Model development for prediction and mitigation of dissolved oxygen sags in the Athabasca River, Canada.

Northern rivers exposed to high biochemical oxygen demand (BOD) loads are prone to dissolved oxygen (DO) sags in winter due to re-aeration occurring within limited open water leads. Additionally, photosynthesis is reduced by decreased daylight hours, inability of solar radiation to pass through ice, and slower algal growth in winter. The low volumetric flow decreases point-source dilution while their travel time increases. The Athabasca River in Alberta, Canada, has experienced these sags which may affect the aquatic ecosystem. A water quality model for an 800 km reach of this river was customized, calibrated, and validated specifically for DO and the factors that determine its concentration. After validation, the model was used to assess the assimilative capacity of the river and mitigation measures that could be deployed. The model reproduced the surface elevation and water temperature for the seven years simulated with mean absolute errors of <15 cm and <0.9 °C respectively. The ice cover was adequately predicted for all seven winters, and the simulation of nutrients and phytoplankton primary productivity were satisfactory. The DO concentration was very sensitive to the sediment oxygen demand (SOD), which represented about 50% of the DO sink in winter. The DO calibration was improved by implementing an annual SOD based on the BOD load. The model was used to estimate the capacity of the river to assimilate BOD loads in order to maintain a DO concentration of 7 mg/L, which represents the chronic provincial guideline plus a buffer of 0.5 mg/L. The results revealed the maximum assimilative BOD load of 8.9 ton/day at average flow conditions, which is lower than the maximum permitted load. In addition, the model predicted a minimum assimilative flow of about 52 m(3)/s at average BOD load. Climate change scenarios could increase the frequency of this low flow. A three-level warning-system is proposed to manage the BOD load proactively at different river discharges. Other mitigation options were explored such as upgrading the wastewater treatment of the major BOD point source and oxygen injection in the effluents. The model can be used as a management tool with updated SOD values to forecast the DO in low flow years and evaluate mitigation measures. As well, the methodology presented here can be applied to manage other ice-covered rivers.