基于Morlet小波的ET0序列多时间尺度分析

Multiple time scale analysis of ET_0 series based on Morlet wavelet

以某地区1984-2003年逐月的参考作物腾发量（ET0）为基本资料，首先对ET0时间序列进行了Dmey小波消噪和距平处理，然后采用Morlet复小波对其进行多时间尺度分析，显示了ET0在各个时间尺度下的演变特征．结果表明，ET0具有120月、55月、40月和16月左右的主要周期变化，这4种主要周期变化主导了ET0在整个时域上的波动变化特性；并且自2003年后期以来，ET0变化的大趋势是从低于平均水平逐渐增大并会持续若干月．

With the monthly reference crop evapotranspiration from 1984 to 2003 of Liuyuankou Irrigation District as the basic information, the reference crop evapotranspiration time series is denoised by Dmey wavelet and dealt with anomalies at first. And then the multiple time scales of it are analyzed using Morlet complex wavelet. The evolution characteristics of it in various time scales are displayed. The result shows that the main periods of the reference crop evapotranspiration are around 120 months, 55 months, 40 months and 16 months respectively. The four kinds of period variations lead the reference crop evapotrans- piration fluctuation in the whole time domain. And the major variant trend of the reference crop evapo- transpiration is increasing gradually from the lower level since the later 2003. It will be sustained for several months.