University of Waterloo

Published purpose

Accurate and timely forecasts of sea ice conditions are crucial for safe shipping operations in the Canadian Arctic and other ice-infested waters. Sea-ice forecasts can be obtained through an optimal combination of sea-ice model output and observational data, in a manner similar to weather forecasting. However, these forecasting systems rely on accurate weather forecasting data to drive the sea-ice model, which is not available at high spatial resolution during the advance tactical planning stage. For this reason, seasonal sea ice forecasts are often obtained through statistical means. This Project will build on a prototype system that uses advanced artificial intelligence (AI) methods to provide state-of-the-art seasonal forecasts of sea ice conditions in the Canadian Arctic. Novel developments in AI will be incorporated into the prototype for enhanced accuracy of sea ice forecasts. The prototype will be evaluated through extensive comparison with independent data and case studies based on community needs. It is anticipated these sea ice forecasts will significantly reduce costs incurred by delayed operations in annual resupply of northern communities. The final system will be delivered to the Government of Northwest Territories and its Marine Transport Services group.