Secular non-linear trends and multi-timescale oscillations of regional surface air temperature in eastern China

Abstract

ABSTRACT: Secular trends and multi-timescale oscillations of global and regional surface airtemperature (SAT) have been widely investigated in recent decades. In this paper, the method ofensemble empirical mode decomposition (EEMD) is first used to decompose 18 station SAT anom-alies in eastern China from 1921 to 2011; 16 stations are then grouped into 4 regions based on thesecular non-linear trend and multi-timescale oscillations longer than 10 yr in these SAT anomalyseries. The analysis showed that multi-timescale oscillations commonly exist in eastern China,while the 2 largest regional warming trends of 3.3 and 1.21°C per century exist in Northeast Chinaand North China, respectively. Four regional secular non-linear trends and multi-timescale oscil-lations were explained and used in the extrapolation for the coming 2 decades. Although there isa secular warming trend, the predicted SAT anomaly for Northeast China is expected to continuethe recent warming slowdown in this decade and then a new warm flat in the following decade.For other regional SAT anomaly predictions, North China will have a slight warming trend in thecoming 2 decades, central China will have a warming process in this decade then a new warmspell in the 2020s, and South China will maintain the current level in this decade then have aquick warming process in the 2020s.KEY WORDS: Secular trend · Multi-timescale variability · Surface air temperature · Eastern China ·Climate prediction