陕北年径流序列的谱分析研究

径流周期识别是径流模拟预测的依据。基于功率谱和极大熵谱分析原理,进行了陕北地区12个测站的年径流序列的周期识别。结果表明：最大熵谱分析具有高分辨率的周期识别的特点,陕北地区的年径流一般具有2－4年、11年左右和18－24年的水文周期。对于年径流序列长度大于35年的测站,功率谱分析和最大熵谱分析的分析结果一致,但是对于序列长度小于35年的序列,功率谱分析表现出分辨率低的不足;最大熵谱分析具有分辨率高的特点,提取的主要周期符合实际。

Spectral Analysis of Annual Runoff Series in Shanbei Area

To identify hydrological runoff period is the basis for runoff simulation and forecast. Based on the principle of power spectrum and maximum entropy spectral analysis, this paper identified the annual runoff series periods of 12 differernt stations in Shanbei area.The results showed that the maximum entropy spectral analysis has high resolution. There were three significant periods of 2～4 years, about 11years and 18～24 years at Shanbei area. If the length of the annual runoff series is greater then 35 years, two methods have the similar analysis results. But the length was less than 35 years, power spectral analysis has the low resolution. Compared with power spectral analysis, maximum entropy spectral analysis has high resolution. The extracted periods are in accordance with reality.