Seasonal prediction of freeze-up dates and ice coverage in the St-Lawrence Seaway using Artificial Intelligence.

Knowledge of river ice conditions from freeze-up to break-up is critical for the safe operation of ships in icy rivers. However, forecasts of freezing onset and ice conditions are almost non-existent for the St. Lawrence Seaway, a major shipping channel for Canada. For Fednav (Canada’s largest international bulk shipping company operating in the Seaway) this lack of information results in large uncertainties, forcing them to adopt a conservative approach to ensure the safe and efficient navigation of their fleet.

While regional ice forecasting typically relies on high-resolution ice-ocean numerical models, this type of model is expensive to develop and run, and none are yet available for the St-Lawrence Seaway. This project proposes to develop a new freeze-up forecast model for the St. Lawrence Seaway based on Machine Learning.