Dynamic and Thermodynamic Control of the Response of Winter Climate and Extreme Weather to Projected Arctic Sea-Ice Loss

A novel sub-sampling method has been used to isolate the dynamic effects of the response of the North Atlantic Oscillation (NAO) and the Siberian High (SH) from the total response to projected Arctic sea-ice loss under 2 degrees C global warming above preindustrial levels in very large initial-condition ensemble climate simulations. Thermodynamic effects of Arctic warming are more prominent in Europe while dynamic effects are more prominent in Asia/East Asia. This explains less-severe cold extremes in Europe but more-severe cold extremes in Asia/East Asia. For Northern Eurasia, dynamic effects overwhelm the effect of increased moisture from a warming Arctic, leading to an overall decrease in precipitation. We show that the response scales linearly with the dynamic response. However, caution is needed when interpreting inter-model differences in the response because of internal variability, which can largely explain the inter-model spread in the NAO and SH response in the Polar Amplification Model Intercomparison Project. The projected loss of Arctic sea-ice under 2 degrees C global warming will cause large warming in the Arctic region and climate and weather anomalies outside the Arctic. The warming in the Arctic will mean warmer airmasses coming from the Arctic and also more moisture from the open Arctic Ocean. Furthermore, it will also change atmospheric circulation. These effects together will determine the impacts of Arctic warming. In this study, we introduce a novel sub-sampling method to isolate atmospheric circulation change in response to the Arctic warming. The method involves selecting members of simulations from the experiment with future Arctic sea-ice conditions, the average of which is equal to the average of the members of simulations in the experiment with present-day Arctic sea-ice conditions. We found that atmospheric circulation change in European regions is relatively weak so that warming effects will dominate the climate and weather response there. On the other hand, atmospheric circulation change will dominate the climate and weather response in East Eurasia. We also found that stronger atmospheric circulation changes will generally increase the response to the Arctic warming. We suggest caution when assessing whether different responses in different models can be interpreted as true differences in model physics. A novel sub-sampling method is introduced to isolate the role of dynamics in the response to projected Arctic sea-ice loss A dynamical Siberian High response dominates the temperature response over East Eurasia while that of the North Atlantic Oscillation is weak Inter-model differences in Polar Amplification Model Intercomparison Project likely contain a large fraction of internal variability due to the unconstrained dynamic effects