Improving 30-Day Great Lakes Ice Cover Outlooks

摘要

Prediction of Great Lakes ice cover is important for winter operations and planning activities. Current 30-day forecasts use accumulated freezing degree-days (AFDDs) to identify similar historical events and associated ice cover. The authors describe statistical models that relate future ice cover to current ice cover, AFDDs, and teleconnection indices, available on the day the forecast is made. These models are evaluated through Monte Carlo simulation and assess the potential of a perfect AFDD forecast in a regression between ice cover and AFDDs between the forecast date (first day of month) and the date for which the forecast is made (first day of next month).