| 水文序列小波分析中分解层数选择方法 |
| |
| 引用本文： | 桑燕芳,王中根,刘昌明.水文序列小波分析中分解层数选择方法[J].水文,2012(4):1-7,73. |
| |
| 作者姓名： | 桑燕芳 王中根 刘昌明 |
| |
| 作者单位： | 中国科学院地理科学与资源研究所陆地水循环与地表过程重点实验室;中国科学院水资源研究中心 |
| |
| 基金项目： | 国家重点基础研究发展计划(973)项目(2009CB421305);国家自然科学基金项目(40971023) |
| |
| 摘 要： | 小波分解层数的合理选择是水文序列小波分析结果的重要影响因素。在详细分析和定量描述不同类型噪声的能量分布规律的基础上,依据水文序列中确定成分和噪声成分的能量分布规律的差异,提出了一个小波分解层数选择方法。通过对不同类型模拟序列和不同特性实测水文序列进行分析,验证了所提方法的有效性和实用性。结果表明:序列组成、噪声含量等因素对小波分解层数的选择结果有较大影响,但噪声类型对小波分解层数选择结果影响较小;应用该方法确定小波分解层数的同时,还可以有效地识别和区分各层上是确定成分或是噪声成分,进而可为序列模拟预测提供依据。由于所提方法基于水文序列不同成分变化特性的差异建立,因此有较好的物理依据且分析结果合理可靠。
|
| 关 键 词： | 水文时间序列 小波分析 分解层数 蒙特卡罗试验 能量分布 |
Wavelet decomposition level choice method for hydrologic series analysis |
| |
| SANG Yanfang,WANG Zhonggen,LIU Changming.Wavelet decomposition level choice method for hydrologic series analysis[J].Hydrology,2012(4):1-7,73. |
| |
| Authors: | SANG Yanfang WANG Zhonggen LIU Changming |
| |
| Institution: | 1,2(1.Key Laboratory of Water Cycle & Related Land Surface Processes,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China;2.Center for Water Resources Research,Chinese Academy of Sciences,Beijing 100101,China) |
| |
| Abstract: | Decomposition level choice is a key factor influencing the wavelet aided hydrologic time series analysis.In this paper,the energy distributions of diverse noise types were quantitatively analyzed and described firstly,and then based on the energy dis tribution difference between deterministic components and noise in hydrologic series data,a decomposition level choice method was proposed.By applying to some typical synthetic series and observed hydrologic series,the effectiveness and applicability of the proposed method have been verified.The results indicate that the stochastic characteristics and noise content of the analyzed hydro logic series have great influence on the choice of decomposition level,but the noise type used has no relation with the decomposi tion level choice results.By using the proposed method suitable decomposition level can be chosen,and also the deterministic components and noise in hydrologic series data under different temporal scales can be accurately identified and distinguished,which can provide useful guide to hydrologic series simulation and prediction.In summary,the proposed method has reliable physical ba sis because of based on the different characteristics of diverse deterministic components in hydrologic series. |
| |
| Keywords: | hydrologic time series wavelet analysis decomposition level Monte-Carlo experiment energy distribution |
| 本文献已被 CNKI 等数据库收录！ |
|
