On Multi-Timescale Variability of Temperature in China in Modulated Annual Cycle Reference Frame

Cheng Qian,Zhaohua Wu,Congbin Fu,Tianjun Zhou
DOI: https://doi.org/10.1007/s00376-009-9121-4
IF: 5.8
2010-01-01
Advances in Atmospheric Sciences
Abstract:The traditional anomaly (TA) reference frame and its corresponding anomaly for a given data span changes with the extension of data length. In this study, the modulated annual cycle (MAC), instead of the widely used climatological mean annual cycle, is used as an alternative reference frame for computing climate anomalies to study the multi-timescale variability of surface air temperature (SAT) in China based on homogenized daily data from 1952 to 2004. The Ensemble Empirical Mode Decomposition (EEMD) method is used to separate daily SAT into a high frequency component, a MAC component, an interannual component, and a decadal-to-trend component. The results show that the EEMD method can reflect historical events reasonably well, indicating its adaptive and temporally local characteristics. It is shown that MAC is a temporally local reference frame and will not be altered over a particular time span by an extension of data length, thereby making it easier for physical interpretation. In the MAC reference frame, the low frequency component is found more suitable for studying the interannual to longer timescale variability (ILV) than a 13-month window running mean, which does not exclude the annual cycle. It is also better than other traditional versions (annual or summer or winter mean) of ILV, which contains a portion of the annual cycle. The analysis reveals that the variability of the annual cycle could be as large as the magnitude of interannual variability. The possible physical causes of different timescale variability of SAT in China are further discussed.
What problem does this paper attempt to address?