The weather and climate of Australia at the Last Glacial Maximum

Abstract

The global climate has experienced four glacial cycles in the last 420,000 years, with each cycle characterised by a prolonged period of cooling culminating in maximal glaciation followed by a brief warm period. The most recent period of maximal glaciation is termed the Last Glacial Maximum (LGM) and occurred about 21,000 years ago. We currently live in one of the warm periods. The global climate is changing, and it is becoming more important to understand the extremes of the climate system and how well our modelling capability can capture those extremes.

There has been a modelling intercomparison project established to examine how global general circulation models compare in simulating past climates, including the LGM. Analysis and comparison of these model results has been presented for many parts of the globe, but there has not been a comparison of the different model results over the Australian region. This thesis aims to fill that gap and explore the simulated LGM weather and climate of Australia and its drivers in more detail. Comparison with proxy evidence is also undertaken, and inconsistencies seen in the literature addressed.

The Australian climate at the LGM was believed to be generally cooler, drier and possibly windier from proxy evidence in the literature. In the comparison done here the mean temperature and precipitation fields from most models show cooler and drier conditions, with some seasonal variability, but there are some strong outliers. It was found that the differences were not dependent on model resolution, but that the surface parameterisations were highly important for these fields.

The shifts in the circulation were examined both in the model results and with a study of the non-linear link between the wind, surface moisture and dunes, which are a proxy for past winds. All the models simulate a southward shift in the westerlies in the Australian region. This is strongly driven byte prescribed sea-surface temperatures. Australia's current wind regime is conducive to dune building. However, the binding effect of soil moisture (or vegetation) is strong enough to limit present day movement, whereas in the drier climate at the LGM there was a capacity for sand movement. The analysis of dune orientations did not produce conclusive evidence for how the westerlies might have shifted at the LGM.

An apparent enigma in the proxy evidence at the LGM is the high lake levels in Australia’s south east, while most inland lakes were dry. Previous authors believed that the precipitation was still low, but the high lake levels were driven by lowered potential evaporation. The hydrological cycle was generally depressed in the LGM simulations, but the potential for evaporation remained high. Thus an alternative hypothesis is posed based on increased run off due to a known shift in the vegetation types and a lag in the timing of the run off due to snowmelt.

The analysis here shows that our capacity to simulate climates quite different from the present is still developing, but that model results can help explain apparent inconsistencies in the reconstruction of past climates from proxies.