Authors

Nicholas Herold, Dietmar Muller, and Maria Seton (University of Sydney)

Abstract

Climate modelling is a fast growing area of scientific research demanding exhaustive computational resources for the generation, post-processing and storage of data. Peering into the history of Earth’s climate (paleoclimatology) gives us valuable insights into future climate change and is a fundamental aspect of climate change research.

The Miocene climatic optimum is a global warming event that occurred approximately 17 to 15 million years ago and is the warmest Earth has been in the past 35 million years. The causes of this warming are enigmatic as the reconstructed concentration of CO2 is approximately the same as modern day concentrations. We use a global model of land, ocean, atmosphere and sea-ice to simulate climate during this warm period in order to elucidate the climate processes contributing to warming. Doing this first requires utilisation of a plate rotation model and geological records to reconstruct the Earth’s physical surface during the Miocene, this includes the location of the continents, the height of topographic features and the global vegetation. These reconstructions then form the boundary conditions for our paleoclimate model.

Our model runs at a relatively low resolution of approximately 3.75° in longitude and latitude in order to allow long integrations, sufficient for a radiative equilibrium to be reached in the simulated climate. To date our simulations show that concentrations of CO2 close to modern are not sufficient to match the warming indicated by geological records, though a warming of greater than 1°C is simulated due mainly to changes in vegetation and bathymetry. Several additional simulations focus on Australia, as the continent’s paleoclimate is not well known. Results show that the Australian monsoon was considerably weaker than present during the Miocene and that on an annual basis the continent was drier than present, contrary to the interpretation of available proxy data.