We present a palaeoclimatological application of the University of Melbourne GCM (MUGCM) to simulate the mid-Holocene (6 kyr BP) global climate. The objective is examination of the mean Australian climate and its variability during the so called `Climatic Optimum'.

Our approach is to employ a sophisticated atmospheric GCM with both the prescribed solar radiation forcing and the atmospheric gas concentrations appropriate for this epoch. A 100 year long palaeo simulation is compared to a control run which represents the present day (PD). The control run successfully replicates the distribution and variability of extratropical cyclone activity in addition to the main features of winds, temperatures, and precipitation of the mean circulation. The statistical significance of the responses to the prescribed changes are estimated, enabling separation of the signal from the noise that is inherent in the model.

The mid-Holocene simulation produces a climate with a reduced seasonal cycle in the Southern Hemisphere, in contrast to the modelled enhanced seasonality in the Northern Hemisphere. Relative to PD, Australian surface air temperatures are lower during Austral summer and autumn and are higher during the winter and spring months. As a result, there is a reduction in the strength of the summer monsoon and its associated precipitation for northern Australia. However, the reduction in rainfall directly associated with the monsoon in the north is more than offset by enhanced summer precipitation associated with a greater number of tropical cyclones in the simulation. In the southern region of the continent the number of extratropical cyclones is unchanged with the exception of spring, during which relatively more systems result in higher precipitation rates over the interior of Australia. However, the winter cyclones in southern Australia coast are less intense and, as a result, lower rainfall amounts are seen compared to PD. In general, Australia experiences less precipitation during each season with the exception of spring. A consequence of the precipitation response, integrated over a year, is that the modelled Australian surface moisture content is drier than today. In contrast, NE Queensland, a location of many palaeo studies, is wetter, due to the enhanced precipitation of the region averaged over an annual cycle.

The simulation results are compared to, and interpreted against the best available Australian palaeo data sets and the consequences of potentially important ocean feedbacks, not represented in the model, are discussed.