The story about anthropogenic global warming doesn't seem to stack up as the best science, despite the "thousands of scientists" who are said to have "consensus" about it. In climate science I see no consensus, only a pretence at a contrived one. In any case, whatever consensus exists in science is always temporary only.

Despite all the hype and the models and the catastrophic predictions, we human beings barely understand "climate". It is too vast a domain.

So to the first AGW proposition. Is our planet warming?

The IPCC has offered an estimated average increase in temperature for the planet over the 20th century of 0.6ºC ± 0.2ºC. On the face of it, there is nothing especially unprecedented about the 20th century temperature rise, given its "agreed" size. In the last million years long ice ages have been followed by much shorter "interglacial" periods, but these relatively brief warm periods have been the times when animal and vegetable life flourished. On balance a shift in temperature downwards of a few degrees will be much more worrying to us than a similar shift upwards.

Is the warming caused by our burning fossil fuels, clearing the forests and other human activities?

The truth is that no one has yet shown that it does. An increase in atmospheric levels of carbon dioxide over the past century is agreed. The physics of the greenhouse effect is straightforward, so that such a relationship is theoretically there. The crucial problem is that we don't know a lot about the positive and negative feedbacks involved with water vapour and clouds.

No dramatically linear relationship between increasing CO2 levels and higher global air temperatures can be shown for the 20th century. Though CO2 levels have been rising for a century, temperature has not done so: one of the warming periods in the 20th century seems to have been at the beginning of the century, when the human production of CO2 was much smaller than it is now. Temperatures seem not to have increased since 1998, though the concentration of CO2 in the atmosphere has gone on increasing.

Two other correlations are much stronger - the relationship between solar energy and temperature, and between ocean movements and temperature. Moreover, actual measurement of temperature does not support the greenhouse gas theory. According to the IPCC's global climate models there ought to be a "hot-spot" in the troposphere at about 10km up. However, the data presently show no sign of it.

On the evidence that is available, I think it has to be said that the assertion that the increase in carbon dioxide has caused the temperature to rise is no more than an assertion. There is simply no evidence that this causal relationship actually exists.

Is the global warming likely to lead to a dangerous increase in sea levels?

The IPCC currently predicts that there will be a rather larger increase in the 21st century, of perhaps 2-3mm per year. Its middle estimate might lead to an increase of perhaps 30cm over this century. Even if that is accepted, there is no warrant here for the claims that Tuvalu, the Maldives and other low-lying island micro-states, or coastal towns in Australia or Florida, will be devastated.

Let us move then to the modelling of climate.

When learning about computing in the US 40 years ago I discovered that if you made a mistake in your input the output from the computer was rubbish, notwithstanding the apparent precision of the numbers in the printout. There was a phrase for it: "GIGO", or "garbage in, garbage out".

Commonsense tells us that if our current knowledge of climate and weather cannot provide forecasts with much accuracy past 24 hours, we don't know enough about the inter-relationships inside the model, no matter how much data we have, even supposing it be perfect data. It is important that the work be done, and modelling is a valuable intellectual activity in its own right. Yet it is these contemporary, unvalidated models that are the basis of the Kyoto Protocol, carbon trading and the climate change policies put forward by our political parties.

To repeat, models are models; they are highly simplified versions of reality, and cannot provide evidence of anything. One recent example, showing a truly catastrophic climate outcome in 20 years' time, was based on the assumption that the central global warming hypothesis is correct. It was then used to show why we must do something now about reducing greenhouse gas emissions. When I was a young undergraduate that kind of argument was known as an "intellectual coup d'êtat": you assumed what it was that you had to prove, surrounded it all with words (or in this case, numbers), and laid it on the table in triumph. I was astounded that no one objected.

What I see is something that the political theorist Paul Feyerabend wrote about a long time ago in Against Method (1975): the tendency of scholars to "protect" their theory by building defences around it. As each new paper that proposes an alternative to one or other aspect of AGW appears, it is as though it has to be ignored, rubbished or noted but dismissed, rather than accepted as a valid contribution to the debate.

Why does this continue? When the Royal Society in London has to issue a paper crushing anyone who asks inconvenient questions, or where the President of our own equivalent body issues a public statement saying that 'those who deny human-induced global warming are in the same camp as those that deny smoking causes lung cancer and that CFCs deplete the ozone layer" you begin to shake your head. This is not at all in the tradition of the best science. It is the language of the boss.

