参考作物腾发量多时间尺度分析的小波变换  

Multi-time-scale analysis of evapotranspiration based on wavelet transform

作　　者：任玉敦[1,2] 崔远来[1] 刘路广[1] 

机构地区：[1]武汉大学水资源与水电工程科学国家重点实验室,湖北武汉430072 [2]浙江水利水电专科学校,浙江杭州310018

出　　处：《武汉大学学报（工学版）》2009年第1期50-55,共6页Engineering Journal of Wuhan University

基　　金：国家自然科学基金项目(编号:50579059);教育部博士点基金项目(编号:20050486002);国家科技支撑计划子题(编号:2006BAD11B08-3)

摘　　要：采用Penman-Monteith方法对国内4个不同气候带气象站点的长系列逐日参考作物腾发量ET0值进行了计算,以昌平站ET0年日均值和月日均值时间序列为例,对比采用不同母小波进行小波变换的结果,确定Morlet小波为ET0多时间尺度分析的最优母小波.以Morlet小波为基础,对4个站点ET0多时间尺度特征进行分析,结果表明:4个站点ET0多年来均存在多个时间尺度上的周期变化,其中以5、10 a最为普遍和显著.小波分析的结果,对于中长期不同时间尺度ET0的预测和灌溉用水预测具有指导意义.Based on the weather data from four meteorological stations which located at different regions in China, long-term daily evapotranspiration （ET0） in each station is calculated by Penman-Monteith equation. The long-term variation of annual daily ET0 are investigated by moving average method and Mann- Kendall test; and it is shown that an increasing trend in Changping Station, but a decreasing trend in Huibei Station. Tuanlin Station and Kunming Station; the trend are also quantified by Mann-Kendall test. Taking the Changping Station for example, two mother wavelet functions are chosen to investigate the multi-time-scale features of ET0 time series, including annually and monthly daily ET0. Morlet wavelet function is identified to be the best one and then used in the other three stations. The results show that ET0 time series in four stations have periods in different time-scales, especially in 5 and 10 years, which can also give advice for middle and long-term ET0 and irrigation water use planning under different timescales.

关 键 词：参考作物腾发量 小波变换 多时间尺度 

分 类 号：S161[农业科学—农业气象学]