Anyone who sets out to discredit a piece of published work can do so by finding a single factual error.

No matter how peripheral the mistake may be, it undermines public confidence in the work. People naturally wonder, if the authors were careless on this point how much else might be wrong?

More than one mistake has been found recently in the work of the Intergovernmental Panel on Climate Change, set up by the United Nations to provide authoritative reports on global warming, and the errors are hardly peripheral.

The IPCC's powerful Fourth Assessment Report in 2007 declared there was a probability of glaciers disappearing from the Himalayas by 2035 or sooner.

This statement - frightening for Indian and Chinese communities depending on Himalayan meltwater - has been investigated and turns out to have been taken from a green campaign group, WWF, which took it from an interview with an Indian glaciologist in the magazine New Scientist in 1999.

He has admitted his prediction was "speculation" not supported by scientific measurement and peer-reviewed.

Worse, the IPCC was notified of this in 2006 and yet the claim appeared in the 2007 report. The glaciologist who alerted the panel to the error can only attribute its appearance to "a kind of amateurism" among those who wrote the offending chapter.

If the Himalayan debacle was bad enough, the panel references to disappearing ice in the Andes, the European Alps and Africa are even more embarrassing.

They turn out to have been based on a student dissertation and an article in a climbing magazine.

Mountaineers may be the only people in a position to notice alpine ice levels but their impressions fall some way short of scientific evidence.

It is not encouraging to hear a New Zealand contributor to the report, climate scientist Jim Salinger, defending it on the grounds that it accords with interviews from somebody as illustrious as the late Sir Edmund Hillary.

Last week, the IPCC's attempts to link natural disasters to global warming was critically examined.

Its claim in 2007 that the world had "suffered rapidly rising costs due to extreme weather-related events since the 1970s" turns out to have been based on a paper that had not been peer-reviewed or published at that time.

When the paper was published in 2008 it included a caveat that the evidence was insufficient to establish a statistical relationship between the global temperature increase and catastrophes.

Yet the panel said nothing about the caveat before last year's Copenhagen Conference where the fear of natural disasters loomed large among African nations in particular.

These errors are not merely academic; they cause real worry in the regions concerned.

New Zealand frequently hears the fear of Pacific states vulnerable to rising sea levels from the predicted melting of polar ice.

Climate scientists are anxious to deny that these "slip-ups" discredit the IPCC's conclusions overall but sceptics of climate change have seized upon them to do exactly that.

The IPCC's reputation is not helped now by the argument of authority its supporters have employed for so long. Criticism was dismissed as conceit in the face of a "scientific consensus" that by implication could not be wrong.

Well the consensus has been wrong, or at least careless on several points. Scepticism has strengthened, but it is only scepticism; human-induced climate change has not been disproved. It remains too worrying to be dismissed.

Governments need dispassionate scientific assessments of it, not anecdotes, unchecked papers and agitators' propaganda.

The IPCC urgently needs new leadership and a return to strict scientific rigour if it hopes to be taken seriously again.