Optimising flow management for ecological response and consumptive use

Abstract

The regulation of flow in river systems and use of water for consumptive and economic purposes has led to detrimental effects on riverine, wetland and floodplain environments in river systems worldwide. In recent years, there has been a concerted effort to develop policies to return water to the environment to minimise these effects. However, there are far fewer instances of actual flows being delivered. One barrier to the delivery of environmental flows is the need to balance environmental and consumptive outcomes, optimising returns for both with a limited volume of water. Various methods are available to help define the flows required to protect specific ecological assets or to mimic aspects of natural flow regimes, but few consider consumptive uses as part of the same set of calculations. In this paper, we present a method of evaluating flow options using ecological response models incorporated into daily hydrology and irrigation river management models. A multi-objective optimisation approach produces the Pareto frontier of non-dominated solutions, which provides decision makers with a range of alternative optimal management options. An integrated water resource model of the Goulburn River, Victoria, Australia, is developed that represents the rivers, water storages, operational constraints, water management and consumptive demands on a daily timescale linked to climate. Models of ecological responses to flow are incorporated into the river model to simulate ecological response and generate environmental flow demands. Storage volumes are used to determine water allocations which in turn determine the area irrigated and relative value of the crop mix. We develop this eco-hydrology model and optimisation approach as a 'proof of concept' example, where the objective functions are to minimise terrestrial vegetation encroachment into the main river channel through the use of environmental flows, while maximising the net relative value of irrigation. The model is run over 35 years and the results indicate that a range of optimal solutions exist. In the best case for irrigation there is a net relative value (over the 35 years) of almost $17B while terrestrial vegetation encroachment could average approximately 13%. In contrast, it would be possible to reduce the terrestrial vegetation encroachment to just 2%, however, this would reduce net relative value of irrigation to around $12B. Interestingly the latter option would also result in some short-term periods of very high vegetation encroachment. This was because of much lower overall storage volumes resulting in 0% water allocations in some years. Our results highlight the importance of hydrological modelling of both consumptive use and ecological response to understand the feedback mechanism of some management decisions. Between the two extremes are a range of results that provide a clear understanding of what outcomes could be expected for both of the objective functions for all optimised solutions. The results presented in this paper are applied to a simple rep...