Interdecadal opposite variation of December–January blocking days between the eastern North Atlantic and Ural region around 2008

Abstract

Based on observational datasets during 1980–2021 and the simulations from the PAMIP (Polar Amplification Model Intercomparison Project), this paper indicates that the eastern North Atlantic–Ural blocking days in December–January is characterized by a zonal dipole pattern which undergoes an interdecadal variation around 2008. Specifically, the blocking days in Ural (eastern North Atlantic–western Europe) regions increases (decreases) after 2008. Further research shows that the variation in the blocking days dipole (BDD) pattern is closely associated with the changes in the simultaneous background atmospheric circulation field including westerly winds, vertical shear of zonal winds, and meridional potential vorticity gradient over the eastern North Atlantic–Urals region. Moreover, after 2008, the positive North Atlantic Oscillation–like phase intensifies, a quasi-stationary eastward-propagating Rossby wave from the North Atlantic is excited, and a quasi-barotropic “−, +, −” tripole circulation anomaly in the eastern North Atlantic–Central Siberia is formed in the mid–upper troposphere. It is also found that, since 2008, there is a remarkable negative correlation between the Barents–Kara sea-ice concentration (SIC) and the BDD. After 2008, the rapid decrease in Barents–Kara SIC, dominated by newly formed SIC, is conducive to an increased local upward surface turbulent heat flux, further exciting Rossby waves and favoring the tripole circulation anomaly. Additionally, the atmospheric internal dynamical processes excited by the tripole circulation anomaly have positive feedback effects on it, contributing to strengthening (weakening) of the local background circulation field in the eastern North Atlantic–western Europe (Urals) region, thereby causing the blocking days to decrease (increase). The responses of the main physical processes to the Arctic sea ice reduction in the PAMIP models are roughly in agreement with the observations.