How Robust is the Atmospheric Response to Projected Arctic Sea Ice Loss Across Climate Models?

We assess the reliability of an indirect method of inferring the atmospheric response to projected Arctic sea ice loss from CMIP5 simulations, by comparing the response inferred from the indirect method to that explicitly simulated in sea ice perturbation experiments. We find that the indirect approach works well in winter, but has limited utility in the other seasons. We then apply a modified version of the indirect method to 11 CMIP5 models to reveal the robust and non-robust aspects of the wintertime atmospheric response to projected Arctic sea ice loss. Despite limitations of the indirect method, we identify a robust enhancement of the Siberian High, weakening of the Icelandic Low, weakening of the westerly wind on the poleward flank of the eddy-driven jet, strengthening of the subtropical jet, and weakening of the stratospheric polar vortex. The surface air temperature response to projected Arctic sea ice loss over midlatitude continents is non-robust across the models.