Hydrological systems are potentially very sensitive to changes in climate. The three key variables are soil moisture, which is a primary control on vegetation and ecosystems; groundwater recharge, which feeds groundwater reserves; and runoff, which feeds rivers and causes floods. Increased temperatures are expected to cause increased potential evaporation and less snow, and possible changes in mean rainfall, rainfall intensity, and rainfall seasonality would affect soil moisture, streamflow, and groundwater recharge and the occurrence of floods and droughts. More frequent high-intensity rainfall would tend to increase the occurrence of flooding. Specific effects will depend on the pattern of change, in rainfall particularly, and the characteristics of catchments and cannot be predicted with confidence. In general, the drier the climate, the greater the sensitivity to climate change (IPCC 1996, WG II, Chapter 10). Although the effects of rainfall changes and sea-level rise on groundwater resources are not adequately understood at present, they cannot be ignored (IPCC 1996, WG II, Section 10.3.6) and may be significant for inland and coastal aquifers in Australia (Ghassemi et al., 1991).
Water resources in the region are strongly affected by the heavy rainfall of major weather events, such as tropical cyclones in northern Australia, and by the ENSO phenomenon, which is the main source of year-to-year variation and contributes both widespread heavy rainfall and widespread drought, depending on its phase. However, climate models are unable to represent these well as yet and therefore cannot represent the resulting major sources of surface runoff and groundwater recharge events. This situation presents a basic difficulty in assessing climate change impacts on hydrological systems and water resources.
A number of preliminary assessments and research studies of hydrological response in the region are available (e.g., Griffiths, 1990; Mosley, 1990; Bates et al., 1994; Chiew et al., 1995; Bates et al., 1996; Fitzharris and Garr, 1996; Minnery and Smith, 1996; Schreider et al., 1996). Most of the research work has been based on available regional scenarios of temperature and rainfall changes (see Section 4.2.3) and has focused on changes in water yield from unregulated rural catchments. Recent studies have begun to consider impacts on groundwater recharge (Green et al., 1997) and on water resources systems as a whole (Hassall and Associates, 1997; see Box 4-1).
A wide range of significant changes in water yield and soil moisture was found by Chiew et al. (1995), who considered the potential impacts of CSIRO (1992) climate scenarios on 28 catchments that represent the large range of climatic, physical, and hydrological regimes experienced in Australia. By 2030, increases in annual runoff of up to 25% and 10% occurred for catchments in the wet tropics of northeastern Australia and in Tasmania, respectively. Decreases of up to 35% occurred for South Australia, and changes of �20% and � 50% occurred for southeastern Australia and the west coast, respectively. Changes in annual soil moisture levels ranged from -25% to +15%. Although the specific figures have high uncertainty (arising from the large scenario uncertainty), their magnitudes provide an indication of the size of changes that may conceivably occur.
Sizable changes in median monthly runoff during the wettest parts of the year and increases in annual maximum monthly runoff were found by Bates et al. (1996), who used the results of a single climate model (CSIRO9, experiment F1, Table 1-1) and a stochastic daily weather generator to represent climate and hydrological variability under current and doubled CO2 conditions. The increases were due to the general increase in rainfall intensity and the increased frequency of heavy rainfall events indicated by the CSIRO9 model.
A similar shift toward greater variability was found by Schreider et al. (1996),
who applied the "most wet" and "most dry" climate change scenarios for 2030
and 2070 (adapted from CSIRO, 1992) to historical daily rainfall and temperature
series for 14 rivers in the Ovens and Goulburn Basins in southeastern Australia.
Figure 4-2 shows that in scenarios in which rainfalls
were projected to decrease ("most dry" scenario), the frequency of high-flow
events substantially decreased, and the drought frequency (as indicated by a
soil wetness index) increased. However, in scenarios in which rainfalls were
projected to increase, there was little increase in average annual runoff, but
the frequency of high flow events increased.
|Figure 4-2: Change in frequencies of runoff and soil wetness for different climate scenarios. White bars are for 2030, and black bars are for 2070. Under the "most dry" scenario (top row), where rainfall decreases and warmings are relatively large, the frequencies of high runoff and high soil wetness substantially decrease. However, under the "most wet" scenario (bottom row), where rainfall increases and warmings are less, there is some increase in the highest runoffs but also some decreases in the highest soil wetness frequencies. For further details, see Schreider et al. (1996).|
Changes in catchment vegetation-either from climate change directly or from adaptation responses (such as afforestation)-would alter catchment hydrological characteristics, including evaporation, runoff, and extreme events (IPCC 1996, WG II, Sections 14.2, 14.4). The effects of changes in rainfall amount and timing on groundwater recharge can be amplified by the dynamic response of vegetation, according to a study by Green et al. (1997) that used a daily soil-vegetation-atmosphere model to determine changes in groundwater recharge beneath North Stradbroke Island in northeastern Australia. It was found that the net recharge increased consistently by amounts greater than the change in rainfall and that the recharge was more affected by vegetation type than by soil type.
The responses to climate change of the large, arid, ephemeral lake systems of interior Australia are difficult to predict. Significant water-level changes may occur for nonephemeral lakes in dry evaporative drainages or small basins where present evaporation is comparable with rainfall inputs (IPCC 1996, WG II, Section 10.3).
Rising sea levels will cause the tidal saltwater wedge to intrude further upstream in estuaries and rivers, with resulting changes in salinity affecting estuarine aquatic ecosystems (Chappell et al., 1996; Waterman, 1996). Similarly, sea-level rise also may lead to greater saltwater intrusion into coastal groundwater, aquifers, and surface waters in some coastal systems (Ghassemi et al., 1991).
Hydrological systems in the future also will be affected by other changes such as deforestation and urbanization, both of which tend to increase runoff amount and runoff speed-increasing the risks of flash flooding, high sediment loadings, and pollution. Changes to water pricing and allocation mechanisms also will affect patterns of water use and demand (Fenwick, 1995; McClintock, 1997), and indeed can be used as adaptation measures.
Other reports in this collection