Greenland ice: Are the Times a-changing?
The part of News Corporation that makes Times Atlases is currently taking the same kind of kicking from scientists that some of its newspapers took from the general public over phone-hacking.
What it's being kicked for is for claiming, in the edition that came out last week, that the Greenland ice sheet has shrunk by 15% over 12 years, necessitating the re-drawing of its boundaries.
Few books receive as much publicity as the Times Comprehensive Atlas of the World.
OK, a new JK Rowling or an undiscovered Wilde might get more - but still, the atlas put up an impressive showing last week, with articles in a lot of major newspapers and executives interviewed on rolling news channels.
And top of the agenda in all of the output I saw was the 15% claim.
The problem is, it's not true; and glaciologists have been queuing up to say why not.
"In the aftermath of 'Himalayagate', we glaciologists are hypersensitive to egregious errors in supposedly authoritative sources," said Graham Cogley from Trent University in Canada.
"Climate change is real, and Greenland ice cover is shrinking. But the claims here are simply not backed up by science; this pig can't fly."
As Professor Cogley was the scientist who raised the alarm over "Himalayagate" - the erroneous Intergovernmental Panel on Climate Change (IPCC) contention that Himalayan glaciers could largely melt away by 2035 - he is well placed to make the comparison.
Jeffrey Kargel from the University of Arizona, principal scientist on the GLIMS project that's trying to improve mapping of ice and glaciers from space, was even more scathing.
"These new maps are ridiculously off base, way exaggerated relative to the reality of rapid change in Greenland," he fumed.
"I don't know how exactly the Times Atlas produced their results, but they are NOT scientific results.
"Just like IPCC's '2035', a number like 15% ice loss used for advertising the book is simply a killer mistake that cannot be winked away."
All this is in addition to the letter from the Scott Polar Research Institute, which I reported on Monday, that concluded: "There is to our knowledge no support for this [15%] claim in the published scientific literature."
Precisely how the Times Atlas team reached its conclusion is not entirely clear.
In a statement issued on Monday, and in a phone call thereafter to Sheena Barclay, MD of the HarperCollins imprint Collins Geo which publishes the atlas, it emerged that the map-makers somehow got the figure from the US National Snow and Ice Data Center (NSIDC).
This Colorado-based institute is one of the world's most respected in terms of polar science.
But how did the NSIDC's data add up to 15%?
I called up Ted Scambos, one of the researchers there with whom I've talked regularly down the years, and found that the Times figures were something of a mystery to him and his colleagues.
At the time of writing (Monday evening UK time, early afternoon in Colorado), their prevailing theory was that the figure may have been derived from a map published in the Atlas of the Cryosphere, an online resource that NSIDC maintains.
If this was the source, it was a big mistake, according to Dr Scambos.
"This was intended to be a map of the thickness of the central portion of the ice sheet - not the peripheral glaciers and... not intended as a definitive outline of the ice sheet," he told me.
"The Atlas of the Cryosphere is intended to be a public resource, a quick-look resource, and not a definitive statement."
When I asked whether a phone call to NSIDC could have cleared up any confusion, he said: "There are probably 1,000 people they could have called that would have been able to steer them clear of the 15% number... anybody working in glaciology, any graduate students working in glaciology, could have steered them clear."
The real proportion of ice sheet area lost over the last 12 years is more like 0.05%, he said.
The Times Atlas may have intended to highlight dramatic changes to the world taking place as a result of warming. But Ted Scambos (and he is not alone) is rather concerned that this episode could have the opposite effect.
"I'm worried that the importance of the changes that are going on will be lost on the public, because the true value of what the ice sheet has lost compared to this 15% number sounds very small.
"Yet if you look at the coastline, if you make measurements along some of these outlet glaciers, you see stunning levels of change - they're losing elevation very rapidly, on the scale of tens of metres, some of them."
Hence all the fears about sea level rise measured in metres as a result of the sheet melting - not next week, not next year, not in the next decade; but possibly irreversible once a threshold of warming has been exceeded.
Whether a mis-interpretation of the NSIDC map is exactly what happened will presumably be clarified at some point; in the meanwhile, the Daily Mail's Michael Hanlon has blogged an account of a longer conversation he had with Collins Geo's Ms Barclay, which sheds a bit more light on the matter.
We should also see the company clarify at some point where it intends to go next. It took the IPCC several weeks to own up to the Himalayan error; but own up it eventually did, and embarked on the process of formally amending its report.
Will HarperCollins do likewise?
It has already - for what reason, I do not speculate - issued a claim about its original claim that does not stand up to scrutiny.
"While global warming has played a role in this [15%] reduction, it is also as a result of the much more accurate data and in-depth research that is now available," the company said on Monday - adding: "Read as a whole, both the press release and the 13th edition of the Atlas make this clear".
Here, I have to admit that I have not read every word on every one of the atlas's pages, so I might have missed something.
But press releases are intended, partially, to condense what's in the thing they're publicising - and often, owing to time pressures, they are all journalists will read.
For the record, here are the glossy document's opening words:
"For the first time, the new edition of The Times Comprehensive Atlas of the World, published on 15 September, has had to erase 15% of Greenland's once permanent ice cover - turning an area the size of the United Kingdom and Ireland 'green' and ice-free.
"This is concrete evidence of how climate change is altering the face of the planet forever - and doing so at an alarming rate.
"Cartographers of the atlas have sourced the latest evidence and referred to detailed maps and records to confirm that in the last 12 years, 15% of the permanent ice cover (around 300,000 sq km) of Greenland, the world's largest island, has melted away."
No mention there of the 15% figure being partially "a result of the much more accurate data and in-depth research that is now available".