Degree Name

Doctor of Philosophy


School of Earth & Environmental Sciences


The availability of surface water resources is of fundamental concern globally, especially in dryland environments where these resources are particularly vulnerable to over-exploitation. Determining the extent and quality of this water also requires some knowledge about the susceptibility of these resources to change. This thesis aims to improve our understanding of the water balance in dryland environments, in particular within the Lake Eyre Basin (LEB) in central Australia, both in terms of modern processes and the factors that may have resulted in changing hydrological conditions during the Late Quaternary.

The dissolved load in the contemporary dryland rivers of the LEB is assessed and indicates that evaporation does not significantly modify the ionic content or solute concentration despite large transmission losses to the surface water budget (~64% of mean annual flow). The source of these solutes is investigated and shows that although rainfall and dust account for the bulk of them, silicate weathering also plays a surprisingly important role. This suggests that dryland environments should also be included estimates of the global silicate-weathering cycle. A comprehensive investigation into the cause and fate of the large catchment transmission losses, and their role in the water budget within the LEB is undertaken. The semi-confined alluvial setting caused by the large mud-dominated multiple channel and floodplain system in many parts of the LEB, and in particular the Channel Country of Cooper Creek, has resulted in at least three distinct shallow groundwater recharge pathways. These control recharge rates, the distribution of freshwater lenses, and the evolution of high groundwater salinities. This highlights the importance of preferential flow in semi-confined alluvial settings, and suggests that where hydrological variability is typically high, such as dryland environments, groundwater recharge is perhaps better considered as a probability of occurrence instead as an average rate. Furthermore, the sum of these probabilities may in turn best account for the observed transmission losses in the water budget.

The current hydrology of the LEB is complex, making it difficult to interpret the conditions that prevailed during the dramatic changes in climate and hydrology associated with the Late Quaternary. This aspect of the study investigates runoff conditions required to maintain the system of terminal lakes. Strzelecki Creek, a distributary of Cooper Creek originating at Innamincka, has a record of fluvial deposition that broadly matches the fillings of the Lake Mega-Frome system. This confirms the ability of the headwaters of the LEB to deliver runoff to both the Mega-Frome and Eyre terminal lakes at various stages of the Late Quaternary. The magnitude of change required to balance the water and energy budgets implied by the existence of much larger lakes throughout the Late Quaternary in central Australia is quantified. The acute sensitivity of surface water hydrology to small changes in climate within dryland environments is highlighted. The results indicate relatively small changes (~ 27 %) in mean basin rainfall can fully account for the existence and maintenance of the largest recorded lake systems. These results have important implications for the interpretation of Late Quaternary climate changes, the impact of early human arrival on the landscape of the LEB, and the role of climate in the extinction of Australia’s megafauna. Previous assessments have largely overlooked the non-linear relationships and potential feedbacks between hydrology and climate, and this has led to some unlikely proposals for palaeo-environmental change during the Late Quaternary. This work can guide future research that combines data-driven and modelling approaches to the investigation of modern and Late Quaternary water balances, especially in the dryland regions of the world.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.