Environmental water assessment at a catchment scale comparing natural and managed conditions

Environmental water is indispensable for promoting and maintaining environmental assets in managed catchments. Water in the Macquarie Catchment is managed by releases from Burrendong Dam, which has played an important role supplying water needs in the Macquarie Valley, and environmental flows to the Ramsar listed Macquarie Marshes. Management decisions tools are necessary to analyze impacts of environmental water at a catchment scale and are critical to preserve ecosystems services under future uncertainties of climate variability and change. Here we implemented WATHNET5, a Network Linear Programming (NLP) tool to analyze effects of environmental water in the Macquarie Catchment. Our semi-distributed model includes storage areas (Dam and wetlands), input flows of the main tributary rivers, irrigation and water consumption demands, routing and conveyance losses. For model setup, rules for operation of the dam were adjusted to current conditions, while tributary rivers, irrigation and water consumption demands were obtained from a hydrological model used by local authorities for the Macquarie River. The ecological outputs of environmental releases were assessed at five locations along the river and following the objectives provided in the Long-Term Water Plan determined by the Environmental Authority (EA). In each of the five locations, our model computed different flows; Base Flows (BF), Small Fresh (SF), Large Fresh (LF) and Overbank flows (OS, OM and OL for small, medium and large respectively), which are associated to different environmental objectives. The EA determined minimum thresholds for each of the flows in terms of timing, duration, frequency and interval between events as indicators of environmental objectives compliance. Our model determines if the different flows met the thresholds and computes the amount for time that the conditions are met during the simulation period. Calibration of the flows over a 30-year period were carried out and the NLP model results were compared with the observations in five gauging station along the catchment. We found that the model adequately represents the flows with Nash–Sutcliffe efficiency coefficients between 0.42 and 0.6. Simulations were carried out for 120 years to analyze the effect of environmental water releases on ecological outcomes compared to natural condition (no dam and irrigation), showing tradeoffs between the different types of flows in different parts of the catchment. Our NLP model can be used as a multi-objective optimization tool to help identify long-term management decisions that can improve system resilience and protect environmental assets under an uncertain future climate.