The predictability of extreme Arctic sea ice variations in a rapidly changing climate

Arctic sea ice has been diminishing dramatically in recent years, reaching a record low in 2012 after a previous sudden drop in 2007. Climate model simulations also generate such large and rapid summertime ice loss events in the near future. However, the natural variability of the Arctic system is very high, and models exhibit instances of increasing sea ice cover even well into the 21st century. Improved understanding of the character, impacts, and potential forecasting of these types of extreme sea ice events has great societal relevance for Arctic marine access, seasonal forecasting, and climate variability.

In this project we will assess the processes responsible for, and the predictability of, rapid Arctic sea ice variations, with an emphasis on the implications for marine navigation and extreme weather. We will utilize the Community Earth System Model 1 (CESM1), one of the better models from the CMIP3/CMIP5 archives. CESM1 simulates 20th-century Arctic sea ice very realistically and will be used to examine the predictability of Arctic sea ice cover, especially rapid and extreme sea ice variations. We will assess the regional nature of this predictability and the impacts of expanding open water coverage on extreme weather. Our proposed research will contribute to the recently funded Sea Ice Prediction Network (SIPN), which is partially supported by NOAA.

This project will center around five hypotheses and questions regarding Arctic sea ice and its changing variability. We expect that in the future: (1) The variability of ice area will increase and its predictability will decrease, (2) Expanding open water area will promote more extreme cyclones, and (3-4) Regional sensitivity to extreme sea ice variations will change, as will the predictability of ice conditions relevant for marine access. We will also address (5) What are the observational network requirements to realize the predictive capability of sea ice conditions?

Through testing these hypotheses this project will advance our knowledge in all three priorities of the CVP competition: identifying the mechanisms responsible for (extreme) sea ice variations (hypothesis 1 and 4); focusing on regional mechanisms and predictability of sea ice variability, which are highly relevant for marine access in the Arctic (hypotheses 3 and 5); and identifying the key driving factors under changing climatic conditions within a fully coupled modeling framework (hypotheses 1, 4, and 5). By improving predictive understanding of Arctic variability and extreme events, our research aligns with the objectives in NOAA’s Next Generation Strategic Plan (NGSP).

We will test these hypotheses with the CESM-CAM5 model and its large ensemble simulations(30+) spanning nearly two centuries (1920-2100), in addition to a comparison with other CMIP5 models. Our research complements and enhances SIPN through the predictability of extreme sea ice variations and their associated drivers/impacts. We will use a variety of metrics to quantify predictability and analyze the associated atmospheric and oceanic conditions Some outcomes of this project will be: (1) Identification of Arctic regions most susceptible to extreme sea ice variations, (2) Improved understanding of the predictability of Arctic marine access, and (3) Assessment of observational network requirements and extreme weather impacts.