Hydrodynamic Modelling of Lake Ontario

Thumbnail Image
Hall, Erin
3D Hydrodynamics Model , Lake Ontario
The 2006 Clean Water Act requires each municipality to come up with science-based plans to protect the quality and quantity of their drinking water. A literature review concerning applicable processes in Lake Ontario along with previous modelling of the lake is presented. The three dimensional Estuary, Lake and Coastal Oceans Model (ELCOM) is used to model Lake Ontario on a 2×2km grid scale. The model is forced using meteorological data from the 2006 summer season, inflows and outflows. The lake-wide model is evaluated using field data from thermistor chains and ADCPs as well as historical water level data. Simulated and observed temperature profiles compared well. However, modelled temperature profiles were slightly cooler than observed. Current results were more variable than temperature profile results but compared better to observed data in the offshore regions. Simulating Lake Ontario water levels proved to be problematic because an accurate water balance is difficult to force with a large drainage basin. A 300×300m nearshore model of the eastern portion of Lake Ontario and the upper St. Lawrence River is also presented. The open boundary is forced using temperature data which is (A) varied with depth, (B) constant with depth and (C) spatially varied over the length of the open boundary and varied with depth. Both spatially varied and non-varied water level data forcing the open boundary is also compared. Spatially varied temperature and water level data is computed from the coarse grid lake-wide model. Lake-wide coarse grid model error appears to propagate through the open boundary negatively affecting nearshore modelled current when coupling the models. It was concluded that lake-wide model results should not be used to force the open boundary for the nearshore model. Nearshore model results using constant temperature with depth forcing files and non-spatially varied water level data agree well with observed temperature profiles, but further analysis is required for better confidence in the model's ability to properly reproduce currents at a 300×300m grid scale.
External DOI