Year-to-year variability of surface air temperature over China in winter

In this study, the year-to-year variability of surface air temperature (SAT) over China in winter was investigated during 1961-2014. The results indicate that the year-to-year SAT variability can explain more than 30% of the SAT variance over most parts of China, with up to 60% variance over southern China, northeastern China and northwestern China. The leading pattern of year-to-year SAT exhibits homogeneous variability over most parts of China, except in small areas over the Tibetan Plateau and southwestern China. The circulation over the northern Pacific is a key factor of this homogeneous variability pattern. An anomalous anticyclonic circulation and weakening midlatitude westerly jet in the middle and high troposphere over the northern Pacific are associated with the homogeneous warmth over China. The second pattern shows a south-north dipole, with variability in northeastern China opposite to that south of 25 degrees N in China and over the Tibetan Plateau. The south-north dipole pattern is part of a global year-to-year SAT anomaly pattern because it exhibits a significant relationship with the year-to-year SAT over large parts north of 50 degrees N over the Eurasian landmass. The north-cold/south-warm pattern is accompanied by a significant weakening of the Arctic Oscillation. In comparison, the atmospheric circulation anomalies associated with the homogeneous variability pattern are much weaker than those with the south-north dipole pattern. The anomalous Indian Ocean dipole in the previous autumn and the snow cover around China in November are the two key causes of the homogeneous variability pattern. Many factors, such as the tropical central Pacific sea surface temperature (SST), stratospheric quasi-biennial oscillation (QBO), Okhotsk sea ice and western Siberia snow cover, can significantly influence the south-north dipole pattern. Compared to the tropical Pacific SST, the impact of the Indian Ocean SST on the winter SAT over China is much more important.