Sceptic rubbishes computer modelling on climate change

Climate change sceptic William Kininmonth has questioned the reliability of long-term computer modelling.

CLIMATOLOGISTS were downplaying the uncertainty of the long-term computer models used to predict climate change, a leading sceptic said yesterday, as repercussions spread from the mistaken IPCC claim that the Himalayan glaciers would melt by 2035.

Climate change sceptic William Kininmonth, a former director of the Bureau of Meteorology's National Climate Centre, questioned the reliability of long-term predictions, given that the limit of accurate forecasts was about 10 days.

"The whole issue about the global warming scenario is that the uncertainty of computer modelling is being downplayed," he said.

"People are saying `we know there are certain aspects of physics the system will respond to and that's what we'll go with', and not recognising the feedback process from weather systems that we just can't control."

Mr Kininmonth said the lead author of the Intergovernmental Panel on Climate Change's 2007 report, Michael Oppenheimer, who was questioned about the glacier mistake on ABC TV's 7.30 Report on Thursday, was hiding behind the excuse that there was a "fundamental uncertainty" in climate projections.

Mr Kininmonth said: "Twenty years ago, at an early climate change convention, he was strongly supportive of anthropogenic global warming", based on modelling and forecasts made with crude computer systems that lacked data about ocean circulation.

For Professor Oppenheimer now to cite uncertainty to defend mistakes in the IPCC report on glaciers was "disingenuous", Mr Kininmonth said.

The flawed glacier claim found its way into a 2008 report by Australian government adviser Ross Garnaut.

UN Framework Convention on Climate Change executive Yvo de Boer has been forced to defend UN climate adviser Rajendra Pachauri against calls that he should resign.

Dr Pachauri said calls for his resignation had come from fossil fuel companies. "This is an organised block of vested interests," he said.

Mr Kininmonth said the current El Nino weather system was a case in point on uncertainty.

"If we can't predict short-term or seasonally, or the extent of an El Nino, how can we make predictions about what might happen 20, 30 or 100 years hence?" he said.

"The best computer models are predictive for six to eight days -- that's the limit of our weather forecasting ability."

NCC climatologist and El Nino specialist Grant Beard said short-range and long-range forecasting were "two different problems" and uncertainty was a given beyond the short term.

"You need to know very precisely the conditions of the atmosphere and ocean . . . but eventually you depart from reality after about 10 days," he said.

"A forecast for 30 or 40 years in the future is not a forecast . . . we're just trying to get a gross measure of the global atmosphere and surface."

ADDITIONAL REPORTING: AFP