摘要
引入子波变换法对香港站29 a月平均降雨1δ8O时间序列进行了分析,并对分解后的序列进行了重构。结果表明,重构后序列与原序列吻合,且香港站降雨氧同位素时间序列存在两个时间尺度上(65、12个月)的周期性变化规律,该法有效可行,可供借鉴。
The method and theory of wavelet packet transform is introduced to decompose and reconstruct time series of monthly mean composition of oxygen-18 in rainfall. Complicated time series of monthly mean composition of oxygen-18 in rainfall can be decomposed to different sub sequences in different time scales by wavelet packet transform, based on which medium and long-term forecasting models of oxygen-18 in rainfall can be built. The paper takes time series of monthly mean composition of oxygen-18 in rainfall in the Vienna station and Hongkong station as examples to conduct time series decomposition and reconstruction by wavelet packet transform. The results showed that the reconstructed time series was consistent with the original time series. Multi-Resolution Analysis （MRA） and periodicity analysis were conducted with db4 wavelet function on the time series, which indicated significant periodic variation at the scale of 5.4 years and 1 year in the time series.
出处
《水电能源科学》
北大核心
2009年第2期11-13,共3页
Water Resources and Power
基金
国家自然科学基金资助项目(50379008)
关键词
降雨
氧同位素
时间序列
子波变换
rainfall
oxygen isotope
time series
wavelet transform
