Development of Real-Time Hydrodynamic Models in the Great Lakes Connecting Channels and Lessons Learned in the Huron-Erie Corridor and St. Lawrence Seaway

Through support from the Great Lakes Observing System (GLOS), two pre-operational, real-time hydrodynamic models have been developed for the Great Lakes connecting channels: (1) Huron-Erie Connecting Waterways Forecasting System (HECWFS), and (2) Upper St. Lawrence River Forecasting System (USL). Both models were designed to meet specific user needs, whether it be state/local governments, water intake operators, US Coast Guard, or recreational boaters, as well as to fill gaps in the Great Lakes Coastal Forecasting System, a pre-operational model research and development system at the NOAA Great Lakes Environmental Research Laboratory. In both cases, three-dimensional hydrodynamics are simulated using FVCOM (Finite Volume Coastal Ocean Model). As a result, real-time hourly predictions of currents and water levels are provided to decision makers and the public via NOAA and GLOS websites. Following development, the models and output have been tailored to provide specialized information for stake holders in the surrounding communities, including spill reference tables for contaminant transport, river plume and bacteria prediction for beach forecasting, water depth for recreational boaters, and mixing conditions for invasive species proliferation.

Due to user needs, model locations, and topography of the connecting channels, we have faced a unique set of problems in model implementation that stem from using FVCOM in an open-boundary river system with both inflows and outflows, and in acquiring observational data from stations along an international border. These conditions can limit both model utility and length or accuracy of forecast. To combat these issues and provide users with the necessary information on river and lake conditions, we have employed several unique approaches including water-level forcing conditions that enable reverse flows in the rivers, pattern recognition for forecasted operating conditions at control structures such as dams, and specialized grid boundaries for river flow development. Through resolving these issues, valuable lessons learned in Great Lakes hydrodynamic modeling and FVCOM implementation have paved the way for future development of operational models in the Great Lakes.