By SIMON COLLINS
In the 1980s, Chris de Freitas wrote several articles in the New Zealand Listener warning about the dangers of global warming.
He was in at the beginning of the debate on climate change after attending one of the first international conferences on the subject.
"We were trying to tell people that human beings could change the world's climate," he says.
"We could affect the ozone layer. That is very real and very serious.
"It was quite possible that we were having a wider impact on the climate. So we argued that we needed to look into that."
To his surprise, the scientists' warnings - channelled through the Intergovernmental Panel on Climate Change (IPCC) - were picked up by politicians and led to today's attempts to control carbon dioxide emissions through the Kyoto Protocol - which, he now thinks, goes too far.
"What I think happened is that it was so successful because it coincided with a conservative environment ... The society took it in hook, line and sinker."
Today, the former greenie features prominently in environmentalist demonology.
De Freitas was one of six "toxic sceptics" listed in June by the London-based New Internationalist.
In the same month, Scientific American reported allegations that de Freitas was "driven by politics" when, as an editor of the journal Climate Research, he published an article by two Harvard researchers questioning the conventional view that the 20th century was the warmest in 1000 years.
Last week, American climatologist Michael Mann told a US Senate committee: "Chris de Freitas ... frequently publishes op-ed pieces in newspapers in New Zealand attacking the IPCC and attacking Kyoto and attacking the work of mainstream climatologists in this area. So that is a fairly unusual editor that we are talking about."
Born in Trinidad, and educated in Canada and at Auckland University since he finished his doctorate in Queensland, de Freitas is now a New Zealand citizen. His artist wife Nancy lectures at AUT.
Sheltering from the storm in his small, paper-laden office in the School of Geography and Environmental Science in Symonds St, de Freitas denies the tag of "climate sceptic".
"I am a global warming agnostic, not a sceptic," he says.
On the balance of the evidence, he believes the average world temperature is likely to rise by about 1C this century - significantly more than the increase of 0.6C in the 20th century, but less than the IPCC forecast of between 1.4C and 5.8C.
But he believes that the main driving forces behind the increase are not human, but natural.
Two graphs sum up the argument. The first, in the IPCC's latest report in 2001, was compiled by the same Michael Mann who attacked de Freitas in the Senate last week.
Mann, a professor of environmental science at the University of Virginia, put together his "hockey-stick" graph from hundreds of reports of "proxy" measures of Northern Hemisphere temperature taken from tree rings, corals, ice cores and evidence of the growth and decline of glaciers, stalactites and stalagmites over the past 1000 years.
It shows a long, slight drop in average temperatures from AD1000 to 1900 (the hockey stick shaft), followed by a sharp kick up (the blade of the stick) in the 20th century.
But the two Harvard scientists whose article got de Freitas into trouble believe Mann's hockey-stick is an oversimplification, driven by the same kind of "politics" of which they and de Freitas are accused.
In particular, they say, Mann wrongly dismissed widespread evidence that the world's climate was warmer than it is today for much of the period from AD800 to 1300 (the "medieval warm period"), and much colder from about AD1300 to 1900 (the "little ice age").
In medieval times, England was warm enough to support 50 vineyards, and Greenland was genuinely "green" enough to be settled by the Vikings in AD986.
The English vineyards and many Greenland settlements were abandoned when the climate got colder after about 1300.
De Freitas has argued that these periods were part of a 1000-year cycle that showed up earlier in the "Roman warm period" around the time of Christ and in the "Dark Ages cold period" later in the first millennium.
The Harvard authors, Sallie Baliunas and Willie Soon, rechecked 240 studies of the same kind of evidence that Mann used, and found signs that it was warmer in most parts of Earth for at least parts of the medieval warm period than it is today.
"The proxies show that the 20th century was not unusually warm or extreme," they concluded.
For New Zealand, de Freitas quotes a 1979 study which found "the temperature curve ... to be broadly similar to England, and such climatic fluctuations as the medieval warm period and the little ice age are not just a local European phenomenon".
The second graph, which de Freitas took last year from tree rings in 14 Northern Hemisphere sites, shows the kind of picture implied by the Harvard work - a warm period 1000 years ago when trees grew by an average of more than 0.6mm a year, followed by a cold period when tree growth dropped to around 0.4mm a year, and then a recovery in the 20th century to the kind of growth that occurred a millennium ago.
"The Mann 'hockey stick' is nothing more than a mathematical construct vigorously promoted in the IPCC's 2001 report to affirm the notion that temperature changes of the 20th century were unprecedented," de Freitas wrote.
"The validity of this has been soundly challenged, and sufficient evidence exists to disprove it."
De Freitas carries the argument further than Baliunas and Soon and argues that the global warming that occurred early in the 20th century may already have levelled off.
Although surface-level temperature gauges show escalating warming, satellite measurements taken higher up in the atmosphere since 1979 show no clear warming trend.
De Freitas believes that the satellite data is more accurate because surface gauges are distorted by local heating effects from cities.
He accepts that there has been an increase in carbon dioxide emissions from industry and cars in the past century or so, but says this has boosted plant growth, absorbing part of the problem.
"Global temperature has not risen appreciably in the last 20 years," he concludes.
"Although the future state of global climate is uncertain, there is no reason to believe that catastrophic change is under way."
This is clearly a minority view. Jim Salinger of the National Institute of Water and Atmospheric Research and an author of the last IPCC report, says there is not enough evidence to prove the medieval warm period was anything more than a local event.
"It might have been as warm as today, but remember, we are only dealing with a bit of the North Atlantic," he says.
In New Zealand, there was evidence that kumara grew more widely when the Maori arrived around AD1300 than it does today, and that the Southern Alps glaciers advanced during the colder period that followed, between about 1600 and 1900.
But this evidence was much less certain than temperature record available from 1860 onwards, which showed a definite warming trend in the last 30 years of the 20th century.
Other critics have noted that Baliunas and Soon did not even try to compare past temperatures with the present day directly, because of the nature of their data.
But Mann told the Senate last week that measurements taken in the past 30 years showed "simultaneous warmth indicated by nearly all the long-term records".
That, he said, could be explained only by human influences.
In some ways, the whole dispute is based on the arbitrary choice of a 1000-year period in Mann's original work and in the Baliunas-Soon rejoinder.
Go back around 5000 to 10,000 years to a time known as the "climatic optimum", and no one disputes that the world was about 2C hotter than it is today. In that context, the 0.6C rise in the 20th century is nothing out of the ordinary.
But the debate about the past is important for what it implies for the future. If the climate has fluctuated often before, any current warming may be due largely to natural factors such as changes in the sun's energy or ocean currents.
But if the changes in our era are much faster than any others in the past few thousand years, human beings may be chiefly to blame.
In that case, we still have something to worry about.