Dynamic water quality modelling and uncertainty analysis of phytoplankton and nutrient cycles for the upper South Saskatchewan River

Abstract

The surface water quality of the upper South Saskatchewan River was modelled using Water Quality Analysis Simulation Program (WASP) 7.52. Model calibration and validation were based on samples taken from four long-term water quality stations during the period 2007-2009. Parametric sensitivities in winter and summer were examined using root mean square error (RMSE) and relative entropy. The calibration and validation results show good agreement between model prediction and observed data. The two sensitivity methods confirmed pronounced parametric sensitivity to model state variables in summer compared to winter. Of the 24 parameters examined, dissolved oxygen (DO) and ammonia (NH3-N) are the most influenced variables in summer. Instream kinetic processes including nitrification, nutrient uptake by algae and algae respiration induce a higher sensitivity on DO in summer than in winter. Moreover, in summer, soluble reactive phosphorus (SRP) and chlorophyll-a (Chla) variables are more sensitive to algal processes (nutrient uptake and algae death). In winter however, there exists some degree of sensitivity of algal processes (algae respiration and nutrient uptake) to DO and NH3-N. Results of this study provide information on the state of the river water quality which impacts Lake Diefenbaker and the need for additional continuous monitoring in the river. The results of the sensitivity analysis also provide guidance on most sensitive parameters and kinetic processes that affect eutrophication for preliminary surface water quality modelling studies in cold regions.

Keywords: Calibration; Local sensitivity analysis; South Saskatchewan River; WASP; Water quality modelling.