Chapter 4 Climate Impacts and Responses

Impacts of Climate Change on Australia > Water resources

Efficient management of water will become increasingly important as we enter times of increasing water use and shrinking sources of supply. This is particularly true for Australia, whose high rainfall variability from year to year and decade to decade necessitates large dams and results in low stream flows in the south. Hence irrigation is extensive for agriculture and summer water restrictions are common in towns and cities even by the coast.

Researchers from the University of Melbourne and CSIRO have found signs that recent warming is affecting evaporation from crater lakes in western Victoria. The lakes have no streams flowing in or out, so their levels are dominated by rainfall and evaporation. In 1841 Lake Bullenmerri, Victoria’s deepest natural lake, was recorded as overflowing into its twin crater. Since then, water levels of this and other lakes have continued to fall, with some now remaining only as dry lake beds. The researchers have found that rainfall is currently 80% of lake evaporation, whereas to maintain the historic lake levels rainfall would have had to have been 95% of lake evaporation. The falling lake levels could be explained by decreases in rainfall and cloud cover, and increases in temperature, but the exact combination is unknown.

Globally, pan evaporation data have shown a decrease in evaporation rates over the past 50 years. Researchers at the Australian National University and Cooperative Research Centre for Greenhouse Accounting concluded that air pollution and cloud increases are diffusing sunlight, blocking some direct sunlight reaching the ground and reducing evaporation rates, despite rising temperatures.

The Cooperative Research Centre for Catchment Hydrology (CRCCH) is investigating the lag relationship between streamflow and El Niño-Southern Oscillation (ENSO), and the serial correlation in streamflow. The streamflow-ENSO connection is strongest in late spring and summer in most parts of Australia, while the streamflow serial correlation is significant for most parts of the year, particularly in southwest and southeast Australia (Figure 4.2). The CRCCH has developed a nonparametric model for forecasting streamflow. The forecasts are expressed as exceedance probabilities so that they can be used to assess the operation of conservative low risk water resources systems. These forecasts can be used to help make decisions on water allocation for competing uses, and to provide a probabilistic indication of likely water allocation in the coming months.

Figure 4.2 Typical streamflow-ENSO relationship and streamflow serial correlation in southeast Australia. (From Chiew et. al. 2003, Journal of Hydrology).

Research is currently underway at the University of New South Wales to develop approaches for predicting hydro-climate variables with the potential to improve the current efficiency of water storage and distribution networks, particularly in the water-scarce regions of the world. The approach involves the use of probabilistic forecasts that allow water managers to predict both the expected input into a water storage system, and the uncertainty associated with each predicted value. The main contribution of the research is the development of an ensemble averaging approach that makes use of multiple model outputs to reduce the chance of model misspecification. The model averaging approach has been developed to make predictions on a seasonal basis, which are then disaggregated in both space and time to each site of interest. Trials of the procedure indicate that the consideration of uncertainty in climatological observations and the use of an ensemble of model outputs result in more reliable and accurate probabilistic forecasts of the Southern Oscillation Index (SOI).