Optimal large-scale wind patterns favoring strong sea ice melting and ice-albedo feedback in the absence of anthropogenic forcing

Arctic sea ice has declined rapidly over the past 40 years, driven by a combination of increasing greenhouse gas emissions and internal variability. Many approaches have attempted to separate these drivers in both observations and modelling studies. However, uncertainties remain regarding the representation of key high latitude processes and reconciling model simulations with observations. To address this, we conduct several simulations in which we constrain atmospheric circulation by nudging mid- to upper tropospheric Arctic (70-90N) winds (U,V) within the Community Earth System Model (CESM1) to those from reanalysis using fixed greenhouse gas concentrations. We find September Arctic sea ice has a sensitivity of ~0.016 million km² per m upper geopotential height (Z300). Additional nudging experiments are run, with individual years from the reanalysis repeated for up to 21 model years, to examine the quasi-equilibrium response to wind forcing and its influence on long-term changes via the ice-albedo feedback. When wind patterns favoring strong sea ice loss are repeated to reach a quasi-equilibrium state, summertime sea ice sensitivity to wind forcing can as much as triple. As sea ice continues to melt and greenhouse gases continue to increase, understanding the impact of these wind patterns over a broad range of timescales will become increasingly important to future projections and forecasts of Arctic sea ice.