Quantifying the internal variability in multi-decadal trends of spring surface air temperature over mid-to-high latitudes of Eurasia

Abstract

This study investigates the influences of internal climate variability (ICV) on Eurasian spring (March–May) surface air temperature (SAT) trends based on a 30-member ensemble of the Community Earth System Model version 1 (CESM1) from 1979 to 2014. The SAT trends over the mid-high latitude Eurasia vary considerably among individual simulations, indicating a large impact of ICV on SAT trends at regional scales. The leading pattern of internally generated SAT trend variability over Eurasia among the 30 SAT trends displays same sign distribution over the mid-high latitudes of Eurasia. This pattern is associated with the Northern Hemisphere annular mode (NAM) that accounts for about 1/3 of total variance of the spring sea level pressure (SLP) trends. The second pattern of internally generated SAT trend variability features a dipole structure, accompanied by prominent circulation changes over the North Atlantic and Europe. Internal dynamics account for approximately 40–70% of the observed warming over the central Eurasia and the Russian Far East. A dynamical adjustment methodology is employed to remove the effects of internal atmospheric circulation variability. Dynamical adjustment greatly reduces the spread of SAT trends within the ensemble and brings both the simulated and observed trends closer to the ensemble mean. Changes in the Barents–Kara Sea ice and Eurasian snow cover contribute to the remaining spread in SAT trends over the north coast of Russia and in the region north of the Caspian Sea, respectively. Both dynamics and thermodynamics contribute to the Barents–Kara Sea ice and Eurasian snow cover trends.