The Mitta Mitta River flows from Dartmouth Dam to the southern arm of Lake Hume and serves as a conduit for water transfers from Dartmouth Reservoir to the lake. Actual and prospective erosion of the banks under historical and alternative flow-release scenarios were modelled at seven sites along the Mitta Mitta River using the dynamic version of the Bank-Stability and Toe-Erosion Model (BSTEM-Dynamic). This approach allowed for identification of erosion-threshold conditions which ranged from about 5,200 Megalitres per day (ML/d) to almost 13,000 ML/d, generally decreasing downstream. This coincided with a general increase in erosion rates downstream. Modelled bank-erosion rates over the period ranged from 0.89 m³ per meter of channel length (m³/m) to 9.8 m³/m. Erosion did not begin at any site until daily flow rates were greater than about 5,000 ML/d, indicating that this discharge would be a conservative, erosion-limiting daily-transfer rate. Previously reported recommendations for drawdown rates (5 mm/hr) were found to be slower than the hydraulic conductivity of any of the materials tested. 

The magnitude and duration of flows above erosion thresholds exert a strong influence on bank erosion rates. Consequently, these parameters were evaluated for their relative influence on bank erosion during the 19 flow-release periods that occurred during the 2006-2016 period. Regression relations (with r²-values greater than 0.9) show that that a metric, defined as the median ratio of the flows to the erosion threshold, times the duration of those flows (in days) explains much of the variation in average erosion rates for a given release type. Reversing these regressions can be used to help guide operational-release scenarios to limit bank erosion to a certain, acceptable value.