Parameter optimisation of river water quality models using genetic algorithms

Rivers can provide valuable supply of drinking water for humans, irrigation water to farmlands, water for hydropower and home for many aquatic ecosystems. However, due to population increase and its adverse effects on the rivers, inappropriate farming activities in the river catchments, and other similar adverse activities, the water quality in rivers has generally declined. Therefore, appropriate river water quality management strategies aimed at controlling and improving water quality should be seriously considered. At least, these strategies should not reduce further degradation of current water quality in rivers. To manage river water quality in the most effective and efficient way, the cause and effect relationships of the river system must first be investigated. One example of this cause and effect relationship is the inappropriate setting of effluent license limits for sewage treatment plants (STPs). Setting low effluent license limits causes poor river water quality with high concentration of nutrients in rivers. Water quality simulation modelling tools are extensively used in water quality management to identify these cause and effect relationships, and to simulate and study the effects of various 'what-if management strategies prior to their implementation. Such a simulation model is not available to model the Yarra River, its tributaries and associated STPs, and therefore, the development of a water quality model for Yarra River is the focus of this thesis.