Abstract
The increased extreme warm and decreased extreme cold temperature events across the Arctic strongly influence the natural environment as well as the societal activities. This study investigates temporal and spatial variability of wintertime extreme high and low temperature events defined by the 95 and 5% percentiles across the Arctic and subarctic regions, respectively (north of 60°N) using data from 238 stations in the Global Summary of the Day for the period 1979–2016. Empirical orthogonal function analyses indicate that the first modes (which account for 30–35% of the total variance) are out‐of‐phase between northern Europe, western and central Russia, and northeastern North America, and that this appears to be related to the Arctic Oscillation ( AO ) and the Northern Atlantic Oscillation. The second modes explain about 8% of the total variance. During the positive phase of the first and second modes the anomalous northeasterly and northerly winds decrease Arctic extreme high and increase extreme low temperature occurrences; while the anomalous southerly and southwesterly winds have the opposite effect. Symmetric and asymmetric effects of the AO index on extreme temperature events refer to the difference and sum between the composite of its positive and negative phases. The symmetric components of the spatial patterns are similar to those of the first modes. The asymmetric components occur mainly over western and central Russia for extreme high and low temperatures, respectively. In addition the impacts of six other large‐scale climate modes are also explored.