Extracting Periodic Components in Temperature Fluctuation Patterns

Based on monthly average air temperature of Shapingba in Chongqing over the period from 1958 to 2007, the cycles on various scales of temperature fluctuation patterns are analyzed by Empirical Mode Decomposition(EMD). After applying EMD to the data, a series of cycle modes and a trend are obtained respectively, including short periodic oscillations and medium periodic oscillations. Their corresponding average periods are calculated by the Fast Fourier Transform Algorithm (FFT). Those components of periods like 1 year, 1.5 years and 2.78 years belong to the former; while the periods of 7.14 years and 10 years belong to the latter. Besides, the monthly average air temperature in Shapingba of Chongqing took on descending trend before 1984, while ascended since then, which coincide with the previously obtained consequence. All the above results indicate the air temperature fluctuate on various time scales and enlighten us to find its possible reason.