The Stern Review, commissioned by the British Government, concluded that the “overwhelming body of scientific evidence now clearly indicates that climate change is a serious and urgent threat”, with climate change as a result of anthropogenic global warming expected to intensify the water cycle, amplifying spatial and temporal variability (Stern, 2007). Any changes to the climate and climate variability will potentially affect hydrological systems and alter the flows in water bodies (Graham et al., 2007). Identifying and quantifying the potential impacts of anthropogenic climate change on hydrological systems allows for the design and implementation of mitigation and adaptation measures.
There are a number of climate models and downscaling techniques available for predicting long-term climate change as a result of anthropogenic global warming. This thesis examined the range and variability of climate change predictions for the south-east Queensland region from climate change scenarios formulated using the SimCLIM software package developed by CLIMsystems. The ‘reliability ensemble average’ (REA) and its uncertainty range was calculated. The value of the REA as a means of predicting climate change and quantifying uncertainty was evaluated.
This thesis explored the possibilities of the available GCMs using downscaling by SimCLIM and presents a thorough analysis of their performance and potential errors. It also provides a
comprehensive estimate of future climate in south-east Queensland as a result of anthropogenic climate change.
This thesis concluded that there are inherent errors and limitations in climate change predictions and there is high variance between projections, which results in low certainty in the potential impacts of anthropogenic climate change on south-east Queensland. It is recommended that hydrological impact assessments consider a number of representative samples along with the uncertainty range calculated using the methods in Giorgi and Mearns (2002).
Results obtained from hydrological impact assessments should be given standing relative to the uncertainties, limitations and errors inherent in the simulation. Where uncertainty and error levels are high, as it is with the current state of science, the potential effects should be viewed more as plausible changes rather than predictions.