A top scientist said Monday he had warned in 2006 that a prediction of catastrophic loss of Himalayan glaciers, published months later by the UN's Nobel-winning climate panel, was badly wrong.

The Intergovernmental Panel on Climate Change (IPCC) report said in 2007 it was "very likely" that the glaciers, which supply water to more than a billion people across Asia, would vanish by 2035 if global warming trends continued.

"This number is not just a little bit wrong, but far out of any order of magnitude," said Georg Kaser, an expert in tropical glaciology at the University of Innsbruck in Austria.

"It is so wrong that it is not even worth discussing," he told AFP in an interview.

The triple-volume Fourth Assessment Report is the scientific touchstone for political action on climate change.

Destruction of Himalayan glaciers by 2035 was questioned in a report by Britain's Sunday Times, which said the reference derived from a news article published in 1999 and had failed to be scrutinized by the IPCC.

Kaser suggested the initial error originated from a misreading of a 1996 Russian study or from findings on a handful of glaciers that were mistakenly extended to apply to the whole region.

In either case, he suggested, the fact that it found its way into the report underpinning global climate negotiations signalled the need for a reform of the way the IPCC collects and reviews data.

"The review community has entirely failed" in this instance, he said.

Kaser was a lead author in Working Group I of the IPCC report, which dealt with the physical science of climate change.

Its conclusions -- that climate change is "unequivocal" and poses a major threat -- remain beyond reproach, he said.

The prediction for the Himalayan glaciers was contained in the separately published Working Group II report, which assessed likely impacts of climate change.

More specifically, the chapter focussed on an assessment of Asia, authored by scientists from the region.

"This is a source of a lot of misunderstandings, misconceptions or failures," Kaser said, noting that some regions lacked a broad spectrum of expertise.

"It is a kind of amateurism from the regional chapter lead authors. They may have been good hydrologists or botanists, but they were without any knowledge in glaciology."

Kaser said some of the scientists from other regional groups took heed of suggestions, and made corrections ahead of final publication in April 2007.

But the Asia group did not. "I pointed it out," he said of the implausible prediction on the glaciers.

"For a reason I do not know, they did not react."

But blame did not rest with the regional scientists alone, Kaser added.

"I went back through the comments afterward, and not a single glaciologist had any interest in looking into Working Group II," he said.

The head of the UN climate panel, Rajendra Pachauri, told AFP his organization would look into the matter.

He has already vowed to probe the so-called Climategate affair involving hacked email exchanges among IPCC scientists that skeptics say points to bias.

The IPCC's Fifth Assessment, scheduled for release in 2013, will probably be adjusted to avoid such problems, said Kaser.

"All the responsible people are aware of this weakness in the Fourth Assessment. All are aware of the mistakes made," he said.

"If it had not been the focus of so much public opinion, we would have said 'we will do better next time.' It is clear now that Working Group II has to be restructured," he said.

There will still be regional chapters, but the review process will be modified, he added.