Predicting an Ice-free Arctic using a Nonlinear Endogenous Co-trending Regression Model

This paper predicts the emergence of an ice-free Arctic using a nonlinear endogenous co-trending regression model. The model captures the nonlinear co-trending relationship between Arctic sea ice extent (SIE) and regional surface temperatures, which are influenced by long-run factors such as radiative forcing from greenhouse gases (GHG) and the Atlantic Multidecadal Oscillation (AMO) index. By conditioning the analysis on various Representative Concentration Pathways (RCPs) for future GHG emissions and projected AMO cyclical patterns, we offer long-term predictions of Arctic SIE. Our findings indicate the likely emergence of an ice-free Arctic before 2050 under the worst-case emission scenario.