Modeling Phosphorus Cycling in a Seasonally Stratified Reservoir (Fanshawe Reservoir, Ontario, Canada)

Abstract

Human activities, such as mining, sewage discharge, fertilizer usage and dam construction for electricity and flood controll, have significantly disturbed the biogeochemical cycling of nutrients, such as carbon, phosphorus, and nitrogen, in atmospheric, terrestrial, and aquatic systems. Globally, negative effects of the excess inputs of nutrients have been observed in freshwater and saline surface water environments. Phosphorus (P) is an essential nutrient for primary production, and due to intensive anthropogenic activities, including rapid agricultural intensification and urban development, excess P has been loaded into the Thames River Watershed (TRW), Ontario, Canada for around 45 years. Water quality in the TRW has been significantly affected by inputs of P and other nutrients. These eutrophic waters could have significant and chronic negative effects on the downstream and nearby aquatic environment, such as Lake St. Clair and Lake Erie. This thesis focuses on Fanshawe Reservoir, located in the Northern TRW, where Fanshawe Dam has been built to control potential flood events that may damage the City of London. However, excess nutrients could accumulate in the reservoir sediments and slowly release over a long period, posing significant difficulties for water quality management. During summertime, blue-green algae and elevated bacterial concentrations have been frequently observed by the Upper Thames River Conservation Authority (UTRCA). However, the existing field data cannot explain the seasonal variation of the algal blooms or the long-term scale interaction between the external loading of P and internal loading of P. To provide a computational framework to analyse existing field data and relate P availability in Fanshawe Reservoir to external and internal P loading, I developed a two-dimensional model for Fanshawe Reservoir using the CE-QUAL-W2 software. The model combines hydrodynamic, water quality, and sediment diagenesis modules. The simulation results imply a major role of internal P loading during the summer when the reservoir stratifies. Retention of P mainly occurs during wintertime, while the reservoir is a source of P during summertime. In a scenario where external P input to the reservoir is instantaneously reduced by 40%, the annual downstream export of P from the reservoir only decreases by 22%, because of continued internal P loading from the sediments. Due to the legacy P stored in the sediments, it would take on the order of 22 years for P export from Fanshawe Reservoir to drop to 36.5% of its current value. In another biomass scenario, the sediment P loading has 40.1% larger effects on algal growth than the external loading of P during summertime. Furthermore, to provide feasible and fast water quality modeling applications, a back propagation artificial neural network (BP-ANN) model was successful developed and calibrated for the future modeling works.