Arctic sea ice: satellite observations, global climate model performance, and future scenarios

Abstract

This thesis examined Arctic sea ice trends through observational records and model-derived scenarios. A regional analysis of Arctic sea ice observations 1980-2008 identified regional trends similar to the pan-Arctic. However, winter maximum (March) extent in the Atlantic quadrant declined faster. Through an analysis of Atlantic Ocean temperatures and Arctic winds, we concluded that melting sea ice extent may result in increased Atlantic Ocean temperatures, which feeds back to further reductions in Atlantic quadrant extent. Further, Arctic winds do not appear to drive Atlantic ice extent. We evaluated performance of 13 Global Climate Models, reviewing retrospective (1980-2008) sea ice simulations and used three metrics to compare with the observational record. We examined and ranked models at the pan-Arctic domain and regional quadrants, synthesizing model performance across several Arctic studies. The top performing models were able to better capture pan-Arctic trends and regional variability. Using the best performing models, we analyzed future sea ice projections across key access routes in the Arctic and found likely reduced ice coverage through 2100, allowing increasingly longer marine operations. This unique assessment found the Northwest and Northeast Passages to hold potential for future marine access to the Arctic, including shipping and resource development opportunities.