which the public hears about constantly — is a different case than the melting of land-based ice sheets. Sea ice melting doesn’t add much to sea-level rise, as the ice is already in the ocean. It’s the melting of land-based ice and ice sheets that is of most concern from the standpoint of sea-level rise. This reality may be obvious to climate watchers and insiders, but it’s not for people who have only a passing awareness of polar melting.

Finally, NBC News, its anchor Brian Williams, and environmental correspondent Anne Thompson deserve special credit not only for leading the evening news with the study’s findings, but for creative data context and analogy: They explain that 344 billion metric tons of ice melting in Antarctica and Greenland is equivalent to “more than one million Empire State Buildings”; and that 11 mm of sea-level rise is “like pouring in more than 26 Lake Tahoes.”

Those are the kinds of verbal images that even the most casual climate change news follower can take to the bank.

* The “darts and laurels” terminology has long been a hallmark of Columbia Journalism Review in rebuking or lauding individual news reports.