Atmospheric Contributions to Observed and Simulated Arctic Sea Ice Loss
Sea ice is in many ways the defining characteristic of the Arctic, exerting a powerful influence on the climate and ecology of this unique region. It has become clear that Arctic sea ice is disappearing, and the rate of recession appears to be accelerating. Recognizing the prominent influence of the atmosphere on ice circulation and the Arctic heat budget, much research into this trend has focused on potential atmospheric drivers. However, establishing robust atmospheric links to declining ice cover has proven difficult. This may be partially due to the treatment of ice as an afterthought to atmospheric variability, the application of linearly constrained statistical methods, and the inadequate representation of regional influences. Nonlinear machine learning applications present alternative approaches to coupled ice/atmosphere analyses that avoid these problems. Given the nonlinear nature of observed and simulated ice loss, these nonlinear methods may be particularly well-suited to this area of study. This project will use a range of machine learning methods as the basis for several related analyses of atmosphere/ice relationships in observations and climate model simulations. New approaches to identifying modes of joint atmosphere and ice variability with be applied, driving forces behind abrupt ice loss will be examined, and the source of common biases in climate model representations of ice cover will be investigated.
30 Nov -0001
Strategic Research Theme
Arctic and Northern Regions
Oceans, Fisheries and Aquaculture