Empirical models for forecasting changes in the phenology of ice cover for Canadian lakes

Abstract

In situ and remote-sensed data on freeze-up and break-up dates for lakes spread over much of Canada were used to develop and validate simple regression models linking lake ice phenology to climatic conditions and lake morphometry. The primary variables affecting fall freeze-up dates were the fall date when 30-day smoothed air temperatures reached 0 °C and lake mean depth; the primary variables affecting spring break-up date were the spring date when 30-day smoothed air temperatures reached 0 °C, solar elevation on that date, and the number of days over winter when 30-day smoothed air temperatures were <0 °C. These models were used to project potential impacts of climate change on ice phenology across Canada; by 2055 (under the Intergovernmental Panel on Climate Change Special Report on Emissions Scenarios (IPCC SRES) A2 emissions scenario), freeze-up dates were projected to be an average of 10 days later. Break-up dates were projected to be from 0 to 16 days earlier, with greater changes occurring at higher latitudes. These projections were similar to those independently derived using a mechanistic ice phenology model.