Arctic sea ice has declined dramatically during the last few decades. Estimating the maximum and minimum sea ice extent, before it occurs, is a important tool for developing and implementing near-term public policy. Examples include: cultural/subsistence planning, biological monitoring, shipping schedules, other commercial activities, regulatory implementation, and national security planning.
Most sea ice projections from 2 to 9 months in advance are generated by dynamic or statistical climate models. These models have widely varying success (verification) rates. An alternative method is to forecast sea ice extent based on similar years in the past. This analog forecast method is increasingly used to forecast geographical areas ranging from hemispheric down to synoptic scales. In an abstract sense, the initiative represents a sophisticated form of pattern matching. In a rapidly changing environment, analog forecasting is a key tool in the climate forecasting toolbox.