We seem to be caught up in an "availability cascade": we judge whether or not something is true by how many examples of it we see reported. Fires, storms, apparently trapped polar bears, floods, cold, undue heat - if these events are authoritatively linked to a single attributed cause, then almost anything in that domain will seem to be an example of the cause, and we become worried. I should say at once that "climate change" has become the offered cause of so many diverse incidents that for me at any rate it ceases to be a likely cause of any. Why does this particular availability cascade have its evident force? I offer some reasons.

One is that some of the senior people in and around the IPCC - one might call them "scientist-activists" - are convinced that unless the world wakes up to itself humanity will not have a future. I would call this a quasi-religious view, and it is the basis of the view that "the end justifies the means", a doctrine that I think has no place in a democracy. A second is that there are now thousands of people, not the least of them scientists, whose work depends on the AGW proposition and the large amounts of money that have flowed to institutes and universities because of it.

National scientific academies are now in the happy position of being powerful, at least in this domain, and they have become political in an apparent attempt to protect that pleasant power, whatever its impact on science.

A third is that the Greens and environmentalists generally welcome the AGW proposition because it fits in with their own world-view. Governments that depend on Green support have found themselves, however willingly or unwillingly, trapped in AGW policies, as is plainly the case with the newly-elected Rudd Government. The hard heads may not buy the story, but they do want to be elected or re-elected.

In short, AGW is now orthodoxy, and orthodoxy always has strong latent support. Because AGW is "science", even well-educated people think it will be too hard for them. David Henderson, a respected British economist and former Treasury official, has called the orthodoxy in climate change a case of "heightened milieu consensus", in which prime ministers and other leaders tell us that nothing could be more serious than this issue. These are not statements of fact; they are no more than conjecture. But they have become, in his phrase "widely accepted presuppositions of policy" Intellectually, AGW is what is known in politics as "a done deal".

So, finally, what should we do about it all?

As someone who has worked closely with ministers in the past, I cannot imagine that I could have advised a minister to go down the AGW path on the evidence available, given the expense involved, the burden on everyone and the possible futility of the outcome.

Some have raised the precautionary principle as an indication that we should, even in the face of the uncertainty about the science, take AGW seriously. The precautionary principle here is very similar to Pacal's wager. Pascal argued that it made good sense to believe in God: if God existed you could gain an eternity of bliss, and if he didn't exist you were no worse off. However, if you didn't believe in God and God did exist, you risked an eternity in Hell. Looking at the risks and benefits, the sensible person would believe in God. Alas, Pascal didn't allow for the possibility that God was in fact Allah, and you had opted for belief in the wrong religion, or that God was a woman with a different set of values, or that an almost infinite set of possibilities existed about the nature of the universe. I am an agnostic about the existence of God, too. The IPCC's account of things seems to me only one possibility, and the evidence for it is not very strong. For that reason, I would counsel that we accept that climate changes, and learn, as indeed human beings have learned for thousands of years, to adapt to that change as rationally and sensibly as we can.

But, because I am still agnostic, and do not dismiss the possibility that the unchecked emission of carbon dioxide may have unexpected and even adverse effects for the environment and for ourselves, I ask for a public inquiry into this matter in which scientists openly argue about the data. I would want it chaired and managed so that it represented what someone has described as "an exemplar of Archimedean science: careful reasoning, humility before nature, understatement, respect for inherent uncertainties, care with language and definitions, grounded in evidence, presentation of theoretical frameworks, respect for contestability, and so on".

The Garnaut Inquiry, set up by the State and Territory Governments on April 30, 2007 and continued by the Rudd Government, is plainly unable to do this work, because it is based on the supposition that the AGW proposition is truth. A Royal Commission has been suggested; if it followed the criteria I have set out, I would be happy with that.

We are, after all, an educated society, and these issues affect every one of us. I would also ask our government do three things:

Develop a professional approach to the preparation and publication of basic observational data about weather and climate; ensure that the funding directed to climate science research be allocated in a disinterested way (that is, without any presupposition that AGW is "settled science"); and wait until there is convincing evidence and argument before it goes ahead with what seem to me to be draconian public policies.

Finally, I would ask that all our governments, as an exercise in much-needed due diligence, look at the existence of the IPCC itself, and ask whether or not it is in Australia's interest to take special notice of its output. For ourselves, if the earth is warming, then we will learn to adapt to that, as human beings have done throughout their history. But it will be important for us to do it rationally.

This is an edited extract of a speech delivered to the Planning Institute of Australia, Canberra, April 2, 2008. The entire speech can be downloaded by clicking here (PDF 258KB).

Don Aitkin has been an academic and vice-chancellor. His latest book, What Was It All For? The Reshaping of Australia was published by Allen & Unwin.