A University of Sydney science student has used supercomputer simulations of an ancient hothouse climate to unravel how the Earth heated up about 13 million years before humans existed, providing insights into Australia's climate future.

Two years ago Sydney University academics Dr John You, an oceanographer, and Associate Professor Dietmar Müller, a geophysicist and founder of their EarthByte group, realised the enormous opportunity - untapped in Australia - in combining high-performance computing with global geological databases to study ancient climate change on Earth.

'Why do we worry about climate change millions of years ago? We worry, because in order to predict future climate trends it is invaluable to understand how climate has behaved in the geological past,' says A/Professor Müller.

Nicholas Herold, a Bachelor of Science student, shared the EarthByters' enthusiasm and decided to study an unusual period of global warming which occurred about 15 million years ago, dubbed the Miocene Climatic Optimum. This period represents just one of a series of global warming events which punctuate the Cainozoic era (a period of time extending to 65 million years ago when the dinosaurs became extinct).

Global mean temperatures during the Miocene Climatic Optimum were as much as 6 °C higher than at present. This resulted in extensive grasslands as well as rainforests throughout much of Australia. Ice caps at the north and south poles were significantly reduced and inland seas occupied low continental areas in much of South America and southern Australia.

But there was a controversy: 'Despite this extreme global warming oceanographers had suggested that atmospheric CO2 levels were possibly as low as 180 parts per million volume (roughly half that of modern levels) whereas others had argued that CO2 levels were much higher than at present,' says Dr You.

If Miocene CO2 levels were low, then what other factors could have caused global warming?

Nicholas set out to construct global digital models of ancient mountain chain elevations and sea level rise for the Miocene to test the impact of palaeogeography on climate. He painstakingly assembled a digital database for Miocene vegetation to check its effectiveness as a driving force of climate change.

All data were geographically reconstructed to their Miocene positions using a global plate tectonic model. For example, due to continental drift Australia was roughly 1000 km further south 15 million years ago. The digital databases were then connected to a state-of-the-art climate model from the National Center for Atmospheric Research (NCAR) in the USA.

For raw computing power, the team utilised a high performance computer at the Australian Partnership for Advanced Computing (APAC) National Facility in Canberra to quantify the amount of warming caused by individual components of the climate system.

Nicholas' study focused on three critical aspects: the warming caused by changes in the distribution of vegetation, the warming caused by reductions in the elevation of mountain chains, and the warming or cooling caused by an increase or decrease in atmospheric CO2 levels.

'The hundreds of gigabytes of model results at first seemed overwhelming,' Nicholas recalls. 'A lot of hard work went into automating the process of trudging through hundreds of model output files and visualizing the results. The supercomputer-powered model outputs revealed some very interesting results.'

'Without increasing some greenhouse gas, either CO2 or, for example, atmospheric water-vapour, the Miocene would have not heated up as it did. But the simulations also demonstrate that changes in Miocene vegetation, mountain chain elevation (Andes and Himalayas) and sea level alone probably caused global warming of up to 1.6 °C,' he concludes.

When analysing his results for the Australian region, Nicholas found that more extreme temperatures started to prevail between northern and southern Australia as vegetation became shorter and more arid from the Miocene to the present. 'This trend is set continue well into the 21st century and most likely beyond,' Nicholas predicts.

'Knowing how different aspects of the climate system will react to future global warming can provide valuable insights into prioritising long-term strategies to mitigating its effects,' says A/Professor Müller.