Modelling effects of land-use and climate changes on catchment streamflow, sediment and nutrient loads by means of alternative and integrated models

Abstract

Changes in land use and climate are key pressures affecting catchment ecosystems worldwide. These pressures are particularly severe in highly developed catchments, in which natural processes are greatly disturbed by anthropogenic activities. More catchment models are being developed to support management decisions related to these critical issues. The capability of models, either single or in combinations, to provide reliable and efficient estimations of effects of locally and globally driven stressors on catchments is of particular importance for decision making. This thesis investigates the capability of modelling tools to address the complex scenarios of management practices, land use and climate change impacts on streamflow and water quality of a suburban catchment in South Australia. The primary modelling tool is based on the ecohydrological model SWAT (Soil and Water Assessment Tool). However, the broader modelling framework also applies the newly developed catchment model SOURCE and the lake model SALMO in order to enable complex scenario analyses using integrated models. Monitoring data collected over three years at twelve sites along the urban part of the Torrens River catchment improved understanding of the catchment characteristics and the calibration of the models. Simulation studies by means of the SWAT model suggest that increases in forest plantations will cause a decrease in water yield, while effects of farm dams depend on their purpose rather than on their volumetric capacities. Scenarios of past and future urbanisation have revealed that it is causing significant increases in flow and phosphate loads. A comparative scenario analysis indicated that the stress from future urbanisation is likely to be of greater concern than the projected effects of climate change on catchments with an already dry climate. Integrated catchment-reservoir modelling showed promising results for better understanding of the cascading effects of catchment processes on drinking-water reservoirs. A comparison of simulation results of the SWAT model and the Australian catchment model SOURCE for the same catchment demonstrated that both models produce similar good results for streamflow, and that SWAT provides better understanding of sediment and nutrient processes. Overall, this thesis offers an improved understanding of catchment processes and responses to natural and anthropogenic stressors in a Mediterranean-climate catchment with mixed land use. The results from this study suggest that decision making about complex catchment issues asks too much of a single model, rather requiring an integrated modelling approach, and that the interpretation of the models’ results needs to be portrayed carefully, taking into consideration the models’ fit-for-purpose designs.