Abstract
Objective predictions of first permanent ice formation and freeze-up on the Great Lakes were made by use of cumulative freezing degree-day totals, by the Lisitzin-Rodhe-Bilello equation, by use of departures from normal air temperature and by use of 30-day temperature outlooks. The four objective methods yield similar improvement over use of the mean date of freeze-up in prediction of these ice events, although freezing degree-day totals appear to represent the best method. Lake Superior ice cover can be predicted using the freezing degree-day method extrapolated to mid-lake locations with better results than a climatological prediction based on the use of long-term mean freeze-up dates.
Ice breakup on the Great Lakes was predicted using thawing degree-day totals and a correlation between stations approach. Both of these predictive techniques are superior to use of the mean date of breakup as a prediction.
